UCLA CENTER FOR HEALTH POLICY RESEARCH HEALTH ECONOMICS AND EVALUATION RESEARCH Second Interim Evaluation of California's Health Homes Program (HHP) Prepared for: California Department of Health Care Services (DHCS) March 2022 UCLA CENTER FOR . HEALTH POLICY RESEARCH eo: ee, % a, of 1 www.healthpolicy.ucla.edu Second Interim Evaluation of California's Health Homes Program (HHP) Nadereh Pourat, PhD Xiao Chen, PhD Brenna O'Masta, MPH Anna Warrick Leigh Ann Haley, MPP Weihao Zhou, MS UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 This evaluation was supported by funds received from The California Endowment and the California Department of Health Care Services. The analyses, interpretations, and conclusions contained within this evaluation are the sole responsibility of the authors. This report contains analysis of data available up to September 30, 2020. Acknowledgments The authors would like to thank Maria Ditter, Wafeeq Ridhaun, Igor Geyn, and Sally Yao for their hard work and support of HHP evaluation activities. Suggested Citation Pourat N, Chen X, O'Masta B, Warrick A, Haley LA and Zhou W. Second Interim Evaluation of California's Health Homes Program (HHP). Los Angeles, CA: UCLA Center for Health Policy Research, March 2022. UCLA Center for Health Policy Research Nae eP es Health Economics and Evaluation Research Program Contents Second Interim Evaluation of California's Health Homes Program (HHP) ..........ssscccssscesssseesseeeeees 1 EX@CUTIVE SUIMIMALY.........cccccccccsssesssssscccccceeesesessnscceassecesesaeesenecaeaesesecesasessuscaesesaeeeesaesesuscaasssaceeneneens 21 Health Homes Program (HHP) Overview ..........scccsssscssseccsssseccsssccessresessncesessnecesssensssesessnesensnanens 21 Evaluation Methods...............:cesecesesccesceceneneeeceneceeeacenenenenenesaeesseneneneneneneaeeseneneeseeceeeasensnenenenenees 21 RESUITS ........ccccsscecssseeesssceeesseetenenenesaneeasauseteneeecesaeeceaeseneneeeeaaeeenanenenaeesecaeeeteneeensnaeecenaesensueneesaeees 22 HHP Implementation and Infrastructure ..........ccssscccsseecesstecesseeeesaesetesaceeseaceeenseeesssaeeesaaes 22 HHP and COVID-19...... ee eeccseceseessreesccseceseesseseaeeseceseeeeesseesacesaeeseesssseaeecaeesesesesseessaeesesenese 22 HHP Enrollment and Enrollment Patterns ...........ccccsccsscsesesessseeeseeceseeseeseeseseessseeeseeseres 23 HHP Enrollee Demographics and Health Status .............ccssssccccesssseesessscecesssseseeesesseeenens 24 HHP Service Utilization among HHP Enrollees................cccssccessnceesscseresereneneneeessecensnenensnees 24 Acute Care Utilization Groups in HHP.............:cccsseccsssceesenenessneeesaneeseneeensnaeecesaesensnenensaeees 24 HHP OUTCOMES... ce eccceeseeeeeseseceeenessoneeenaseceeeeessononesenasecaeensansnesesesesceesenesecaeeseseceetasenoseeaeenass 26 Estimated Med-Cal Payments for HHP Enrollees and HHP Coste............:.ccsssssecessssseeeeees 32 Conclusions and NeXt Steps ..........ccessssccccssssscecsssssceeessssceceesssaceesesseaeeceeceseseceesssneeeceeseaaseseeseaaees 33 INtPOCUCTION ......cceceseseseseseceseceeseneseceseensneesesaseseeaceensesesesasesacaceeeesesascesececsceeseseseseeacedeneesesesasesaeanees 36 Health Homes Program OVErview .......:ccccsscccssscecsssecessseeeesseeessanecesseeeecaneeseneeecenaeecenaesecsuaeeesnaeees 36 HHP Implementation Plan ...........cccccccsssssscccssssstcceesssseccecssssaeeecesssaeeceessseeeceseeseaeeeeesseseceessseaeenees 36 HHP ServiCes ......csesecscssccecsccesesseessseeesssesesssseessseesescsesecseacecscecseseeasaseeseceteeseseeesssseeesceeseseseeesoeeens 38 HHP Target Populations. ...........cccsssscccsssssssccesssscceesssscececessneceesessssseeenessneesesessaaseseceuausesesenanaqenens 39 Funding and Payment Methodology ..........::ccssccccssecessseeesssecessenenenenenesaneetenenensneeeeesaesensnenensneeees 40 Transition to CalAIM .........eesceseeseceeecceseceneneneseeeseeneneseseneeaceeeasenenenenaeesaeeesesenenensaeeeseesenereneneenens 41 UCLA HHP Evaluation ...........ceceeccececccesceceeeesecescceesacensneneseeesaeesseseseceesaeeesaesenesesaeesaceeeasensnasenaeesaes 41 Conceptual Frame@work..........cccssccccsssecsssnecsssncecssscecssaceesnsceeeneeessneecssneecssanecseaeeeseseeeessaeeesns 41 Evaluation Questions and Data SOUrCES ..........sssssecesstesseeseceseseseseeeeseeeseceseeseseeseseeeseeeeenens 43 HHP Implementation and Infrastructure............ccccsssssccessssscccessssseeesesssecceeessseeeecessaeeesesssasecesenaes 45 HHP IMpleMentation ........ccccesccsssecesscecssecessscecessecessnesensnesecssnesenneteneseeseneeecsnesenesaasensnasensnaeens 46 HHP Delivery Models ..........ccscsccessrccesenceceseeeessaceccsnecensneeeesseeessanetenanesasaneetenenenenaeecenaesensnenensaeens 47 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program HHP Delivery N@tworkS..........c:scccsssccsssncecesececssacecsseeesseeeessaeessanenenaeeeacaneeteneeeceuaeecenaesenseeeesaeens 47 CB-CMEs by Organization Type.............:::::ccssceceeeeeseceeeceeseeeesceeeseeeaeneseensaceeaeeeeaseesaaeeneeeaes 47 CB-CMEs and Projected HHP Capacity.............cccsssscccssssssccecessseceesssssesceesssseeeeesssaeeseeseanees 48 Changes in CB-CME Networks Over Time.........cssssccccsssssecessssscceeecssseseeeesssnccetessssasecesssenes 49 HHP Staffing ...........cccssssscccsssssscccsssssececesssseeecesssseeceessssneeecessaceesessnasesenessusesesessnasesenesausesensnasenecens 49 HHP Datta SHaring...........ccsccccseneneseneeesansensnenensnacecenaeeensnesensseeessanesenenetasaneeteeeenenesecenaesensnesensnaeens 50 Communication with HHP Enrollees...............::secccseceseceeeseeeeeeceneeeseeeeeceeeaeensneneeeeseeeeseeeneneeenetens 50 HHP and COVID-19...........cccccssssecesseesssneeesseeeessaeecssaeeessaeeeesaaeessanesenaeecasauesesaasenenaeesenaeeceaaeeecssaeeesaaes 51 Progression of COVID-19 Cases and Hospitalizations in HHP Counties..............ccsssscceeesseeeeees 52 Impact of COVID-19 on HHP Implementation and Infrastructure...........ccccsssssccecssstseceessseeeeees 54 Impact of COVID-19 on MCP Processes, Procedures, and Policies .............ccssscccsessseceeees 54 Transition to Telephonic Delivery of HHP Outreach, Engagement, and Services. ............ 58 Contribution of HHP to MCPs' COVID-19 Pandemic RESPONSE ..........::sccecesecessserersseeeeees 60 Estimated Prevalence of and Trends in COVID-19 among HHP Enrollees and their Controls . 60 COVID-19 Related Service Use of HHP Enrollees and their Controls................:esecesseesererererees 61 Changes in Use of Health Services Before and During the COVID-19 Pandemic ................006 62 HHP Enrollment and Enrollment Patterns ..............c:cccesccesceseceeseceseceeeceeeseeseneneseeesnensaeensneseeeeesaes 65 Trends in Enrollment............::cscccseceeecceseceeeeeeeseeesaeseeeesaeeesceesaaeeaeeesaensaeseneeeeaeeeeaeeeeaeneneeaeaeeenens 67 Growth in HHP Enrollment Overall and by SPA .Q..........ccccccccssssececsssssecccessssreeeeesssseeetsessanees 67 Growth in HHP Enrollment among Homeless by SPA............sssccccssssssceesssscccetesssseseeesssenees 69 Enrollment Size by Group and COunty...........ccssscsccssssscccsesssececesessscesesesssesesessssesesensssnesenens 70 Enrollment from the Target Engagement List .............ccccsescceserecesssetesenenenseeeesserensnenensaees 71 Enrollment Patterns ............::ccsecccecceeseeceneneseeesaceeeaseneneneneeesaeeseseneneesaeneaeeseaeneeeesaceeeasensnenenaeesaes 71 Enrollment CHUII..........:cc2sccescceseceeseceeeceeeeeeeeeneseensaeesaceesasensaesesaeesaeseseeeeaeeeeaesenaeseesesaeeenees 71 Enrollment Length ...........ccsssscccsssssscecssssscceeesssscceesssceesesesssaeeceessseeeeesessaseseessesecessssaeeeess 72 MCP Exclusions of Specific HHP Eligible Populations ..............ccssscccesscsecesessseseeesssssecesesseeeetens 72 HHP Enrollee Demographics and Health Status .............cccssccccssssseecsssseeceeesssseceesesseeeesesssesesenenaes 74 Demographics of HHP Enrollees at Time of EnrollMent...........:.:cccccessseeseseeeneseneeeseesessnerensneeees 75 Health Status of HHP Enrollees Prior to Enroll Ment.............ccccecsesesccececccccsesenecececcscessrenerercseneee 76 UCLA Center for Health Policy Research Nae TeP es Health Economics and Evaluation Research Program HHP Service Utilization among HHP Enrollees.............:ccccsssccssseecessnecessneeesansereneceescaeecenneeenssaeeesnees 78 HHP Service ......cceccceseseceeeseeseeaeasescceeenaesseaeaeaseneeaeeasesaeasesasecaeaesasseasasesesecaeseuseeaeeseseceesasesesecasenase 79 Estimated Overall HHP Service Delivery to HHP Enrollees.............ccccssssccccessssceceesssnsecessssceeeeees 81 Estimated Types of HHP Services REC@IVEd ............csssscccesssstcecessssecceeessecesesessaeececsssnseeeeesseaeetens 83 Estimated HHP Core Services by Modality and Staff Type ...............ccccsssssccsssssseeesesstsesensssseeecees 84 HHP Housing ServiCess ............ceceeeeeeesecceceneeeneneneneceneneneneneneenaauaneneneneneseeneseneneneseneeeneenensneneneneeess 85 Acute Care Utilization Groups in HHP............:cccssscecessecessnecessseeeesanenenaeeeeeaneeteneeenseaeecenaesensueneesaeens 87 Acute Care Utilization of HHP Enrollees ...............:cessceseceeeceeeecereceeeseeeseeeeseseseceeeceeeeaesenesesaeetenens 88 Acute Care Utilization of HHP Implementation Groups ............ccscccsstecssseecsssecssseeeseseceesseneeees 89 Demographics of HHP Enrollees by Acute Care Utilization Group .............csscccecssstsecesessreeeees 91 Health Status of HHP Enrollees by Acute Care Utilization GroupS..............cssssccessssrsecesessseeenees 92 Health Service Utilization Trends of Acute Care Utilization GroupS............::eseccesscessserenenenees 93 Trends in Primary Care Services ..........::cccsscccsssecessnecesseetenanenenaneeesaesetenenenseaeeeeaueeensneneesaaes 94 Trends in Specialty SErvice ...........:ccccssccssseessneeecseeeesseenenenecesaneeecesensneeeeseaeeeecseeensneeeesaaes 95 Trends in Emergency Department ViSits ............cccsssccccsssssrecessssreceesssssececessseseesesssaseeensnaes 96 Trends in Hospitalizations.............ccccssececsssceessneeessnecessneenesenenenaeeeesauseteneeenseaeeeeauenensneneesaaes 97 Trends in Admissions to a Long-Term Care Facility from the Community..............:c0ss0 98 HHP OuUtCOMES ......eeceeseeceneeeseeeeeesesececenaseceeeeasenaeeseseceesaseaasecacesaseueneaaseceeasesessoneeasasecaeeneenonseaeeneneea 101 HHP Utilization MEetrics..........cccccssccseseesseessseseseeesseseeseessceseseeessecsseceseeesseeesssasesaseseeeseeseseeneeesens 102 Outpatient Utilization 0.0... ecsssscceeessssececessssceeecesscseeecssssaeececensaeeeesesseaeeceeseseseeesesaes 103 Emergency Department Utilization .............cccscsccsssscccessssccceeesscceesssssasesenessnseeesesssenenens 108 Hospital UtiliZation............cccesssccsseeeneseeenssaceessnesensneeessaeetenanenenenecesansenenesensnesecenuenensnenensas 110 Institution Utilization «20.0... ee eeeeeceeeeeecceeeeeceneneneeesceeeacenenensneeesaeesseseseeenaceneaeetenenenenetaes 114 HHP Process Metrics ........ccscsececesesececeeeessnoeeeesesecaeensensnaeesaseneeaetansnecacesaseeesenaseceeasenesecnaeenasenanene 118 Adult Body Mass Index ASSESSMENT ..........cccssccccssscesssscessnscessneeesaneeceeneecseatecssaecessnaeeesnaes 118 Screening for Depression and Follow-Up Plan ..............sssscccssssscceesssesccesessseceeeeseseeeeeesaes 120 Follow-Up After Hospitalization for Mental IIN@SS .............ccccssssccessssseecesesseceeesesseeenens 121 Follow-Up After Emergency Department Visit for Alcohol and Other Drug Abuse or DEPONenCe.........cccccssssececssssncccsesssesececsssceeecssssacecsessaeeececesaeaeecesseaesesesseaseceessausecesesaeaeenens 123 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Initiation and Engagement of Alcohol and Other Drug Abuse or Dependence Treatment 125 Use of Pharmacotherapy for Opioid Use Disorder .............ccccsssssccesesssnceceessssecesesseaeeeees 128 HHP OutcOMe MEUrics ......ccsessccsesscecsssesessseeseseeescseseccccessceceesseesasseeseceeeesessesssseeseseeseseseeesoeses 129 Controlling High BloOd PreSSure.........sscccssssssccesssssccecsssssececesssaeeeeeesseesenessaeeeeeesssesesenenees 129 Plan All-Cause Readmission .............:::cceccccscecesereseseneceeeaceneneneneneaeeseeseneeeneneneaeetenenenenenees 131 Prevention Quality Indicator (PQI) 92: Chronic Conditions Composite ................::0c00 132 Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs ............:ccssssccccssssererenesees 133 Estimated Payments for HHP Services............::ccccssecesssecessneceessnenenenenenneeteneeensnenenenaesensnenenenes 134 Total Estimated Medi-Cal PayMent............ccsscccsssscssseetesenenenaneeesansenenenenseaeecesseeenseneeses 134 Estimated Payments for Outpatient Services ...........cccscccesescesstecessseneseceesseeeenserensseeeeees 135 Estimated Payments for Outpatient Medication. ..............:::cccsssscsecssssscecesessseseessseeeenees 136 Estimated Payments for Emergency Department ViSItS.............:ccccccssssccessssrccesessteeerees 138 Estimated Payments for Hospitalizations ..............ccccccccsssssscecesscsccesessseeecesesseceeesesseeeesens 139 HHP Program Expenditures ...........sccccccssssscccsssssccccsesssecesesssssesessnsnesesessseesesensnaseeecesssasenessnaness 140 CONCIUSIONS ........:ccccsececsseceessaceesenenensnenessaeetenenenenenecenaesensuesensneceeeauesenauenessasasneneneuesenenaeeetenenensnenens 141 Appendix A: Data Sources and Analytic Methods ...........:ccccsssccssenenesseeeesaneetenesenssaeeessaesensnereesaaes 142 Readiness DOCUMEMMS............::::cccecceesceeeseeeeceeseeeesceenseeeeaeeesaensaeneseceeaceesaeennesesaeesaeeteeeeneeeneaetens 142 Analytic Methods............ccscccsssccssstcecsssccsssccesssececssaceesseeeessuecsssusesssusesesaeecseaueessaesecseaeeessaes 142 LIMITATIONS... eeececessccesssseesesseescnseeessssesssseeeesesecessceeesseeesesseaseseeaseseceussseesssssesscneeesoseseees 142 Enrollment Reports and MCP Quarterly Reports ..........cccsssscccsssssececessssecesesssseceeessscsesesesenenees 143 Analytic Methods.............c:cccsesccesscecesenensneeeeessesensnesensneseesanenenanenenaneeteneueneneeenenaesensnenensnens 143 LiPMitATIONS. 202. e eee ceeeesecee eee eeeeneeneneneneneeeeeeeesenenenenacaeaeeesenenenenaaacaeeesenesenenseesceeesenenenensenes 145 Medi-Cal Enrollment and Claims Data ...............:cc:ceeececescceeeeeeseeeseceeceeseeeneneneaeeeseeeeeeeenaneneeetens 145 Analytic Methods. ...........ccscccsssccessncecsscecssaceeensececssaeeesaeeessaaecasuaeasaneeseneeecenaeeesuaeseeseaeeeenaes 145 LiMMitatiONS........ccescssseessssesesseesssseessseeseseeessssaceesseeecseeesssseasssseesesseessssaseseseeesesseseeseaeerees 158 Attributing Estimated Medi-Cal Payments to Claims ..............csssccccssscccccsessseceeesssenceeesesseeeeees 158 BaCkgrounG ........ssccccsssssscccesssscccecessscccenssscceeceesssasenessnenseeesenaneeecessaeeeesessnasesenenansesesensaenenens 158 Service Category Specifications ............ccccsssssccecssssncecsesesececeessseeeesssseaeecsessaeeeseseseeeerenaes 160 UCLA Center for Health Policy Research Nae TeP es Health Economics and Evaluation Research Program Attributing Payments to Specific Services ...........cscccsssccssssecesstcessseecessecessnecesecesssaeeessees 164 Comparison of Estimated Payments with Medi-Cal Paid Amounts ..............ccssssscceseseee 170 LEVITATION. 20.2... eee eee eeeseeee cece eeeneeneneneneneneceeeesenenenenenaeeeeesenenennenacaeeesenenenensnacecesesenenenenaes 171 HHP RatteS..........ceccecececeeeeeeeeee nena eeneneneneneeeeeeeneaeaeeeaeeeseaeaeseneaeceneasesesasasasasaeaeneneaaeaeasanaeeeueueaeaenenene 172 Appendix B: UCLA HHP Evaluation DeSign........... cc ccsssscccssssccccecesssaceceesssnsecesssacceeeesscaeseesessaaees 173 INtPOGUCTION ooo... eee ectcstneneeeeseseseeeseceneccesenaneesaneeececsesecenssansneeccsaseneeeneseneceasesasenesaneeeceesesesenenanens 173 HHP Evaluation Conceptual Framework and QuESTIONS .............::cccccsssessesesssetceecesssnesenessnaeees 174 Data SOUICES .0...ccceceseseseceeeeeeseeesenesecaeeneeseeeaeesencaeeanessaaeesaseceeeetanssacasenaseuesenaseseeasesesecnesenaeenanene 176 Methods. ........ccccccssscssecesesecesseceseeececeesceeseeaseseseseeeseceasacesecasesesecsasaseeasedesecadseeaseseseseessesensaeeseeasens 177 LIMITATIONS... cece estenescceesenenenennenscesensnenensneseccesesenssenensuececseeenencenssedacsesesensneseascedseseneseneneas 180 Tie line...........cccccseneeessneeesaneeteneeenenacecenaesecsneeensnaceccauenensnenaeaueeeenanenenauedasaeeeteneeensuaeecenaesensnenensaaes 181 Appendix C: HHP Enrollees Enrolled Less Than 31 DayS............ccsccssssccsssteeceseeessaeeessaeseessneeesaaes 182 Appendix D: Homeless Enrollment by Group ............cccssssccecssscsccecesssececessssesecesssseeeetesseseeeesesseeees 183 Appendix E: Survey: COVID-19 Impact on the Health Homes Program (HHP)...........:csseceseeesees 185 Appendix F: MCP-Level Descriptives and Unadjusted HHP Core Metrics .............ccccsssssseesssseees 195 Appendix G: Enrollees with More than One Year of HHP Enrollment.............::csecccesssessseresenees 236 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Table of Figures Exhibit 1: General Health Homes Program Acronyms and Definitions .................csssccesessssteeeesees 18 Exhibit 2: Managed Care Plans Acronyms/Abbreviations and Definitions.............cscsscssseceseeeses 20 Exhibit 3: Outcomes for SPA 1 HHP Enrollees as of September 30, 2020............ccsscccssececssseeesees 27 Exhibit 4: Outcomes for SPA 2 HHP Enrollees as of September 30, 2020.............ccssccccsssssreeeeesees 30 Exhibit 5: Timeline of HHP Implementation by Group and SPA wuu........cssccccsssssccceesssseeesessseeeeenenees 37 Exhibit 6: MCPs that Implemented HHP across California, by Group and Countty..............ccceeee 38 Exhibit 7: HHP Services Provided through MCPs and CB-CMES .u.........cssccccsssssscecssssseeesessseeeesenees 39 Exhibit 8: HHP Eligibility Inclusion and Exclusion Criteria..............cccccsssessceceseeenssenecessecensnerensnees 40 Exhibit 9: HHP Evaluation Conceptual FraMe@work............ccscccccssssscessssseccceesssseceesessceeeesessseeeeeeesens 42 Exhibit 10: Health Homes Program Evaluation Questions and Data Sources ..........::cccsssescceeesees 43 Exhibit 11: Distribution of California Counties by Health Homes Program Implementation Group and MCPs Implementing Health Homes Program by COuNty...........:ccccsssssccessssseceeesssstseeessssceeetens 46 Exhibit 12: Health Homes Program CB-CME Network by Organization Type as of September Exhibit 13: Total Projected CB-CME Capacity for Health Homes Program Enrollment by CB-CME Organization Type as of September 2020 ..........cccccccecssscecsseceessnenecaeeeesseeeesseeesesaeeeeeaeeeseneeeessaeeeses 49 Exhibit 14: Cumulative COVID-19 Cases per 100,000, April 2020 to September 2020, HHP Counties And California... .ecesssssccessccecesscesssececessnseessececeseeecsceeeesaeasecaeesessesesssaeesesaeensssaeeeseeserses 52 Exhibit 15: 14-day Average COVID-19 Hospitalization Rate per 100,000, April 2020 to September 2020, Statewide and HHP CounttieS .........ccccccsscsssssssscsssssccsscssesssnsensassccssessessscseceaseeessessessssseceaaaeens 53 Exhibit 16: MCP Reports of Impact of COVID-19 Pandemic on HHP Processes, Procedures, ANG/Or POLICIES ........ccccccccsesscsesseesesscesaccceevenseeueseeesanseeecsaeeaseeeacceueueseasacceasacseesuacaeeeasceaeceeeacecaeeces 55 Exhibit 17: COVID-19 Pandemic Impact on HHP Processes, Procedures, and Policies, from the CB-CME Perspective, Overall and Range by COUNnty............::ccccsseeesenersssteeesseeeneneeeeesaeeetsnetensaeeeeas 56 UCLA Center for Health Policy Research Nae TeP es Health Economics and Evaluation Research Program Exhibit 18: Illustrative Quotes on COVID-19 Pandemic Impact on HHP Processes, Procedures, and Policies, from the CB-CME Perspective ..........ccscccsssccsssceesssececaeeesaeeeesseeeessaeeeeeaeeeseneeeesnaeeeses 57 Exhibit 19: Illustrative Quotes from MCPs on Transition to Predominantly Telephonic Contact during the COVID-19 Pandemic............ccscccesesecesscecsscecersseeecssaeeeennenesenecesaeeeeeseeenenaeeeenaeeetenesensaeeesas 59 Exhibit 20: Illustrative Quotes from MCPs on How HHP Facilitated MCP Response to COVID-19 PANCEMIC.......ccccccsessessescncenccessevencusucucueuseusueveucusucncesasencueueusususcaesaueueueveusuenseusneueueueueusesacencneneneuensenss 60 Exhibit 21: Proportion of HHP Enrollees and their Controls with a COVID-19 Related Service by month, April 2020 to September 2020 ...........ccsssccccssssscccessssccceessssecenesssensceesssacseesessaaeeesesssasecenenaes 61 Exhibit 22: Proportion of HHP Enrollees and their Controls with a COVID-19 Diagnosis that Received Specific COVID-19-Related Services ..........ccsccssssccsssetesenenessneeesansenenenenssaeecesueeensnaneesaees 62 Exhibit 23: Comparing Monthly Service Utilization Rates in the Year Before the COVID-19 Pandemic (2019) versus the Year During (2020) for HHP Enrollees and the Control Group....... 63 Exhibit 24: Proportion of Primary Care and Specialty Care Services Provided through Telehealth by HHP Enrollees and Control groups, March 2020 to September 2020...............cssscccsssssreeeeesees 64 Exhibit 25: Unduplicated Monthly and Cumulative Enrollment in HHP, July 1, 2018 to September BO, 2020 .........sccsseccssessscuccessseuccssseeccessseucessseeceesassuccassssocessseeceassssscusssssceusssesccsssesecesssesccsssesocesssssocess 67 Exhibit 26: Unduplicated Quarterly Enrollment in HHP by SPA, July 1, 2018 to September 30, Exhibit 27: Enrollment of Individuals Reported as Homeless or At-Risk of Homelessness each Quarter in HHP by SPA, July 1, 2019 to September 30, 2020.............cccssssccsessssceeeesssseeesssseeenees 69 Exhibit 28: Unduplicated Cumulative HHP Enrollment by Group and County as of September 30, Exhibit 29: Proportion of HHP Enrollees that were identified in the Target Engagement List (TEL) as of September 2020, Overall and Dy Group ..........ccccssscecsssceessnecesaeeesaneeesseeecssaeeeeeaeeeseneeeessaeeeses 71 Exhibit 30: Enrollment and Disenrollment Patterns in HHP as of September 30, 2020............... 71 Exhibit 31: Average Length of Enrollment in Months in HHP by Group as of September 30, 2020 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 32: Percent of Eligible Beneficiaries Excluded by MCPs by Reason for Exclusion in the First Year of HHP Implementation .............:ccccsscccssseesssecessseeeesnetesaeecasanecesaesecesaeeeseaeeeeeneeeessaeeesnees 73 Exhibit 33: HHP Enrollee Demographics, Overall, and by SPA, at the Time of HHP Enrollment as Of September 30, 2020 ........cccescccssseessteetesenenssacecseaesensneeecsnaeeeeneneneneeeeaeeeesneeetenaeeeecaeeesenesensuaeeeses 75 Exhibit 34: Top Ten Most Frequent Physical and Mental Health Conditions among HHP Enrollees, 24 Months Prior to HHP Enrollment ...............ccccsccsssseceuenensnssscnsnsueueuensusssesensnenenenensenss 76 Exhibit 35: Complexity of HHP Enrollees' Health Status by SPA, 24 Months Prior to HHP Enrollment as of September 30, 2020 ............ccsssccssssssccesssscceesssseeecessneeceeessseesecsssaaeeesessnesesenenaes 77 Exhibit 36: HHP Services ........cc.cccccccseccsccccconcnesenecececcccceneneueceeesaaseneueuececesuacdueuerenesaceaecaesescsercneaecas 80 Exhibit 37: Estimated Overall HHP Units of Service Received by HHP Enrollees by SPA, July 1, 2018 to September 30, 2020 u.........ccccsssssscccssssseccessssscccenssseecceessasesensnassceesesaneeesessaaseesessnasesenenaes 82 Exhibit 38: Estimated Average Number of HHP Units of Service Provided to HHP Enrollees in Months HHP Services were Received by Service Type and SPA, July 1, 2018 to September 30, Exhibit 39: Estimated Average Number of HHP Core Unites of Service Provided to HHP Enrollees in Months those HHP Services were received by Modality and SPA, July 1, 2018 to September BO, 2020 0... .eeeecssccsseessccssccessccesecesacecsneesnsecenesesacesaceesaeeeseeceneceaesenacesaeeceneneasecenesesacesacensaeesasenenenens 84 Exhibit 40: Homeless Status and Receipt of Housing Services by HHP Enrollees, July 1, 2019 to September 30, 2020............ccccccssssscecessseccecesssceceesssceccesssssaseceessseseeesesaeeeesessaeeesecesseaeeeessaaaeeseeseaaees 86 Exhibit 41: Proportion of HHP Enrollees in Acute Care Utilization Groups at HHP Enrollment, Overall and by SPA Qu... eccccccsssesesseesssscesenenecsnenenenacensnesesenesessesesesenesensnesecsesesenasesenaseassnesensnasenens 88 Exhibit 42: Average Number of ED Visits and Hospitalizations by Acute Care Utilization Group, 24 months prior to ENrollMent ............ccssssscecesssscccecssssceeesssccecsessseseceessensceesesaeeeesessuaeeesesssaeecesenses 89 Exhibit 43: HHP Acute Care Utilization Groups by HHP Implementation GroupS..............scccesee 90 Exhibit 44: Demographics of HHP Acute Care Utilization Groups at the Time of HHP Enrollment Exhibit 45: HHP Acute Care Utilization Groups by Chronic Condition Eligibility Criteria, 24 Months Prior to Enrollment..........cccccsescccssscesesscercsceesssccesssecesscesesacesesaeesessesesesasesesaeesessnsessaeersnees 92 UCLA Center for Health Policy Research Rane eP es Health Economics and Evaluation Research Program Exhibit 46: Top Ten Most Frequent Physical and Behavioral Health Conditions among HHP Enrollees by Acute Care Utilization Group Prior to HHP Enrollment .............-:::cececeeeeceeeeeeeseeeeees 93 Exhibit 47: Primary Care Services per 1,000 Member Months Before and During HHP Enrollment by Acute Care Utilization Group ..........cccccsesenesenecsssneeeenecenssacecsnaesensnesecseeeesaaseneneneeenaeecenuesenenaeeesaaes 94 Exhibit 48: Specialty Services per 1,000 Member Months Before and During HHP Enrollment by Acute Care Utilization Group ..........ceccccsssseceseeenensceeesseeensnesessseeeesanenenanenssaneetenesenenesecenaesensnenensnaeess 95 Exhibit 49: Emergency Department Visits per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group ...........ccescccsssssececessssecenessseccceesssseeeenessnaceesessnenesenesaes 96 Exhibit 50: Hospitalizations per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group ..........cccccsescessneeessneeeesecenssacecseaesensnenecsaeeesaasenenenecenaeenecueeensnaeeesaees 97 Exhibit 51: Admissions to a Long-Term Care Facility Resulting in a Short-Term Stay per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group....... 98 Exhibit 52: Admissions to a Long-term Care Facility Resulting in a Medium-Term Stay per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group....... 99 Exhibit 53: Admissions to a Long-term Care Facility Resulting in a Long-Term Stay per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group..... 100 Exhibit 54: Trends in Primary Care Services per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 ............:ccccccssssscceeesseecesessneeeeesssnscesenssaneeeesessuesenensneeesenssanesenees 103 Exhibit 55: Trends in Specialty Services per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 ...........ccccccccssssscecessssccccessseccenessncesesessasececssaasesesenasesecessasesenenenanans 105 Exhibit 56: Trends in Mental Health Services per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 ..........cccssscccsssstccecsssscccesessseeceeesssaeeeesessceceesessaeeceessensesesesanaeetens 106 Exhibit 57: Trends in Substance Use Disorder Services per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 ............scccssscccsseecesstecsssteesssesececeecsseeeseseceessaeeeses 107 Exhibit 58: Trends in Ambulatory Care: Emergency Department Visits per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 .0............sssccesssscceeessseeeenenees 108 Exhibit 59: Trends in Percentage of Patients with Any ED Visits Before and During HHP by SPA as Of September 30, 2020 .0........eecssscccesssscecssssscccesssseeceesessacecsessaeseceseseaeeseeseaesecesssaeeeeeessaeseeesseagees 109 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 60: Trends in Inpatient Utilization per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 ..........cccccssccecssscecsssceesssececsseeeesseeessanecacaueeacaneeseneeeeeeaeececaesensaaeeesaaes 110 Exhibit 61: Trends in Percentage of Patients with Any Hospitalization Before and During HHP by SPA as of September 30, 2020 ........ceccccsssceceseeecsssceccssecessneeeessaeessanenenaneeasaeeetenesenscaeecesaesensneneesaaes 111 Exhibit 62: Trends in Inpatient Length of Stay Before and During HHP by SPA as of September BO, 2020 00... cceccccccstetetetetececescssenenenesenesasaesenenenenenssssscsesesesesessssssdseseserenenesseceteneneneneneeeeeetenenenenenens 113 Exhibit 63: Trends in Admissions to an Institution from the Community (Short-Term Stay) Before and During HHP by SPA as of September 30, 2020 .............ccsssscccssssscecesssseseeessssecceeessaseeesesenenes 115 Exhibit 64: Trends in Admissions to an Institution from the Community (Medium-Term Stay) Before and During HHP by SPA as of September 30, 2020..............c:ssscccccssesscesesssnceeeesssseerenesaes 116 Exhibit 65: Trends in Admissions to an Institution from the Community (Long-Term Stay) Before and During HHP by SPA as of September 30, 2020 ............ccssssccccssssececsssseccecesssseceeessssesesessneees 117 Exhibit 66: Trends in Adult Body Mass Index Assessment Before and During HHP by SPA for HHP Enrollees and the Control group as of September 30, 2020 .............csssccccessssceeessseceeeessseeeesesees 118 Exhibit 67: Trends in Screening for Depression and Follow-Up Plan Before and During HHP for SPA 1 HHP Enrollees and the Control group as of September 30, 2020 ...........cssecesssseceeseeseees 120 Exhibit 68: Trends in Follow-Up After Hospitalization for Mental Illness within 7 Days Before and During HHP by SPA for HHP Enrollees and the Control group as of September 30, 2020......... 121 Exhibit 69: Trends in Follow-Up After Hospitalization for Mental Illness within 30 Days Before and During HHP by SPA for HHP Enrollees and the Control group as of September 30, 2020.. 122 Exhibit 70: Trends in Follow-Up After ED Visit for Alcohol and Other Drug Abuse and Dependence within 7 Days Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020...........ccsssscccssssseccecessseccessssececesssseeeessesaesecesssaeeeeesssaeeesesssages 123 Exhibit 71: Trends in Follow-Up After ED Visit for Alcohol and Other Drug Abuse and Dependence within 30 Days Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020............cccssssccesssscceecsssscecesssnsecesssaeaeeeseseaeseceessaeeeeesssaesessessaaes 124 Exhibit 72: Trends in Initiation of Alcohol and Other Drug Abuse or Dependence Treatment Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, UCLA Center for Health Policy Research Nae TeP es Health Economics and Evaluation Research Program Exhibit 73: Trends in Engagement of Alcohol and Other Drug Abuse or Dependence Treatment Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, Exhibit 74: Trends in Use of Pharmacotherapy for Opioid Use Disorder Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020 ...........:ccssecessseees 128 Exhibit 75: Trends in Controlling High Blood Pressure Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020...........cccssscccssteteserenssseeeesserensneneeees 129 Exhibit 76: Trends in Plan All-Cause Readmission Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020...........cccescessceteseeeteneneeesserensnenenens 131 Exhibit 77: Trends in Prevention Quality Indicator (PQI) 92: Chronic Conditions Composite Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, Exhibit 78: Trends in Total Estimated Payments Before and During HHP by SPA as of September BO, 2OZO ........cccccccccccssssscneeececcccaasneneeeccesccaasnseseeceaenaassesasecaeaenaaseesesesesseaaaseeeteesesacasaenesteesegeganseneaeoes 134 Exhibit 79: Trends in Payments for Outpatient Services Before and During HHP by SPA as of September 30, 2020..........:ccccsssssscccessssncceecesseececssssaeseceessaeeceessscaesesseseaeesseseaesececssaeaeeeseseaesessessaaees 135 Exhibit 80: Trends in Outpatient Medication Payments Before and During HHP by SPA as of September 30, 2020..........:ccccccssssscccssssccccesssssccecessssccesesssseecesensaeesesessasecessnaasesensnaeasenensasasenensnanens 136 Exhibit 81: Trends in Payments for Emergency Department Visit Before and During HHP by SPA as Of September 30, 2020...........ccccssssccssssssccesessccceesesssacecesssensesesesseaeesecsnaesecesesanececessaeeeeesssanees 138 Exhibit 82: Trends in Payments for Hospitalizations Before and During HHP by SPA as of September 30, 2020..........ccssscccsstecssstccssscecssscccsssccesssececsseeeesseeeesseaesssueesesusesesaeecseaeeesenesecseaeeesnaes 139 Exhibit 83: Estimated HHP Supplemental Expenditures by Enrollees Type and Implementation Group, as of September 30, 2020 .........cccccesececssscecsseeeecsseceesseceesnnenesaaeeesaeeeesseeetesaeeceeaeeeteneneesaeees 140 Exhibit 84: Evaluation Questions and Data SOUICES...........cccccccssscssssssssscssssscsescscscscscscnsnssssssssseaes 142 Exhibit 85: Beneficiary-Level Variables ...............ccccssssscccesssscecessssececessseeceessseceeesesseeeeeesssaeeeesesaes 144 Exhibit 86: HHP Service Utilization INGiCAtOIS ......... cee ceesesccccecccccccssencececccecceuscseececeseccaucesneseeeseseae 145 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 87: HHP Services ...........-:::cccesceeseeceneeeseeeseceeeecensneeeaeeeaeesseseseceesceeeaeeseaeneseeasaeeeeesenaneneaaeens 146 Exhibit 88: Demographic INdiCAators............cccsscccsssccecsscesssccessnnecesseeessaneeessneeceeasecseatecseseeeessaeeesns 147 Exhibit 89: Health Status INGicators..............:eseceeecceeeceeeeeneseeeeceeeseceneeeesceeeaeeseneneneeesaeeeeesenaneneaetens 147 Exhibit 90: Healthcare Utilization INGiCators ............ ce eeeeesceeeeeeeeeeseceeeceeeneeteneeeeeesaeeeeeseneeeeeaeeens 149 Exhibit 91: HHP Core Metrics, Definitions, and Reporting Status ...............ccscccsssssceeesssssreceeesees 150 Exhibit 92: Variables Used to Select the Control Group..........ccccssssecsssteeesseeteseceessaeeeenserensseeeeees 153 Exhibit 93: Comparison of Select Characteristics of HHP SPA 1 Cohort 5 Enrollees (Enrolled July to September 2019) and Matched Control Beneficiaries.................::ccccssssssecesssesececesssnesenessnaeees 156 Exhibit 94: Description of Mutually Exclusive Categories of Service® ...........cccccssscecessereesseeeeees 161 Exhibit 95: Percentage of 2019 Total Estimated Payments by Category of Service for HHP Medi- Cal Claims .........cccccccseecssececsseceessnenensneeesaaeenenenenenaeeseaesecsuesensaaeeseauesenauenasaneseeueeeteneseceuaeeetenenensnaeens 163 Exhibit 96: Category of Service and Payment DeSCriptions............:ccccssccsssrereseeesseecesserersneeeees 164 Exhibit 97: Payment Data SOUICES...........ccccssssscccesssstsccecsssecceescsssececeesssnceeeseseeeesseseaeeesesesaeeceesnaes 165 Exhibit 98: Comparison of Estimated Fee-for Service Payments and Paid Amounts for 2019 HHP Medi-Cal Claims ............:ccccsssccccesssssccecessncececssssneeeesssaesesecasacesesessauesecassuaeseesessanececasaaaeeceeseussecenesaes 171 Exhibit 99: Comparison of Average Fee- for-Service and Managed Care Payments per Claim for 2019 HHP Medi-Cal Claims .............::cceceeecceeeceeeseeeeceeseneseeenaceneaeesnenenasesaeeeacsenecenenesenaeessetenesennees 171 Exhibit 100: Evaluation Conceptual Fra mewolrk..........:ccccsssccssseesesseecesaneeesaeserssaeeessaeeeseseneessaeeeses 174 Exhibit 101: Evaluation Questions and Data SOUICES ............:::ccseceseceeeeeeeseeseneneeteneceeeesenenenenenees 175 Exhibit 102: Evaluation Timeline and Deliverables. .................:ccs:ceseceeeeceeeeeeeseneseeeseeeeeeeeseneneaeeees 181 Exhibit 103: Unduplicated Monthly Cumulative Enrollment of HHP Homeless Enrollees by Group, July 1, 2018 to September 30, 2020 ...........ccccccsseessssceeesenessnenecseeeesseeeneneneessaeeetsnesensneeees 183 Exhibit 104: HHP Implementation and Enrollee Demographics for Aetna, Alameda Alliance, Blue Shield, and CA Health and Wellness as of September 30, 2020...............cssscccsssssceeeessssesensssnenens 196 UCLA Center for Health Policy Research Rane eP es Health Economics and Evaluation Research Program Exhibit 105: HHP Enrollee Health Status and Utilization Prior to Enrollment and Service Delivery for Aetna, Alameda Alliance, Blue Shield, and CA Health and Wellness as of September 30, 2020 Exhibit 106: Trends in HHP Metrics for Aetna, Alameda Alliance, Blue Shield, and CA Health and Wellness as of September 30, 2020 .........cccssccssscecsssecessseeessseeessnecesaeeeesaneeseneeeeseaeeceeesensuaeeesaaes 198 Exhibit 107: Trends in Estimated Payments for Aetna, Alameda Alliance, Blue Shield, and CA Health and Wellness as of September 30, 2020...........:cssscccsseecesenenesseeeesaesetsneeeessaeecesseeensneeeeees 200 Exhibit 108: HHP Implementation and Enrollee Demographics for Anthem Blue Cross as of September 30, 2020.............:cccsssecccessssccceesessececessncececessnasenensssecesecesaneseneeenaeseseueuaeesecessanesenessnenees 202 Exhibit 109: HHP Enrollee Health Status and Utilization Prior to Enrollment and Service Delivery for Anthem Blue Cross as of September 30, 2020...........:cccccssssscesesssscecesssssecesssscesenessssesenessnanens 203 Exhibit 110: Trends in HHP Metrics for Anthem Blue Cross as of September 30, 2020............. 204 Exhibit 111: Trends in Estimated Payments for Anthem Blue Cross as of September 30, 2020 206 Exhibit 112: HHP Implementation and Enrollee Demographics for LA Care, Community Health Group, Kern Health Systems, and CalOptima as of September 30, 2020..............:ccc:sssccecessseeees 208 Exhibit 113: HHP Enrollee Health Status and Utilization Prior to Enrollment and Service Delivery for LA Care, Community Health Group, Kern Health Systems, and CalOptima as of September BO, 2020 ......ecccseccccsseeeessecenssaceessnenensnenesaeeneneneneneneseaesensnesensneceeeauesenauenenaeasaauanetenesensnaeeetenenensnenens 209 Exhibit 114: Trends in HHP Metrics for LA Care, Community Health Group, Kern Health Systems, and CalOptima as of September 30, 2020 ............ccccccssssscccesesscecessssecensesneseseseneeceessssesesessnanens 210 Exhibit 115: Trends in Estimated Payments for LA Care, Community Health Group, Kern Health Systems, and CalOptima as of September 30, 2020.............cscssccccessssceeesssssecessssecetesseseceesesseeees 212 Exhibit 116: HHP Implementation and Enrollee Demographics for Inland Empire Health Plan and Kaiser as of September 30, 2020 .............cssccccsscccsssccesssececsseceesseecessuecsssueeessuseceeaeecseaueeseseceeseaceeses 214 Exhibit 117: HHP Enrollee Health Status and Utilization Prior to Enrollment and Service Delivery for Inland Empire Health Plan and Kaiser as of September 30, 2020 ..............csssssccesssseceeesssees 215 Exhibit 118: Trends in HHP Metrics for Inland Empire Health Plan and Kaiser as of September BO, 2020 00... eeececssccessccssecesecessccesaceescccsacesseeesacesaceeseeecsseeeseeesacesencesaeeeseeeeasesseeesacesaceessatensesenaeenaes 216 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 119: Trends in Estimated Payments for Inland Empire Health Plan and Kaiser as of September 30, 2020............:cccssssccccssssccceessssencecesssnesecssssneesesesaesececssaaaeeeessnaeseceeasacesecassaneeecaessaeees 218 Exhibit 120: HHP Implementation and Enrollee Demographics for Molina Healthcare Plan as of September 30, 2020............cccccsssecccessssccceesssscececasssneeecsssnseecesssaneesecsssneeeeessnaesessusuaeesecassanesenesssaeees 219 Exhibit 121: HHP Enrollee Health Status and Utilization Prior to Enrollment and Service Delivery for Molina Healthcare Plan as of September 30, 2020..........::cccsssccessrensssteeteseeneneeecesesensnenensnees 220 Exhibit 122: Trends in HHP Metrics for Molina Healthcare Plan as of September 30, 2020.....221 Exhibit 123: Trends in Estimated Payments for Molina Healthcare Plan as of September 30, Exhibit 124: HHP Implementation and Enrollee Demographics for Health Net as of September BO, 2020 .......ecccccccccccssssssececcecccesassnececcasscaaassnasecescsansasesaceaeeeeausssaeesesesenanaseaseeeseeauanaseaeseesacouaseneaesens 225 Exhibit 125: HHP Enrollee Health Status and Utilization Prior to Enrollment and Service Delivery for Health Net as of September 30, 2020...........csccccssecesssecessseceesenenessneeesaneeteneeecenaeecssaesensneeeesaaes 226 Exhibit 126: Trends in HHP Metrics for Health Net as of September 30, 2020.............:::cseceeees 227 Exhibit 127: Trends in Estimated Payments for Health Net as of September 30, 2020.............. 229 Exhibit 128: HHP Implementation and Enrollee Demographics for San Francisco Health Plan, Santa Clara Family Health Plan, and United Healthcare as of September 30, 2020................... 231 Exhibit 129: HHP Enrollee Health Status and Utilization Prior to Enrollment and Service Delivery for San Francisco Health Plan, Santa Clara Family Health Plan, and United Healthcare as of September 30, 2020.........ccsscccesereneseneeesensensseeecsneeecessecensnesensaaeeesanenenanenesaneetenesenenenesesaesensnenensnees 232 Exhibit 130: Trends in HHP Metrics for San Francisco Health Plan, Santa Clara Family Health Plan, and United Healthcare as of September 30, 2020 ...........cscsssssccceessssecesessnceeeesssesesenenees 233 Exhibit 131: Trends in Estimated Payments for San Francisco Health Plan, Santa Clara Family Health Plan, and United Healthcare as of September 30, 2020............ccssssscecesssececeessseeeeeesees 235 Exhibit 132: Count of SPA 1 Enrollees by Number of Months of HHP Enrollment as of September UCLA Center for Health Policy Research Tyr) Health Economics and Evaluation Research Program Exhibit 133: Count of SPA 2 Enrollees by Number of Months of HHP Enrollment as of September March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 1 defines acronyms and terms referenced throughout the report. Exhibit 1: General Health Homes Program Acronyms and Definitions Acronym Definition AB Assembly Bill ACO Accountable Care Organization AHF AIDS Healthcare Foundation AHS Alameda Health Systems AOD Alcohol and Other Drug ASC Ambulatory Surgical Center ASP Average Sales Price BMI Body Mass Index CB-CME Community-Based Care Management Entity CBO Community Based Organizations CBAS Community-Based Adult Services CCA Clinical Care Advance CCW Chronic Condition Warehouse CDPS Chronic Illness and Disability Payment System Risk Score CKD Chronic Kidney Disease CM Care Management CMS Centers for Medicare and Medicaid Services COPD Chronic Obstructive Pulmonary Disease CPT Current Procedural Terminology CSH Corporation for Supportive Housing DD Difference-in-Difference DHCS California Department of Health Care Services DME Durable Medical Equipment DRG Diagnosis Related Grouping E&M Evaluation & Management ED Emergency Department EHR Electronic Health Record ER Emergency Room FFS Fee-for-Service FMAP Federal Medical Assistance Percentage FQHC Federally Qualified Health Center GRM General Risk Model HAP Health Action Plan HCPCS Healthcare Common Procedure Coding System HCSA Alameda County Health Care Services Agency HEDIS Healthcare Effectiveness Data and Information Set HH/HCBS Home Health and Home and Community-Based Services HHP Health Homes Program HIE Health Information Exchange HIT Health Information Technology HMIS Homeless Management Information Session ICD International Classification of Diseases LA Los Angeles LCSW Licensed Clinical Social Worker LTC Long-Term Care MCP Managed Care Plan Glossary | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Acronym Definition MFT Marriage and Family Therapist MM Member months NADAC National Average Drug Acquisition Cost NPI National Provider Identifier NPPES National Plan and Provider Enumeration System NUCC National Uniform Claims Committee OPPS Outpatient Prospective Payment System OUD Opioid Use Disorder PACE Program of All-Inclusive Care for the Elderly PCP Primary Care Provider PMPM Per Member per Month POS Place of Service Pal Prevention Quality Indicator RHC Rural Health Center RN Registered Nurse SCAN Senior Care Action Network SFTP Secure File Transfer Protocol SMI Severe Mental Illness SNF Skilled Nursing Facility SNOMED CT Systematized Nomenclature of Medicine-Clinical Terms SPA State Plan Amendment SUD Substance Use Disorder SW Social Worker TAR Treatment Authorization Request TEL Targeted Engagement List UBREV Revenue Code UCLA University of California, Los Angeles Center for Health Policy Research UOS Unit of Service she) March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 2 defines acronyms and full names of participating Managed Care Plans. Exhibit 2: Managed Care Plans Acronyms/Abbreviations and Definitions Acronym/Abbreviations Managed Care Plan Full Name ABHCA Aetna Better Health of California AAH Alameda Alliance for Health Anthem Anthem Blue Cross of California Partnership Plan, Inc. BSCPHP Blue Shield of California Promise Health Plan CHW California Health & Wellness CalOptima CalOptima CHG Community Health Group Partnership Plan HNCS Health Net Community Solutions, Inc. IEHP Inland Empire Health Plan Kaiser Kaiser Permanente KHS Kern Health Systems L.A. Care L.A. Care Health Plan MHC Molina Healthcare of California Partner Plan, Inc. SFHP San Francisco Health Plan SCFHP Santa Clara Family Health Plan UnitedHealthcare UnitedHealthcare Community Plan of California, Inc. Glossary | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Executive Summary Health Homes Program (HHP) Overview The California Department of Health Care Services (DHCS) implemented the Medi-Cal Health Homes Program (HHP) to serve eligible Medi-Cal beneficiaries with complex needs and chronic conditions. HHP was authorized under California Assembly Bill 361 and approved by Centers for Medicare and Medicaid Services under Section 2703 of the 2010 Patient Protection and Affordable Care Act. HHP was designed to provide six core services for eligible enrollees: (1) comprehensive care management; (2) care coordination; (3) health promotion; (4) comprehensive transitional care; (5) individual and family support; and (6) referral to community and social support services. DHCS selected 12 California counties where all 16 Medi-Cal managed care plans (MCPs) operating in those counties would implement HHP for their enrollees who met certain chronic condition and acuity criteria. HHP was implemented in phases by county groupings and two subsets of enrollees, with the first group implementing in July 2018 and the last group implementing in July 2020. Subsets of enrollees included those with chronic physical health conditions or substance use disorders (SUD) referred to as SPA 1 (State Plan Amendment 1) and those with severe mental illness (SMI) referred to as SPA 2. MCPs implemented SPA 2 six months after SPA 1 within each county grouping. DHCS published a program guide to ensure uniform HHP implementation, delivery of services, and reporting across all MCPs. MCPs contracted with Community-Based Care Management Entities (CB-CMEs) to deliver HHP services. MCPs enrolled eligible beneficiaries from a Targeted Engagement List (TEL) provided by DHCS but had discretion in enrolling other eligible beneficiaries. Evaluation Methods The UCLA Center for Health Policy Research was selected to evaluate HHP and developed a conceptual framework and evaluation questions to conduct a rigorous assessment of the program. UCLA used all available data for the evaluation. These included MCP readiness documents that contained MCP's HHP policies and procedures for implementation and delivery of services; Targeted Engagement Lists (TEL) created every six months by DHCS to identify potentially eligible HHP enrollees per MCP; MCP enrollment and quarterly reports that included beneficiary level enrollment data and homeless status; and Medi-Cal enrollment and claims data for all HHP enrollees with information on demographics, health status, and use of health services. UCLA used readiness documents to describe HHP implementation including composition of HHP networks, types of staff, data sharing, enrollee outreach and engagement, and HHP service delivery approaches. UCLA used TEL, MCP enrollment and utilization reports, UCLA Evaluation | Executive Summary March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program and Medi-Cal data to assess HHP enrollment patterns, demographics, health status, HHP service use, and health care service utilization. UCLA attributed a dollar amount to all claims and assessed change in estimated payments. Results HHP Implementation and Infrastructure HHP was implemented by all 16 MCPs operating in 12 California counties, with six MCPs implementing HHP in more than one county. MCP HHP implementation plans outlined in readiness documents at the beginning of HHP indicated that 15 MCPs used HHP delivery Model |, where CB-CMEs were typically medical providers that hired and housed HHP staff, including care coordinators. In addition, MCPs ensured that CB-CMEs had adequate staffing to deliver HHP services; utilized data sharing technologies including SFTP, dedicated email, electronic health records (EHR), care management platforms, or health information exchange (HIE); and used predictive modeling and risk grouping of eligible beneficiaries to identify and target beneficiaries for HHP enrollment. See the first interim evaluation report for more details. e In their Quarterly HHP Reports, MCPs reported that they had developed HHP delivery networks with 244 unique CB-CMEs by September 2020. These CB-CMEs were primarily community health centers (41%), followed by community based social service organizations or local government entities (28%) and community based primary care or specialty physicians (19%). Six MCPs indicated that they acted as a CB-CME for a portion of their HHP enrollees in an effort to expand service capacity in regions where community based infrastructure was insufficient. e MCPs reported that they anticipated that contracted CB-CMEs had an enrollment capacity of approximately 79,370 enrollees with 34% of that capacity in community health centers. The median capacity per CB-CME was 180 enrollees. HHP and COVID-19 e The evaluation timeframe for this interim report encompasses activities and data from July 2018 through September 2020. The COVID-19 pandemic began during this time and led to a statewide shelter in place order in mid-March 2020, 20 months following the first HHP enrollment. The COVID-19 hospitalizations in HHP counties peaked near the end of July 2020 with 18 hospitalizations per 100,000. e MCPs reported that the COVID-19 pandemic had impacted HHP processes, procedures, and/or policies, with the greatest impact on housing and homeless support services, comprehensive transitional care, and delivery of care coordination by frontline staff. MCPs were able to establish effective workflows and infrastructure to support their own and CB- Executive Summary | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program CME's operation by transitioning to telehealth and strategically coordinating with shelters and other short-term housing services. UCLA estimated that 4.3% of both HHP enrollees and a group of similar Medi-Cal beneficiaries not enrolled in HHP, the control group, had at least one service with COVID-19 as the primary or secondary diagnosis. This rate was highest in July 2020. HHP enrollees and controls with a COVID-19 diagnosis most commonly had used COVID-19 related primary care services (53% for HHP enrollees vs 47% for the control group), followed by emergency department visits (28% vs 29%) and hospitalizations (28% vs 28%). Examining the overall service utilization patterns from 2019 and 2020 showed a limited decline in use of primary care services for HHP enrollees in 2020 compared to 2019. In contrast, specialty care services, ED visits, and hospitalizations declined in 2020 compared to 2019. Specialty care services utilization returned to 2019 levels by September 2020 but the rates of ED visits and hospitalizations remained below 2019 levels through September 2020. Telehealth service use was under 0.2% before March 2020 but rapidly increased to 19% of primary care services in April and declined to 13% by September among HHP enrollees. A similar pattern was observed for specialty telehealth services. HHP Enrollment and Enrollment Patterns A total of 48,925 individuals enrolled in HHP between July 1, 2018 and September 30, 2020, with 38,228 enrolled in SPA 1 and 10,697 enrolled in SPA 2. As of September 2020, three- quarters of the current enrollment were in SPA 1. The number of enrollees experiencing homelessness or at risk of homelessness increased over time and represented 10% of all HHP enrollees by September 2020; a likely underestimate due to data limitations. The rate of enrollment varied by when each group implemented HHP. Groups 2 and 3 had the highest levels of enrollment (14,426 and 32,530, respectively) and Group 4 had the lowest levels of enrollment (759), by September 2020. Los Angeles County had the highest level of enrollment with 18,919 enrollees. DHCS identified eligible Medi-Cal beneficiaries in the Targeted Engagement List (TEL) and shared it with MCPs. Overall, 78% of HHP enrollees were reported on the TEL prior to enrollment. The highest rate of enrollment from the TEL was 90% in Groups 1 and 4. Most (70%) of HHP enrollees were continuously enrolled through September 2020, 30% were disenrolled by September 2020, and 0.2% enrolled multiple times through September 2020. The average length of enrollment in Group 1 was 10.7 months for SPA 1 enrollees and 8.4 months for SPA 2 enrollees. Overall, the average length of enrollment was 9.4 months for Group 2, 6.7 months for Group 3, and 4.3 months for Group 4 enrollees. UCLA Evaluation | Executive Summary March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program The most common reason MCPs reported for not enrolling from the TEL in Groups 2 and 3 was that an eligible beneficiary was not an MCP member, indicating the data informing the TEL did not always reflect current enrollment status (members are permitted to change MCPs every 30 days). The most common reason for Group 1 was eligible enrollee declined to participate and for Group 4 it was the eligible enrollee was already well managed. HHP Enrollee Demographics and Health Status The majority of HHP enrollees were between 50 and 64 years old, female, and spoke English. Nearly half of enrollees were Latinx. SPA 2 enrollees were more often between 18 and 49 years old and more often female in comparison to SPA 1 enrollees. Prior to enrollment, the most common chronic conditions among all HHP enrollees and SPA 1 enrollees were hypertension (67%) and diabetes (49%). The most common condition among SPA 2 enrollees was depression (72%). MCPs enrolled Medi-Cal managed care beneficiaries with multiple chronic health conditions, consistent with HHP's requirements. For example, 55% had hypertension along with chronic obstructive pulmonary disease, diabetes, coronary artery disease, and/or chronic or congestive heart failure and 40% had a combination of complex conditions such as chronic renal (kidney) disease, chronic liver disease, and traumatic brain injury. HHP Service Utilization among HHP Enrollees MCPs reported challenges and significant lag with reporting of HHP services by way of encounter data, which led to 24% of enrollees without any HHP service codes during this time frame. Existing data showed that MCPs reported 412,463 HHP units of service (UOS) to HHP enrollees from July 2018 through September 2020. In months where encounter data for HHP services were present, enrollees averaged 2.1 HHP UOS per month. Enrollees had a higher average use of core HHP services (1.7 UOS per month) and other HHP services (1.6) compared to engagement services (1.3). Average number of services was higher for services provided through telehealth (1.6 UOS per month) compared to in-person (1.3) and by non- clinical providers (1.8) compared to clinical providers (1.6). Among enrollees at risk of or experiencing homelessness in the third quarter of 2020, 68% received housing services and 7% were reported as no longer homeless by September 2020. Acute Care Utilization Groups in HHP UCLA examined the HHP population by their level of acute care utilization in the 24 months prior to HHP by creating five groups; enrollees with super utilization (6% of all enrollees), high utilization (15%), moderate utilization (35%), low utilization (32%), and enrollees at risk for high utilization (13%). These rates were similar for SPA 1 and SPA 2. Enrollees with super Executive Summary | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program utilization had 14.9 ED visits and 4.1 hospitalizations on average per year compared to 2.7 and 0.5 on average per year among those with moderate utilization. Group 4, which consisted of Orange County and the latest to implement HHP, included the highest share of enrollees with super utilization (18%) and high utilization (28%) and the lowest share of enrollees at risk for high utilization (3%). Enrollees with super utilization were more often younger than 65 (96%), male (49%), white (26%), and were experiencing homelessness (14.6%) compared to other acute care utilization groups. The super utilization group had the largest proportion of homeless enrollees (14.6%) and the at-risk group had the smallest proportion of homeless enrollees (5.6%). The prevalence of HHP chronic condition eligibility criteria varied by acute utilization groups. Criteria 1 was the second most prevalent eligibility criteria (two specific chronic conditions) and the majority of enrollees with super utilization met this criterion (65%) vs. 49% of those with high and 35% of those with moderate utilization. Furthermore, hypertension was the most common chronic condition across all groups, followed by chronic kidney disease among enrollees with super utilization and diabetes among all other groups. An examination of the unadjusted rates of service use of HHP enrollees showed the lowest number of primary care services per 1,000 member months 19 to 24 months before enrollment and the largest increase during months 1 to 6 of HHP enrollment. This usage declined 7 to 12 months after enrollment but remained at above pre-enrollment numbers. Enrollees with super utilization included a cohort of 14.6% enrollees who were experiencing homelessness. Those with super utilization had the highest magnitude of primary care service use, peaking at 1,346 services per 1,000 member months. The same pattern was observed for specialty services; enrollees with super utilization peaked at 928 specialty services per 1,000 member months. From 19-24 months before enrollment to 7-12 months during enrollment, the unadjusted ED visits followed by discharge decreased among enrollees with super utilization (from 921 to 635), high utilization and moderate utilization; and hospitalizations declined among enrollees with super utilization (from 284 to 227) as well as those with high and moderate utilization. Admissions to an institution from the community were calculated annually before enrollment (Pre-Year 2 and Pre-Year 1) and the first year of enrollment (HHP Year 1). When comparing Pre-Year 2 to HHP Year 1, the unadjusted number of short-term stays in long- term care facilities declined for enrollees with super utilization, high utilization, and moderate utilization. However, the number of medium-term stays increased for enrollees with super, high, and moderate utilization prior to enrollment and then declined post- enrollment. The number of long-term stays increased for all acute care utilization groups. UCLA Evaluation | Executive Summary March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program HHP Outcomes Exhibit 3 and Exhibit 4 display an overview of changes in HHP core metrics and optional measures of health care utilization and estimated Medi-Cal payments. For each measure, UCLA hypothesized an intended direction consistent with HHP goals. UCLA did not hypothesize a direction for outpatient utilization and related payments because unmet need for outpatient care is likely at enrollment but the utilization level and subsequent payments are likely to drop over time as unmet need is addressed. Similarly, UCLA did not hypothesize a direction for outpatient prescription estimated payments. The exhibits show (1) if the trend changed significantly in the intended direction during HHP for enrollees, (2) if the trend changed significantly in the intended direction from before to during HHP for enrollees, and (3) whether the trend from before to during HHP was significantly changed in the intended direction for enrollees compared to the control group. Data indicated a significant and overall decline in primary care and specialty care services compared to the control group. More in-depth analysis showed that there was an initial increase early in enrollment followed by a decline later on. Analysis also showed that the rate of use of these services remained higher than before HHP enrollment and compared to the control group. Improvements in comparison to the control group were observed for some HHP core metrics measuring utilization, process, and outcomes of care, but trends in some metrics did not change. Specifically, the HHP core metric of Ambulatory Care: Emergency Department (ED) Visits declined significantly during the first year of HHP in the intended direction (SPA 1: 17 visits per six months; SPA 2: 25 visits) and this decline was greater compared to before HHP (SPA 1: 20; SPA 2: 29) and in comparison to the control group for both SPA 1 (DD: 9) and SPA 2 (DD: 15) enrollees. In contrast, there was a significant increase (rather than the intended decrease) in the HHP core metric of Admissions to an Institution from the Community (Long-Term Stay) during the first year of HHP for SPA 1 and SPA 2 enrollees. The trend in this HHP core metric from before to during HHP was significantly greater (DD: 0.4 admission per year) for SPA 2 enrollees vs. their control group. Among HHP core metrics reflecting processes of care, Adult BMI Assessment metric increased in the intended direction during the first year of HHP and in comparison to the control group for SPA 1 (DD: 1.1% per year) and SPA 2 (DD: 1.0%) enrollees. Similarly, Screening for Depression and Follow-Up Plan improved more for SPA 1 (DD: 1.6%) enrollees than the control group. Initiation of Alcohol and Other Drug Treatment metric declined (in the wrong direction) for SPA 1 (DD: 2.7%) HHP enrollees and in comparison to the control group. There were no other significant changes for the remaining metrics for SPA 1 Executive Summary | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program IV Elacale para enrollees. However, the Engagement of Alcohol and Other Drug Treatment metric improved in the intended direction for SPA 2 (DD: 10.9%) enrollees vs. their control group. e Among HHP core metrics reflecting outcomes of care, the Controlling High Blood Pressure metric increased (in the intended direction) significantly during HHP for SPA 1 enrollees but decreased significantly (in the wrong direction) for SPA 2 enrollees. However, there were no differences in trends with the respective control groups. In addition, the Prevention Quality Indicator (PQI 92) significantly improved (in the intended direction) during HHP for SPA 1 and SPA 2 enrollees. However, when compared to the control group, the change for SPA 1 enrollees was significantly smaller (in the wrong direction) than the control group (DD: 0.9 per year) but the rate was similar to the control group for SPA 2 enrollees. Exhibit 3: Outcomes for SPA 1 HHP Enrollees as of September 30, 2020 Trend during Trend from before Trend for HHP Intended HHP changed to during HHP patients was direction significantly in changed better than the intended significantly in the control group direction? intended direction? (DD)? UTILIZATION MEASURES Primary Care Services Not No Direction No Direction No Direction per 1,000 MM Specified (-126) (-159) (-101) Specialty Services per Not No Direction No Direction No Direction 1,000 MM Specified (-40) (-114) (-60) Mental Health Services Not No Direction No Direction per 1,000 MM Specified Not Significant (-420) (-236) Substance Use Disorder Not No Direction No Direction No Direction Services per 1,000 MM Specified (-24) (-26) (-19) Ambulatory Care: ED Visits per 1,000 MM Decrease Yes (-17) Yes (-20) Yes (-9) Percentage of HHP Enrollees with Any ED Visits Resulting in Discharge Decrease Yes (-2.40%) Yes (-2.30%) Yes (-1.10%) Hospitalizations per 1,000 MM Decrease Yes (-10) Yes (-15) Yes (-7) Percentage of HHP Enrollees with Any Hospitalizations Decrease Yes (-3.1%) Yes (-4.9%) Yes (-1.9%) Average Length of Stay for Hospitalizations Decrease Not Significant Not Significant Not Significant Admission to an Institution from the Community (Short-Term Stay) Decrease Y (-0.30) Y (-0.60) Not Significant UCLA Evaluation | Executive Summary March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Intended direction Trend during HHP changed significantly in the intended direction? Trend from before to during HHP changed significantly in the intended direction? Trend for HHP patients was better than control group (DD)? Admission to an Institution from the Community (Medium- Term Stay) Decrease Not Significant Not Significant Not Significant Admission to an Institution from the Community (Long-Term Stay) Decrease No (0.30) No (0.40) Not Significant PROCESS METRICS Adult Body Mass Index Assessment Increase Yes (5.40%) No (-4.90%) Yes (1.10%) Screening for Depression and Follow- Up Plan Increase Yes (9.00%) Yes (3.10%) Yes (1.60%) Initiation of Alcohol and Other Drug Treatment Increase No (-3.40%) No (-4.70%) No (-2.70%) Engagement of Alcohol and Other Drug Treatment Increase Not Significant Not Significant Not Significant Follow-Up After Hospitalization for Mental Illness within 7 days Increase Not Significant Not Significant Not Significant Follow-Up After Hospitalization for Mental Illness within 30 days Increase Not Significant Not Significant Not Significant Follow-Up After Emergency Department Visit for Alcohol and Other Drug Abuse or Dependence within 7 days Increase Not Significant Not Significant Not Significant Follow-Up After Emergency Department Visit for Alcohol and Other Drug Abuse or Dependence within 30 days Increase Not Significant Not Significant Not Significant Use of Pharmacotherapy for Opioid Use Disorder Increase Not Significant Not Significant Not Significant Executive Summary | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Trend during Trend from before Trend for HHP Intended HHP changed to during HHP patients was direction significantly in changed better than the intended significantly in the control group direction? intended direction? (DD)? OUTCOME METRICS Plan All-Cause Readmissions Decrease Not Significant Not Significant No (1.20%) Controlling High Blood Pressure Increase Yes (3.10%) No (-0.90%) Not Significant Prevention Quality Indicator (PQI) 92: Chronic Conditions Composite Decrease Yes (-1.90) Yes (-4.60) No (0.90) ESTIMATED PAYMENTS Estimated Total Payments per Enrollee per 6 Months Decrease No ($331) No ($163) Yes (-S96) Estimated Payments for Outpatient Services per Not No Direction No Direction No Direction Enrollee per 6 Months specified (S258) (S148) (-S23) Estimated Payments for Outpatient Medications per Enrollee per 6 Not No Direction No Direction No Direction Months specified (S50) (S25) (-S7) Estimated Payments for ED Visits Resulting in Discharge per Enrollee per 6 Months Decrease Yes (-S7) Yes (-S9) Yes (-S29) Estimated Payments for Hospitalizations per Enrollee per 6 Months Decrease No ($7) Yes (-S53) Yes (-S7) Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: Yes indicates significant in the intended direction. No indicates significant in the unintended direction. Green indicates change in the intended direction. Red indicates change in the unintended direction. Yellow indicates significant change when direction is not specified. MM indicates member months. ED indicates Emergency Department. UCLA Evaluation | Executive Summary March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 4: Outcomes for SPA 2 HHP Enrollees as of September 30, 2020 Trend during HHP Trend from before Trend for HHP to during HHP . changed patients was Intended sags . changed - ae significantly inthe | ._... : better than direction . significantly in the intended . control group direction? intended (DD)? direction? UTILIZATION MEASURES Primary Care Services Not No Direction No Direction No Direction per 1,000 MM Specified (-80) (-102) (-83) Specialty Services per Not No Direction No Direction 1,000 MM Specified Not Significant (-47) (-49) Mental Health Services per 1,000 Not No Direction No Direction No Direction MM Specified (-578) (-1,354) (-957) Substance Use Disorder Services per Not No Direction No Direction No Direction 1,000 MM Specified (-56) (-78) (-62) Ambulatory Care: ED Visits per 1,000 MM Decrease Yes (-25) Yes (-29) Yes (-15) Percentage of HHP Enrollees with Any ED Visits Resulting in Discharge Decrease Yes (-3.20%) Yes (-3.50%) Not Significant Hospitalizations per 1,000 MM Decrease Yes (-12) Yes (-16) Yes (-10) Percentage of HHP Enrollees with Any Hospitalizations Decrease Yes (-4.2%) Yes (-5.7%) Yes (-1.9%) Average Length of Stay for Hospitalizations Decrease Not Significant Not Significant Not Significant Admission to an Institution from the Community (Short- Term Stay) Decrease Not Significant Yes (-0.40) Not Significant Admission to an Institution from the Community (Medium-Term Stay) Decrease Not Significant Not Significant Not Significant Admission to an Institution from the Community (Long- Term Stay) Decrease No (0.40) No (0.50) No (0.40) PROCESS METRICS Executive Summary | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program IV Elacale para Intended direction Trend during HHP changed significantly in the intended direction? Trend from before to during HHP changed significantly in the intended direction? Trend for HHP patients was better than control group (DD)? Adult Body Mass Index Assessment Increase Yes (2.90%) No (-8.50%) Yes (1.00%) Initiation of Alcohol and Other Drug Treatment Increase No (-3.40%) No (-5.80%) Not Significant Engagement of Alcohol and Other Drug Treatment Increase Yes (8.30%) Not Significant Yes (10.90%) Follow-Up After Hospitalization for Mental Illness within 7 days Increase Not Significant Not Significant Not Significant Follow-Up After Hospitalization for Mental Illness within 30 days Increase Not Significant Not Significant Not Significant Follow-Up After Emergency Department Visit for Alcohol and Other Drug Abuse or Dependence within 7 days Increase Not Significant Not Significant Not Significant Follow-Up After Emergency Department Visit for Alcohol and Other Drug Abuse or Dependence within 30 days Increase Not Significant Not Significant Not Significant Use of Pharmacotherapy for Opioid Use Disorder Increase Not Significant No (-5.30%) Not Significant OUTCOME METRICS Plan All-Cause Readmissions Decrease Not Significant Not Significant Not Significant Controlling High Blood Pressure Increase No (-1.80%) No (-6.40%) Not Significant Prevention Quality Indicator (PQI) 92: Decrease Yes (-1.30) Yes (-2.50) Not Significant UCLA Evaluation | Executive Summary March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Intended direction Trend during HHP changed significantly in the intended direction? Trend from before to during HHP changed significantly in the intended direction? Trend for HHP patients was better than control group (DD)? Chronic Conditions Composite ESTIMATED PAYMENTS Estimated Total Payments per Enrollee per 6 Months Decrease No ($1,277) No ($1,116) Yes (-$121) Estimated Payments for Outpatient Services per Enrollee per 6 Months Not Specified No Direction ($529) No Direction (S436) No Direction ($18) Estimated Payments for Outpatient Medications per Enrollee per 6 Months Not Specified No Direction ($311) No Direction ($302) Not Significant Estimated Payments for ED Visits Resulting in Discharge per Enrollee per 6 Months Decrease No ($55) No (S50) Yes (-S20) Estimated Payments for Hospitalizations per Enrollee per 6 Months Decrease No ($200) No ($136) Yes (-$127) Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: Yes indicates significant in the intended direction. No indicates significant in the unintended direction. Green indicates change in the intended direction. Red indicates change in the unintended direction. Yellow indicates significant change when direction is not specified. MM indicates member months. ED indicates Emergency Department. Estimated Med-Cal Payments for HHP Enrollees and HHP Costs e UCLA developed estimated payments for Medi-Cal services and these estimated payments are intended for measuring whether HHP led to efficiencies by reducing the total payments for HHP enrollees before and after the program, and in comparison to a group of similar patients in the same timeframe. Executive Summary | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program e Exhibit 3 and Exhibit 4 show that the total estimated Medi-Cal payments continued to increase during the first year of HHP for enrollees compared to before HHP, for both SPA 1 ($163 per enrollee per 6 month) and SPA 2 ($1,116) enrollees. However, the growth in payments for HHP enrollees was smaller than the growth for the respective control groups (DD) by $96 for SPA 1 and $121 for SPA 2. This is likely due to savings associated with receipt of HHP services. o Among SPA 1 enrollees, the trends in estimated payments for all categories of service examined showed a significantly slower growth in total payments for HHP enrollees than the control group, including ED visits (DD: $29) and hospitalizations (DD: $7). o For SPA 2 enrollees, the trends in estimated payments for ED visits (DD: $20) and hospitalizations (DD: $127) decreased significantly more than the control group but trends in outpatient payments increased significantly more (DD: $18) than the control group. e Total estimated HHP expenditures were $189,737,702 and the average expenditures per enrollee per month was $479 by September 30, 2020. Conclusions and Next Steps The findings in this report build on the findings of the first interim evaluation report, which described early progress in building CB-CME networks by MCPs; delivery of HHP services; enrollment size; and health and utilization profile of HHP enrollees prior to enrollment. This report has highlighted the progress made by MCPs in the same areas through September 2020 and additional comparisons that highlight the early impact of HHP. The updated information on the CB-CME networks indicated a substantial growth commensurate with the growth in HHP enrollment over time and continued challenges in reporting of HHP services in claims data. The growth in enrollment may have slowed down and the ability of MCPs and their contracted CB-CMEs to provide HHP services were likely diminished by the onset of the COVID-19 pandemic and subsequent statewide shelter in place order in mid-March 2020. Some of this impact was mitigated by MCP efforts to adapt their workflows and use infrastructure such as telehealth capacity to address challenges. HHP enrollees were complex and high need as highlighted previously in the first interim evaluation report. A closer look at use of acute care services further indicated that a notable proportion of enrollees had super utilization of acute care in emergency departments and hospitals but most had moderate or low utilization. The higher prevalence of enrollees with UCLA Evaluation | Executive Summary March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program super utilization who were also experiencing homelessness and had conditions such as chronic kidney disease and depression confirmed the level of complexity of these enrollees and the reasons for their high use of acute services. This report also included analyses of changes in HHP core metrics and additional utilization and estimated Medi-Cal payment measures. Findings indicated some improvements in metrics that reflected processes and outcomes of care compared to the control group but no change in others. Among the latter were several process metrics related to enrollees with substance use disorders. Lack of change in these metrics during the first year of HHP may reflect challenges of engaging this population in treatment particularly for those who also have SMI in SPA 2. Lack of improvement in outcome metrics compared to the control group may similarly reflect challenges of improving outcomes for enrollees with multiple comorbidities. In contrast to limited findings for core HHP process and outcome metrics, the findings indicated greater declines from before HHP in core metrics for ED visits and hospitalizations in the intended direction and significantly greater declines compared to the control group. These findings provided evidence that enrollment in HHP had the desired effect of reducing the use of acute and high cost services. While estimated payments for ED visits and hospitalization grew more slowly among HHP enrollees than the control group, the payments during HHP increased. These findings likely reflect a higher average cost per service due to a reduction in avoidable and lower cost ED visits and hospitalizations. The findings of increased reductions during HHP among those with super and high utilization of acute care services are consistent with these findings. UCLA further examined use of outpatient services and their associated payments, including payments for outpatient medications to further highlight how reductions in acute care and associated payments may have been realized. The analyses indicated that reductions in acute services occurred concurrently with provision of more primary, specialty, mental health, and SUD services as well as outpatient medications in the first six months following HHP enrollment likely to address the needs of enrollees. These increases were followed by reductions in use of these services in the second half of the year likely because those early needs were addressed. The DD findings are not likely to have been impacted by changes in service utilization due to the COVID-19 pandemic because the pandemic appeared to impact the HHP enrollee and the control groups similarly. Nevertheless, use of telehealth in lieu of in-person outpatient visits may have restricted the ability to provide certain types of services that require an in-person visit. For example, it may have been more challenging to initiate treatment for alcohol and drug use and more difficult to provide outpatient care through telehealth for patients following mental health hospitalizations for patient with more severe conditions such as SMI. Executive Summary | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Certain data limitations prevented a comprehensive assessment of the impact of HHP. For example, UCLA lacked adequate data on delivery of HHP services and could not assess the role of CB-CMEs and specific services they provided on enrollee outcomes. It is possible that outcomes varied by CB-CMEs or type of services they provided. UCLA also lacked data on specific approaches employed by MCPs in selecting eligible beneficiaries or approaches to program implementation beyond their preliminary plans highlighted in their readiness documents. The next evaluation report will include data for the final year of HHP, including changes in the HHP core metrics and measures of utilization and estimated Medi-Cal payments. The report will also discuss the implications of the findings for improving the health of Medi-Cal beneficiaries with complex conditions and high utilization of health care services. UCLA Evaluation | Executive Summary March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Introduction Health Homes Program Overview The Health Homes Program (HHP) was created and implemented under the statutory authority of California Assembly Bill (AB) 361. The legislation authorizes the California Department of Health Care Services (DHCS) to create HHP under Section 2703 of the 2010 Patient Protection and Affordable Care Act. Section 2703 allows states to create Medicaid health homes to coordinate the full range of physical health, behavioral health, and community-based long-term services and supports needed by Medi-Cal enrollees with chronic conditions. HHP is implemented in 12 California counties for Medi-Cal Managed Care Plan (MCP) enrollees who meet certain chronic condition and acuity criteria. All Medi-Cal MCPs in the 12 participating counties were required to participate in HHP. HHP has a phased implementation schedule, and individuals with chronic physical health conditions or substance use disorders (SUD) are included in State Plan Amendment (SPA) 1 (i.e., Phase 1) and those with severe mental illness (SMI) are included in SPA 2 (i.e., Phase 2). The primary goals of HHP are to improve member outcomes through care coordination and reduce avoidable health care costs. MCPs are expected to deliver HHP services directly or through contracted community-based care management entities (CB-CMEs), which could include primary care providers (PCPs), Federally Qualified Health Centers (FQHCs), and other service providers. CB-CMEs work with Community Based Organizations (CBOs) to provide linkages to community and social support services, as needed. HHP Implementation Plan The HHP implementation schedule is displayed in Exhibit 5. The 12 counties implementing HHP were divided into four groups, with Group 1 scheduled to begin implementation on July 1, 2018, and Group 4 to implement the final phase on July 1, 2020. Each Group would first implement HHP for SPA 1 enrollees (those with chronic physical health conditions and/or SUD), followed six months later by SPA 2 enrollees (those with SMI). Introduction | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Exhibit 5: Timeline of HHP Implementation by Group and SPA July 2019 Group 3 Alameda Imperial Kern Los Angeles January 2019 Sacramento duly 2018 Group 2 San Diego January 2020 Group 1 Riverside Santa Clara Group 4 San Francisco San Bernardino Tulare Orange A ite) January 2019 duly 2019 January 2020 duly 2020 Group 1 Group 2 Group 3 Group 4 San Francisco Riverside Alameda Orange San Bernardino Imperial Los Angeles Sacramento San Diego Santa Clara Tulare Source: Adapted from HHP Implementation Schedule. HHP Managed Care Plans. Note: SPA is State Plan Amendment. A total of 16 MCPs implemented HHP across the 12 counties (Exhibit 6). MCPs were responsible for the overall administration of HHP and expected to fulfill HHP requirements by leveraging existing infrastructure, communication, and reporting capabilities. MCP responsibilities included (1) performing regular auditing and monitoring activities; (2) training, supporting, and qualifying CB-CMEs; (3) providing CB-CMEs with timely information on admissions, discharges, and other key utilization and health condition information; (4) when possible, providing access to immediate housing post discharge and permanent housing for the homeless; and (5) fulfilling HHP care management requirements. UCLA Evaluation | Introduction UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 6: MCPs that Implemented HHP across California, by Group and County Group County Managed Care Plan 1 San Francisco Anthem Blue Cross of California Partnership Plan, Inc. San Francisco Health Plan 2 Riverside Inland Empire Health Plan Molina Healthcare of California Partner Plan, Inc. San Bernardino Inland Empire Health Plan Molina Healthcare of California Partner Plan, Inc. 3 Alameda Alameda Alliance for Health Anthem Blue Cross of California Partnership Plan, Inc. Imperial California Health & Wellness Molina Healthcare of California Partner Plan, Inc. Kern Health Net Community Solutions, Inc. Kern Health Systems Los Angeles Health Net Community Solutions, Inc. L.A. Care Health Plan Sacramento Aetna Better Health of California Anthem Blue Cross of California Partnership Plan, Inc. Health Net Community Solutions, Inc. Kaiser Permanente Molina Healthcare of California Partner Plan, Inc. San Diego Aetna Better Health of California Blue Shield of California Promise Health Plan Community Health Group Partnership Plan Health Net Community Solutions, Inc. Kaiser Permanente Molina Healthcare of California Partner Plan, Inc. UnitedHealthcare Community Plan of California, Inc. Santa Clara Anthem Blue Cross of California Partnership Plan, Inc. Santa Clara Family Health Plan Tulare Anthem Blue Cross of California Partnership Plan, Inc. Health Net Community Solutions, Inc. 4 Orange CalOptima Source: DHCS. Notes: MCP is Managed Care Plan and DHCS is the California Department of Health Care Services. HHP Services The overarching goal of HHP was to achieve the "triple aim" of better care, better health, and lower costs. To achieve these goals, MCPs provided HHP services most often through community-rooted CB-CMEs. These services included (1) comprehensive care management, (2) care coordination, (3) health promotion, (4) comprehensive transitional care, (5) individual and family support services, and (6) referrals to community and social support services. Exhibit 7 displays detailed descriptions of these services. Introduction | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 7: HHP Services Provided through MCPs and CB-CMEs Service Description Comprehensive care management | e¢ Engage MCP members to participate in HHP e Collaborate with HHP enrollees and their family/support persons to develop a Health Action Plan (HAP) within 90 days of enrollment that is comprehensive and person-centered e Reassess HAP as needed and track referrals e Case conferencing to support continuous and integrated care among all service providers Care coordination e = Provide enrollee support to implement HAP and attain enrollee goals e = Coordinate referrals and follow-ups, share information to all involved parties, and facilitate communication e Frequent, in-person contact between HHP enrollees and care coordinators e Appointment with primary care physician within 60 days of enrollment encouraged e Identify and address enrollee gaps in care ® Maintain an appointment reminder system for enrollees as appropriate e Link eligible enrollees who are homeless or experiencing housing instability to permanent housing Health promotion e Encourage and support HHP enrollees to make lifestyle choices based on health behavior e Encourage and support health education e Assess and motivate enrollees and family/support person understanding of health condition and motivation to engage in self- management Comprehensive transitional care e Facilitate HHP enrollees' transition from and among treatment facilities Provide medication information and reconciliation e Plan follow-up appointments and anticipate care or place to stay post-discharge Individual and family support e Ensure HHP enrollees and family/support persons are educated about services the enrollee's conditions to improve treatment and medical adherence Referrals to community and social | ¢ Determine appropriate services to meet HHP enrollee's needs support services e Identify and refer enrollees to available community resources Source: Adapted from Health Homes Program Guide. Notes: MCP is Managed Care Plan and CB-CME is Community-Based Care Management Entity. HHP Target Populations The eligibility criteria defined by DHCS for HHP was based on the presence of specific chronic conditions and evidence of high acuity (Exhibit 8). These criteria aimed to identify the Medi-Cal population who may benefit the most from HHP services. DHCS identified a Targeted Engagement List (TEL) of Medi-Cal MCP enrollees in the 12 participating counties who were likely to be eligible for HHP services based on specific inclusion and exclusion criteria. UCLA Evaluation | Introduction March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program The exclusion criteria were designed to limit enrollment to eligible enrollees who were not receiving similar services in other programs and were more likely to benefit from HHP than other interventions, among other reasons. The TEL did not capture the inclusion criteria of chronic homelessness or some exclusion criteria, such as enrollees who would benefit from alternative care management programs, due to data limitations. DHCS delegated this responsibility to MCPs, and allowed MCPs to use other eligibility identification strategies, subject to DHCS approval. Exhibit 8: HHP Eligibility Inclusion and Exclusion Criteria Eligibility Requirement Criteria Details Met at least one chronic condition e =©At least two of the following: chronic obstructive pulmonary disease, criteria diabetes, traumatic brain injury, chronic or congestive heart failure, coronary artery disease, chronic liver disease, chronic renal (kidney) disease, dementia, substance use disorders e Hypertension and one of the following: chronic obstructive pulmonary disease, diabetes, coronary artery disease, chronic or congestive heart failure e One of the following: major depression disorders, bipolar disorder, psychotic disorders (including schizophrenia) Asthma Met at least one acuity/complexity e Has at least three or more of the HHP eligible chronic conditions criteria e =At least one inpatient hospital stay in the last year e Three or more emergency department (ED) visits in the last year e Chronic homelessness Did not meet one of the exclusion e Hospice recipient or skilled nursing home resident criteria e Enrolled in specialized MCPs (e.g., Program of All-Inclusive Care for the Elderly (PACE), Senior Care Action Network (SCAN) and AIDS Healthcare Foundation (AHF)) e Fee-for-service rather than managed care e = Sufficiently well managed through self-management or another program e More appropriate for alternative care management programs e Behavior or environment is unsafe for CB-CME staff Source: Adapted from Health Homes Program Guide. Funding and Payment Methodology Under federal rules, DHCS would receive a 90% enhanced Federal Medical Assistance Percentage (FMAP) for HHP services for the first two years of each phase of implementation. However, the federal portion will revert to the 50% FMAP after this period. DHCS used grant funds provided by The California Endowment to pay for the state's share of HHP services. MCPs received a supplemental per member per month (PMPM) payment for HHP services and reimbursed CB-CMEs based on claims for services under contractual agreements. DHCS also created an HHP-specified Healthcare Common Procedure Coding System (HCPCS) procedure Introduction | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program code and modifiers to report HHP services. These codes are described later in this report in the HHP Service Utilization among HHP Enrollees chapter. Transition to CalAIM Services provided under HHP will be incorporated into CalAIM, a multi-year initiative by DHCS designed to use HHP approaches to improve beneficiaries' health outcomes. Under CalAIM, Medi-Cal managed care plans are expected to provide Enhanced Care Management and Community Supports through contracts with community-based providers, including CB-CMEs participating in HHP. Members receiving HHP will be transitioned to Enhanced Care Management when CalAIM is expected to begin implementation in January 2022. UCLA HHP Evaluation AB 361 required an independent evaluation of HHP and submission of a report to the legislature after two years of implementation; this requirement was met by way of submission of the first HHP Evaluation Report in October 2020. This is the second interim evaluation report and a final evaluation report will be developed after the HHP program ends at the end of 2021, and Members are transitioned to Enhanced Care Management as part of CalAIM in January 2022. The UCLA Center for Health Policy Research (UCLA) was selected as the evaluator of the HHP program. Conceptual Framework UCLA developed a conceptual framework for the evaluation of HHP (Exhibit 9). Following the HHP program goals and structure, the framework indicated that better care is achieved when MCPs establish the necessary infrastructure and deliver HHP services. Delivery of HHP services will in turn lead to better health indicated by reduced utilization of health care services that are associated with negative health outcomes as well as improvements in population health indicators. Better care and better health will lead to lower overall health care expenditures. UCLA Evaluation | Introduction UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 9: HHP Evaluation Conceptual Framework eInfrastructure: HHP network composition, organization model of community-based care management, care coordination staffing, health information technology (HIT) and data sharing approach, patient enrollment approach ¢Process: provide comprehensive care management, coordinate care, deliver health promotion services, provide comprehensive transitional care, provide individual and family support Better Care ' , ' services, refer to community and social support services eHealth care utilization: reduce emergency department visits, reduce inpatient hospitalizations, reduce length of stay, increase outpatient follow-up care post admission, reduce nursing facility admissions, increase use of substance use treatment ¢Patient outcomes: control blood pressure, screen for depression, assess body mass index BTlaeclg (BMI), reduce all-cause readmissions, reduce inpatient admission for ambulatory care sensitive Steel Ith chronic conditions eHealth care expenditures: reduce overall expenditures by lower spending on acute care services and higher spending on needed outpatient services ¢Cost neutrality: maintain cost neutrality by insuring HHP service expenditures do not lead to higher overall expenditures eReturn on investment: show return on investment due to HHP program implementation Introduction | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Evaluation Questions and Data Sources Exhibit 10 displays the evaluation questions and data sources that were used to answer those questions. The evaluation questions were aligned with the components of the conceptual framework. Questions 1-7 examined the infrastructure established by MCPs including the composition of their networks, populations enrolled, and the services delivered. Questions 8-13 examined the impact of HHP service delivery on multiple indicators of health services utilization as well as patient health indicators. Questions 14 and 15 examined the impact of HHP on lowering costs of the Medi-Cal program. Exhibit 10: Health Homes Program Evaluation Questions and Data Sources Evaluation Questions | Data Sources Better Care Infrastructure 1. What was the composition of HHP networks? e MCP Readiness Documentation 2. Which HHP network model was employed? e MCP Quarterly HHP Reports 3. When possible, what types of staff provided HHP services? 4. What was the data sharing approach? 5. What was the approach to targeting patients for enrollment per HHP network? Process 6. What were the demographics of program enrollees? What was the acuity level of the enrollees including health and health risk profile indicators, such as aggregate inpatient, ED, and rehab skilled nursing facility (SNF) utilization? What proportion of eligible enrollees were enrolled? How did enrollment patterns change over time? What proportion of enrollees are homeless? 7. Were Health Home services provided in-person or telephonically? Were Health Home services provided by clinical or non-clinical staff? How many enrollees received engagement services? How many homeless enrollees received housing services? Better Health Health care utilization 8. How did patterns of health care service use among HHP e Medi-Cal Enrollment and Claims Data enrollees change before and after HHP implementation? 9. Did rates of acute care services, length of stay for hospitalizations, nursing home admissions and length of stay decline? 10. Did rates of other services such as substance use treatment or outpatient visits increase? Patient outcomes MCP Enrollment Reports MCP Quarterly HHP Reports TEL Medi-Cal Enrollment and Encounter Data UCLA Evaluation | Introduction March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Evaluation Questions Data Sources 11. How did HHP core health quality measures improve e MCP Quarterly HHP Reports before and after HHP implementation? @ Medi-Cal Enrollment and Claims Data 12. Did patient outcomes (e.g., controlled blood pressure, screening for clinical depression) improve before and after HHP implementation? 13. How many homeless enrollees were housed? Lower Costs Health care expenditures 14. Did Medi-Cal expenditures for health services decline e Medi-Cal Enrollment and Claims Data after HHP implementation? 15. Did Medi-Cal expenditures for needed outpatient services increase? Note: TEL is Targeted Engagement List. Detailed descriptions of the data sources and analytic methods used in the evaluations can be found in Appendix A: Data Sources and Analytic Methods and Appendix B: UCLA HHP Evaluation Design. Introduction | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program HHP Implementation and Infrastructure This section addresses the following HHP evaluation questions: What was the composition of HHP networks? Which HHP network model was employed? When possible, what types of staff provided HHP services? What was the data sharing approach? What was the approach to targeting patients for enrollment per HHP network? Ww PWnr UCLA relied on three data sources to address these questions: (1) MCP readiness documents, which outlined MCPs' plans to develop and implement HHP under the guidelines set by DHCS; (2) the MCP Quarterly HHP Reports, which detailed the networks developed by the MCP during each quarter of the program; and (3) a one-time self-report by MCPs in September 2020 to provide additional detail on their CB-CME networks. A total of 16 MCPs implemented HHP across California, submitting both readiness documents and Quarterly HHP Reports. The time period of this report covers data through September 30, 2020. UCLA aimed to answer the HHP evaluation questions by identifying and analyzing the strategies that each MCP planned to implement and by providing selected illustrative examples of these strategies. Since the first interim report, the data available through MCP readiness documents remain the same and UCLA provides a summary of these findings from the first interim report in this section. The HHP Delivery Networks section is updated with new information. Further analytic approach details can be found in Appendix A: Data Sources and Analytic Methods. UCLA Evaluation | HHP Implementation and Infrastructure March 2022 HHP Implementation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 11 displays the participating HHP counties by their respective implementation groups and the MCPs implementing HHP in each county. Of the 12 counties implementing HHP, four counties were in Northern California, two in Central California, and the remaining six were in Southern California. A total of 16 MCPs were operating across the state with six MCPs (Aetna, Anthem, Health Net, Inland Empire, Kaiser Permanente, and Molina) operating in multiple counties. Exhibit 11: Distribution of California Counties by Health Homes Program Implementation Group and MCPs Implementing Health Homes Program by County Legend [ | Group 1 |_| Group 2 il Group 3 Li Group 4 PNET ce] Pacific Dye Rel le- San Bernardino Fa) ere CoM cL oft ' Los Angeles Source: Adapted from Health Homes Program Guide. Note: MCP is Managed Care Plan. Aetna Better Health of California Sacramento, San Diego Alameda Alliance for Health Alameda Anthem Blue Cross of California Partnership Plan, Inc. Alameda, Sacramento, Santa Clara, Tulare, San Francisco Blue Shield of California Partnership Plan San Diego Promise Health Plan "ei California Health & I ial mperia Wellness P CalOptima Orange Community Health Group San Diego Health Net Community Solutions, Inc. Kern, Los Angeles, Sacramento, San Diego, Tulare Inland Empire Health Plan Riverside, San Bernardino Kaiser Permanente Sacramento, San Diego Kern Health Systems Kern LA. Care Health Plan Los Angeles Molina Healthcare of California Partner Plan, Inc. Imperial, Riverside, San Bernardino, Sacramento, San Diego San Francisco Health Plan San Francisco Santa Clara Family Health California, Inc. plan Santa Clara United Healthcare Community Plan of San Diego HHP Implementation and Infrastructure | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program MCP HHP implementation plans outlined in readiness documents were used to examine MCP intentions at the beginning of HHP, even though the plans may have changed during implementation. These plans indicated that 15 (of 16) MCPs used delivery Model |, where CB- CMEs were typically medical providers that hired and housed HHP staff, including care coordinators. When HHP enrollees' medical providers were not able to take on these responsibilities, MCPs utilized Models II and III to deliver services centrally or regionally. See the first interim evaluation for more details. HHP Delivery Models HHP Delivery Networks HHP delivery networks were composed of CB-CMEs who either used their own staff or sub- contracted with other community-based organization to deliver care management (CM) services. CB-CMEs were certified by the MCPs using DHCS general guidelines and requirements. CB-CMEs were required to maintain a strong and direct connection with the HHP enrollee and their primary care physician, the latter being applicable when CB-CMEs were not medical providers. Goals in developing a MCP's CB-CME network included: (1) ensuring CM delivery at point of care, (2) experience with high utilizing and homeless populations, and (3) building upon existing CM infrastructure within the county. Six MCPs indicated that they acted as a CB-CME for a portion of their HHP enrollees; these MCPs included Blue Shield, CalOptima, Inland Empire, Kern, LA Care, and San Francisco Health Plan. In their Quarterly HHP Reports, and as verified through self-reports to UCLA, MCPs reported developing contracts with 244 unique CB-CMEs (as identified by organization name per MCP) by September 2020. CB-CMEs by Organization Type MCPs identified the organization type of their CB-CMEs. Of the 244 unique reported CB-CMEs, MCPs most commonly identified them as community health centers (includes Federally Qualified Health Centers, rural health centers, Indian health centers, and other similar organizations; 41%; Exhibit 12). The next most common organizational type of CB-CMEs included community-based social service organizations or local government entities (28%). CB- CMEs were also commonly identified as community based primary care or specialty physicians (19%). UCLA Evaluation | HHP Implementation and Infrastructure March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 12: Health Homes Program CB-CME Network by Organization Type as of September 2020 Community health centers [iE 41% Community based social service organization or local a 23 government entity Community based primary care or specialty physician Po 19% Hospital or hospital-based physician group m7 6% Specialty mental health, behavioral health, or substance use | 6% treatment center ° Source: MCP Quarterly HHP Reports up to September 2020 and MCP Self-Reports to UCLA in September 2020. Note: CB-CME is Community-Based Care Management Entity, MCP is Managed Care Plan, and NPI is National Provider Identifier. A total of 244 CB-CMEs were reported and MCPs clarified CB-CME type in self reports to UCLA in September 2020. Community health centers included Federally Qualified Health Centers, rural health centers, Indian health centers, and other similar organizations. CB-CMEs and Projected HHP Capacity MCPs reported the projected number of enrollees each CB-CME would serve under their contracts (referred to as capacity) in MCP Quarterly HHP reports. DHCS required MCPs to report capacity criteria such as the HHP care manager ratios and certification requirements. For example, CB-CMEs had to have the ability to provide appropriate and timely in-person care coordination, telephonic communication, and accompany HHP enrollees to critical appointments. As of September 2020, MCPs reported 224 CB-CMEs with capacity for a minimum of 11 or more enrollees. These CB-CMEs collectively had a projected capacity for managing the needs of approximately 79,370 HHP enrollees, with a median of 180 enrollees per CB-CME (Exhibit 13). The median capacity was largest among community based primary care or specialty physicians and hospital or hospital-based physician groups (240 enrollees). Community health centers reported the smallest median capacity (122 enrollees). An additional 20 CB-CMEs with less than 11 enrollees were reported, but not included in the analysis below. HHP Implementation and Infrastructure | UCLA Evaluation UCLA Center for Health Policy Research NTE Health Economics and Evaluation Research Program Exhibit 13: Total Projected CB-CME Capacity for Health Homes Program Enrollment by CB-CME Organization Type as of September 2020 CB-CME Type N Total Capacity Median Projected Capacity Total 224 79,370 180 Community health centers 95 26,974 (34%) 122 Other entity (e.g., community based social 56 17,935(23%) 150 service organization, homeless service provider) Community based primary care or specialty 45 28,722 (36%) 240 physician Hospital or hospital-based physician group 15 3,713 (5%) 174 Specialty mental health, behavioral health, or 13 2,026 (3%) 132 substance use treatment center Source: MCP Quarterly HHP Reports up to September 2020 and MCP Self-Reports to UCLA in September 2020. Notes: CB-CME is Community-Based Care Management Entity, MCP is Managed Care Plan, and NPI is National Provider Identifier. A total of 224 CB-CMEs were reported to have 11 or more enrollees assigned and MCPs self-reported CB-CMEs into distinct organization types in self reports to UCLA. Community health centers included Federally Qualified Health Centers, rural health centers, Indian health centers, and other similar organizations. CB-CMEs in the "Other" category included community based social service organizations, homeless service providers, and local government entities. Changes in CB-CME Networks Over Time Since September 2019, an additional 54 CB-CMEs were reported among all MCPs as of September 2020 (previously 190 CB-CMEs). MCPs most often classified these new CB-CMEs as community based social service organizations (41 of 54). From September 2019 to September 2020, most MCPs gained CB-CMEs (9 of 16), ranging from one to 13 additional CB-CMEs. Few MCPs (3 of 16) lost CB-CMEs or had no change (four of 16) within their CB-CME network. CalOptima and L.A. Care reported the greatest increase in CB-CMEs (13), whereas Inland Empire Health Plan had the greatest decrease in CB-CMEs (only two). HHP Staffing MCPs ensured that CB-CMEs had adequate staffing to deliver HHP services by requiring certain staffing types such as care coordinators, HHP directors, clinical consultants, and housing navigators. In readiness documents, 11 MCPs (of 16), including all of the MCPs that implemented in more than one County, indicated that they planned to hire certain HHP staff internally to improve efficiency and effectiveness. These roles most often included directors, program managers, and housing specialists. See the first interim evaluation for more details. UCLA Evaluation | HHP Implementation and Infrastructure ca March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program HHP Data Sharing Seven MCPs planned to use a SFTP or dedicated email and six MCPs planned to use electronic health records (EHR), care management platforms, or health information exchange (HIE) data sharing technologies. Both CB-CMEs and MCPs planned to use data sharing technologies to provide timely access to information. Eight MCPs (of 16) planned to provide access to a dynamic Health Action Plan (HAP) to allow access to up-to-date information and five MCPs planned to provide real-time and automated notifications of HHP hospital admissions or emergency department visits to CB-CMEs. See the first interim evaluation for more details. Communication with HHP Enrollees MCPs developed plans for identifying and targeting individuals for HHP enrollment including use of predictive modeling and risk grouping of eligible beneficiaries. MCPs most often planned to use newsletters (nine of 16) and websites (nine) to communicate with eligible beneficiaries and developed plans on how often they would outreach to eligible beneficiaries. MCPs planned to use a mix of approaches to target individuals experiencing homelessness. These approaches included collaborating with CB-CMEs or community-based organizations that specialized in working with these individuals and leveraging existing infrastructure developed under Whole Person Care to provide outreach. See the first interim evaluation for more details. HHP Implementation and Infrastructure | UCLA Evaluation UCLA Center for Health Policy Research NTE Health Economics and Evaluation Research Program HHP and COVID-19 This section addresses the following evaluation questions, included in response to the COVID- 19 pandemic: 1. How did the COVID-19 pandemic impact HHP implementation? How many HHP enrollees had COVID-19 related services? How did healthcare utilization patterns change among HHP enrollees during the COVID- 19 pandemic compared to the year prior to the pandemic? The COVID-19 pandemic began during HHP enrollment. HHP Group 1, Group 2 and Group 3/SPA 1 were implemented between 6 and 18 months prior to the first reports of COVID-19 in the United States in January 2020. HHP Group 3/SPA 2 and Group 4 implemented just as these first cases were reported. In this chapter, UCLA examines the likely impact of the pandemic on HHP implementation. The progress of the pandemic in counties where HHP was implemented was examined using data on COVID-19 cases and hospitalizations from April 2020, when such data were first available, through September 2020, the last month of this evaluation. These data, along with population counts from the Census Bureau, were used to calculate cases and hospitalizations per 100,000. The impact of COVID on MCP implementation efforts was examined through a COVID-19 Impact Survey (Appendix E) of all participating MCPs (n=16, response rate of 100%) in September 2020. MCPs respondents included HHP program managers and directors who were most informed about HHP implementation at their respective organizations. The impact of COVID-19 on CB- CMEs that had contracted with MCPs was assessed from a survey administered by the Corporation for Supportive Housing (CSH) in August 2020. UCLA submitted survey questions that were similar to those asked from MCPs to CSH who then distributed the survey to all contracted CB-CMEs at the time and collected the data. The 59 CB-CMEs (response rate of 24%) that responded were unevenly distributed by county with six CB-CMEs operating in more than one county. In addition, respondents ranged from frontline staff such as care coordinators, to program managers, to chief operating officers and were likely to represent different points of view. This data was included in this section to provide a general overview of CB-CME experiences during the pandemic, but represents a convenience sample that may not be generalizable to all CB-CMEs participating in HHP. UCLA further used Medi-Cal enrollment and claims data to (1) identify HHP enrollees and their controls that have services with COVID-19 as the primary or secondary diagnosis and (2) report UCLA Evaluation | HHP and COVID-19 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program changes in overall health care utilization pre- and post-pandemic for HHP enrollees and their controls. COVID-19 cases were identified using the COVID-19 ICD diagnosis code, which was first introduced in late March 2020. Therefore, these cases were likely to be underreported early in the pandemic. Progression of COVID-19 Cases and Hospitalizations in HHP Counties UCLA assessed the progression of the COVID-19 cases by examining cumulative case rates and 14-day average hospitalization rates in HHP counties and California overall. Among all Californians, the cumulative case rate of COVID-19 reached 2,074 per 100,000 by the end of September 2020 (Exhibit 14). Rates remained low across the state and HHP counties, with the exception of Imperial County, until June 2020 and then began to climb. Cumulative case rate per 100,000 as of September 2020 among HHP counties ranged from a low of 1,105 in Santa Clara to a high of 6,571 in Imperial. The cumulative case rates for all Group 2 (Riverside and San Bernardino) HHP counties were above that of the entire state. Changes in these rates over time represent both the progression of the pandemic as well as changes in testing and reporting. Exhibit 14: Cumulative COVID-19 Cases per 100,000, April 2020 to September 2020, HHP Counties and California Imperial: 6,571 Kern: 3,569 Tulare: 3,489 Los Angeles: 2,693 San Bernardino: 2,532 Riverside: 2,405 California: 2,074 Orange: 1,693 Sacramento: 1,460 San Diego: 1,413 Alameda, 1,279 San Francisco: 1,279 Santa Clara, 1,105 oan ee oe] - \pril 2020 May 2020 June 2020 July 2020 August 2020 September 2020 Source: UCLA analysis of daily COVID-19 cases reported from April 1, 2020 to September 30, 2020 by the LA Times. State and County population numbers were collected through Census data. Cases per 100,000 were calculated by multiplying cases by 100,000 then dividing by the population. HHP and COVID-19| UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program UCLA also assessed COVID-19 hospitalization rates as an indicator of the burden of disease on the healthcare system. From April to September 2020, the 14-day average hospitalization rate across California peaked near the end of July with 18 hospitalizations per 100,000 before returning to around 7 hospitalizations per 100,000 as seen early in the pandemic (Exhibit 15). While most HHP counties had a similar burden of disease, notable exceptions included Imperial County that had an extended peak from May 2020 through August 2020 and two peaks in Los Angeles County in late April 2020 and late July 2020. Exhibit 15: 14-day Average COVID-19 Hospitalization Rate per 100,000, April 2020 to September 2020, Statewide and HHP Counties 50 45 Alameda 40 em |mperial eee Kern 35 Statewide Peak, 18 === Los Angeles Hospitalizations per 30 100,000 Orange 35 o== Riverside oe Sacramento 20 eee San Bernardino 15 ee San Diego eee San Francisco 10 e===== Santa Clara 5 em Tulare = = California April-2020 May-2020 June-2020 July-2020 August-2020 September- 2020 Source: Daily COVID-19 hospitalizations reported from April 1, 2020 to September 30, 2020 through the California Department of Public Health. State and County population numbers were collected through Census data. Hospitalizations per 100,000 were calculated by multiplying hospitalizations by 100,000 then dividing by the population. UCLA Evaluation | HHP and COVID-19 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program UCLA also assessed the cumulative death rate per 100,000 and new daily deaths from COVID-19 in California, as reported by local public health departments, to estimate the burden of highly resource intensive, severe disease. By the end of September 2020, there were 40 COVID-19 deaths per 100,000 in California (data not shown). The death rate among HHP counties was highest in Imperial (177 deaths per 100,000), followed by Los Angeles (66 per 100,000). The new daily deaths from COVID-19 in California had an initial peak on April 22, 2020 at 115 new deaths and a second peak at 219 new deaths on July 31, 2020 (data not shown), which aligned with the peaks in hospitalizations in Los Angeles County. Impact of COVID-19 on HHP Implementation and Infrastructure UCLA assessed the impact of COVID-19 on HHP implementation using the MCP and CB-CME surveys. At the time of these surveys, all HHP counties were at or beyond their first peak in COVID-19 hospitalizations as shown in Exhibit 15. Impact of COVID-19 on MCP Processes, Procedures, and Policies MCPs were asked to indicate if any of their processes, procedures, or policies changed and how they were impacted due to the COVID-19 pandemic. They were further asked to rate the overall impact of the pandemic on these processes, procedures, or policies. Responses showed that only two processes, (1) identifying eligible HHP enrollees and (2) reporting, were not changed by any MCP due to the pandemic (Exhibit 16). However, other activities, such as (1) housing and homeless support services, (2) comprehensive transitional care, and (3) delivery of care coordination by frontline staff, were largely impacted with mean impact scores of 7.5, 6.9, and 6.8, respectively. Overall, there was no difference between SPA 1 and SPA 2 in changes to processes, procedures, or policies, except for the ability to provide health promotion and individual/family support services (eight MCPs reported a change in SPA 1, while seven MCPs reported a change in SPA 2). Only Anthem and Molina noted variation in impact at the county level (data not shown). HHP and COVID-19| UCLA Evaluation eS 6T-GIAOD pue dHH | voenjeaq yION "Wd§ JayUa Ul (0 < 9409s) JDedWI Ue payOUap OYM SSOU} JO4» '6T-GIAOD 0} anp saaiuas Yoddns Ajiwey/jenpiaipul pue uoljowodd yyeay apiAosd 07 Aqjiqe ay} uo Joedul ue pajou Z Wd§ Ul UAVS PUe T Yds Ul 1419 - Yds Aq asuodsa, ul UONeWeA, "Oedw! pue sadueyd ay} aquasap Ayag asea|d "SAIJAIIIE paye/aJ-dqHH SUIMO]|O} dYy1 Wiopad 07 Ajiqe (S,JIND-dD pa}2e1}U09 UNOA JO) S$ ,Uol}ezIUeZIO UNOA UO DILUapUed GT-CIAOD 24} JO edu! Uy} a]e1 aseald 'OT-O Jo a]eds e UC,, :uoI]saNb Oo} asuodsay :310N 'QT=U '0Z0Z JaqUIaAON-saquiaidas 'AaAins yedWI| GT-GIAOD dHH VION :@2unos es 6 9 SJIND-dD JO 1UZIsiaAo pue SuOWUOW qIW Ee v 0 (e1ep Jo ssauanisuayasdwos Jo Adeunave 'SFIAID-qD Woy eyep BUIAIadaI ul SAejap "3'a) Sulioday vv S € (sadueys ssado0id/Adijod 0} asuodsay ul! $]9e11U0D FIAD-g2 SUIISIXa 0} SUOISIAAI 'SJIND-dD Mau YUM ZulDeI}UOD SaBua]|eyd "3"3) SIIAID-GD YUM spesqU0?) "Lg TT S (saounosa. poo 'suisnoy ""3"3) soddns |eld0s pue AWUNWWOD 0} SIIAID-gD 40/pue daW Aq jeajay GL TT L SddIAJas Oddns ssajawoy pue suisnoy 6'9 8 S (uoieyodsuew} 'ssauueld azseudsip jeyidsoy YUM ZUNEUIPJOODS 'SUOIIedIJ170U UOIssiwipe ""3°a) ased |euOl Sue] SAISUBYSIGWO7 LS TT 8 x (UOIJESSAD BUPj|OWS se YINS SaduNOSad 0} Sjeajas 'UOIJEONpa YWeay saoaya "d'3) SadI|AJas Woddns Ajiwey/jenpiaipul pue uoljowold yyeay apiaoid 0} Aqjiqy 6S TL 6 (saouasajuod ased 'Ue|q UOlDY YESH JO UO!}eJUaWAa]dWI ':3"~a) JyeyS SuI[WUOIy Aq UOMeUIPIOOS a4ed Jo AJaAI[aq 89 TT OT (pepiaoid YydIYyM ul UO! e30] 'Ayjepow 'AQuanbay "3"a) yyeys auljyuos Aq JUaWaZeUeW Sued DAISUBYaIGLUOS Jo AJaAI|[aQg T'S ST TT (sacred pue salayjod uouajad JO JUBWINIIAI 'Jdd JO BSN pue SUSIA UOSJad -U! 10} $[090} 01d 'yJOMaIIa} 0} YIYs "3"d) Saunpadoud pue saloyjod suijyeys auljyUol4 a9 €T OT (uossad-ul 'yyeayajay 'olUoYydaja} "3*d) saa]joJUa qHH YUM Suo!yed|UNWWO) 61 ST 8 (yoeasjno "3'3) HH OJ! Saelijauaq a[qIsi|a Jo JUaW||OJUa pue JUaWasesuUy os € 0 (sjesuajas 'eyep SAIes}sIuIWpe "3'a) Sda|JO1Ua gHH aqisi|a SulApUap| (pa}2eduy Ajawias3x3 6T-GIAOD 6T-GIAOD Ad\j0d/ainpas014/ss9201d = OT pue pazoedui je e 0} ang pedwu| 0} ang asuey)D ION = 0 'OT-0 2]2S) a109S ue payoday e payioday ypeduy ueayy edu] ue JEY} Sd JO 34} SdOIN payoday oym SddW 404 | (9T=N) sequinn 40 (9T=N) Jaqunn SalD1|Od J0/pue 'Sainpad0Jd 'SAaSSaI01d dHH UO WWapUed GT-GIAOD JO Ped Jo sioday dIW :9T UqIYXI cc0¢ YEW wei301qg yoseasay UoNenjeaq pue soiu0Uodg YyeaH youeasay Adljod YeaH JOJ Ja]UaD ION March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Overall, the majority of MCPs (n=9) noted a relatively low impact of COVID-19 on their plan's overall ability to achieve desired HHP outcomes (e.g., enrollment, care management, outreach and engagement), while three MCPs reported more extensive impacts (data not shown). In general, there was an initial adjustment period to the new restrictions and policies put into place for the MCPs, but many were able to establish new workflows and infrastructure to support their own and CB-CME's operation in the pandemic environment (e.g., acquiring the necessary personal protective equipment, transitioning services to telehealth, strategic coordination with housing services and shelters). CB-CMEs ratings of the impact of COVID-19 on their (1) ability to engage enrollees in HHP, (2) infrastructure and HHP implementation, (3) delivery of HHP core services, and (4) linkages and referral activities are shown in Exhibit 17. On a scale of 0-10, average CB-CMEs ratings to each question ranged from 5.7-6.4 and there was variability by county. For example, CB-CMEs reported an average rating of 6.3 on delivery of HHP core services, but this rating was as low as 3.3 in some counties and as high as 10 in others. Exhibit 17: COVID-19 Pandemic Impact on HHP Processes, Procedures, and Policies, from the CB-CME Perspective, Overall and Range by County Engagement of enrollees in HHP Infrastructure (e.g., policies and procedures, staffing, data sharing) and HHP implementation Delivery of HHP Linkages and referrals (i.e., community and (e.g., outreach, (e.g., shift to core services (e.g., social support communication, telework, | care management, services, housing follow-up) telehealth) care coordination) services) Average Rating 5.9 6.4 6.3 5.7 Range of Ratings by County 4.7-10.0 4.1-8.3 3.3-10.0 2.7-10.0 Source: Corporation for Supportive Housing (CSH), Health Home Program & Homeless Training Survey, August-September 2020 (n=59). Note: Response to question: "On a scale of 0-10, please rate the impact of the COVID-19 pandemic on your organization's ability to perform the following HHP-related activities." HHP and COVID-19| UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 18 highlights illustrative quotes from CB-CME respondents on the impact of COVID-19 on the same four questions. CB-CMEs transitioned to predominantly telephonic engagement and service provision with mixed success, indicating that some enrollees have been more available to engage telephonically and others have been challenging to engage meaningfully. Throughout the pandemic, data sharing and electronic health records had proved beneficial to coordinating care and ensuring timely communication with enrollees and amongst the care team. Many CB-CMEs noted how community resources have been stretched thin as a result of the pandemic, which has created challenges for timely linkages and referrals. Exhibit 18: Illustrative Quotes on COVID-19 Pandemic Impact on HHP Processes, Procedures, and Policies, from the CB-CME Perspective Themes Quotes Organization Type Engagement of enrollees in HHP (e.g., outreach, communication, follow-up) "We have transitioned to a predominately phone engagement process. This has helped with some patients; however, overall, it has not helped with outreach. We have had both surprising months, but we've also had some months with very little interaction." Community health center "Contacts with patients shifted primarily to telephone engagements while staff worked remotely. We have experienced an increase in successful contacts with patients through telephone encounters. More patients at home. More time to answer telephone calls. More time to talk." Community health center "Due to COVID 19, members services had to be via telehealth and not in person. Members have felt disconnected from their providers and team. The inability to fully utilize the services has made life challenging for them. We have weekly calls via telephone, monthly calls with physician to ensure their basic needs are being met. Our team continues to educate members and provide resources available to them." Community based primary care medical group Infrastructure (e.g., policies and procedures, staffing, data sharing) and HHP implementation (e.g., shift to telework, telehealth) "increased hiring to support telehealth services for HHP." Substance use treatment center ".. When the COVID-19 lock down was initiated, had to temporarily suspend the face to face visits. With this valued process not an option, additional engagement with phones, emails, mailing out of letters and adoption of a HIPPA compliant texting platform for communication between members and case managers... impacted by HHP staff reduction of almost 30%... Data sharing continues as an important measure of internal and external communication, capability to filter resources and information with other organizations, and monitoring and evaluation..." Community health center Delivery of HHP core services (e.g., care management, care coordination) "Additional time to allocate and find resources for our member has had to be set. With COVID-19 many Substance use treatment center UCLA Evaluation | HHP and COVID-19 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Themes Quotes Organization Type resources we previously used shut down and or are at capacity." "Prior to COVID, face-to-face and accessibility of providers | Community were notable components of HHP. As a result of COVID- based primary 19, our team experienced challenges and limitations, but | care we continue to provide coordination of care using alternative modes. We use the internal electronic medical record (EMR) system called GuidingCare and data tracking systems, as well as telephonic coordination of services through our computers and phones. We have implemented trainings on telephonic engagement in order to establish rapport, build trust, use motivational interviewing (MI) techniques, and create a supportive mutual connection. We have established desktop procedures (DTPs) to outline the process from an initial call to engagement in appropriate level of case management. We also have regular online staff meetings to ensure we are all operating cohesively." Linkages and referrals (i.e., "We continue to link to all required services, however, Behavioral community and social turnaround time for referrals has increased. Internally we | health provider support services, housing are adding more staff to support but, connection to services) outside agencies has become challenging." "It has impacted in the way that there are less resources Community available during the pandemic. For example, less food health center pantries, less shelters are open. And this makes it more difficult to assist patients." Source: Corporation for Supportive Housing (CSH), Health Home Program & Homeless Training Survey, August-September 2020 (n=59). Note: Response to question: "On a scale of 0-10, please rate the impact of the COVID-19 pandemic on your organization's ability to perform the following HHP-related activities. What changes have you made to the above process as a result of COVID-19?" Transition to Telephonic Delivery of HHP Outreach, Engagement, and Services In the UCLA survey and as a result of COVID-19, many MCPs reported that they transitioned to delivering HHP services through telehealth and electronic modalities (n=13; data not shown). The majority of MCPs (n=10) noted that they would continue to use telehealth for physical and behavioral health services after the COVID-19 pandemic, and recognized many tangible benefits. Some MCPs (n=7) noted improvements in enrollee engagement as a result of the pandemic, as individuals were more likely to be available by phone due to shelter in place orders and staff had additional time to contact a larger number of enrollees. However, some MCPs (n=5) noted limitations in not being able to engage in face-to-face contact with enrollees and an inability to have meaningful and consistent encounters. Exhibit 19 provides illustrative quotes that highlight these findings. HHP and COVID-19| UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 19: Illustrative Quotes from MCPs on Transition to Predominantly Telephonic Contact during the COVID-19 Pandemic Themes Quotes Managed Care Plan Continued use of "UnitedHealthcare has offered greater flexibility United Healthcare telehealth regarding modes of engagement via telephonic and telemedicine, in addition to in-person encounters. As we move forward post COVID-19 pandemic, we plan on continuation of this greater flexibility based on our Member feedback to support their preferences. Our focus remains to support continuous engagement and Member choice." "We are considering the possibility of on-going use of telehealth given the positive impacts. We are continuing to compile feedback and will continue to look at support and services that are impacting health in a positive way." Blue Shield of California Improvements in enrollee engagement through telehealth "The direct impact from COVID-19 on HHP desired outcomes was primarily evident in the conversion from in-person to telephonic and virtual visits. Pandemic stay-at-home orders increased the likelihood of members being available by phone and engaging with care managers. So far there is no evidence that COVID prevented us from achieving any HHP desired outcome." Kaiser Permanente "CB-CMEs utilized telephonic out-reach to enroll, engage, and provide resources during the pandemic. We have seen a steady enrollment throughout the pandemic." Community Health Group Partnership Plan Limitations due to lack of face-to-face contact "The ability to engage with members who don't have access to phones and/or computers limits the ability to engage all eligible folk for the intervention/program. The technology gap has really been highlighted by COVID." San Francisco Health Plan "Some in-home visits were replaced with telephonic check-ins which seem to be adequate, yet not optimal. Telephonic check-ins are more effective when there has been a previous, well established relationship." Kaiser Permanente Source: UCLA HHP COVID-19 Impact Survey, September-November 2020, n=16. Note: Quotes are directly taken from open ended responses to survey questions. UCLA Evaluation | HHP and COVID-19 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Contribution of HHP to MCPs' COVID-19 Pandemic Response Three-fourths of MCPs reported that HHP contributed in positive and synergistic ways to their plan's overall COVID-19 response (data not shown). More specifically, HHP had established infrastructure and partnership networks for providing care management and coordination to Medi-Cal's most vulnerable populations. Many MCPs found that there was a significant overlap between HHP enrollees and those most at risk for COVID-19 complications. Exhibit 20 highlights illustrative quotes from MCPs regarding how HHP contributed to their response to COVID-19. Exhibit 20: Illustrative Quotes from MCPs on How HHP Facilitated MCP Response to COVID-19 Pandemic Quotes Managed Care Plan "Many members who were eligible for our COVID outreach program were also San Francisco Health enrolled or eligible for Health Homes. This mean we could combine efforts for Plan those members." "Because we had an established communications infrastructure within our CB- Inland Empire Health CME network, we were able to immediately pivot our communications and Plan training strategies to address concerns related to Covid-19." "HHP helped with overall response as our CB-CMEs enabled us to ensure outreach | Blue Shield of that included COVID-19 screening and support to enrollees who may need California additional support during this time." "Medi-Cal Health Home significantly helped provide support during the pandemic | United Healthcare by having an established infrastructure, clinically rigorous framework and network to provide care coordination services in our most vulnerable Medicaid Members. Through the Health Home Program, UnitedHealthcare has been able to identify and address homelessness by providing housing support and services to a targeted population greatly impacted by COVID-19." Source: UCLA HHP COVID-19 Impact Survey, September-November 2020, n=16. Note: Quotes are directly taken from open ended responses to survey questions. Estimated Prevalence of and Trends in COVID-19 among HHP Enrollees and their Controls The diagnosis code for COVID-19 was developed and utilized by providers starting in late March 2020. UCLA analyzed Medi-Cal claims starting in March 2020 and identified services used that had COVID-19 as the primary or secondary diagnosis in order to estimate the prevalence of the disease among HHP enrollees and the control group. A total of 2,088 HHP enrollees (4.3%) had at least one COVID-19 related service. The same proportion of the control group, 4.3%, had at least one COVID-19 related service (data not shown). UCLA examined monthly trends in the proportion of enrollees and their controls with at least one COVID-19 related service in that month. Data showed a peak in July 2020 (Exhibit 21), similar to the peak in COVID-19 hospitalizations seen in California and HHP counties during this ze HHP and COVID-19| UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program timeframe (Exhibit 15). The estimated incidence of COVID-19 per month was similar between HHP enrollees and their controls. Exhibit 21: Proportion of HHP Enrollees and their Controls with a COVID-19 Related Service by month, April 2020 to September 2020 2.0% 1.8% 1.5% 1.5% 1.1% 1.1% 1.0% 0.9% 0.8% 0.7% 0.4% 0.4% Apr-20 May-20 Jun-20 Jul-20 Aug-20 Sep-20 @ HHP Enrollees Control Group Source: UCLA analyses of Medi-Cal enrollment and claims data from April 2020 to September 2020. Notes: COVID-19 diagnosis was identified using ICD code U07.1 in primary or secondary diagnosis per claim. March 2020 was not included because of limited reporting using U07.1 that month. COVID-19 Related Service Use of HHP Enrollees and their Controls UCLA examined the type of COVID-19-related health services used by HHP enrollees and their controls with at least one COVID-19-realted service in 2020. Both enrollees and their controls most commonly used primary care services (53% for HHP enrollees vs 47% for controls), followed by emergency department visits (28% vs 29%) and hospitalizations (28% vs 28%). Less common services included lab tests (21% vs 22%), specialty services (19% vs 19%), and long- term care stays (7% vs 4%; Exhibit 22). UCLA Evaluation | HHP and COVID-19 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 22: Proportion of HHP Enrollees and their Controls with a COVID-19 Diagnosis that Received Specific COVID-19-Related Services 47% Primary Care Services ns 537 29% Emergency Department SS 22%, 28% Hospitalizations nn 22% 22% 'eb Tests nn 21% 19% Specialty Services TT 19% Long Term Care Stays i 4% 7% 0 Control Group = M HHP Enrollees Source: UCLA analyses of Medi-Cal enrollment and claims data from March 2020 to September 2020. Notes: Services with COVID-19 as primary or secondary diagnosis (identified using ICD code U07.1) only. Emergency department visits only include visits that did not result in hospitalization. Changes in Use of Health Services Before and During the COVID-19 Pandemic UCLA assessed service utilization patterns among all HHP enrollees and their controls before and during the pandemic and found a decline around April 2020 compared to April 2019 for all service types except for primary care services (Exhibit 23). Data showed that the rate of primary care services per month for HHP enrollees had not declined in March or when the shelter in place order was issued in late March 2020 in California but this rate declined in May and increased by June 2020 above the 2019 levels. The rate of primary care services per month for the control group was generally lower in 2020. Specialty care services, emergency department visits, and hospitalizations declined around April 2020, corresponding to a statewide shelter in place order. By September 2020, however, rates of specialty service utilization were similar to those observed in September 2019 for both enrollees and controls. In contrast, the number of ED visits and hospitalizations declined in April 2020 (relative to April 2019) and stayed lower by September 2020 (relative to September 2019) for both enrollees and the control group. HHP and COVID-19| UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 23: Comparing Monthly Service Utilization Rates in the Year Before the COVID-19 Pandemic (2019) versus the Year During (2020) for HHP Enrollees and the Control Group COVID-19 Pandemic 1100 in 2020 1000 900 800 700 600 700 Primary Care Services 600 Specialty Care Serives 500 400 250 200 a em ee Emergenc 5 Y 150 Department Visits 100 50 100 as = TM. = = Hospitalizations 50 a . -- 25 Jan Feb Mar Apr May Jun Jul Aug Sep e====e HHP Enrollees 2019 eee HHP Enrollees 2020 «== Control Group 2019 =< = Control Group 2020 Source: UCLA analyses of Medi-Cal enrollment and claims data from January 2019 to September 2020. Notes: Emergency department visits only include visits that did not result in hospitalization. UCLA Evaluation | HHP and COVID-19 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Further analyses (data not shown) found that less than 0.2% of primary care and specialty services were delivered via telehealth before the pandemic. In response to the pandemic, California's Department of Managed Health Care required that MCPs must reimburse telehealth visits at the same rate as in-person visits starting March 18, 2020. Starting in March 2020, rates of telehealth primary care and specialty care services increased substantially for both enrollees and the control group, peaking in April 2020 (Exhibit 24). Exhibit 24: Proportion of Primary Care and Specialty Care Services Provided through Telehealth by HHP Enrollees and Control groups, March 2020 to September 2020 19% 20% 18% 17% oes 16% /% 16% cog 149,490.4% 39, p Il r | | HHP Enrollees Control Group Primary Care Services 24% 21% 18% 17% 1595%5% ,, of AG 14% 1392:4944% 30, ; . | . p . | HHP Enrollees Control Group Specialty Care Services March mApril BMay June MJuly MAugust m September Source: UCLA analyses of Medi-Cal enrollment and claims data from March 2020 to September 2020. HHP and COVID-19| UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program HHP Enrollment and Enrollment Patterns This section addresses the following HHP evaluation questions: 1. What proportion of eligible enrollees were enrolled? 2. What proportion of enrollees were homeless? 3. How did enrollment patterns change over time? From July 1, 2018 to July 31, 2019, MCPs reported data on individual-level enrollment in ad hoc Enrollment Reports requested by DHCS. Beginning in the third quarter of 2019, DHCS requested for MCPs to report on member level enrollment data in their Quarterly HHP Reports. Both reports included monthly enrollment status by individual, along with individual level SPA data. Homeless status was reported by MCPs at the member level in Quarterly HHP Reports beginning in Quarter 3 of 2019. Therefore, enrollment growth and patterns among homeless enrollees was not available for enrollees who had disenrolled prior to this time. UCLA used these data from July 1, 2018, to September 30, 2020, to examine how enrollment changed over time for the overall HHP population, by SPA, and for homeless enrollees. Data was available for counties for all implementation groups (Groups 1, 2, 3, and 4) at the time of this report. Further details can be found in Appendix A: Data Sources and Analytic Methods. A small number of HHP enrollees (1,439) were enrolled for less than 31 days and were excluded from these analyses. MCPs received PMPM payments for one month which allowed for MCPs and CB-CMEs to work together to verify HHP eligibility, however MCPs did not receive payments if those individuals could no longer be enrolled in the program. MCPs did not provide other services to this group. Comparison of these enrollees with those enrolled for longer than 30 days indicated the groups had similar demographics, health status, and health care utilization prior to HHP. Further detail about this group can be found in Appendix C: HHP Enrollees Enrolled Less Than 31 Days. DHCS defined inclusion and exclusion eligibility criteria for HHP enrollees and used these criteria to identify eligible Medi-Cal beneficiaries to be included in the TEL, which was then distributed to MCPs in six-month intervals. However, DHCS did not have access to all eligibility criteria in Medi-Cal enrollment and claims data. Specifically, DHCS lacked information on three exclusion criteria including "sufficiently well managed through self-management or another program", " environment is unsafe for CB-CME staff". In addition to lack of data, the TEL was based on retrospective claims data used to define acuity criteria of "at least one inpatient hospital stay in the last year" and "three or more emergency department (ED) visits in the last year". Nearly all more appropriate for alternative care management programs", and "behavior or UCLA Evaluation | HHP Enrollment and Enrollment Patterns March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program the exclusion criteria were also retrospective and may have changed prior to enrollment by the MCPs. For example, individuals in a skilled nursing facility, enrolled in specialized MCPs, or enrolled in fee-for-service Medi-Cal may have been discharged back to the community, disenrolled from a specialized MCP, or enrolled in managed care outside of the TEL defined timeline, respectively. In addition, DHCS issued the TEL every six months based on adjudicated Medi-Cal claims data, while MCPs had and used more recent data on diagnoses and service utilization. MCPs were likely to have access to electronic medical records that contain more comprehensive diagnoses and information on health problems and needs of patients. Furthermore, MCPs had the option to enroll members that were referred by providers that may not have matched the HHP eligibility criteria in Medi-Cal data. Ultimately, MCPs prioritized some TEL enrollees based on severity, complexity, or risk-status using information not available to DHCS ze HHP Enrollment and Enrollment Patterns| UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Trends in Enrollment Growth in HHP Enrollment Overall and by SPA A total of 48,925 enrollees had ever enrolled in HHP by the end of September 2020 (Exhibit 25). Enrollment in HHP began with Group 1, SPA 1 in San Francisco in July 2018 and expanded rapidly when Groups 2 and 3 began enrollment. The growth in enrollment continued steadily after enrollment and when Group 4 started. Monthly new enrollment into the program varied between a high of 3,625 in July 2019 and a low of 26 in November 2018, averaging at 1,839 new enrollees per month (data not shown). Total monthly enrollment (new enrollment plus existing enrollment) increased each month except for July 2020. Exhibit 25: Unduplicated Monthly and Cumulative Enrollment in HHP, July 1, 2018 to September 30, 2020 44,089 Enrollees Group 4 SPA 2 Starts Mam New Enrollees per Month Mmmm Existing Enrollees per Month 28,383 Enrollees Group 3 SPA 2 and Group 4 SPA 1 Start 48,925 e@=-= Cumulative Enrollment 10,660 Enrollees Group 2 SPA 2 and Group 3 SPA 1 Start 2,313 Enrollees Group 1 SPA 2 and Group 2 SPA 1 Start 68 Enrollees Group 1 SPA 1 Starts | oo 0 oOo oO 0 Nana Dn HD HD VN nn DD Hn n CO Oo Oo oO Oo oO Oo oO oO Sissy Ap appt pay apt ay sty st pat at typ typ ty yy a a a ey Ss woe F oO fC oD SF FEO BWM WH FO fC aD Fe FS SF CZ wM a = 2S So se sSsPsetgsrzgsozs ses lgsegs72Zz8 2018Q3 2018Q4 2019Q1 2019Q2 201993 2019Q4 2020Q1 2020Q2 2020Q3 Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2019 to September 2020. HHP enrollment was limited to available data for the period between July 2018 and September 2020. Notes: MCP is managed care plan. Groups of MCPs implemented at different time points. Those enrolled for less than 31 days were excluded from this analysis. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. UCLA Evaluation | HHP Enrollment and Enrollment Patterns UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Examining HHP enrollment by SPA revealed a total cumulative enrollment of 38,228 in SPA 1 and 10,697 in SPA as of September 2020 (data not shown). In the third quarter of 2020, there were 26,659 SPA 1 enrollees and 8,962 SPA 2 enrollees (Exhibit 26). In the first two quarters of the program, MCPs only enrolled in SPA 1 but enrollment grew over time. Exhibit 26: Unduplicated Quarterly Enrollment in HHP by SPA, July 1, 2018 to September 30, 2020 mSPA1 mSPA2 1,447 942 139 2 " 2018 Q3, 2018 Q4 2019 Q1, 2019 Q2 2019 Q3 2019 Q4 2020 Q1 2020 Q2 2020 Q3 SPA 1 SPA 2 Enrollment Enrollment Begins Begins Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2019 to September 2020. HHP enrollment was limited to available data for the period between July 2018 and September 2020. Notes: MCP is managed care plan. Those enrolled for less than 31 days were excluded from this analysis. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. ze HHP Enrollment and Enrollment Patterns| UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Growth in HHP Enrollment among Homeless by SPA MCPs began reporting homeless data per enrollee in Quarter 3 of 2019 (Q3; July 1 to September 30) through HHP Quarterly Reports. UCLA used the identifier indicating enrollees who were ever homeless or at risk of homelessness during each quarter to show the patterns of enrollment over time. However, these data underestimate the size of homeless enrollees in HHP because they excluded homeless enrollees that disenrolled prior to July 2019 and did not reenroll in HHP. During the third quarter of 2020, 2,322 SPA 1 and 966 SPA 2 enrollees were homeless or at risk of homelessness (Exhibit 27). Enrollees experiencing homelessness or at risk of homelessness represented 10% of HHP enrollees overall by September 2020 (data not shown). The variation in number of homeless enrollees by Group can be seen in Appendix D: Homeless Enrollment by Group. Exhibit 27: Enrollment of Individuals Reported as Homeless or At-Risk of Homelessness each Quarter in HHP by SPA, July 1, 2019 to September 30, 2020 mSPA1 ; ; 2019 Q3 2019 Q4 2020 Q1 2020 Q2 2020 Q3 Source: Quarterly HHP Reports from July 2019 to September 2020. Homeless enrollees that disenrolled prior to July 2019 are not included. Notes: MCP is Managed Care Plan. Those enrolled for less than 31 days were excluded from this analysis. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Monthly enrollment of less than 11 was recorded as 11. Excludes HHP enrollees that were designated as homeless and were disenrolled prior to Q3. Includes homeless enrollees that were included in Q3 HHP Quarterly Reports. UCLA Evaluation | HHP Enrollment and Enrollment Patterns En UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Enrollment Size by Group and County Exhibit 28 shows enrollment by group and county as of September 2020. Enrollment varied by county. Los Angeles had the largest enrollment, reaching 18,919 in September 2020. Other counties with large enrollment include Riverside (7,885) and San Bernardino (6,541), from Group 2. Exhibit 28: Unduplicated Cumulative HHP Enrollment by Group and County as of September 30, 2020 18,919 7,885 6,541 3,676 2,976 3,839 5 a 8 a zZ e 8 s Fe < Z bs 2 a 5 iw Ww Zz uw <s < Ww x 2 S ce 9 =" S a oO 5 = o s < > s z g z < F i] S O oc oc nd n co <x 5 <x = o oc o O YW é z a = <x a <= z " " < Yn Group 1 Group 2 Group 3 Group 4 Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2019 to September 2020. HHP enrollment was limited to available data for the period between July 2018 and September 2020. Notes: MCP is Managed Care Plan. Those enrolled for less than 31 days were excluded from this analysis. Group 1 implemented HHP on July 1, 2018, Group 2 implemented HHP on January 1, 2019, Group 3 implemented HHP on July 1, 2019, and Group 4 implemented HHP on January 1, 2020 (SPA1) and June 1, 2020 (SPA2). HHP Enrollment and Enrollment Patterns| UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Enrollment from the Target Engagement List UCLA assessed the concordance between Medi-Cal enrollees identified by DHCS as eligible for HHP, based on their prior claims and communicated to MCPs biannually in the Targeted Engagement List (TEL), and Medi-Cal beneficiaries enrolled in HHP. The analyses showed that 78% of HHP enrollees were identified in the TEL as of September 2020 and this proportion varied by group (Exhibit 29). The proportion of enrollees identified in the TEL did not differ by SPA (data not shown). Exhibit 29: Proportion of HHP Enrollees that were identified in the Target Engagement List (TEL) as of September 2020, Overall and by Group Total Enrollment Proportion Identified in TEL Overall 48,375 78% Group 1 1,110 90% Group 2 14,426 82% Group 3 32,630 75% Group 4 759 90% Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2019 to September 2020. Target Engagement Lists from May 2018 to May 2020. Notes: Those enrolled for less than 31 days were excluded from this analysis. Group 1 implemented HHP on July 1, 2018, Group 2 implemented HHP on January 1, 2019, Group 3 implemented HHP on July 1, 2019, and Group 4 implemented HHP on January 1, 2020. Individuals identified on the TEL supplemental list were not included as part of TEL. Enrollment Patterns Enrollment Churn Most HHP enrollees (70%) remained continuously enrolled from enrollment date to September 2020, with a higher share for SPA 2 enrollees (82%) than SPA 1 enrollees (67%; Exhibit 30). Disenrollment rates increased since September 2019 for each of the two SPAs (data not shown). Overall, nearly one-third of enrollees (30%) have disenrolled once and stayed disenrolled from the program. Re-enrollment rates are low across both SPA 1 (0.2%) and SPA 2 (0.1%). Exhibit 30: Enrollment and Disenrollment Patterns in HHP as of September 30, 2020 Total Enrollment | Continuously Enrolled | Disenrolled Once | Enrolled Multiple Times Overall 48,925 70% 30% 0.2% SPA1 38,228 67% 33% 0.2% SPA 2 10,697 82% 18% 0.1% Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2019 to September 2020. HHP enrollment was limited to available data for the period between July 2018 and September 2020. Notes: MCP is Managed Care Plan. Those enrolled for less than 31 days were excluded from this analysis. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. UCLA Evaluation | HHP Enrollment and Enrollment Patterns March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Enrollment Length Average length of enrollment by Group and SPA was commensurate with implementation date. The length of enrollment was shorter for Groups 2 through 4 relative to Group 1, and was shorter for SPA 2 than for SPA 1 (Exhibit 31). Exhibit 31: Average Length of Enrollment in Months in HHP by Group as of September 30, 2020 Overa L oS SPA 1 § 10.7 Group 1 SPA 2 en § Sf Overall a 9.4 SPA Ls «10.2 Group 2 SPA 2 Ss 6.5 Ove ra - k= 7 SP 7 Group 3 SPA 2 A 4.2 Overall Es 4.3 SPA 1 Es 4.8 SPA2 ES 1.3 Group 4 Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2019 to September 2020. HHP enrollment was limited to available data for the period between July 2018 and September 2020. Notes: MCP is managed care plan. Those enrolled for less than 31 days were excluded from this analysis. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. MCP Exclusions of Specific HHP Eligible Populations MCPs were able to use standardized criteria to exclude some of the eligible beneficiaries identified in their respective TELs and were required to report the reason for such exclusions in their Quarterly HHP Reports in the aggregate and for the first year of implementation. Ten MCPs only reported for the first three quarters of implementation and one MCP did not report at all. Exhibit 32 displays the percent of eligible beneficiaries in the TEL that were excluded by reasons for such exclusions. For Groups 2 and 3 the most common reason was that an eligible beneficiary was not an MCP member. At the time the TEL was constructed, these individuals may have been members of the MCP, but were no longer members when the MCP began enrollment either due to enrollment in another MCP or disenrollment from Medi-Cal. Other most common reasons for exclusion were eligible enrollee declined to participate (Group 1) and eligible enrollee was already well managed (Group 4). HHP Enrollment and Enrollment Patterns| UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 32: Percent of Eligible Beneficiaries Excluded by MCPs by Reason for Exclusion in the First Year of HHP Implementation Group Exclusion Rationale 1 2 3 4 Excluded because well-managed 0.4% 0.5% 0.4% 7.2% Excluded because declined to participate 3.1% 1.9% 2.2% 2.2% Excluded because of unsuccessful engagement 0.9% 3.0% 2.5% 4.8% Excluded because duplicative program 0.5% 0.3% 1.0% 0.6% Excluded because unsafe behavior or environment n/a <0.0% <0.0% n/a Excluded because not enrolled in Medi-Cal at MCP 0.3% 7.4% 3.1% 1.8% Externally referred but excluded <0.0% 0.1% <0.0% n/a Source: MCP Quarterly HHP Reports from September 1, 2018 to September 30, 2012. Groups 1 and 2 reported excluded beneficiaries for the first year of implementation. Group 3 MCPs reported 3 or 4 quarters of excluded beneficiaries. Group 4 only reported 3 quarters of excluded beneficiaries. HealthNet counties (Kern, Los Angeles, Sacramento, San Diego and Tulare) were excluded from analysis due to insufficient reporting. Eligible beneficiaries were identified on the targeted engagement lists created prior to the last quarter of reporting for each MCP and County. Notes: MCP is Managed Care Plan and TEL is Targeted Engagement List. n/a indicates small cell size. UCLA Evaluation | HHP Enrollment and Enrollment Patterns UCLA Center for Health Policy R h ee HHP Enrollee Demographics and Health Status This section addresses the following HHP evaluation questions: 1. What were the demographics of program enrollees? What was the acuity level of the enrollees including health and health risk profile indicators, such as aggregate inpatient, ED, and rehab SNF utilization? 3. What proportion of enrollees are homeless? UCLA used demographic information from the Medi-Cal enrollment data, homeless status from MCP Quarterly HHP Reports, and Medi-Cal claims data to construct measures of health status and healthcare utilization prior to enrollment in HHP. Medi-Cal data included both managed care and fee-for-service encounters. UCLA used a look-back period of 24 months for these measures in line with the HHP Program Guide. The exception to this was calculation of enrollee demographics, which was based on an enrollee's HHP enrollment date. Measures of chronic conditions and acuity eligibility criteria were created based on definitions in the HHP Program Guide and the Centers for Medicare and Medicaid Service's Chronic Condition Warehouse condition categories, using primary and secondary diagnosis codes in each Medi-Cal claim. Further details can be found in Appendix A: Data Sources and Analytic Methods. UCLA reported demographics and health status for (1) all enrollees, (2) SPA 1 enrollees, and (3) SPA 2 enrollees. Of the 48,922 HHP enrollees (see HHP Enrollment and Enrollment Patterns), three enrollees were missing Medi-Cal data prior to HHP enrollment and were not included in these analyses. HHP enrollees enrolled for less than 31 days (1,436 enrollees) were excluded from these analyses. DHCS defined inclusion and exclusion eligibility criteria for HHP enrollees and used these criteria to identify eligible Medi-Cal beneficiaries to be included in the TEL, which was then distributed to MCPs in six-month intervals. However, DHCS did not have access to all eligibility criteria in Medi-Cal enrollment and claims data. Specifically, DHCS lacked information on the "chronic homelessness" acuity criteria. Conclusions | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Demographics of HHP Enrollees at Time of Enrollment As of September 2020, MCPs had enrolled 48,922 individuals for over 30 days, with 38,225 in SPA 1 and 10,697 in SPA 2. Overall, HHP enrollees were most often 50 to 64 years old, female and Latinx. When comparing SPA 1 and SPA 2 enrollees, the former group were more often older, less likely to be White, and less likely to speak English. Some (8%) of HHP enrollees were reported as experiencing homelessness at any point during HHP enrollment, and rates varied by SPA with 7% for SPA 1 and 9% for SPA 2 (Exhibit 33). The overall demographics of enrollees as of September 2020 did not differ greatly from the demographics of enrollees as of September 2019 (data not shown). Exhibit 33: HHP Enrollee Demographics, Overall, and by SPA, at the Time of HHP Enrollment as of September 30, 2020 Total SPA 1 Enrollees SPA 2 Enrollees Enrollment N 48,922 38,225 10,697 Age (at time of % 0-17 7% 8% 5% enrollment) % 18-34 13% 11% 22% % 35-49 22% 21% 26% % 50-64 50% 51% 44% % 65+ 8% 9% 4% Gender % male 41% 42% 35% Race/Ethnicity % White 21% 20% 26% % Latinx 46% 47% 41% % African American 18% 18% 17% % Alaskan Native or <1% <1% <1% American Indian % Asian 5% 5% 3% % Hawaiian, 1% 1% 1% Guamanian, Samoan, Other Asian or Pacific Islander % other 4% 4% 7% % unknown 5% 5% 5% Language % English proficient 72% 70% 78% Enrolled in Medi- Average number of 12 12 12 Cal full-scope months during the year prior to enrollment Homelessness Experienced 8% 7% 9% homelessness during enrollment Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2019 - September 2020. HHP enrollment was limited to available data for the period between July 1, 2018 and September 30, 2020, and homelessness is only reported for enrollees who were active as of July 2019. Demographics at the time of HHP enrollment were obtained from Medi- Cal enrollment data from July 1, 2016 to September 30, 2020. UCLA Evaluation | Conclusions March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Notes: MCP is Managed Care Plan. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Homeless data was not reported for 720 enrollees. Health Status of HHP Enrollees Prior to Enrollment UCLA examined the proportion of enrollees with the top ten most frequent physical health and mental health conditions in the 24 months prior to enrollment. Data showed high rates of hypertension (67%) and diabetes (49%) among HHP enrollees (Exhibit 34). When comparing SPA 1 and SPA 2, SPA 2 enrollees were more likely to have mental health conditions, including depression (72%), anxiety (50%), and bipolar disorder (27%) compared to SPA 1. Exhibit 34: Top Ten Most Frequent Physical and Mental Health Conditions among HHP Enrollees, 24 Months Prior to HHP Enrollment Total SPA 1 Enrollees SPA 2 Enrollees N=48,922 N=38,225 N=10,697 Hypertension (67%) Hypertension (71%) Depression (72%) Diabetes (49%) Diabetes (54%) Depressive Disorders (68%) Hyperlipidemia (42%) Hyperlipidemia (45%) Hypertension (52%) Obesity (40%) Chronic Kidney Disease (41%) Anxiety (50%) Depression (38%) Obesity (40%) Obesity (37%) Chronic Kidney Disease (37%) Asthma (31%) Hyperlipidemia (33%) Depressive Disorders (36%) Depression (29%) Diabetes (30%) Anxiety (30%) Depressive Disorders (27%) Fibromyalgia, Chronic Pain and Fatigue (30%) Asthma (28%) Fibromyalgia, Chronic Pain and Bipolar (27%) Fatigue (26%) Fibromyalgia, Chronic Pain and Rheumatoid Arthritis / Drug Use Disorders (25%) Fatigue (27%) Osteoarthritis (25%) Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2020. HHP enrollment was limited to available data for the period between July 1, 2018 and September 30, 2020. Chronic and other chronic health, mental health, and potentially disabling condition categories were identified using the Chronic Condition Warehouse methodology using Medi-Cal claims data from July 1, 2016 to September 30, 2020. Notes: MCP is managed care plan. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. In order to further examine the level of complexity of health status of HHP enrollees, UCLA examined the proportion of HHP enrollees that met each of the four HHP eligibility criteria outlined in the HHP Program Guide in the 24 months prior to enrollment. Exhibit 35 shows that 55% of HHP enrollees had hypertension along with chronic obstructive pulmonary disease, diabetes, coronary artery disease, chronic or congestive heart failure (Criteria 2). A greater proportion of enrollees had serious mental health conditions (Criteria 3) compared to a combination of very complex conditions such as chronic renal (kidney) disease, chronic liver disease, traumatic brain injury and a more common condition (Criteria 1). A smaller proportion of HHP enrollees (28%) had asthma (Criteria 4). Consistent with HHP program goals, more SPA 2 enrollees had major depression disorder, bipolar disorder, or psychotic disorders (Criteria 3) Conclusions | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 than SPA 1 enrollees (83% versus 32%). The composition of enrollees by eligibility criteria did not differ greatly as of September 2020 compared to September 2019 (data not shown). Exhibit 35: Complexity of HHP Enrollees' Health Status by SPA, 24 Months Prior to HHP Enrollment as of September 30, 2020 Total SPA 1 Enrollees | SPA 2 Enrollees Number of HHP Enrollees N=48,922 N=38,225 N=10,697 Two specific conditions (Criteria 1) 40% 44% 23% Hypertension and another specific condition (Criteria 2) 55% 61% 32% Serious mental health conditions (Criteria 3) 43% 32% 83% Asthma (Criteria 4) 28% 31% 16% Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2019 - 2020. HHP enrollment was limited to available data for the period between July 1, 2018 and September 30, 2020. Utilization data was calculated using Medi-Cal claims data from July 1, 2016 to September 30, 2020. Chronic condition categories were based on definitions from the HHP Program Guide. Notes: Criteria 1 includes any two of the following conditions: chronic obstructive pulmonary disease, diabetes, traumatic brain injury, chronic or congestive heart failure, coronary artery disease, chronic liver disease, chronic renal (kidney) disease, dementia, substance use disorders. Criteria 2 includes hypertension and one of the following: chronic obstructive pulmonary disease, diabetes, coronary artery disease, chronic or congestive heart failure. Criteria 3 includes one of the following: major depression disorders, bipolar disorder, psychotic disorders including schizophrenia. Criteria 4 includes asthma. HHP enrollees may meet multiple criteria. UCLA Evaluation | Conclusions UCLA Center for Health Policy R h ee HHP Service Utilization among HHP Enrollees This section addresses the following HHP evaluation questions: 1. Were HHP services provided in-person or telephonically? 2. Were HHP services provided by clinical or non-clinical staff? 3. How many homeless enrollees received housing services? MCPs were required to report HHP services to DHCS in Medi-Cal claims data starting on July 1, 2018. Two different procedure codes with unique modifiers that further indicated type and modality of services as well as type of providers were used. DHCS required HCPCS code G0506 from July 1, 2018 to September 30, 2018, but discontinued it because it led to denial of claims where a provider had submitted more than one unit of service per date of service. Therefore, DHCS adopted HCPCS code G9008 starting on October 1, 2018. Both codes were used to report HHP services in this report. Prior to Q3 2019, MCPs reported on the number of HHP enrollees experiencing or at risk for homelessness and the provision of housing services to these beneficiaries in the aggregate and per quarter. This data could not be used to assess trends since it lacked information on each individual member and changes in their status. MCPs began reporting this data at the member level starting in Q3 2019, representing July 1 through September 30, 2019, and reported homeless status during each quarter, receipt of housing services during each quarter, and whether a person was no longer homeless by the end of each quarter. Therefore, this report describes the homeless status and receipt of housing services for homeless and at-risk-of- homelessness beneficiaries for each quarter from Q3 2019 to Q3 2020. UCLA used all available data to examine the type and frequency of HHP services received by enrollees at the SPA level. Further details can be found in Appendix A: Data Sources and Analytic Methods. HHP Service Utilization among HHP Enrollees | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program MCPs were required to report HHP services under HCPCS code G9008, defined as "coordinated care fee, physician coordinated care oversight services." MCPs were required to use HCPCS code modifiers (U1 - U7) to identify three unique service types, service provider, and service modality (Exhibit 36). MCPs were expected to use at least one modifier per claim to define an HHP service. For example, a single visit where an enrollee receives HHP core services in-person by both clinical and non-clinical staff would use two modifiers (U1 and U4). Multiple units of service (UOS) were allowed, where one UOS was equivalent to 15 minutes of time to provide the service. Clinical staff included licensed medical professionals such as physicians, nurse practitioners, LCSWs, and medical assistants, while non-clinical staff included employees working in administrative or technical roles. In-person visits could occur at a variety of locations (e.g., home, office, or clinic). Telehealth allowed for remote patient monitoring (e.g., vitals and blood pressure), allowing enrollee care, reminders, and education to occur through telephone HHP Services and electronic communications. UCLA Evaluation | HHP Service Utilization among HHP Enrollees UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 36: HHP Services Provider Type hosdiner Modality Definition Engagement Services Provider Type Not | U7 Not specified Active outreach such as direct communications with Specified member (e.g., face-to-face, mail, electronic, and telephone), follow-up if the member presents to another partner in the HHP network or using claims data to contact providers the member is known to use. Providers must show active, meaningful, and progressive attempts at member engagement each month until the member is engaged. Examples of acceptable engagement include: (1) letter to member followed by phone call to member; (2) phone call to member, outreach to care delivery partners and social service partners; (3) and street level outreach, including, but not limited to, where the member lives or is accessible. Core Services Provided by U1 In-person Comprehensive care management, care coordination, Clinical Staff health promotion, comprehensive transitional care, U2 Telehealth individual and family support services, and referral to community and social supports Provided by Non- | U4 In-person Clinical Staff U5 Telehealth Other Services Provided by U3 Not specified Case notes, case conferences, tenant supportive services, Clinical Staff and driving to appointments Provided by Non- | U6 Not specified Clinical Staff Source: Adapted from Health Homes Program Guide. Notes: HCPCS is Healthcare Common Procedure Coding System, MCP is Managed Care Plan, and UOS is Unit of Service. Service use was reported by MCPs in claims data. Each service (UOS) represented a 15-minute interaction between HHP staff and HHP enrollee. Multiple UOS' were allowed within a single visit. Modifiers U1-U7 accompanied both HCPCS code GO506 (July 1, 2018 to September 30, 2018) and HCPCS code G9008 (October 1, 2018 to September 30, 2020) to specify the service. Telehealth includes phone and other forms of remote communication. re HHP Service Utilization among HHP Enrollees | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program UCLA's examination of claims data revealed that HHP-specific HCPCS codes were not yet reported for 24% of HHP enrollees and that enrollees without these codes came from all 16 MCPs (data not shown). DHCS reported identifying deficiencies in reporting of HHP services both in claims and in MCP reports. MCPs reported to DHCS that CB-CMEs had challenges in reporting of HHP services that were included in claims. DHCS provided technical support to MCPs to address these problems. MCPs also reported to DHCS that they were providing technical assistance to CB-CMEs to improve reporting for all data. An examination of the extent of this under-reporting showed that 24% of HHP enrollees lacked any HHP-specific HCPCS modifier codes and 38% of HHP enrollees lacked HCPCS codes for some months during their enrollment (data not shown). Further analysis showed that the rate of under-reporting varied by type of service with a higher rate for engagement services and a lower rate for core services. Therefore, UCLA calculated the average number of HHP services during months when HHP-specific HCPCS codes were present for each enrollee rather than calculating HHP services across all months of enrollment. The latter methodology would have been based on the incorrect assumption that HHP enrollees did not receive HHP services when HCPCS modifier codes were missing. Due to the limitations of data on HHP services and the methodology employed by UCLA, the data presented in this chapter are considered estimates of HHP services received by enrollees. Estimated Overall HHP Service Delivery to HHP Enrollees Exhibit 37 shows estimated service utilization for any HHP service (HCPCS modifiers U1-U7), regardless of provider type and modality between July 1, 2018 and September 30, 2020. Available data showed that a total of 412,463 UOS (in 15-minute increments) were received during this time period, averaging to 2.1 UOS per enrollee per month in months where services were received. Comparison of services received by HHP enrollees by SPA showed a higher number of total UOS delivered to SPA 1 enrollees corresponding to more enrollees in this SPA. However, SPA 2 enrollees had a slightly higher average number of UOS than SPA 1 enrollees (2.2 UOS versus 2.1 UOS per month per enrollee in months that HHP services were received). The median UOS per enrollee was similar between SPAs. UCLA Evaluation | HHP Service Utilization among HHP Enrollees UCLA Center for Health Policy Research Health Economics and Evaluation Research March 2022 Program Exhibit 37: Estimated Overall HHP Units of Service Received by HHP Enrollees by SPA, July 1, 2018 to September 30, 2020 All HHP Enrollees SPA 1 Enrollees SPA 2 Enrollees (n=48,922) (n=38,225) (n=10,697) Total number of units of service received 412,463 348,959 63,504 Average number of units of service per enrollee per month in months where HHP services were received 2.1 2.1 2.2 Median number of units of service per enrollee per month in months where HHP services were received 1.0 1.0 1.0 Source: Medi-Cal Claims data from June 1, 2018 to September 30, 2020. Notes: HCPCS is Healthcare Common Procedure Coding System, MCP is Managed Care Plan. Service use was under-reported by MCPs in claims data. Each unit of service (UOS) represented a 15-minute interaction between HHP staff and HHP enrollee. Multiple UOS' were allowed within a single visit. Modifiers U1-U7 accompanied both HCPCS code G0506 (July 1, 2018 to September 30, 2018) and HCPCS code G9008 (October 1, 2018 to September 30, 2020) to specify the service. Data are based on the number of months during HHP enrollment where HCPCS codes were present. HHP Service Utilization among HHP Enrollees | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program IV Elaca ley Estimated Types of HHP Services Received Exhibit 38 shows estimated average number of UOS per enrollee per month in months where HHP services were received by type of service from July 1, 2018 to September 30, 2020. The average number of UOS received was higher for core HHP services (1.7) than engagement services (1.3) or other HHP services (1.6). Also, the average number of UOS for engagement and other HHP services was higher for SPA 2 than SPA 1 enrollees. Exhibit 38: Estimated Average Number of HHP Units of Service Provided to HHP Enrollees in Months HHP Services were Received by Service Type and SPA, July 1, 2018 to September 30, 2020 Service Type All HHP Enrollees SPA 1 Enrollees SPA 2 Enrollees (n=48,922) (n=38,225) (n=10,697) Engagement Services (U7) 1.3 13 1.4 Core HHP Services (U1, U2, U4, or U5) 1.7 1.7 1.7 Other Health Homes Services (U3 or U6) 1.6 1.6 1.7 Source: Medi-Cal Claims data from July 1, 2018 to September 30, 2020. Notes: Data show estimated average number of units of services (UOS) per enrollee during months that specific service was received. HCPCS is Healthcare Common Procedure Coding System, MCP is Managed Care Plan. Service use is under-reported by MCPs in claims data. Each UOS represented a 15-minute interaction between HHP staff and HHP enrollee. Multiple UOS' were allowed within a single visit. Core HHP services include claims with HCPCS code G0506 (July 1, 2018 to September 30, 2018), HCPCS code G9008 (October 1, 2018 to June 30, 2019), and modifier U1, U2, U4, or U5. HHP engagement service includes claims with HCPCS code G0506 (July 1, 2018 to September 30, 2018), HCPCS code G9008 (October 1, 2018 to June 30, 2019), and modifier U7. Other HHP service includes claims with HCPCS code G0506 (July 1, 2018 to September 30, 2018), HCPCS code G9008 (October 1, 2018 to September 30, 2020), and modifier U3 or U6. Data are based on the number of months during HHP enrollment where HCPCS codes were present. UCLA Evaluation | HHP Service Utilization among HHP Enrollees March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Estimated HHP Core Services by Modality and Staff Type MCPs were required to report the modality of HHP core services including in-person or through telehealth. However, DHCS did not require reporting modality for other HHP services or engagement services. Exhibit 39 shows the average number of telehealth UOS received per enrollee during months that telehealth services were received (1.6 UOS) was higher than the average number of in-person services received per enrollee (1.3 UOS). MCPs were required to report the types of staff that provided core and other HHP services. The average number of services received from non-clinical staff (1.8 UOS) were higher than clinical staff (1.6 UOS) for SPA 2. Exhibit 39: Estimated Average Number of HHP Core Unites of Service Provided to HHP Enrollees in Months those HHP Services were received by Modality and SPA, July 1, 2018 to September 30, 2020 All HHP Enrollees SPA 1 Enrollees SPA 2 Enrollees (n=48,922) (n=38,225) (n=10,697) Modality In-Person UOS (U1 or U4) 1.3 1.3 1.3 Telehealth UOS (U2 or U5) 1.6 1.6 1.7 Staff Types Who Delivered the Service Clinical Staff UOS (U1, U2, or U3) 1.6 1.6 1.7 Non-Clinical Staff UOS (U4, U5, or U6) 1.8 1.8 1.9 Source: Medi-Cal Claims data from July 1, 2018 to September 30, 2020. Notes: Data show estimated average number of units of services per enrollee during months that service was received. HCPCS is Healthcare Common Procedure Coding System, MCP is Managed Care Plan, and UOS is Unit of Service. Service use was under-reported by MCPs in claims data. Each service (UOS) represented a 15-minute interaction between HHP staff and HHP enrollee. Multiple UOS' were allowed within a single visit. Modifiers U1-U7 accompanied both HCPCS code G0506 (July 1, 2018 to September 30, 2018) and HCPCS code G9008 (October 1, 2018 to September 30, 2020) to specify the service. Data are based on the number of months during HHP enrollment where HCPCS codes were present. HHP Service Utilization among HHP Enrollees | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Housing navigation and transition services included activities such as conducting tenant HHP Housing Services screenings, developing an individualized housing plan, assisting with move-in, and assisting with the housing search and application process. MCPs began reporting enrollee level data on homeless status and delivery of housing services in Q3 2019 (July 1 through September 30, 2019). In this period and onward, MCPs reported on enrollees who were homeless or at risk for homelessness during each quarter, those who were no longer homeless by the end of the quarter, and those who received housing services during the quarter. They also reported on whether an enrollee had ever been homeless during HHP, although this measure was not examined due to data inconsistencies. MCPs communicated challenges in reporting for provision of housing services. DHCS provided technical support to MCPs to address these problems, and MCPs reported to DHCS that they were providing technical assistance to CB- CMEs to improve reporting for all data. The table below is considered an estimation of homeless status and receipt of housing services due to inconsistent reporting across these variables. Inconsistencies were present when an enrollee was reported as no longer homeless while that enrollee was never reported as homeless or at risk; an enrollee was reported as receiving housing services although they were never reported as homeless or at risk; and an enrollee was not reported as homeless or at risk during the same quarter when they first reported as being homeless at some point during the program. One reason for such discrepancies may have been that CB-CMEs had 90 days to assess an enrollee's homeless status and may not have done so when the quarterly report had to be submitted 60 days after the end the quarter. Using data from the MCP Quarterly Reports, UCLA estimated that the percentage of enrollees who were homeless or at risk for homelessness in a given quarter grew during HHP, from 4% of the population in Q3 2019 to 9% of the population in Q3 2020 (Exhibit 40). The percentage of homeless or at risk enrollees who received housing services also increased over time, starting at 38% in Q3 2019 and increasing to 68% in Q3 2020. This percentage did not include an additional 118 enrollees who were not identified as homeless or at risk but who received housing services. Of those who were homeless or at risk during a given quarter, 3% were no longer homeless by the end of Q3 2019, and this number peaked in Q2 2020 at 10%. This percentage does not include 330 enrollees who reported as no longer homeless, but were not identified as homeless or at risk. UCLA Evaluation | HHP Service Utilization among HHP Enrollees NETSB Opy) UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 40: Homeless Status and Receipt of Housing Services by HHP Enrollees, July 1, 2019 to September 30, 2020 Percentage of Enrollees Percentage of Enrollees Percentage of Enrollees _ _ - Experiencing Homeless or Experiencing Homeless or Experiencing Homelessness . . . . . were at Risk who Received were at Risk who were No or were at Risk During . . . Quarter Housing Services During Longer Homeless by End of Quarter Quarter Q3 2019 4% 38% 3% Q4 2019 6% 44% -- Q1 2020 7% 47% 4% Q2 2020 8% 54% 10% Q3 2020 9% 68% 7% Source: MCP Quarterly Reports from July 1, 2019 to September 30, 2020. Notes: "--" indicates samples of less than 11 enrollees. Housing services data is shown only for enrollees who were reported as homeless or at risk for homelessness. za HHP Service Utilization among HHP Enrollees | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Acute Care Utilization Groups in HHP This section examines characteristics and health utilization of enrollees given their level of acute care service use. The data specifically inform the following HHP evaluation questions: 1. What was the acuity level of the enrollees including health and health risk profile indicators, such as aggregate inpatient, ED, and rehab SNF utilization? 2. How did patterns of health care service use among HHP enrollees change before and after HHP implementation? UCLA examined the number of ED visits and hospitalizations of HHP enrollees prior to enrollment and identified five categories of enrollees including those with super utilization (10 or more ED visits or 4 or more hospitalizations a year), high utilization (5 or more ED visits or 2 or more hospitalizations), moderate utilization (2 or more ED visits or 1 or more hospitalization), low utilization (less than 2 ED visits or less than 1 hospitalization); and those at risk for high utilization (no ED visits or hospitalizations, but were eligible for HHP mostly due to multiple chronic conditions). UCLA examined the demographics, health status, and service utilization of these five groups. UCLA used demographic information from the Medi-Cal enrollment data, enrollment information and homeless status from MCP ad hoc Enrollment Reports and Quarterly HHP Reports, and Medi-Cal claims data to construct measures of health status prior to enrollment in HHP, healthcare utilization prior to enrollment in HHP, and metric trends before and during HHP. Medi-Cal data included both managed care and fee-for-service encounters. UCLA used a look-back period of 24 months for the measures of health status in line with the HHP Program Guide. The calculation of enrollee demographics was based on an enrollee's HHP enrollment date. Measures of chronic conditions and acuity eligibility criteria were created based on definitions in the HHP Program Guide and the Centers for Medicare and Medicaid Service's Chronic Condition Warehouse condition categories, using primary and secondary diagnosis codes in each Medi-Cal claim. Utilization measures were constructed following the HHP Technical Specifications. Further details can be found in Appendix A: Data Sources and Analytic Methods. HHP enrollees enrolled for less than 31 days (1,436 enrollees) were excluded from these analyses. Of the 48,922 HHP enrollees (see HHP Enrollment and Enrollment Patterns), three enrollees were missing Medi-Cal data prior to HHP enrollment and were not included in these analyses. UCLA Evaluation | Conclusions UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Acute Care Utilization of HHP Enrollees Exhibit 41 shows that the majority of HHP enrollees had moderate utilization (35%) or low utilization (32%). A small proportion of enrollees had super utilization (6%). The proportion of enrollees in each acute care utilization group in SPA 1 and SPA 2 was similar. Exhibit 41: Proportion of HHP Enrollees in Acute Care Utilization Groups at HHP Enrollment, Overall and by SPA @ Super Utilization @ High Utilization & Moderate Utilization @ Low Utilization At Risk for High Utilization All Enrollees SPA 1 SPA 2 Source: UCLA analysis of MCP Enrollment Reports from August 2019 and Quarterly HHP Reports through September 2020. HHP enrollment was limited to available data for the period between July 1, 2018 and September 30, 2020. Utilization data was calculated using Medi-Cal claims data from July 1, 2016 to September 30, 2020. Notes: At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Conclusions | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Further analysis showed that the average annual number of ED visits and hospitalizations for enrollees with super utilization were 14.9 and 4.1 vs. 2.7 and 0.5 for moderate utilization, respectively (Exhibit 42). Enrollees who were at risk for high utilization but had no ED visits or hospitalizations in the 24 months prior to enrollment are not shown. Exhibit 42: Average Number of ED Visits and Hospitalizations by Acute Care Utilization Group, 24 months prior to Enrollment 14.9 °7 4.1 2.7 . | as | Average Annual ED Visits Average Annual Hospitalizations Super Utilization High Utilization Moderate Utilization Low Utilization Source: MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2020. HHP enrollment was limited to available data for the period between July 1, 2018 and September 30, 2020. Utilization data was calculated using Medi-Cal claims data from July 1, 2016 to September 30, 2020. Notes: Low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Acute Care Utilization of HHP Implementation Groups Exhibit 43 shows the acute care utilization groups by implementation group. Group 1, which consisted of San Francisco County and the earliest HHP enrollees, included a higher proportion of enrollees with super utilization (12%) than Group 2 and Group 3. Group 4, which consisted of Orange County and the latest HHP enrollees, included the highest share (18%) of those with super utilization and high utilization (28%) and the lowest share of enrollees at risk for high utilization (3%). UCLA Evaluation | Conclusions cE UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 43: HHP Acute Care Utilization Groups by HHP Implementation Groups m Super Utilization @ High Utilization m@ Moderate Utilization @ Low Utilization W At Risk for High Utilization Group 1 (n=1,109) Group 2 (n=14,426) Group 3 (n=32,625) Group 4 (n=759) Source: UCLA analysis of MCP Enrollment Reports from August 2019 and Quarterly HHP Reports through September 2020. HHP enrollment was limited to available data for the period between July 1, 2018 and September 30, 2020. Notes: At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Group 1 began SPA 1 enrollment in July 2018, Group 2 in January 2019, Group 3 in July 2019, and Group 4 in January 2020. SPA 2 enrollment began six months after the start of enrollment for each group. The average length of enrollment varied by acute care utilization group and implementation group. Group 2, the largest implementation group through September 30, 2020, had the shortest average enrollment for enrollees with super utilization (257 days) and the longest average enrollment for enrollees at risk for high utilization (212 days; data not shown). Group 4, the smallest implementation group, had mostly an inverse pattern; enrollees with low utilization had the shortest enrollment (113 days) and enrollees with super utilization had the longest enrollment (154 days). Group 3 had consistent enrollment across all acute care groups except the at-risk group, which was longer than the other groups, and Group 1 had variable enrollment across all acute care groups, with enrollees with high utilization enrolled for the shortest period of time and enrollees with low utilization enrolled for the longest period of time. Further analysis of acute care utilization of new enrollees showed few differences in patterns over time through September 2020 (data not shown). En Conclusions | UCLA Evaluation UCLA Center for Health Policy Research Tey Health Economics and Evaluation Research Program Demographics of HHP Enrollees by Acute Care Utilization Groups Exhibit 44 shows demographics of HHP enrollees by acute care utilization groups. Enrollees with super utilization were most often younger than 65 (96%), male (49%), white (26%), and were experiencing homelessness (14.6%). Those at risk for high utilization were more often 50 years of age or older (78%), Asian (11%), and had a primary language other than English (40%). The super utilization group had the largest proportion of homeless enrollees (14.6%) and the at-risk group had the smallest proportion of homeless enrollees (5.6%). Exhibit 44: Demographics of HHP Acute Care Utilization Groups at the Time of HHP Enrollment At risk for Super High Moderate Low High Utilization | Utilization | Utilization | Utilization | Utilization Enrollees N 2,967 6,875 17,303 15,634 6,140 % 0-17 2% 5% 11% 7% 1% (atti f % 18-34 18% 18% 17% 10% 5% e (at time o AB % 35-49 30% 26% 23% 19% 16% enrollment) % 50-64 47% 45% 44% 54% 62% % 65+ 4% 5% 6% 10% 16% Gender % male 49% 41% 39% 40% 43% % White 26% 24% 21% 20% 20% % Latinx 37% 42% 47% 48% 44% % African American 23% 20% 19% 16% 14% % Alaskan Native or _ . . 0% 0% 0% 0% 0% Race/Ethnicity American Indian % Asian 2% 2% 3% 5% 11% % Native Hawaiian . 1% 1% 1% 1% 2% and Pacific Islander % other/unknown 10% 9% 9% 10% 10% Primary . % speak English 86% 81% 75% 67% 60% Language Enrolled in Medi- Cal full-scope ; Average number of during the year 11.49 11.78 11.88 11.94 11.78 . months prior to enrollment Experienced Homelessness homelessness during 14.6% 10.9% 7.9% 6.2% 5.6% enrollment Source: UCLA analysis of Medi-Cal enrollment data from July 1, 2016 to September 30, 2020 and Quarterly HHP Reports. Demographics were reported at the time of enrollment into HHP. Notes: Homeless data was not reported for 720 enrollees. At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, UCLA Evaluation | Conclusions UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Health Status of HHP Enrollees by Acute Care Utilization Groups Exhibit 45 shows the proportion of enrollees in each acute care utilization group that met a given HHP chronic condition eligibility criteria. Data showed highest prevalence of enrollees who met criteria 2 (hypertension and another specific condition) and lowest prevalence of criteria 4 among all acute care utilization groups. There were variations in the criteria as well. For example, the majority of enrollees with super utilization met criteria 1 (two specific chronic conditions; 65%) but 49% of enrollees with high utilization and 35% of enrollees with moderate utilization met that criteria. Exhibit 45: HHP Acute Care Utilization Groups by Chronic Condition Eligibility Criteria, 24 Months Prior to Enrollment 65%65% 65% 57% 47% Super Utilization High Utilization Moderate Utilization Low Utilization At Risk for High Utilization Two specific conditions (Criteria 1) @ Hypertension and another specific condition (Criteria 2) @ Serious mental health conditions (Criteria 3) m@ Asthma (Criteria 4) Source: UCLA analysis of Medi-Cal claims data from July 1, 2016 to September 30, 2019. Notes: At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Criteria 1 includes any two of the following conditions: chronic obstructive pulmonary disease, diabetes, traumatic brain injury, chronic or congestive heart failure, coronary artery disease, chronic liver disease, chronic renal (kidney) disease, dementia, substance use disorders. Criteria 2 includes hypertension and one of the following: chronic obstructive pulmonary disease, diabetes, coronary artery disease, chronic or congestive heart failure. Criteria 3 includes one of the following: major depression disorders, bipolar disorder, psychotic disorders including schizophrenia. Criteria 4 includes asthma. HHP enrollees may meet multiple criteria. Conclusions | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 46 shows the most frequent physical and behavioral health conditions of acute care utilization groups. Hypertension was the most common condition for all enrollee groups. However, chronic kidney disease was the second most common condition among enrollees with super utilization (55%) followed by mental health conditions like depression (53%) and anxiety (53%). Diabetes was the second most common condition in all other acute care utilization groups. Exhibit 46: Top Ten Most Frequent Physical and Behavioral Health Conditions among HHP Enrollees by Acute Care Utilization Group Prior to HHP Enrollment Super Utilization High Utilization Moderate Utilization Low Utilization At risk for high utilization (N=2,967) (N=6,875) (N=17,303) (N=15,634) (N=6,140) Hypertension (78%) Hypertension (67%) Hypertension (59%) Hypertension (69%) Hypertension (78%) Chronic Kidne' . y Diabetes (45%) Diabetes (42%) Diabetes (52%) Diabetes (63%) Disease (55%) . Chronic Kidney . Hyperlipidemia . . Depression (53% Obesity (39% Hyperlipidemia (55% P (53%) Disease (45%) ¥ (39%) (46%) yperp (55%) Anxiety (53%) Depression (43%) Depression (38%) Obesity (41%) Obesity (39%) D ive Disord Chronic Kid Di Diabetes (51%) epressive misoraer Hyperlipidemia (37%) Depression (36%) ronic eney misease (40%) (38%) Depressive Disorder Depressive Disorder Chronic Kidne pressive ™ Obesity (39%) pressive ™ . remeney Depression (34%) (50%) (35%) Disease (35%) Anemia (48%) Anxiety (39%) Chronic Kidney Disease | Depressive Disorder Depressive Disorder (32%) Y (34%) (33%) P Hyperlipidemia . Rheumatoid Arthritis / Drug (48% Asthma (32% Anxiety (25% us ( (38%) ( riety ( Osteoarthritis (24%) Fibromyalgia, Fibromyalgia, Rheumatoid Chronic Pain and Chronic Pain and Anxiety (30%) Arthritis / Anxiety (21%) Fatigue (47%) Fatigue (34%) Osteoarthritis (25%) Tobacco (42%) Anemia (33%) Fibromyalgia, Chronic Pain and Fatigue (26%) Fibromyalgia, Chronic Pain and Fatigue (24%) Fibromyalgia, Chronic Pain and Fatigue (20%) Source: UCLA analysis of Medi-Cal claims data from July 1, 2016 to September 30, 2019. Physical and Behavioral Health condition categories were identified using the Chronic Condition Warehouse methodology. Notes At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Health Service Utilization Trends of Acute Care Utilization Groups UCLA examined the unadjusted trends in a number of different types of health services used per 1,000 member months for acute care utilization groups using Medi-Cal claims data. These measures were constructed per the HHP Technical Specifications when possible. UCLA examined trends in these measures for each enrollee in six month increments up to 24 months (1-6, 7-12, 13-18, and 19-24) before HHP enrollment and up to 12 months (1-6 and 7-12) during HHP. UCLA Evaluation | Conclusions UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Trends in Primary Care Services Exhibit 47 shows that enrollees with super utilization had the lowest number of primary care services per 1,000 member months 19 to 24 months before enrollment (886);this number notably increased during months 1 to 6 of enrollment (1,346) with a decline 7 to 12 months during enrollment (1,157). Rates of primary care services were higher in the first year of HHP compared to before HHP. The same pattern was observed for other utilization groups though the numbers were lower. The number of services were relatively similar for enrollees with moderate or low utilization, and those at risk for high utilization. Exhibit 47: Primary Care Services per 1,000 Member Months Before and During HHP Enrollment by Acute Care Utilization Group 1-6 es =-1,120 1-6 es 1346 1-6 Mm 519 1-6 Me 770 1-6 Mm 482 1-6 Ms = 791 7-12 Mm 697 N N ao 1 iow 1 1 19-24 es S886 13-18 Mi «1,015 7-12 Me =-1,036 7-12 ms §=(722 7-12 "es 529 7-12 "em 676 7-12 Mm 485 19 - 24 13-18 7-12 es 1157 619 687 739 783 995 892 19-24 mu 491 13-18 M554 7-12 ME (580 1-6 588 1-6 Ds «=-808 19-24 mus 464 13-18 "eee 510 19-24 mum «(459 13-18 Mm 486 Before (mos.) During Before (mos.) During Before (mos.) During Before (mos.) During Before (mos.) During (mos.) (mos.) (mos.) (mos.) (mos.) Super Utilization High Utilization Moderate Utilization Low Utilization At Risk for High Utilization Source: UCLA analysis of Medi-Cal Claims data from July 1, 2018 to September 30, 2020. Notes: Service rates are unadjusted. Primary care services were identified as services with a primary care physician, physician assistant, or nurse practitioner per NUCC's Taxonomy code set. At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Before (mos.) is the number of months before HHP. During (mos.) is the number of months after HHP enrollment. Conclusions | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Trends in Specialty Services Exhibit 48 shows that enrollees with super utilization had the lowest number of specialty services per 1,000 member months 19 to 24 months before enrollment (559) and this number continued to increase through 1 to 6 months during enrollment (928) with a decline 7 to 12 months during enrollment (895). Rates of specialty services were higher in the first year of HHP compared to before HHP. The same pattern was observed for other utilization groups though the numbers were lower. The number of services were relatively similar for enrollees with moderate or low utilization, and those at risk for high utilization. Exhibit 48: Specialty Services per 1,000 Member Months Before and During HHP Enrollment by Acute Care Utilization Group 3 wo nN A + a oOo oO ' ' ' ac a 500 7-12 «742 NO ON cd 1 1 a ' ' 19-24 = 559 13-18 «667 7-12 «895 572 646 719 785 752 19-24 mu 404 13-18 Es (463 7-12 = 516 1-6 «601 7-12 = «(568 19-24 "ee «412 13-18 "em 456 7-12 "en 512 1-6 Me «516 1-6 Mn «556 7-12 Se ~540 19-24 ME «388 13-18 ME «(408 7-12 Mees 453 1-6 Ms «(472 1-6 ME «526 7-12 ME §=505 19 - 24 13-18 Before (mos.) During Before (mos.) During Before (mos.) During Before (mos.) During Before (mos.) During (mos.) (mos.) (mos.) (mos.) (mos.) Super Utilization High Utilization Moderate Utilization Low Utilization At Risk for High Utilization Source: UCLA analysis of Medi-Cal Claims data from July 1, 2018 to September 30, 2020. Notes: Service rates are unadjusted. Specialty care services were identified as services with a specialty physician, physician assistant, or nurse practitioner per NUCC's Taxonomy code set. At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Before (mos.) is the number of months before HHP. During (mos.) is the number of months after HHP enrollment. UCLA Evaluation | Conclusions UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Trends in Emergency Department Visits Exhibit 49 shows a decline in number of ED visits that did not result in a hospitalization per 1,000 member months among enrollees with super utilization from 19 to 24 months before enrollment (921) to 1 to 6 months during enrollment (635). Enrollees with high, moderate, and low utilization also had a decline in ED visits followed by discharge. Exhibit 49: Emergency Department Visits per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group qr nt+aad an a st oO nw wn * o yo DA Oo © V0 Oo mn UD mM MM zg o FT GB w& NoaanR a aA ata ts a 4am wo er ~ SCoCGEA G&G B® a | od cA oO Oo N Heaeeaetdia == ST ON WO WON TAN WON FON WO WON TAON DO O N TF OC oon Noo, ss AND De AND DI A ND DI lh hl AD ND I ll mmm mmm mm met an mn ~ nm nnn nam KR nan m nN dd acd did | did Before (mos.) During Before (mos.) During Before (mos.) During Before (mos.) During Before (mos.) During (mos.) (mos.) (mos.) (mos.) (mos.) Super Utilization High Utilization | Moderate Utilization Low Utilization At Risk for High Utilization Source: UCLA analysis of Medi-Cal Claims data from July 1, 2018 to September 30, 2020. Notes: Only includes ED visits that did not result in hospitalization. Service rates are unadjusted. At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Before (mos.) is the number of months before HHP. During (mos.) is the number of months after HHP enrollment. Ea Conclusions | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Elec 40 yar Trends in Hospitalizations Exhibit 50 shows that hospitalizations per 1,000 member months declined among enrollees with super utilization from 19 to 24 months before enrollment (284) to 1 to 6 months during enrollment (227). The same pattern was observed among those with high and moderate utilization. Exhibit 50: Hospitalizations per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group oO wa qn +7 nnn om t+ % R N N ~m N N mM oO N N Sua & toa 4 in a 00 nOoR n QQ om ot TT Tm yo A KR Y A On aa Nn os N a & eee eeece so ep ope EF = os tT ON DO ON TON WO WON TFT WON DO ON TFTWON HO ON TAWON O ON Nod 5s. pp AN eae es Ss th A NO WD lhl th hl RA NO DI lt RA ND wD ll mmm mm mmm mm met nm Mh» nn ™"M nm Mm na MM dS n~ nM BS ~ | ad did | dod Before (mos.) During Before (mos.) During Before (mos.) During Before (mos.) During Before (mos.) During (mos.) (mos.) (mos.) (mos.) (mos.) Super Utilization High Utilization | Moderate Utilization Low Utilization At Risk for High Utilization Source: UCLA analysis of Medi-Cal Claims data from July 1, 2018 to September 30, 2020. Notes: Service rates are unadjusted. At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Before (mos.) is the number of months before HHP. During (mos.) is the number of months after HHP enrollment. UCLA Evaluation | Conclusions UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Trends in Admissions to a Long-Term Care Facility from the Community Exhibit 51 shows the number of admissions to a skilled nursing or intermediate care facility from the community that resulted in short-term stays of up to 20 days per 1,000 member months. Trends for this metric were examined on an annual rather than semi-annual basis due to an enrollment requirement requiring one year of observation. The number of short-term stays declined from 3.64 in Pre-Year 2 to 3.16 in HHP Year 1 among enrollees with super utilization and the same pattern was observed among those with high and moderate utilization. Exhibit 51: Admissions to a Long-Term Care Facility Resulting in a Short-Term Stay per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group 3.64 3.74 3.16 2.32 2.11 1.58 1.36 1.04 0.83 0.40 0.54 0.53 a i i 0.06 0.10 0.18 = = ml N a ol N a a N a a N So a N a a o o o o o o c 3 o o o o o o o ) ov ) wo ov oO co) vo ) oa vo ) wv eo) ) ~~ > oe = > * > > * 7 > ~ > > v v v 2 & L 2 a 2 2 2 v £ Super Utilization High Utilization Moderate Utilization Low Utilization At Risk for High Utilization Source: UCLA analysis of Medi-Cal Claims data from July 1, 2018 to September 30, 2020. Notes: Short-term stay is defined as up to 20 days. Service rates are unadjusted. At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Ea Conclusions | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 52 shows trends in the number of admissions to a long-term care facility from the community that resulted in a medium-term stay (21 to 100 days) per 1,000 member months. Medium-term stays increased from 3.61 in Pre-Year 2 to 4.94 in Pre-Year 1 for enrollees with super utilization. By HHP Year 1, the rate had declined to 4.43 for this group. Similar increases prior to enrollment and declines during enrollment were observed for enrollees with high and moderate utilization. Exhibit 52: Admissions to a Long-term Care Facility Resulting in a Medium-Term Stay per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group 4.94 4.43 2.98 1.85 Ww Pre-Year2 = 2 Pre-Year1 HHP Year 1 ST, N oO Co o Pre-Year2 eS an Pre-Year1 ° KB HHP Year1 Sa 9 ° Ww a oO oy oxy ° ao SL oO Bb N ° oO NSN oO & wo oe oe ee | N a - N a - N a a - - - - - - - - - oO oO oO oO oO oO oO oO oO Vo ) Vo Vo co) wo ) ov cy) 7 7 = > * = a > = L L L L L L £ a a x a a x a a = Super Utilization High Utilization Moderate Utilization Low Utilization At Risk for High Utilization Source: UCLA analysis of Medi-Cal Claims data from July 1, 2018 to September 30, 2020. Notes: Medium-term stay is defined as 21 to 100 days. Service rates are unadjusted. At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. UCLA Evaluation | Conclusions ze UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Exhibit 53 shows trends in the number of admissions to a long-term care facility from the community that resulted in a long-term stay (101 days or more) per 1,000 member months. Long-term stays increased from 2.27 in Pre-Year 2 to 3.77 in HHP Year 1 for enrollees with super utilization, and increased for all other acute care utilization groups except enrollees with high utilization. Exhibit 53: Admissions to a Long-term Care Facility Resulting in a Long-Term Stay per 1,000 Member Months from Before to During HHP Enrollment by Acute Care Utilization Group 3.77 3.47 N Pre-Year2 I . NSN Pre-Yea 1 DD! HHP Year 1 B oO Oo R Ow wo R Ww N oO N N oO Ww Ww ° HHP Year1 on oO oO aay oO be oO oO N & oO oO N oO oO ~N oO N £& =o & _ -_- & - _ N a a N Se N a a N a i . . . . . . . . . . . oO Oo oO oO oO oO oO oO oO oO Oo 7) oO oO oO vo oO oO oO oO wo oO > > > > > > > > > on > L L L L L v £ v L a a x a a a a x a a x Super Utilization High Utilization Moderate Utilization Low Utilization At Risk for High Utilization Source: UCLA analysis of Medi-Cal Claims data from July 1, 2018 to September 30, 2020. Notes: Long-term stay is defined as 101 or more days. Service rates are unadjusted. At risk for high utilization is defined as no ED utilization or hospitalizations 24 months prior to enrollment, low utilization is less than 2 ED visits and less than 1 hospitalizations per year, moderate utilization is 2 or more ED visits or 1 or more hospitalizations per year, high utilization is 5 or more ED visits or 2 or more hospitalizations per year, and super utilization is 10 or more ED visits or 4 or more hospitalizations per year. Conclusions | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program HHP Outcomes This section addresses the following HHP evaluation questions: 1. How did patterns of health care service use among HHP enrollees change before and during HHP implementation? 2. Did rates of acute care services, length of stay for hospitalizations, nursing home admissions and length of stay decline? 3. Did rates of other services such as substance use treatment or outpatient visits increase? 4. How did HHP core health quality measures improve before and after HHP implementation? 5. Did patient outcomes (e.g., controlled blood pressure, screening for clinical depression) improve before and after HHP implementation? UCLA used Medi-Cal claims data, which included both managed care and fee-for-service encounters, to construct HHP metrics per the HHP Technical Specifications. UCLA measured trends before and during HHP for each metric based on the date of an individual HHP enrollee's enrollment. UCLA did not examine trends in the second year of HHP enrollment because as of September 2020, only 20% of enrollees had enrollment longer than 12 months and 74% of those enrollees had less than six months of enrollment in the second year (further details can be found in Appendix G: Enrollees with More than One Year of HHP Enrollment). UCLA restricted the sample to enrollees with a minimum 1 month of HHP enrollment and calculated all metrics per member month by SPA and overall. UCLA examined trends for all HHP metrics for SPA 1 and SPA 2 per HHP metric specifications and further created and examined the trend for seven optional measures to further describe changes in utilization of services during HHP. UCLA examined changes in trends before and during HHP using a difference-in-difference (DD) analysis. The DD analyses differed for HHP specified metrics that required one year of observation from metrics that did not require one year of observation and for optional measures. For HHP specified metrics with a one year requirement, the DD analyses measured changes from Pre-HHP Year 2 to Pre-HHP Year 1 for both HHP enrollees and the control group; the change from Pre-HHP Year 1 to HHP Year 1 for both HHP enrollees and the control group; and the difference between the changes for HHP enrollees vs. the control group. For the remaining metrics and measures, UCLA examined changes in six month increments up to 24 months (1-6, 7-12, 13-18, and 19-24) before HHP enrollment and up to 12 months (1-6 and 7-12) during HHP. For these, the DD analysis measured the change from 19-24 vs. 1-6 months before HHP for both HHP enrollees and the control group; the change during HHP from UCLA Evaluation | HHP Outcomes [ilehl March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program 1-6 to 7-12 months for both HHP enrollees and the control group; and the difference between the changes in HHP enrollees vs. the control group. The shorter timeframe for examining metrics allowed for a clearer assessment of changes during the early phase of HHP implementation. The findings were not subject to potential seasonality in service utilization due to rolling enrollment throughout the year and measuring change following the date of enrollment per beneficiary. Further details can be found in Appendix A: Data Sources and Analytic Methods. HHP Utilization Metrics Trends in three HHP specified metrics and six optional measures were examined on a semi- annual basis. Trends in one HHP specified metric were examined on an annual basis. HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Outpatient Utilization Primary Care Services UCLA calculated the number of primary care services per 1,000 member months as an optional measure of service utilization under HHP. There is no intended direction for this measure. Primary care services are likely to increase due to unmet need and increased access, but this use is likely to decrease once health needs are addressed. Exhibit 54 shows an increase in the number of primary care services before HHP by 33 services per 1,000 member months every 6 months for SPA 1 enrollees. The rate further increased following enrollment and during the first six months of HHP (879 primary care services per 1,000 member months). This rate declined by 126 services per 1,000 member months in the next 6 months but still remained higher than the control group. The decline from before to during HHP was significantly greater for HHP enrollees than the control group by 101 (DD). A similar trend was observed for SPA 2 enrollees. Exhibit 54: Trends in Primary Care Services per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 gs BR ln oD woo wn 00 WO stm ~ Nun LN 00 ~ aon an NAN D mao BX no © > S N © ow m am Lo iF A " in m| QQ mn 1H Win 8 : in + 0 N oO oO N + 0 N Ww Wo N N a a] 1 1 a N a a ' ' a ! ! ! - a ! ' ! t a - ' oO mM Ss oa ~ S » a | a | Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 SPA 2 m HHP Enrollees Control Group Difference Between (eT <-03 Dacascos Difference (DD) fer ay 43 feat 343) Before HHP During HHP SPA1 HHP Enrollees | 33* -126* -159* Control Group | 32* -26* -58* -101* SPA 2 HHP Enrollees | 23* -80* -102* Control Group | 22* 3 -20* -83* Overall | HHP Enrollees | 31* -116* -147* Control Group | 30* -20* -50* -97* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes ps<0.05, a statistically significant difference. Primary care services were identified as services with a primary care physician, physician assistant, or nurse practitioner per NUCC's Taxonomy code set. SPA 1 includes enrollees with chronic UCLA Evaluation | HHP Outcomes [lee March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 -- 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 - 12 months of HHP minus 1 -6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). Specialty Care Services UCLA calculated the number of specialty care services per 1,000 member months as an optional measure of service utilization under HHP. There is no intended direction for this measure. Specialty care services are likely to increase due to unmet need and increased access, but this use is likely to decrease once health needs are addressed. Exhibit 55 shows an increase in the number of specialty care services before HHP by 74 services per 1,000 member months every 6 months for SPA 1 enrollees. The rate further increased following enrollment and during the first six months of HHP (714 specialty care services per 1,000 member months). This rate declined by 40 services per 1,000 member months in the next 6 months but still remained higher than the control group. The decline from before to during HHP was significantly greater for HHP enrollees than the control group by 60 (DD). For SPA 2 enrollees, there was no change in specialty service use after its initial increase during HHP but the number of specialty services declined significantly from before HHP and in comparison to the control group (49, DD). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Exhibit 55: Trends in Specialty Services per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 ro ta Bo 8 a, 8, & aN go Bg SB FS 5S "Sloe fe BR SE FF Fa in Oo wn WO T L L | L L : : + 00 N oO wo N s+ 0 N WO Ww N N a A 1 1 a N a a 1 1 a ! ! ' ae | ae | ' ' ! ' a ae ' oa ~m Ss oO) ~ a a a a Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 SPA 2 m HHP Enrollees Control Group ery 43 feat 342) sales balaie Difference-in- Before HHP During HHP Difference (DD) (erat <-05 SPA 1 HHP Enrollees | 74* -40* -114* Control Group | 63* 9* -54* -60* SPA 2 HHP Enrollees | 54* 7 -47* Control Group | 47* 49* 2 -49* Overall | HHP Enrollees | 69* -30* -100* Control Group | 60* 17* -42* -57* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. Specialty care services were identified as services with a specialty physician, physician assistant, or nurse practitioner per NUCC's Taxonomy code set. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1-6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 - 12 months of HHP minus 1 -6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). Mental Health Services UCLA calculated the number of mental health services per 1,000 member months as an optional measure of service utilization under HHP. There is no intended direction for this measure. Mental health services are likely to increase due to unmet need and increased access, but this use is likely to decrease once health needs are addressed. Exhibit 56 shows that mental health services further increased following enrollment for SPA 1 enrollees and remained above the control group, but there was no significant change in the number of mental health services during HHP. Rates were declining by 420 services per 1,000 member months compared to before HHP and the decline was significantly greater for HHP enrollees than the control group by 236 (DD). For SPA 2 enrollees, data show a significant increase before HHP, a significant UCLA Evaluation | HHP Outcomes [les March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program decline during HHP (578 services per 1,000 members), and a significantly greater decline from before to during HHP compared to the control group (957 services, DD). Exhibit 56: Trends in Mental Health Services per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 oO a oO N on 00 ma Oa AD Cc N aN x ~ ~O + aD a eo " Ro TS OR xR WH Ln oO + So Uy Moo OLN min S on wn on wo aun + 0 1A mo an << wo on aN aon OY v4 = N nS - wm LS N co < < wv s N aN at a N N N ; : l L i + 0 N Oo oO t+ 0 N oO oO N N Ae a 1 1 a N A a 1 ' a ! ' ! a a ! ' ' 1 a a ! a mM ~n a ~ nm So a a a Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA1 SPA 2 m HHP Enrollees Control Group Change Difference Difference-in- fey 1 ay) ; Before Peat Between Parse) lala Changes (DD) SPA 1 HHP Enrollees 404* -16 -420* Control Group _| 308* 124* -184* -236* SPA 2 HHP Enrollees 776* -578* -1,354* Control Group =| 545* 148 -398* -957* Overall | HHP Enrollees 482* -133* -615* Control Group 358* 129* -229* -386* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. Mental health services were identified as services with a mental health procedure code. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 - 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 - 12 months of HHP minus 1 - 6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). Substance Use Disorder Services UCLA calculated the number of substance use disorder (SUD) services per 1,000 member months as an optional measure of service utilization under HHP. There is no intended direction for this measure. SUD services are likely to increase due to unmet need and increased access, but this use is likely to decrease once health needs are addressed. Exhibit 57 shows a small but significant increase (2 services per 1,000 member months) every 6 months before HHP for SPA 1 enrollees. During HHP this rate declined significantly by 24 services and the change from HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 before to during HHP was a significantly greater decline for HHP enrollees than the control group by 19 services (DD). A similar pattern was observed for SPA 2 enrollees, though the magnitude of change before and during HHP was greater. Exhibit 57: Trends in Substance Use Disorder Services per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 200 Ml 235 19 - 24 N oO N wt LON t+ No Ne WN li li 00 N Oat Se ' ! ~ BS ad Before HHP (months) SPA 1 1-6 206 195 189 Ml 250 Ml 226 N ' = ! During HHP (months) m@ HHP Enrollees Change Before HHP During HHP M408 288 19 - 24 Oo Ln rT) ~ © v t + 00 wo om Ln a 00 ~" | m "N 00 N wo a oan 1 ' ! a on a Before HHP (months) SPA 2 Control Group Change Difference Between «458 297 7-12 During HHP (months) Difference-in- Dace a-es) Ot (DD) SPA 1 HHP Enrollees | 2* -24* -26* Control Group | 2* -6* -8* -19* SPA 2 HHP Enrollees | 22* -56* -78* Control Group | 16* 0 -16* -62* Overall | HHP Enrollees | 7* -31* -37* Control Group | 5* -4* -9* -28* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes ps0.05, a statistically significant difference. SUD services were identified as services with a SUD treatment procedure code or an NDC for pharmacotherapy. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1- 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 -- 12 months of HHP minus 1- 6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | HHP Outcomes 107 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Emergency Department Utilization Ambulatory Care: Emergency Department Visits Ambulatory Care: Emergency Department Visits is an HHP core metric that measures the rate of emergency department (ED) visits that do not result in hospitalization per 1,000 member months. The intended direction of the metric and DD is decrease. Exhibit 58 shows an increase in the number of ED visits before HHP by 2 visits per 1,000 member months every 6 months for SPA 1 enrollees. This rate declined during HHP by 17 visits and the decline from before to during HHP was significantly greater than the control group by 9 visits (DD). A similar trend was observed for SPA 2 enrollees with a greater decline compared to the control group (15 visits, DD). Exhibit 58: Trends in Ambulatory Care: Emergency Department Visits per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 wo a 0 st w a t+ N BS ow Doo oo ns 0 oo oO oO as -R aay KR © Kh 0 a N a 149 141 160 150 Ml 155 MN 138 2 MN 203 Mn «190 Ms «178 Mm 154 + 0 N oO wo N + 0 oO wo N N - a 1 1 4 N a = 1 1 al 1 I 1 Sa at 1 1 1 1 So | S| 1 a ~m S Oo) m a = = = Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA1 SPA 2 m HHP Enrollees Control Group (OFT ati) OF Tati) yh arsastara= , , ; yh ar-te-arx lie Before During Between Re Dy) lala eat <-43 SPA1 HHP Enrollees | 2* -17* -20* Control Group | 2* -8* -10* -9* SPA 2 HHP Enrollees | 4* -25* -29* Control Group | 4* -10* -14* -15* Overall | HHP Enrollees | 3* -19* -22* Control Group | 3* -8* -11* -11* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: Includes ED visits that do not result in hospitalization. * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 - 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research Tee Health Economics and Evaluation Research Program HHP is calculated as: (7-12 months of HHP minus 1 - 6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). Any Emergency Department Visit UCLA created a second measure to assess the likelihood of any ED visit, which is distinct from the HHP core metric of number of ED visits. The intended direction of the measure and DD is decrease. Exhibit 59 shows a significantly greater decline in the proportion of enrollees with any ED visit during HHP for SPA 1 (2.4%) and SPA 2 (3.2%). For SPA 1 enrollees, the decline in this proportion compared to before HHP was greater than that of the control group by 1.1% (DD). Exhibit 59: Trends in Percentage of Patients with Any ED Visits Before and During HHP by SPA as of September 30, 2020 So .o oS .<o NP eX NN x © x0 xe xX XX se 30 nA 2M MH SH Re wel SH 2S V2 HoH Re xy ay F9 BS mt AH SHl an YF SL wm Sm TH ey ¢ tt aoa LEX | t+ TT TT FQ ON | | | | tl | | | | | | + 0 N oO oO N t+ © N oO oO N N a 4 1 1 a N a a ' ' a ! ! ! Se | c ! ' L 1 ci S| ! oa m oO) m » a a S| | Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 SPA 2 m HHP Enrollees Control Group Oy T ay =i2) ibaa Phare) Pyare le Before HHP er Between Changes Difference (DD) SPA1 | HHP Enrollees | 0.0% -2.4%* -2.3%* Control Group | 0.0% -1.3%* -1.2%* -1.1%* SPA2 | HHP Enrollees | 0.3%* -3.2%* -3.5%* Control Group | 0.3%* -1.9%* -2.2%* -1.3% Overall | HHP Enrollees | 0.0% -2.5%* -31.5%* Control Group | 0.0% -1.4%* -30.8%* -1.1%* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: Includes ED visits that do not result in hospitalization. * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 - 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 --12 months of HHP minus 1 -6 months of HHP). Difference between changes is calculated as: (Change UCLA Evaluation | HHP Outcomes March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program During HHP -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). Hospital Utilization Inpatient Utilization Inpatient Utilization is an HHP core metric that measures the rate of acute inpatient care and services per 1,000 member months. The intended direction of the metric and DD is decrease. Exhibit 60 shows an increase in the number of hospitalizations before HHP by 6 stays per 1,000 member months every 6 months for SPA 1 enrollees. During HHP, this rate declined by 10 stays and the decline from before to during HHP was significantly greater for HHP enrollees than the control group by 7 (DD). A similar trend was observed for SPA 2 enrollees. Exhibit 60: Trends in Inpatient Utilization per 1,000 Member Months Before and During HHP by SPA as of September 30, 2020 aN a ol on a 6 y BO Fe Bn Be 2 i an of i | | 'i : ii + 0c + 0c oO wo N N - Ss 1 1 N S| S| 1 1 a 1 1 t ae ae 1 ! I 1 ae | I a mM NS ns a ~m S| - a S| Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 SPA 2 m HHP Enrollees Control Group PyhaKasalexe) fer Tay 43 feat) Between Difference-in- Before HHP During HHP Difference (DD) (eT <-03 SPA 1 HHP Enrollees | 6* -10* -15* Control Group | 6* -3* -9* -7* SPA 2 HHP Enrollees | 4* -12* -16* Control Group | 5* -2 -7* -10* Overall | HHP Enrollees | 5* -10* -15* Control Group | 6* -2* -8* -7* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes ps<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 - 6 months before HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research Neyer) Health Economics and Evaluation Research Program HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 - 12 months of HHP minus 1 -6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in- difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). Any Hospitalization UCLA created a second measure to assess the likelihood of any hospitalization, which is distinct from the HHP core metric of number of hospitalizations. The intended direction of the measure and DD is decrease. Exhibit 61 shows a significantly greater decline in the proportion of enrollees with any hospitalization during HHP for SPA 1 (3.1%) and SPA 2 (4.2%). The decline in this proportion compared to before HHP was greater than that of the control group by 1.9% (DD) for both SPA 1 and SPA 2 enrollees. Exhibit 61: Trends in Percentage of Patients with Any Hospitalization Before and During HHP by SPA as of September 30, 2020 © 4 © x xk Yo NRE x am . nN': NN ao 3 NJ nm ok ° -c - 0 ° wee N xg Oo eX ao on 7c RG Sy N on oN RO aN "I cg Ns xe SRK * 00 a Ko at > 7 xe oS Na 58 © > : ~S 4 oO X xo + 4 mo Be | Ce a 4 AG Os a 4 a NG aa a ao =" 4oO + 0 N oO oO N + 0 N oO oO N N a S| 1 I! a -N a a 1 ' S| ! 1 1 c a 1 ' 1 1 a - 1 a mM oO m Ss a a a S| Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 SPA 2 m HHP Enrollees Control Group yh aceasta) Between (era <-05 Difference-in- Difference (DD) Change Change Before HHP During HHP SPA 1 HHP Enrollees | 1.8%* -3.1%* -4.9%* Control Group | 1.9%* -1.1%* -3.0%* -1.9%* SPA 2 HHP Enrollees | 1.5%* -4.2%* -5.7%* Control Group | 1.6%* -2.2%* -3.8%* -1.9%* Overall | HHP Enrollees | 1.8%* -3.3%* -18.1%* Control Group | 1.8%* -1.3%* -17.1%* -1.9%* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 - 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 - 12 months of HHP minus 1-6 UCLA Evaluation | HHP Outcomes (Ral UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 months of HHP). Difference between changes is calculated as: (Change During HHP --Change Before HHP). Difference-in- difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Inpatient Length of Stay is an HHP core metric that measures the average length of stay per hospitalization. The intended direction of the metric and DD is decrease. Exhibit 62 shows that lengths of stay were increasing before HHP for both SPA 1 and SPA 2, but these rates did not change during HHP, and the trends were similar with the control group. Inpatient Length of Stay Exhibit 62: Trends in Inpatient Length of Stay Before and During HHP by SPA as of September 30, 2020 00 oO ~_ + 4 oO NW w 00 M un WI ao a Wa) oO t+ On al a ° o ° a Ln On nm un an mW : Ln : ie) - . : I " L + 0 N Oo oO N + 0 N oO oO N N ol a 1 1 So N a i ' ' a 1 1 t a P| 1 ' I 1 P| a 1 oa ~m ns ao ~m So a a S| Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 SPA 2 m HHP Enrollees Control Group erly x- fear Tay <3 sabe yh acle-lale= le Before HHP During HHP Difference (DD) e423 SPA1 | HHP Enrollees | 0.13* 0.11 -0.03 Control Group | 0.13* 0.04 -0.09 0.06 SPA2 | HHP Enrollees | 0.11* 0.15 0.04 Control Group | 0.11* -0.01 -0.12 0.16 Overall | HHP Enrollees | 0.13* 0.11 -0.01 Control Group | 0.13* 0.03 -0.09 0.08 Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 - 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 - 12 months of HHP minus 1 -6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in- difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | HHP Outcomes [#Er! March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Institution Utilization Admission to an Institution from the Community Admission to an Institution from the Community is an HHP core metric that measures the number of admissions to an institutional facility among individuals age 18 and older residing in the community for at least one month. The rate is reported for short stays (<20 days), medium stays (21-100 days) and long stays (>100 days). The criteria that determines whether admissions come from the community requires a full year of data. The intended direction of the metric and DD is decrease. HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Exhibit 63 shows a significant decrease in short-term admissions between the change before HHP and the change Pre-Year 1 to HHP Year 1 for SPA 1 enrollees (0.6 admissions per 1,000 member months) and SPA 2 enrollees (0.4 admissions), but these trends were similar to that of Short Term the respective control groups. Exhibit 63: Trends in Admissions to an Institution from the Community (Short-Term Stay) Before and During HHP by SPA as of September 30, 2020 1.4 1.3 1.1 1.1 1.1 ~ 1.0 1.0 1.0 0.9 . 0.8 0.9 0.9 Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m HHP Enrollees Control Group (Only 4- aa DTT a=) 5 5 Difference-in- Year 1 to HHP Between Difference (DD) Year 1 Changes SPA1 HHP Enrollees | 0.3* -0.3* -0.6* Control Group | 0.3* -0.3* -0.6* 0.0 SPA 2 HHP Enrollees | 0.2* -0.2 -0.4* Control Group | 0.2* -0.1 -0.3 -0.2 Overall | HHP Enrollees | 0.3* -0.3* -0.5* Control Group | 0.3* -0.2* -0.5* 0.0 Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | HHP Outcomes [hi March 2022 Medium Term UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 64 shows no significant changes in medium-term admissions from Pre-Year 1 to HHP Year 1 for HHP SPA 1 and SPA 2 enrollees and the trends between HHP enrollees and their respective control groups were similar. Exhibit 64: Trends in Admissions to an Institution from the Community (Medium-Term Stay) Before and During HHP by SPA as of September 30, 2020 1.9 2.0 16 45 1.7 1.6 3 La 1.7 1.1 1.1 1.2 Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m HHP Enrollees Control Group tO YT ati) Before yao Year 1 to HHP Difference Between Dat aseso la Difference HHP Year 1 eye) (DD) SPA1 HHP Enrollees | 0.2* 0.1 -0.2 Control Group | 0.2* -0.1 -0.3* 0.2 SPA 2 HHP Enrollees | 0.1 0.3 0.3 Control Group | 0.0 0.1 0.1 0.2 Overall | HHP Enrollees | 0.2* 0.1 -0.1 Control Group | 0.2* -0.1 -0.3* 0.2 Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes ps<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 -Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Exhibit 65 shows a small but significant increase in long-term admissions from Pre-Year 1 to HHP Year 1 for HHP SPA 1 (0.3 admissions per 1,000 member months) and SPA 2 (0.4 admissions) enrollees. The change from before to during HHP among SPA 1 HHP enrollees was Long term similar to the control group but among SPA 2 enrollees was significantly greater than the control group by 0.4 admissions (DD). Exhibit 65: Trends in Admissions to an Institution from the Community (Long-Term Stay) Before and During HHP by SPA as of September 30, 2020 1.1 1 0 1.2 1.1 1.0 0.9" 0.9 0.7 0.8 0.6 . 06 2 0.6 Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m@ HHP Enrollees Control Group Change DThatya-Tia= Difference-in- Change Pre-Year ; Yaa cola re eel Between Difference HHP Changes (DD) SPA 1 HHP Enrollees | -0.1 0.3* 0.4* Control Group | -0.1 0.2 0.2* 0.1 SPA 2 HHP Enrollees | -0.1 0.4* 0.5* Control Group | -0.1 0.0 0.1 0.4* Overall | HHP Enrollees | -0.1* 0.3* 0.4* Control Group | -0.1* 0.1 0.2* 0.2* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | HHP Outcomes [ky March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program HHP Process Metrics Trends in six HHP specified metrics were examined on an annual basis. Adult Body Mass Index Assessment Adult Body Mass Index Assessment is an HHP core metric that measures the percentage of beneficiaries between the ages of 18 and 74 who had an outpatient visit and whose body mass index (BMI) was documented during the measurement year or the year prior to the measurement year. The intended direction of this metric and DD is increase. Exhibit 66 shows a significant increase in documented BMI from Pre-Year 1 to HHP Year 1 for HHP SPA 1 enrollees (5.4%) and the control group (4.3%). These were slower rates of increase for both groups compared to the changes before HHP (10.3% for both HHP enrollees and the control group). However, the decline in BMI screening for the HHP enrollees was significantly smaller than the control group (1.1%, DD). The same pattern was observed for SPA 2 enrollees. Exhibit 66: Trends in Adult Body Mass Index Assessment Before and During HHP by SPA for HHP Enrollees and the Control group as of September 30, 2020 x 2 . 4d Oo © © . ao wo A 3 8 8 F 8 = 8 8 8 8 in ) in in of Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP Control Group SPA2 SPA 1 m@ HHP Enrollees Change Pre- Difference Difference-in- Oy T ay =i) ; eat Year 1 to Between DTT] a=T A=) alae dtl eet (DD) SPA1 | HHP Enrollees 10.3%* 5.4%* -4.9%* Control Group 10.3%* 4.3%* -6.0%* 1.1%* SPA 2 | HHP Enrollees 11.4%* 2.9%* -8.5%* Control Group 11.4%* 2.0%* -9.5%* 1.0%* Overall | HHP Enrollees 10.5%* 4.9%* -5.6%* Control Group 10.5%* 3.8%* -6.7%* 1.1%* HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes ps<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 -Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | HHP Outcomes JRE) March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Screening for Depression and Follow-Up Plan Screening for Depression and Follow-Up Plan is an HHP core metric that measures the percentage of beneficiaries age 12 and older with an outpatient visit in the measurement year who were screened for depression and had a documented follow-up plan on the date of the positive screen. This metric was not reported for SPA 2 because the metric specifications excludes enrollees with an active diagnosis of depression or bipolar disorder, which were very common conditions among the SPA 2 enrollees. An increase in this metric and DD is intended. Exhibit 67 shows a significant increase in this metric from Pre-Year 1 to HHP Year 1 for both HHP SPA 1 enrollees (9.0%) and the control group (7.2%). These increases were greater for both groups compared to before HHP (5.9% for HHP enrollees and 5.7% for the control group). This rate of increase was significantly greater (1.6%, DD) for HHP enrollees than the control group. Exhibit 67: Trends in Screening for Depression and Follow-Up Plan Before and During HHP for SPA 1 HHP Enrollees and the Control group as of September 30, 2020 x © x oO) Ww a nN a 4 ° e9 a a S as x oO o + + Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP SPA 1 m HHP Enrollees Control Group Cent Tay << a ee Difference Difference-in- eee Year 1 to Between Difference HHP Year 1 Ot (DD) SPA1 HHP Enrollees 5.9%* 9.0%* 3.1%* Control Group 5.7%* 7.2%* 1.5%* 1.6%* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes ps<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Follow-Up After Hospitalization for Mental Iliness Follow-Up After Hospitalization for Mental Illness is an HHP core metric that measures the percentage of beneficiaries age 6 and older who were hospitalized for treatment of selected mental illness in the measurement year and who had a follow-up visit within 7 and 30 days with a mental health practitioner. The intended direction of the metric and DD is increase. Exhibit 68 shows that the trends for 7-day follow-up did not change significantly for SPA 1 or SPA 2 enrollees during HHP or between HHP enrollees and the control group. Exhibit 68: Trends in Follow-Up After Hospitalization for Mental Illness within 7 Days Before and During HHP by SPA for HHP Enrollees and the Control group as of September 30, 2020 e, &y &s | Se 88 88 oo gg GS eg a nS Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m HHP Enrollees Control Group Change Oli <-e ee Difference Pha Ta-yal eo Before Year 1 to HHP Between in-Difference HHP (Lae Changes (DD) SPA 1 HHP Enrollees 3.0% -1.1% -4.0% Control Group 2.7% -1.2% -3.9% -0.1% SPA 2 HHP Enrollees 4.0% -0.6% -4.6% Control Group 3.7% -0.6% -4.3% -0.3% Overall | HHP Enrollees 3.5%* -0.8% -4.3% Control Group 3.2%* -0.9% -4.1% -0.2% Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes ps<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | HHP Outcomes JPA! March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 69 shows that that the trends for 30-day follow-up also did not change significantly for SPA 1 or SPA 2 enrollees during HHP or between HHP enrollees and the control group. Exhibit 69: Trends in Follow-Up After Hospitalization for Mental Illness within 30 Days Before and During HHP by SPA for HHP Enrollees and the Control group as of September 30, 2020 Se A x xe 3 i (} at 3 N oS o oO ur xe - oO 5 x ~ om 3 8 S § oe 3B <u 8 ~ ~ § =. 8 wo Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m HHP Enrollees Control Group creme ent 4- aaa DT Ta-R-T are) DThatya-tar a= ett Year 1 to HHP Between in-Difference Year 1 Changes (DD) SPA 1 HHP Enrollees | 7.4%* 2.8% -4.6% Control Group | 7.0%* 1.3% -5.6% 1.0% SPA 2 HHP Enrollees | 2.8% 2.1% -0.7% Control Group | 2.7% -2.6% -5.3% 4.6% Overall | HHP Enrollees | 5.1%* 2.4% -2.6% Control Group | 4.8%* -0.7% -5.4%* 2.8% Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Follow-Up After Emergency Department Visit for Alcohol and Other Drug Abuse or Dependence Follow-Up after Emergency Department Visit for Alcohol and Other Drug Abuse or Dependence is an HHP core metric that measures the percentage of emergency department (ED) visits in the measurement year among individuals age 13 and older with a principal diagnosis of alcohol and other drug (AOD) abuse or dependence who had a follow-up visit for AOD abuse or dependence. The measure is reported for follow-up within 7 days and within 30 days. The intended direction of the metric and DD is increase. Exhibit 70 shows that no significant trends were observed for follow-up after ED visit for AOD abuse or dependence within 7 days during HHP for HHP enrollees and no difference in trends with the control group in SPA 1 or SPA 2. Exhibit 70: Trends in Follow-Up After ED Visit for Alcohol and Other Drug Abuse and Dependence within 7 Days Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020 xe x 8 x to ° ° 5 x x Oo x x x xf &S o a4 x oa wx nN a ~" GS nO ON o F wo wo wo wo wn Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 m@ HHP Enrollees Control Group SPA 2 ee yo ae Difference Difference-in- ean Year 1 to HHP Between Date) (Tae | Changes (DD) SPA1 HHP Enrollees | 0.9% 0.6% -0.3% Control Group | 0.8% 0.0% -0.9% 0.5% SPA 2 HHP Enrollees | 1.7% -0.9% -2.6% Control Group | 1.8% -2.3% -4.1%* 1.4% Overall | HHP Enrollees | 1.2%* 0.0% -1.2% Control Group | 1.2%* -0.8% -2.0% 0.8% Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | HHP Outcomes [R! March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 71 shows that no significant trends were observed for follow-up after ED visit for AOD abuse or dependence within 30 days during HHP for HHP enrollees and no difference in trends with the control group in SPA 1 or SPA 2. Exhibit 71: Trends in Follow-Up After ED Visit for Alcohol and Other Drug Abuse and Dependence within 30 Days Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020 x & am as x SOS In 38 > aes 1 x x o a Sf + NS 4 = so "i a 8 S a Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m HHP Enrollees Control Group eee Ota yh arseae= Pyare aaa eatin Year 1 to HHP Between in-Difference Year 1 Changes (DD) SPA1 HHP Enrollees | 1.2% 1.4% 0.2% Control Group | 1.1% 0.1% -1.0% 1.2% SPA 2 HHP Enrollees | 4.6%* -1.3% -5.9% Control Group | 4.7%* -4.7%* -9.3%* 3.5% Overall | HHP Enrollees | 2.4%* 0.4% -2.0% Control Group | 2.4%* -1.6% -4.0%* 2.0% Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Initiation and Engagement of Alcohol and Other Drug Abuse or Dependence Treatment Initiation of AOD Abuse or Dependence Treatment is an HHP core metric that measures the percentage of individuals age 13 and older with a new episode of AOD abuse or dependence in the measurement year who received initiation of treatment within 14 days of the diagnosis. The intended direction of this metric and DD is increase. Exhibit 72 shows that initiation of AOD treatment declined significantly for SPA 1 HHP enrollees Pre-Year 1 to HHP Year 1 (3.4%) after increasing before HHP (1.3%). While the change in initiation rates also declined for the control group, the decline was larger for HHP enrollees (2.7%, DD). SPA 2 HHP enrollees also experienced a decline of 3.4% during HHP and after an increase of 2.4% before HHP enrollment but this decline was not significantly larger than that of the control group. Exhibit 72: Trends in Initiation of Alcohol and Other Drug Abuse or Dependence Treatment Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020 ° x x xX e x we x x xe Qo 8 no 5 3 x XN nm Oo sem MN WN N ON do ON a GS Ss Sa "ON NN a4 & NN 5 a Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m HHP Enrollees Control Group ee alae Pyare ae Phare ala=oe aati Year 1 to HHP Between in-Difference Year 1 eet (DD) SPA 1 HHP Enrollees | 1.3%* -3.4%* -4,.7%* Control Group | 1.3%* -0.7% -2.0%* -2.7%* SPA 2 HHP Enrollees | 2.4%* -3.4%* -5.8%* Control Group | 2.3%* -2.5%* -4.9%* -0.9% Overall | HHP Enrollees | 1.6%* -3.4%* -5.0%* Control Group | 1.6%* -1.2% -2.8%* -2.2%* UCLA Evaluation | HHP Outcomes [R5 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Engagement of AOD Abuse or Dependence Treatment is an HHP core metric that measures the percentage of beneficiaries age 13 and older that initiated AOD abuse or dependence treatment and who were engaged in ongoing treatment within 34 days of the initiation visit. The intended direction of the metric and DD is increase. Exhibit 73 shows that trends engagement in AOD treatment did not change for SPA 1 or the control group. However, trends increased for SPA 2 from Pre-Year 1 to HHP Year 1 by 8.3% after no significant increase before HHP and the increase in engagement compared to before HHP was significantly larger (10.9%, DD) in comparison to the control group. Exhibit 73: Trends in Engagement of Alcohol and Other Drug Abuse or Dependence Treatment Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020 x o x " o oO x x & t 3 x ee SS & 6S a 5 Qf 60 in © o 1 Nor Qt Qs ie) : oOo © nan " 5 8 am Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m@ HHP Enrollees Control Group yo ae Darla) , , Change Difference-in- Year 1 to HHP Between i Before HHP Difference (DD) Year 1 ery 453 SPA1 HHP Enrollees | 1.5% 1.6% 0.1% Control Group | 1.5% 1.0% -0.5% 0.6% SPA 2 HHP Enrollees | 0.2% 8.3%* 8.1% Control Group | 0.2% -2.6% -2.8% 10.9%* Overall | HHP Enrollees | 1.1% 3.7% 2.7% Control Group | 1.1% -0.1% -1.2% 3.9% Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | HHP Outcomes -ihy, March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Use of Pharmacotherapy for Opioid Use Disorder Use of Pharmacotherapy for Opioid Use Disorder is an HHP core metric that measures the percentage of beneficiaries aged 18 to 64 with an opioid use disorder (OUD) who filled a prescription or were administered a medication for the disorder during the measurement year. The intended direction of the metric and DD is increase. Exhibit 74 does not show a change in this metric for SPA 1 enrollees and their control group during HHP. There was a significant decline in the rate of pharmacotherapy from before HHP for SPA 2 enrollees (5.3%) but there was no significant difference in change with the control group. Exhibit 74: Trends in Use of Pharmacotherapy for Opioid Use Disorder Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020 ° Sx Ss Se Ss xs x in © co ON © © ° m + ™ vn oo x 5 & = § i 3 mm mom nyo Co Oo wo + t+ N N N N N N Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m HHP Enrollees Control Group Change Pre- Difference Pyare hae Year 1 to Between in-Difference HHP Year 1 Changes (DD) SPA 1 HHP Enrollees 1.7%* 0.3% -1.4% Control Group | 1.7%* 0.2% -1.5% 0.1% SPA 2 HHP Enrollees 3.6%* -1.8% -5.3%* Control Group | 3.5%* 0.4% -3.1%* -2.2% Overall HHP Enrollees 2.2%* -0.3% -2.4%* Control Group | 2.2%* 0.2% -1.9%* -0.5% Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Trends in three HHP specified metrics were examined on an annual basis. HHP Outcome Metrics Controlling High Blood Pressure Controlling High Blood Pressure is an HHP core metric that measures the percentage of beneficiaries aged 18 to 85 who had a diagnosis of hypertension and whose blood pressure was adequately controlled during the measurement year. The intended direction is increase. Exhibit 75 shows that there was a significant increase in SPA 1 HHP enrollees with controlled high blood pressure both before HHP (4.0%) and from Pre-Year 1 to HHP Year 1 (3.1%), however the latter increase was significantly less than the increase before HHP by 0.9%. The decline was not significantly different from a similar decline observed in the control group. SPA 2 enrollees had a significant decline (1.8%) in the percentage of enrollees with controlled high blood pressure from Pre-Year 1 to HHP Year 1 after an increase (4.6%) before HHP, but the decline did not significantly differ from the control group. Exhibit 75: Trends in Controlling High Blood Pressure Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020 Ss x Sx Ss so oS . x ~ 8 2 8 q 3 sf 85 | $8 BS e8 x & = eg ~ . og | Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 m@ HHP Enrollees ControlGroup SPA2 fy 1 ay 34) Oya Difference yh a =te-tala=oe Before Year 1 to HHP Between in-Difference lala Year 1 Ory 453 (DD) SPA 1 HHP Enrollees | 4.0%* 3.1%* -0.9%* Control Group | 4.2%* 3.0%* -1.2%* 0.3% SPA 2 HHP Enrollees | 4.6%* -1.8%* -6.4%* Control Group | 4.6%* -0.3% -4.9%* -1.5% Overall HHP Enrollees | 4.1%* 2.3%* -1.8%* Control Group | 4.3%* 2.5%* -1.8%* 0.0% UCLA Evaluation | HHP Outcomes JR) March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 -Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Plan All-Cause Readmission Plan All-Cause Readmission is an HHP core metric that measures the percentage of acute inpatient and observation stays during the measurement year that were followed by an unplanned acute readmission for any diagnosis within 30 days for beneficiaries ages 18 to 64. The intended direction is decrease. Exhibit 76 shows that readmission rates did not significantly change from Pre-Year 1 to HHP Year 1 and the change in rate from before HHP was not significantly different for SPA 1 or SPA 2 enrollees. However, SPA 1 enrollees had a significantly greater increase in the rates from before to during HHP than the control group by 1.2% (DD) more readmissions. Exhibit 76: Trends in Plan All-Cause Readmission Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020 ° oO x xs & °o o o x x x 8 in cS x 3 3 3 ao oo . oj 3 Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 m@ HHP Enrollees Change Before HHP Control Group ay. ae Year 1 to HHP Difference Between Dac asaesollae Difference Year 1 et (DD) SPA1 HHP Enrollees | 0.8%* 0.5% -0.2% Control Group | 0.8%* -0.7%* -1.5%* 1.2%* SPA 2 HHP Enrollees | -1.0%* -0.3% 0.7% Control Group | -1.0%* 0.2% 1.2% -0.5% Overall | HHP Enrollees | 0.4%* 0.4% -0.1% Control Group | 0.4%* -0.5% -1.0%* 0.9%* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | HHP Outcomes RH UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Prevention Quality Indicator (PQI) 92: Chronic Conditions Composite PQI 92 is an HHP core metric that measures the number of inpatient hospital admissions for ambulatory care sensitive chronic conditions per 100,000 member months for individuals age 18 and older. The intended direction of the metric and DD is decrease. Exhibit 77 shows that PQI was significantly increasing before HHP in SPA 1 (2.7) and SPA 2 (1.2) enrollees. The rates then declined significantly from Pre-Year 1 to HHP Year 1 for both SPA 1 (1.9) and SPA 2 (1.3), but SPA 1 rates declined less from before to during HHP compared to the control group (0.9, DD). Exhibit 77: Trends in Prevention Quality Indicator (PQI) 92: Chronic Conditions Composite Before and During HHP by SPA for HHP Enrollees and the Control Group as of September 30, 2020 ao + So 96 S| S| + o o NON ~ N ©0 a ow nN on ' Qo Pre-Year 2 Pre-Year 1 Year 1 Pre-Year 2 Pre-Year 1 Year 1 Before HHP During HHP Before HHP During HHP SPA 1 SPA 2 @ HHP Enrollees Change Before stale Control Group Difference Between ently 4-05 (ent Tay << a eee Year 1 to HHP Year 1 DTT ai-aaTa x= Tae PT ar) a-T11a= (DD) SPA 1 HHP Enrollees | 2.7* -1.9* -4.6* Control Group | 2.7* -2.8* -5.5* 0.9* SPA 2 HHP Enrollees | 1.2* -1.3* -2.5* Control Group | 1.4* -1.6* -3.0* 0.5 Overall HHP Enrollees | 2.4* -1.8* -4.2* Control Group | 2.4* -2.5* -4.9* 0.8* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (Pre-Year 1 - Pre-Year 2). Change Pre-Year 1 to HHP Year 1 is calculated as: (Year 1 - Pre-Year 1). Difference between changes is calculated as: (Change Pre-Year 1 to HHP Year 1 -Change Before HHP). Difference-in-difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). HHP Outcomes | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs This section addresses the following HHP evaluation questions: 1. Did Medi-Cal expenditures for health services decline after HHP implementation? 2. Did Medi-Cal expenditures for needed outpatient services increase? UCLA calculated estimated payments for all services provided to HHP enrollees and the control group before HHP and during HHP using Medi-Cal claims and encounter data. Payments were estimated by creating mutually exclusive categories of service and attributing a fee to each Medi-Cal claim in that category (Appendix A: Attributing Estimated Medi-Cal Payments to Claims). This methodology allowed UCLA to estimate payments for HHP enrollees and the control group before each enrollee's HHP enrollment and during HHP and assess if payments for HHP enrollees declined more than for the control group using the DD methodology. UCLA developed DD models to measure changes in total estimated payments and in specific categories of services including ED visits, hospitalizations, outpatient medication, and outpatient services. UCLA examined changes in six month increments up to 24 months (1-6, 7-12, 13-18, and 19-24) before HHP enrollment and up to 12 months (1-6 and 7-12) during HHP. The DD analysis measured the change from 19-24 vs. 1-6 months before HHP for both HHP enrollees and the control group; the change during HHP from 1-6 to 7-12 months for both HHP enrollees and the control group; and the difference between the changes in HHP enrollees vs. the control group. The shorter timeframe for examining payments allowed for a clearer assessment of change during the early phase of HHP implementation. The findings were not subject to potential seasonality in service utilization due to rolling enrollment throughout the year and measuring change following the date of enrollment per beneficiary. The payment amounts reported in this section are estimates and are not equivalent to overall Medi-Cal expenditures for multiple reasons, including significant differences between this attribution methodology vs. per member per month payments to managed care plans for enrolled beneficiaries. These estimated payments are primarily intended to compare change in trends between HHP enrollees and the control group. See (Appendix A: Attributing Estimated Medi-Cal Payments to Claims) for further detail and limitations. UCLA Evaluation | Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs (EE) March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Estimated Payments for HHP Services Total Estimated Medi-Cal Payments UCLA measured total estimated Medi-Cal payments before and during HHP. The payment estimates were generated using the methodology described above and detailed further in Appendix A: Data Sources and Analytic Methods. These estimates are intended for measuring whether HHP led to efficiencies and do not represent actual Medi-Cal expenditures for HHP enrollees. Examples of Medi-Cal expenditures include inpatient and outpatient services, pharmaceuticals, imaging and laboratory services, behavioral health services, and long-term care stays. The intended direction of the measure and DD is decrease. Exhibit 78 shows that total estimated payments were significantly increasing for SPA 1 ($168 per enrollee per six month) and for SPA 2 ($161) before HHP. The total estimated payments continued to increase during HHP by $331 and $1,277 for SPA 1 and SPA 2 enrollees, respectively. However, payments from before HHP to during HHP increased significantly less than the control groups by $96 (DD) for SPA 1 enrollees and $121 (DD) for SPA 2 enrollees. Exhibit 78: Trends in Total Estimated Payments Before and During HHP by SPA as of September 30, 2020 32 awn oN s+ com na om tw in tn On <s a5 88 8S gh aR gales ga Re $8 ae Se OO GN AA soy BR OY OD 8A Ge ta vu aN UU ia ict UV) U} ul ! / L L nT | | L + © N Ww wo N + co N WO WO N N - - ° oO a N - = ° Oo - ° ° ° + + °o ° ° ° + - °o ~ ~ ~ <a el ~ ~ ~ ~ a a ~ oO) " » ~s Oo m Ss » S| oo a S| Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 SPA 2 m HHP Enrollees Control Group Tax fey <3 saab Difference-in- Before HHP During HHP Difference (DD) Ot SPA1 | HHP Enrollees | $168* $331* $163* Control Group | $173* $432* $259* -$96* SPA2 | HHP Enrollees | $161* $1,277* $1,116* -$121* Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs | UCLA Evaluation UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program Control Group | $167* $1,404* $1,237* Overall | HHP Enrollees | $167* $528* -$1,253* Control Group | $172* 5634* -$1,205* -$101* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 - 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 -12 months of HHP minus 1-6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in- difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). Estimated Payments for Outpatient Services UCLA estimated Medi-Cal payments for outpatient services. There is no intended direction for this measure. Payments for outpatient services are likely to increase due to unmet need and increased access to these services, but payments are likely to decrease once health needs are addressed and service use declines. Exhibit 79 shows that after an initial increase in estimated payments at the start of HHP, estimated payments continued to increase significantly for SPA 1 and SPA 2 enrollees during HHP. Compared to control groups, the increase from before HHP to during HHP was significantly smaller for SPA 1 ($23, DD) and significantly greater for SPA 2 ($18, DD) per HHP enrollee per six months. Exhibit 79: Trends in Payments for Outpatient Services Before and During HHP by SPA as of September 30, 2020 wore oN © DO D moO ON Go © A 09 att Sts ve m™ 00 * wn NS wn an WY oO ay a om a Vv) On Vii a4 ~~ wm wn oe V/) st 00 + > Win oO aS HR o So FBS BO Aa Ss ot wR uo Re ROS Bo MH st 0 N Wo wo N + 0 N oO oO N N S| S| °o oO a N a Sa oO fo) S| ° ° ° + - ° ° ° ° + + ° ~ ~ ~ a al ~ ~ ~ ~ a a ~ a m nN Nn oa Mm ~N nN S| - - - Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 SPA 2 m@ HHP Enrollees Control Group lita (ery 43 sabi Difference-in- HHP DIVE a tay Mala eee Difference (DD) SPA1 | HHP Enrollees | $110* $258* $148* Control Group | $103* $274* $172* -$23* UCLA Evaluation | Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs [BEL March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program SPA2 | HHP Enrollees | $93* $529* $436* Control Group | $87* $506* $419* $18* Overall | HHP Enrollees | $107* $315* -S489* Control Group | $99* $322* -$427* -$15* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes ps<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1-6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 - 12 months of HHP minus 1-6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in- difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). Estimated Payments for Outpatient Medication UCLA estimated Medi-Cal payments for outpatient medication. There is no intended direction for this measure. Payments for outpatient medication are likely to increase due to unmet need and increased access to these medications, but payments are likely to stabilize or decrease once health needs are addressed. Exhibit 80 shows a significant increase in estimated payments during HHP for both SPA 1 and SPA 2. Compared to their respective control groups, the change in estimated payments from before HHP to during HHP increased significantly less for SPA 1 ($7, DD) per HHP enrollee per six months and was not significant for SPA 2. Exhibit 80: Trends in Outpatient Medication Payments Before and During HHP by SPA as of September 30, 2020 W 00 Mm Oo &) 00 U--V)- AD wn =) + Ln to OS OR FG RS BS BO tt SSF nun un ao Va An IH uy to t+ ty Trt s+ © N Ww wo N + 0c N oO Ww N N a a oO oO | N a a Oo oO el oO oO ° - + °o ° oO °o + + oO » ce ~ a a ~ ~ ~ ~ | a ~ Oo m » ~~ 0) foe] s ~~ a a a a Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 m@ HHP Enrollees ControlGroup SPA2 Difference Dacascos icine arene Between Difference Before HHP During HHP SER ere) SPA1 | HHP Enrollees | $25* S$50* $25* Control Group | $25* $58* $32* -$7* Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 SPA2 | HHP Enrollees | $9* $311* $302* Control Group | $9* $304* $295* S7 Overall | HHP Enrollees | $22* $104* -$258* Control Group | $22* $109* -$256* -S4* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes ps<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1-6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 -12 months of HHP minus 1-6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in- difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs (BEY, March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Estimated Payments for Emergency Department Visits UCLA estimated Medi-Cal payments for emergency department (ED) visits. The intended direction of the measure and DD is decrease. Exhibit 81 shows that these estimated payments were increasing significantly before HHP for both SPA 1 and SPA 2. During HHP, the estimated payments for ED visits decreased by $7 per SPA 1 enrollee per six months and increased by $55 per SPA 2 enrollee. The decline in the estimates for SPA 1 enrollees from before HHP to during HHP (S9) was significantly greater than the control group by $29 (DD) and the increase for SPA 2 enrollees from before HHP to during HHP ($50) was significantly smaller than the control group by $20 (DD). Exhibit 81: Trends in Payments for Emergency Department Visit Before and During HHP by SPA as of September 30, 2020 oO 5a 6 XN m ne ns oO ~ wo Wt NW a oO + + in + a wo ~ 4 fad oF ¢% AGH aa an AY] oO N on An wm Sn SCH at AO a on wh ow a un | | | | s+ 0 N oO wo N s+ 0 N oO oO N N a a oO oO S| N =o =" ° oO = ° ° ° r - ° ° ° ° - + ° ~ ~ » a Pel ~ ~ ~ ~ a a ~ oO m ~ ~~ oO) oe] NS » a a a a Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 SPA 2 m HHP Enrollees Control Group Oy T ay :<2) rT ay =<) salabali Difference-in- Before HHP During HHP Difference (DD) Oy 4-54 SPA1 | HHP Enrollees | $2* -S7* -S9* Control Group | $3* $23* $20* -S29* SPA2 | HHP Enrollees | $5* S55* $50* Control Group | $6* $76* $70* -$20* Overall | HHP Enrollees | $3* S6* -$73* Control Group | $4* $34* -S65* -S27* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 - 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 -12 months of HHP minus 1-6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in- difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs | UCLA Evaluation UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Estimated Payments for Hospitalizations UCLA estimated Medi-Cal payments for hospitalizations. The intended direction of the measure and DD is decrease. Exhibit 82 shows that the change in estimated payments for hospitalization declined significantly for SPA 1 enrollees and increased significantly for SPA 2 enrollees from before HHP to during HHP. These changes were significantly less for both SPAs compared to the control group ($7 and $127, respectively, per HHP enrollee per six months, DD). Exhibit 82: Trends in Payments for Hospitalizations Before and During HHP by SPA as of September 30, 2020 s+ 0 S sl no ~ +7 NU + ow N oO a ~" oJ 2 af *y gh RS 88 » 2° oh 88 2 nin Dun oF nr wt mo {0 Qu rw on wp VW Sn mf UV +U On + I : I I : + 0 N oO wo N + oo N oO oO N N | S| oO oO a N Se = oO oO 4 oO oO ° - - °o ° oO ° + - oO ~ ~ ~ a a » ~ ~ ~ al 4 ~ a mM Ss Ss oa m S s a oa a a Before HHP (months) During HHP Before HHP (months) During HHP (months) (months) SPA 1 m HHP Enrollees Control Group = spa 2 Dy at=le=ta) ty 1 at 342) eye T ay 42) "a nn py arse-aa mle Before HHP During HHP Difference (DD) Oy et4 SPA1 | HHP Enrollees | S60* S7* -$53* Control Group | $72* $26* -S46* -S7* SPA2 | HHP Enrollees | $64* $200* $136* Control Group | $83* $347* $264* -$127* Overall | HHP Enrollees | $61* S47* -$437* Control Group | $74* $93* -S$496* -$32* Source: Medi-Cal claims data from July 1, 2016 through September 30, 2020. Notes: * Denotes p<0.05, a statistically significant difference. SPA 1 includes enrollees with chronic conditions and substance use disorders. SPA 2 includes enrollees with severe mental illness. Change Before HHP is calculated as: (1 - 6 months before HHP minus 19 - 24 months before HHP divided by 3). Change During HHP is calculated as: (7 -12 months of HHP minus 1-6 months of HHP). Difference between changes is calculated as: (Change During HHP -Change Before HHP). Difference-in- difference is calculated as: (Difference between changes for HHP enrollees - Difference between changes for control group). UCLA Evaluation | Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs 139 March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program HHP Program Expenditures UCLA examined HHP supplemental payments based on per-member per-month (PMPM) rates to participating MCPs and calculated the estimated total and average per-enrollee HHP expenditures per month from July 1, 2018 to September 30, 2020. PMPM payments varied by MCP and county and were changed each fiscal year. PMPM rates were higher at the start of the program in order to account for anticipated start-up costs, and were lowered as the program went on. Rates were consistently lower for enrollees covered by both Medicare and Medi-Cal (Duals) compared to those covered by Medi-Cal only. Exhibit 83 shows that total estimated HHP expenditures by September 30, 2020 were $189,737,702 and the average expenditures per enrollee per month was $479. The overall estimated expenditures for duals were lower ($3,237,651) than those covered by Medi-Cal only ($186,500,051), and the average monthly per person expenditures were lower as well ($123 for duals, $504 for Medi-Cal only). Exhibit 83: Estimated HHP Supplemental Expenditures by Enrollees Type and Implementation Group, as of ember 30, 2020 $189,737,702 $479 $4,574,677 $396 $64,404,038 $433 $120,758,987 $512 $3,237,651 $123 $147,550 $102 $735,532 $122 $2,354,569 $125 $186,500,051 $504 $4,427,127 $439 $63,668,506 $446 $118,404,418 $545 Source: UCLA Analysis of MCP Enrollment Reports from August 2019 and Quarterly HHP Reports from September 2019 to September 2020. Per-member, per-month rates by MCP and dual-status were provided by the California Department of Health Care Services. Estimated Medi-Cal Payments among HHP Enrollees and HHP Costs | UCLA Evaluation UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Conclusions This report has highlighted the continued progress made by MCPs through September 2020 and since the first interim evaluation report. This report contains additional comparisons that highlight the early impact of HHP and updated information on the CB-CME networks. The COVID-19 pandemic and subsequent statewide shelter in place order likely impacted enrollment and the ability of MCPs and their contracted CB-CMEs to provide HHP services, but some of this impact was mitigated by MCP efforts to adapt workflows and increase telehealth capacity. The findings indicated a substantial growth in CB-CME networks; a notable proportion of enrollees who had super utilization of acute care in emergency departments and hospitals, particularly enrollees who were also experiencing homelessness and had conditions such as chronic kidney disease and depression; improvements in selected metrics that reflected processes and outcomes of care compared to the control group; an initial increase in use of outpatient services soon after enrollment; greater declines in ED visits and hospitalizations than the control group; and a slower growth in estimated payments for ED visits and hospitalization than the control groups. Collectively, the findings implied challenges of engaging HHP enrollees in treatment and improving outcomes for enrollees with multiple comorbidities. Yet, the findings also implied early success of HHP in achieving reductions in acute services concurrently with provision of more primary, specialty, mental health, and SUD services as well as outpatient medications in the first six months following HHP enrollment to address the needs of enrollees. The next and final evaluation report will include additional data for the final 15 months of HHP, including changes in the HHP core metrics and measures of utilization and estimated Medi-Cal payments. UCLA Evaluation | Conclusions [@Z5) UCLA Center for Health Policy R h ee Appendix A: Data Sources and Analytic Methods Readiness Documents UCLA used the readiness documents from 16 MCPs submitted to DHCS to report on MCP implementation of HHP. In these readiness documents, MCPs reported on topics including organizational model, staffing, health information technology, HHP services, HHP network, and HHP operations. Analytic Methods UCLA reviewed all readiness documents to answer the UCLA evaluation questions detailed in Exhibit 84. MCPs varied in the level of detail in their documents. UCLA identified and tabulated relevant information to the extent possible given this variation by MCP. Information from readiness documents were cross-checked with other data including MPC Quarterly HHP Reports to improve accuracy when possible. Exhibit 84: Evaluation Questions and Data Sources Evaluation Question Location in Readiness Documents 1. Which HHP network model was employed? Organizational Model 2. What was the composition of HHP networks? Organizational Model MCP Duties/Responsibilities 3. What types of staff provide HHP services? Organizational Model Staffing 4, What was the data sharing approach? Health Information Technology/Data and Information Sharing 5. What was the approach to targeting patients for enrollment | Member Engagement into HHP? Member Notices Risk Grouping Housing Services Source: UCLA Health Homes Program Evaluation Design, 2019. Limitations The MCP readiness documents represented MCP plans for HHP implementation and may not reflect the final implementation approach by MCPs. Several MCPs submitted periodically revised readiness documents during HHP implementation. These documents included drafts, revisions, and communications with DHCS regarding further revisions and/or clarifications. In addition, MCPs provided variable amounts of detail on planned implementation, which may have led to a limited understanding of MCPs' final approach. The MCPs maximum estimated HHP enrollment overall and by CB-CME in readiness documents and their responsibilities are unlikely to align with actual quarterly enrollment data. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research Nae eP es Health Economics and Evaluation Research Program Enrollment Reports and MCP Quarterly Reports UCLA used MCP Enrollment Reports and Quarterly HHP Reports to analyze HHP enrollment. Enrollee-level HHP enrollment data was only available in MCP Enrollment Reports prior to July 2019. All four MCPs (Anthem Blue Cross of California Partnership Plan, San Francisco Health Plan, Inland Empire Health Plan, and Molina Healthcare of California Partner Plan) that implemented HHP by July 2019 submitted an Enrollment Report to DHCS in August 2019, covering the period of July 1, 2018 to June 30, 2019. All MCPs submitted Quarterly HHP Reports during the time they had implemented HHP from July 1, 2018 to September 30, 2020. Starting in July 2019, MCP Quarterly HHP Reports included enrollee-level data on both enrollment, homelessness, and housing status. These two data sources had some differences, which resulted in UCLA only being able to analyze enrollment at a monthly level. Staggered implementation of the program by county resulted in MCPs with different reporting lengths. Homeless and housing statuses on an enrollee-level were examined quarterly, from July 1, 2019 when enrollee-level homeless data was first reported, through September 30, 2020. Analytic Methods Exhibit 85 shows the enrollment data obtained from these reports. Monthly enrollment data from the MCP Enrollment Reports and Quarterly HHP Reports were combined to determine monthly enrollment status by individual enrollee. If there were conflicting data for individual enrollees between the two data sources, UCLA used the more recent data from the Quarterly HHP Reports. Forty-three enrollees that switched counties or plans during their enrollment were excluded from further analysis. Beneficiaries who were enrolled on any date during a given month were considered enrolled for the whole month. Beneficiaries that were disenrolled for less than 30 days in between enrolled months were considered enrolled in the program for that month. However, 1,439 beneficiaries who were only enrolled for less than 31 days were excluded from the analyses of enrollment patterns. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods NETSB lOpy) UCLA Center for Health Policy Research Health Economics and Evaluation Research Program UCLA used the MCP Quarterly HHP Reports to analyze data on enrollee's housing status and housing service utilization. Enrollee-level housing services data were included in the Quarterly HHP Reports starting in July 2019, which limited the analysis of housing services to July 1, 2019 through September 30, 2020. Exhibit 85: Beneficiary-Level Variables Data Elements Definitions SPA Enrolled in SPA 1 vs. SPA 2. Dual Status Ever enrollee in both Medicare and Medi-Cal during HHP enrollment. County County in which enrollee is enrolled. Monthly Enrollment Status Indicator for HHP enrollment status for a particular month. Enrollment Date The date an enrollee starts to enroll in HHP. Enrollment date reported prior to 2019 Quarter 3 always begins on the first day of the initially enrolled month. Enrollment date reported after June 30, 2019 is the exact date. Disenrollment Date The date an enrollee disenrolled from HHP. Disenrollment date reported prior to July 1, 2019 is the last day of the month. Disenrollment date reported after June 30, 2019 is an exact date. Number of Times Disenrolled The number of times each enrollee disenrolled from the MCP throughout their enrollment. Length of Enrollment The differences between disenrollment date and enrollment date. If an enrollee enrolls in and disenrolls from HHP on the same date, the length of enrollment will be one day. Day count was divided by 30 to estimate length of enrollment in months. Ever Homeless during HHP Data only available from Quarterly HHP Reports. Indicates whether enrollee was ever homeless during HHP enrollment. Data only available from Quarterly HHP Reports. Enrollee is homeless or at risk for homelessness from July 1, 2019 to September 30, 2020. Homeless or at Risk for Homelessness Received Housing Services Data only available from Quarterly HHP Reports. Enrollee received housing services from July 1, 2019 to September 30, 2020. Housed by September 2019 | Data only available from Quarterly HHP Reports. Indicator of whether enrollee was housed by September 30, 2020. Notes: Data from MCP Enrollment Reports from July 1, 2018 to September 30, 2020 and MCP Quarterly HHP Reports from July 1, 2019 to September 30, 2020. From the MCP Quarterly HHP Reports, UCLA reported on CB-CMEs by organization type as of September 2020. MCPs reported individual CB-CMEs, identified by the National Plan and Provider Enumeration System (NPPES) NPI, serving HHP enrollees and the projected capacity of each CB-CME. UCLA used the NPI Registry to identify characteristics of unique CB-CMEs in MCP networks. In addition, UCLA reported on the percentage of eligible beneficiaries by implementation group excluded from HHP for seven exclusion rationales defined by DHCS and reported in the MCP Quarterly Reports. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research Nae eP es Health Economics and Evaluation Research Program UCLA analyzed the enrollment data provided by MCPs. Given that enrollee-level data in the MCP Quarterly Report were not required until July 2019, UCLA had to combine these data with MCP Enrollment Reports from July 1, 2018 to June 30, 2019 to examine enrollment and enrollment patterns. These two data sources had some differences, which resulted in UCLA only being able to analyze enrollment at a monthly level. Staggered implementation of the program by county resulted in MCPs with different reporting lengths. Limitations Medi-Cal Enrollment and Claims Data UCLA used Medi-Cal enrollment and claims data from July 1, 2016 to September 30, 2020 to create demographic health status indicators, health care utilization indicators, and preliminary metrics used in this report. Claims data included both managed care and fee-for-service encounters. Analytic Methods HHP Services HHP services were reported for all MCPs, although reporting varied by MCP. Kaiser reported that none of their enrollees received services while Alameda Alliance reported that 98% of their enrollees received services. All MCPs reported that less than 100% of their enrollees received any HHP service, although every HHP enrollee should have received at least one service. Exhibit 86 displays indicators of utilization of HHP services reported by MCPs in Medi-Cal claims data. Exhibit 86: HHP Service Utilization Indicators Indicators Definitions Proportion of enrollees that ever received an HHP The percent of enrollees that ever received the service service. Proportion of enrolled months that services were The percent months with services received out of the provided per enrollee number of months enrolled in HHP among HHP enrollees that have ever received the service. Average number of units of service per enrollee per The average of each HHP enrollee's monthly average month during months that services were provided number of service units for the received service each month among HHP enrollees that have ever received the service. Units of service are defined as 15-minutes of service; multiple units of service are possible. Median number of units of service per enrollee during | The median of each HHP enrollee's monthly number months that service was provided of service units for the received service each month among HHP enrollees that have ever received the service. Units of service are defined as 15-minutes of service; multiple units of service are possible. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program UCLA used the HHP designated HCPCS codes and modifiers to identify encounters that included HHP services, defined in Exhibit 87. HCPCS code G0506 and modifier codes U1 to U7 were used July 1, 2018 through September 30, 2018, and HCPCS code G9008 and modifier codes U1 to U7 were used October 1, 2018 through September 30, 2020. Exhibit 87: HHP Services Provider Type Modifier Modality Definition Engagement Services Clinical Staff Provider Type Not | U7 Not specified Active outreach such as direct communications with Specified member (e.g., face-to-face, mail, electronic, and telephone), follow-up if the member presents to another partner in the HHP network or using claims data to contact providers the member is known to use. Providers must show active, meaningful, and progressive attempts at member engagement each month until the member is engaged. Examples of acceptable engagement include: (1) letter to member followed by phone call to member; (2) phone call to member, outreach to care delivery partners and social service partners; (3) and street level outreach, including, but not limited to, where the member lives or is accessible. Core Services Provided by U1 In-person Comprehensive care management, care coordination, Clinical Staff health promotion, comprehensive transitional care, U2 Telehealth individual and family support services, and referral to community and social supports Provided by Non- | U4 In-person Clinical Staff US Telehealth Other Services Provided by U3 Not specified Case notes, case conferences, tenant supportive services, Clinical Staff and driving to appointments Provided by Non- | U6 Not specified Demographic Indicators Exhibit 88 displays demographic indicators created by UCLA using Medi-Cal monthly enrollment data. UCLA calculated age based on an enrollee's HHP enrollment date. On the rare occasion enrollment data included more than one birthday for an enrollee, UCLA used the latest birthday reported. While not common, if the Medi-Cal enrollment data contained conflicting data for gender, race, or language for an HHP enrollee, UCLA used the most frequently reported category. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research Nae eP es Health Economics and Evaluation Research Program Exhibit 88: Demographic Indicators Indicators Definitions Age Enrollee's final age in years at the time of HHP enrollment. Gender Indicates whether an enrollee is male or female. Race The race label for an enrollee: White, Hispanic, African American, Asian American and Pacific Islander, American Indian and Alaska Native, other, or unknown. English as Primary Indicating whether an enrollee's primary language is English or not. Language Number of Months Full scope coverage is defined as at enrollment in at least one dental MCP and another with Full Scope non-dental MCP during the eligible date period. The number of months that an enrollee Coverage is full scope is reported for the year prior to the enrollee's initial enrollment in HHP. Health Status Indicators UCLA used Medi-Cal claims data from July 1, 2016 to September 30, 2020 to assess health status of HHP enrollees prior to their enrollment in HHP. UCLA followed chronic condition and acuity eligibility criteria developed by DHCS for HHP as described in the HHP Program Guide (Exhibit 89). According to these criteria, chronic conditions were present if an enrollee had two or more services on different dates for the specified condition during the two years prior to HHP enrollment. UCLA also used the criteria set by CMS's Chronic Condition Warehouse to obtain a complete list of chronic condition and potentially chronic or disabling condition categories. Exhibit 89: Health Status Indicators Indicators | Definition Chronic Conditions Chronic Condition | The percentage of enrollees that meet chronic condition criteria 1. An enrollee satisfies Criteria 1: Two chronic condition criteria 1 if the enrollee has at least two of the following HHP eligible specific chronic conditions: chronic obstructive pulmonary disease (COPD), chronic kidney disease conditions and (CKD), diabetes, traumatic brain injury, chronic or congestive heart failure, coronary artery SUD disease, chronic liver disease, dementia, substance use disorder. Chronic Condition | The percentage of enrollees that meet chronic condition criteria 2. An enrollee satisfies Criteria 2: chronic condition criteria 2 if the enrollee has hypertension and one of the following HHP Hypertension and | eligible chronic conditions: chronic obstructive pulmonary disease, diabetes, coronary another specific artery disease, chronic or congestive heart failure. comorbidity Chronic Condition | The percentage of enrollees that meet chronic condition criteria 3. An enrollee satisfies Criteria 3: Serious | chronic condition criteria 3 if the enrollee has one of the following HHP eligible chronic Mental Illness conditions: major depression disorders, bipolar disorder, psychotic disorders (including (SMI) schizophrenia. Chronic Condition | The percentage of enrollees that meet chronic condition criteria 4. An enrollee satisfies Criteria 4: Asthma | chronic condition criteria 4 if the enrollee has the HHP eligible chronic condition asthma. Acuity Acuity Criteria 1: The percentage of enrollees that meet acuity criteria 1. An enrollee satisfies acuity criteria Three or more 1 if the enrollee has at least three of the following HHP eligible chronic conditions: chronic chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD), diabetes, traumatic conditions brain injury, chronic or congestive heart failure, coronary artery disease, chronic liver disease, dementia, substance use disorder. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Indicators Definition Acuity Criteria 2: One or more Hospitalizations The percentage of enrollees that meet acuity criteria 2. An enrollee satisfies acuity criteria 2 if the enrollee has at least one inpatient hospital stay during one year prior to HHP enrollment. Acuity Criteria 3: Three or more ED Visits The percentage of enrollees that meet acuity criteria 3. An enrollee satisfies acuity criteria 3 if the enrollee has at least three or more emergency department visits during one year prior to HHP enrollment. Chronic Condition The percentage of enrollees meeting each of the CCW condition category criteria in the Warehouse period prior to HHP enrollment. (CCW) Conditions CDPS (Chronic The mean, median, and standard deviation of CDPS among all enrollees. The CDPS is Illness and calculated based on the International Classification of Diseases (ICD) diagnosis codes in Disability Medi-Cal claims data. Payment System Risk Score) UCLA Evaluation | Appendix A: Data Sources and Analytic Methods Health Economics and Evaluation Research Program Healthcare Utilization Indicators UCLA Center for Health Policy Research Tyr) UCLA also created healthcare utilization indicators using Healthcare Effectiveness Data and Information Set (HEDIS) 2019 Volume 2 definitions, National Uniform Claim Committee taxonomy designations, the Chronic Conditions Warehouse, and the American Medical Association's Current Procedure Terminology (CPT) Codebook. Exhibit 90 displays these indicators. Exhibit 90: Healthcare Utilization Indicators Indicators Definitions Improvement Measured by Increase or Decrease Number of Hospitalizations per 1,000 | The number of inpatient hospitalization visits Decrease Member Months during the service month. Length of hospitalization (days) The total lengths measured in number of total Decrease days of all hospitalizations during the service month. Percentage of Enrollees with Any The percentage of enrollees who ever had at Decrease Hospitalizations least one hospitalization Percentage of Enrollees with Any ED The percentage of enrollees who ever had at Decrease Visits Resulting in Discharge least one ED visit resulting in discharge Number of Primary Care Services per | The number primary care provider services Increase or 1,000 Member Months during the service month. Decrease Number of Specialty Services per The number of specialty services during the Increase or 1,000 Member Months service month. Decrease Number of Mental Health Services The number of mental health services during the | Increase or per 1,000 Member Months service month. Decrease Number of Substance Use Disorder The number of substance use disorder services Increase or Services per 1,000 Member Months during the service month. Decrease UCLA Evaluation | Appendix A: Data Sources and Analytic Methods ee ePes UCLA Center for Health Policy Research Health Economics and Evaluation Research Program HHP Metrics and Additional Mesures HHP metrics were calculated based on HHP metric specifications in CMS's Core Set of Health Care Quality Measures for Medicaid Health Home Programs. HHP metrics were grouped by whether they measured process of care delivery or patient outcomes. All metrics were reported in the aggregate and included data for two years prior to and one year following each individual's enrollment in HHP when possible. UCLA assessed any length of enrollment or required number of months of enrollment on Medi-Cal enrollment rather than HHP enrollment in order to be consistent between HHP enrollees and the control group. A limited number of metrics were reported semi-annually rather than annually in order to calculate the change in the measure during HHP when there was only one year of data. Exhibit 91 includes descriptions of all HHP metrics and how changes in the metric are to be interpreted. Exhibit 91: HHP Core Metrics, Definitions, and Reporting Status . . Improvement Measured by Metric Description Increase or Decrease Adult Body Mass Percentage of Health Home enrollees ages 18 to 74 who | Increase Index (BMI) had an outpatient visit and whose body mass index Assessment (BMI) was documented during the measurement year or the year prior to the measurement year. Follow-Up After Percentage of discharges for Health Home enrollees age | Increase Hospitalization for 6 and older who were hospitalized for treatment of Mental Illness within selected mental illness diagnoses and who had a follow- 30 days up visit with a mental health practitioner within 30 days. Follow-Up After Percentage of discharges for Health Home enrollees age | Increase Hospitalization for 6 and older who were hospitalized for treatment of Mental Illness within selected mental illness diagnoses and who had a follow- 7 days up visit with a mental health practitioner within 7 days. Follow-Up After ED Percentage of ED visits for Health Home enrollees age Increase Visit for Alcohol and 13 and older with a principal diagnosis of alcohol or Other Drug Abuse or other drug (AOD) abuse or dependence who had a Dependence within 7 | follow-up visit for AOD abuse or dependence with 7 days days. Follow-Up After ED Percentage of ED visits for Health Home enrollees age Increase Visit for Alcohol and 13 and older with a principal diagnosis of alcohol or Other Drug Abuse or other drug (AOD) abuse or dependence who had a Dependence within follow-up visit for AOD abuse or dependence with 30 30 days days. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Metric Description Improvement Measured by Increase or Decrease Screening for Percentage of Health Home enrollees age 12 and older | Increase Depression and screened for clinical depression on the date of the Follow-Up Plan encounter, and if positive, a follow-up plan is documented on the date of the positive screen. Initiation of Alcohol Percentage of enrollees who initiate treatment through | Increase and Other Drug Abuse | within 14 days of the diagnosis. or Dependence Treatment Engagement of Percentage of enrollees who initiate treatment and who | Increase Alcohol and Other had two or more additional AOD services or MAT within Drug Abuse or 34 days of the initiation visit. Dependence Treatment Controlling High Blood | Percentage of Health Home enrollees ages 18 to 85 who | Increase Pressure had a diagnosis of hypertension (HTN) and whose blood pressure (BP) was adequately controlled during the measurement year. Plan All-Cause For Health Home enrollees ages 18 to 64, the number of | Decrease Readmissions acute inpatient stays during the measurement year that were followed by an unplanned acute readmission for any diagnosis within 30 days and the predicted probability of an acute readmission. Prevention Quality Number of inpatient hospital admissions for ambulatory | Decrease Indicator (PQ|) 92: care sensitive chronic conditions per 100,000 member Chronic Conditions months for Health Home enrollees age 18 and older. Composite This measure includes adult hospital admissions for diabetes with short-term complications, diabetes with long-term complications, uncontrolled diabetes without complications, diabetes with lower extremity amputation, chronic obstructive pulmonary disease, asthma, hypertension, or heart failure without a cardiac procedure. Ambulatory Care: Rate of emergency department (ED) visits resulting in Decrease Emergency Department (ED) Visits discharge per 1,000 member months among Health Home enrollees. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Metric Description Improvement Measured by Increase or Decrease Institution from the Community (Long- Term Stay} (skilled nursing facility or intermediate care facility) from the community that result in a long-term stay (more than 100 days) during the measurement year per 1,000 member months. Inpatient Utilization Rate of acute inpatient care and services (total, Decrease maternity, mental and behavioral disorders, surgery, and medicine) per 1,000 member months among Health Home enrollees Inpatient Length of All approved days from admission to discharge. Decrease Stay Use of Percentage of enrollees ages 18 to 64 with an opioid Increase Pharmacotherapy for | use disorder who received buprenorphine, oral Opioid Use Disorder naltrexone, long-acting injectable naltrexone, or methadone for the disorder. Admission to an The number of admissions to an institutional facility Decrease Institution from the (skilled nursing facility or intermediate care facility) Community (Short- from the community that result in a short-term stay (1 Term Stay} to 20 days) during the measurement year per 1,000 member months. Admission to an The number of admissions to an institutional facility Decrease Institution from the (skilled nursing facility or intermediate care facility) Community (Medium- | from the community that result in a medium-term stay Term Stay} (21 to 100 days) during the measurement year per 1,000 member months. Admission to an The number of admissions to an institutional facility Decrease Source: Detailed information for each metric is available in HHP Metric Specifications. Control Group Construction UCLA obtained administrative Medi-Cal monthly enrollment and claims data from July 2016 to September 2020 for 48,925 individuals reported as enrolled into HHP and for 802,670 individuals that were potentially eligible for HHP on a targeted engagement list (TEL). The TEL was produced bi-annually and UCLA used all TELs through May 2020. These data included two years prior to the start of HHP enrollment (July 2016 to June 2018) and up to the first 27 months of HHP enrollment (July 2018 to September 2020). UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program UCLA used 48 variables indicating demographic, health status, service utilization, and cost to select the control group (Exhibit 92). Demographic variables were constructed from Medi-Cal enrollment data. Health status variables were constructed from claims data and reflected the HHP chronic condition eligibility criteria and measures of chronic health conditions (e.g., asthma, diabetes, hypertension, chronic kidney disease). The chronic condition eligibility criteria and indicators were constructed following the specifications developed to create the TEL by DHCS (HHP Program Guide). UCLA created and included a measure of acute care utilization by grouping enrollees based on their number of ED visits and hospitalizations. Slopes and intercepts in monthly utilization of ED visits and hospitalizations were also included in the model. Cost variables include estimated Medi-Cal payments, overall and for specific categories of service, and indicators of trends in those payments. Exhibit 92: Variables Used to Select the Control Group Indicator Description Demographics (9 indicators and variables) Age Group Age at the start of HHP enrollment (0-17, 18-34, 35-49, 50-64, or 65+ years) Gender Reported Gender in Medi-Cal Enrollment (Male or Female) Race/Ethnicity Reported Race/Ethnicity in Medi-Cal (White, Hispanic, Black, Asian or Pacific Islander, or Native American/Other/Unknown) Language English as the preferred language Homelessness UCLA developed indicator that uses address-based and claim-based indicators to predict homelessness WPC enrollment Indicator of whether or not individual was ever enrolled in Whole Person Care County County of residence Managed Care Plan Medi-Cal Managed Care Plan Full Scope Months in Medi-Cal | Number of months in the reported as having full-scope Medi-Cal coverage Health Status (4 indicators) HHP Chronic Condition At least two of the following: Chronic Obstructive Pulmonary Disease (COPD), Eligibility Criteria 1 Chronic Kidney Disease (CKD), Diabetes, Traumatic Brain Injury, Chronic or Congestive Heart Failure, Coronary Artery Disease, Chronic Liver Disease, Dementia, Substance Use Disorder. HHP Chronic Condition Hypertension and one of the following: COPD, Diabetes, Coronary Artery Eligibility Criteria 2 Disease, Chronic or Congestive Heart Failure. HHP Chronic Condition One of the following: Major Depression Disorders, Bipolar Disorder, or Eligibility Criteria 3 Psychotic Disorders (including Schizophrenia). HHP Chronic Condition Asthma Eligibility Criteria 4 Service Utilization (20 indicators and variables) Acute Care Utilization Groups UCLA created indicators that groups individuals by their baseline emergency (5 indicators) department and hospital utilization: super utilization, high utilization, moderate utilization, low utilization or at-risk-for high utilization UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Average Number of Average annual number of service use in the baseline period, normalized by Hospitalizations and the number of months enrolled in Medi-Call: Hospitalization and Emergency Emergency Department Visits Department Visits (2 variables) Utilization Slopes (4 variables) | Slope of monthly service utilization in the baseline period for emergency department visits, hospitalizations, primary care services and specialty care services. Utilization Intercepts (4 Intercept of monthly service utilization in the baseline period for emergency variables) department visits, hospitalizations, primary care services and specialty care services. Primary Care Organization type | Number of primary care services by organization type: health centers, group (3 variables) organizations, and individual practices Behavioral Health Services (2 Use of behavioral health services in the baseline period: mental health and indicators) substance use disorder Cost (15 variables) Estimated Medi-Cal Payments | Estimated payments for total costs, emergency department visits, (5 variables) hospitalizations, outpatient services, and outpatient prescriptions. Estimated Payment Slopes (5 Slope of monthly estimated Medi-Cal payments in the baseline period for variables) total costs, emergency department visits, hospitalizations, outpatient services, and outpatient prescriptions. Estimated Payment Intercepts | Intercept of monthly estimated Medi-Cal payments in the baseline period for (5 variables) total costs, emergency department visits, hospitalizations, outpatient services, and outpatient prescriptions. Due to the phased implementation of HHP, UCLA grouped HHP enrollees into nine cohorts based on the quarter in which they enrolled and selected a potential pool of control beneficiaries for each cohort. This method ensured that the control group beneficiaries had a similar baseline period to their matched enrollee. To select the final matched control group, UCLA first estimated a propensity score in generalized additive models for modeling non-linear effect and avoiding overfitting including the variables in Exhibit 92. HHP enrollees and two control beneficiaries were further matched within each MCP and county based on a combination of nearest neighborhood match and exact match, including propensity scores, acute care utilization groups, and HHP chronic condition eligibility criteria. UCLA used sampling with replacement due to unavailability of similar matches per MCP. The final control group to HHP enrollee ratio was 1.47. To balance the sample, each control group beneficiary was matched to multiple HHP enrollees. Exhibit 93 shows the characteristics of the final control group for the largest HHP SPA 1 enrollee cohort (cohort 5; n=8,595), which consisted of those enrolled from July to September 2019 from Groups 1, 2, and 3 for SPA 1. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Data show that the control group was similar to the HHP enrollees for all indicators and measures. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Exhibit 93: Comparison of Select Characteristics of HHP SPA 1 Cohort 5 Enrollees (Enrolled July to September 2019) and Matched Control Beneficiaries SPA1HHP Enrollees | After Match in Cohort 5 Control Group Age (at time of enrollment) % 0-17 6% 5% % 18-34 13% 13% % 35-49 25% 26% % 50-64 49% 51% % 65+ 6% 5% Gender % male 41% 42% Race/Ethnicity % White 21% 19% % Latinx 44% 46% % African American 20% 21% % Asian 6% 6% % Other or Unknown 9% 9% Language % English proficient 73% 73% Medi-Cal full-scope months Average number of months 11.5 11.5 year prior to enrollment Homelessness UCLA-constructed indicator 23% 25% WPC enrollment Any enrollment in WPC through 6% 6% September 2020 Two specific conditions (Criteria 1) 44% 43% Hypertension and another specific 60% 60% HHP Chronic Condition condition (Criteria 2} Criteria Serious mental health conditions 41% 40% (Criteria 3) Asthma (Criteria 4) 31% 30% Hypertension 72% 70% . was Diabetes 53% 52% Select Chronic Conditions Major Depressive Disorders 35% 33% Substance Use Disorders 9% 10% Emergency Department Visit Normalized Annual Rate 2.9 3.0 U hiiection. partment visi. ED Intercept 0.23 0.23 ED Slope 0.005 0.004 Normalized Annual Rate 0.7 0.7 Hospitalization Utilization Hospitalization Intercept 0.05 0.05 Hospitalization Slope 0.004 0.004 PCP slope 0.05 0.03 Outpatient Services PCP intercept 0.59 0.56 Utilization Specialty slope 0.06 0.04 Specialty intercept 0.47 0.38 ED cost slope 0.06 0.05 ED cost intercept 3.1 2.8 Estimated Med-Cal Payment | Hospitalization cost slope 0.29 0.23 Trends Hospitalization cost intercept 3.8 3.4 Outpatient cost slope 0.52 0.46 Outpatient cost intercept 6.9 6.1 UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program UCLA developed unique matched control groups by different outcomes. For metrics that restricted the sample to specific subpopulations, such as follow-up after hospitalization for mental illness, UCLA developed a control group within groups based on whether individuals met the denominator criteria (i.e., hospitalized for mental illness) before HHP, during HHP or is both time periods. In addition, the match models for utilization metrics and measures did not include slopes and intercepts for costs due to collinearity of those variables with utilization indicators. Similarly, the match models for cost measures did not include slopes and intercepts for utilization variables. Difference-in-Difference Models UCLA assessed the impact of HHP for the overall HHP population and for SPA 1 and SPA 2 separately, using the difference-in-difference (DD) modeling approach. All models were controlled for demographics (gender, age, race/ethnicity, primary language, months of Medi- Cal enrollment), utilization indicators (acute care utilization group), and health status indicators (baseline CDPS risk scores and HHP chronic condition eligibility criteria). The model additionally included an indicator for having at least one primary or secondary diagnosis of COVID-19 in the claims data and the number of months spent enrolled in HHP during the pandemic. The models predicted changes in metrics before and during HHP for HHP enrollees and the matched control group and differences in these differences. The baseline and enrollment periods for each HHP enrollee and their matched controls were based on their enrollee's date of enrollment, and the sample included only HHP enrollees with at least one years of baseline data and at least one month of enrollment in HHP. UCLA used logistic regression models for binary metrics (e.g., Controlling High Blood Pressure) and a zero-inflated count model with Poisson distribution for count metrics (e.g., Primary Care Visits per 1,000 Member-Months, Specialty Care Visits per 1,000 Members-Months) and HHP estimated Medi-Cal payments. The exposure option within a Generalized Linear Model (GLM) was used to adjust for different number of months of Medi-Cal enrollment and the subsequent different lengths of exposure to HHP. All analyses of individual-level metrics were analyzed based on Medi-Cal member months. The DD analyses differed for HHP specified metrics that required one year of observation from metrics that did not require one year of observation and for optional measures. For HHP specified metrics that required one year of observation, the DD analyses measured changes from the Pre-HHP Year 2 to Pre-HHP Year 1 for both HHP enrollees and the control group; the UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program change from Pre-HHP Year 1 to the HHP Year 1 for both HHP enrollees and the control group; and the difference between the changes for HHP enrollees vs. the control group. For the remaining metrics and measures, UCLA examined changes in six month increments up to 24 months (1-6, 7-12, 13-18, and 19-24) before HHP enrollment and up to 12 months (1-6 and 7-12) during HHP. For these, the DD analysis measured the change from 19-24 vs. 1-6 months before HHP for both HHP enrollees and the control group; the change during HHP from 1-6 to 7-12 months for both HHP enrollees and the control group; and the difference between the changes in HHP enrollees vs. the control group. The shorter timeframe for examining metrics allowed for a clearer assessment of changed during the early phase of HHP implementation. The findings were not subject to potential seasonality in service utilization due to rolling enrollment throughout the year and measuring change following the date of enrollment per beneficiary. Limitations One of the acuity criteria set by DHCS in the HHP Program Guide was chronic homelessness. However, Medi-Cal Enrollment and Claims data do not provide sufficient data to identify individuals that experience chronic homelessness. As a result, UCLA could not report on this acuity criteria. The data in this report are restricted to September 2020 due to a minimum lag of six months for relatively complete claims data. The identification of chronic conditions relied on the primary and secondary diagnoses associated with each service. Any error in original reporting of these diagnoses by providers may have resulted in under or over reporting of chronic conditions. HHP services may have been underreported due to missing HCPCS code modifiers by MCPs. MCPs that did not report any encounters with the HHP HCPCS code included Aetna Better Health of California, UnitedHealthcare Community Plan of California, Community Health Group Partnership Plan, and Kaiser Permanente. Attributing Estimated Medi-Cal Payments to Claims Background The great majority of services under Medi-Cal are provided by managed care plans that receive a specific capitation amount per member per month and do not bill for individual services received by Medi-Cal beneficiaries. While managed care plans are required to submit claims to Medi-Cal, these claims frequently include payment amounts of unclear origin that are different from the Medi-Cal fee schedule. A small and unique subset of Medi-Cal beneficiaries are not enrolled in managed care and receive care under the fee-for-service (FFS) reimbursement methodology and have claims with actual charges and paid values. FFS claims are reimbursed UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program primarily using fee schedules developed by Medi-Cal. The capitation amounts for managed care plans are developed using the same fee schedules by Mercer annually, using complex algorithms and other data not included in claims. To address the gaps in reliable and consistent payment data for all claims, UCLA estimated the amount of payment per Medi-Cal claim under HHP using various Medi-Cal fee schedules for services covered under the program. The methodology included (1) specifying categories of service observed in the claims data, (2) classifying all adjudicated claims into these service categories, (3) attributing a dollar payment value to each claim using available fee schedules and drug costs, and (4) examining differences between these and available external estimates. UCLA estimated payments for both managed care and FFS claims to promote consistency in payments across groups and to avoid discrepancies due to different methodologies. The payment estimates generated using this methodology are not actual Medi-Cal expenditures for health care services delivered during HHP. Rather, they represent the estimated amount of payment for services and are intended for measuring whether HHP led to efficiencies by reducing the total payments for HHP enrollees before and after the program, and in comparison to a group of comparison patients in the same timeframe. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Service Category Specifications Data Sources UCLA used definitions from multiple sources to categorize and define different types of services. These sources included Medi-Cal provider manuals, HEDIS value set, DHCS 35C File, American Medical Association's CPT Codebook, National Uniform Code Committee's taxonomy code set, and other available sources. e DHCS's Medi-Cal provider manuals included billing and coding guidelines for provider categories and some services. e The HEDIS Value Set by the National Committee for Quality Assurance used procedure codes (CPT and HCPCS), revenue codes (UBREV), place of service codes (POS), and Systematized Nomenclature of Medicine-Clinical Terms (SNOMED CT) to define value sets that measure performance in health care. For example, the HEDIS value set "ED" is a combination of procedure codes that describe emergency department services and revenue codes specifying that services were provided in the emergency room. e DHCS Paid Claims and Encounters Standard 35C File (DHCS 35C File) provided specifications to managed care plans on how claims must be submitted and contained detailed information about claims variables and their meaning and utility, such as vendor codes describing the location of services and taxonomy codes describing the type of provider and their specializations. e The American Medical Association's Current Procedure Terminology (CPT) Codebook contained a list of all current procedural terminology (CPT) codes and descriptions that are used by providers to bill for services. e The National Uniform Claim Committee's (NUCC's) Health Care Provider Taxonomy code set identified provider types such as Allopathic and Osteopathic Physician and medical specialties such as Addiction Medicine defined by taxonomy codes. UCLA also used other resources to address gaps in definitions. For example, hospice codes that were used in claims submitted before 2016 were not included in the Medi-Cal provider manual, but UCLA collected the pre-2016 hospice codes from other DHCS guidelines. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research March 2022 Health Economics and Evaluation Research Program UCLA constructed eighteen mutually exclusive categories of service (Exhibit 94). Available Methods claims data included managed care, fee-for-service, and Short-Doyle. Some categories were defined using complementary definitions from more than one source. UCLA assigned claims to only one of the eighteen service categories to avoid duplication when calculating total estimated HHP payments. The outpatient services category may include claims included in other categories and therefore is not included in calculation of the total estimated payment in this report. UCLA assigned claims to the first service category a claim meets the criteria for as ordered in Exhibit 94. All services, apart from primary care visits, provided on the day of an ED visit were grouped as part of the ED visit to represent the total cost of the visit. For example, patients may have received transportation to an emergency department and laboratory tests during the emergency department visit, and these services were included in the ED category rather than the transportation or laboratory services categories. This approach may have included lab or transportation services in the ED category that were not part of the ED visit, and may have undercounted lab and transportation in their respective categories. However, this was necessary because claims data lacked information on the specific time of day when services were rendered. Similarly, all claims for services received during a hospitalization were counted as part of the same stay and were excluded from other categories of service, except for primary care visits on the day of admission. Other categories were identified solely by the procedure code or place of service and were not bundled with other services occurring on the same day, such as long-term care, home health/ home and community-based services, community-based adult services, FQHC services, labs, imaging, outpatient medication, transportation, and urgent care. Some claims lacked the information necessary to be categorized and were classified under an "Other Services" category. These frequently included physician claims without a defined provider taxonomy and durable medical equipment codes that were billed separately and could not be associated with an existing category. Exhibit 94: Description of Mutually Exclusive Categories of Service* Service category Definition DYeXYol df old fey) Tol UT Ca) 1 Emergency HEDIS Place of service is hospital emergency Department Visits room and procedure code is emergency (ED) service 2 Hospitalizations DHCS 35C File Place of service is inpatient and admission and discharge dates are present and are on different days UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 Order UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Service category Definition source DY=XYol df old (ela) 3 Hospice Care DHCS 35C File, Provider is hospice or procedure code is HEDIS, and hospice service DHCS Medi-Cal Provider Manuals 4 Long-Term Care DHCS 35C File Claim is identified as LTC or provider is (LTC) Stays LTC organization; stays one day apart are counted as one visit, stays two or more days apart are separate stays 5 Home Health and DHCS 35C File Provider is a home health agency or Home and and DHCS Medi- | home and community-based service Community-Based | Cal Provider waiver provider, procedure is home Services (HH/HCBS) | Manuals health or home and community-based service 6 Community-Based | DHCS 35C File Provider is adult day health care center or Adult Services and DHCS Medi- | procedure code is community-based (CBAS) Cal Provider adult service, which are health, Manuals therapeutic and social services in a community-based day health care program 7 Federally Qualified | DHCS 35C File Provider is an FQHC or RHC (FQHC) and Rural Health Center (RHC) Services 8 Laboratory Services | DHCS 35C File Claim is identified as clinical laboratory, laboratory & pathology services, or laboratory tests 9 Imaging Services DHCS 35C File Claim is identified as portable x-ray services or imaging/ nuclear medicine services 10 Outpatient DHCS 35C File Claim is identified as pharmacy Medication 11 Transportation DHCS 35C File Claim is identified as medically required Services transportation 12 Primary Care National Provider is allopathic and osteopathic Services Uniform Claim physician (with specialization in adult Committee medicine, adolescent medicine, or geriatric medicine, family medicine, internal medicine, pediatrics, or general practice), or physician assistant or nurse practitioner (with specialization in UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Service category Definition DY=XYol df old (ela) Tol b TC) medical, adult health, family, pediatrics, or primary care) 13 Specialty Care National Provider is allopathic and osteopathic Services Uniform Claim physician or physician assistant or nurse Committee practitioner (with all specializations not captured in the Primary Care Services category) 14 Outpatient Facility | DHCS 35C File Claim is identified as outpatient facility Services 15 Dialysis Services DHCS 35C File Provider is a dialysis center and and CPT procedure is dialysis Codebook 16 Therapy Services DHCS Medi-Cal | Procedure code is occupational, physical, Provider Manual | speech, or respiratory therapy 17 Urgent Care National Provider is ambulatory urgent care facility Services Uniform Claim Committee 18 Other Services N/A Provider, procedure, or place of service is not captured above N/A Outpatient Services | HEDIS Claim type is outpatient and procedure code, revenue code, or place of service code is outpatient Source: UCLA Methodology. Notes: * indicates categories are mutually exclusive except for outpatient services category UCLA examined four of the above categories that made up 69% of total payments for HHP claims in 2019 (Exhibit 95). Exhibit 95: Percentage of 2019 Total Estimated Payments by Category of Service for HHP Medi- Cal Claims Percentage of Total agent Category of Service aaa lesvom Naat All Categories 100% Outpatient Services 22% Outpatient Medication 17% Emergency Department Visits 5% Hospitalizations 25% All other categories 31% Source: UCLA analysis of Medi-Cal Claims data from July 1, 2018 to September 30, 2020 UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Attributing Payments to Specific Services To attribute payments to each category of service, UCLA developed methods to calculate an estimated payment for each category based on available data. Exhibit 96 displays the categories of service and what is included in the calculation of estimated payments for each category. Exhibit 96: Category of Service and Payment Descriptions Category of Service Calculation of Estimated Payment Emergency Department Visits (ED) Payments for all services taking place in the emergency department of a hospital, including services on the same day of the ED visit, excluding services by PCPs and FQHCs and RHCs. Two sub-categories are reported: ED visits followed by hospitalizations and all other ED visits that are followed by discharge. Hospitalizations Payments for all services that take place during a hospitalization, excluding visits with primary care providers on the first or last day of the stay, FQHC visits on the first or last day of the stay, or ED visits that preceded hospitalization Hospice Care Payments for hospice services in an LTC facility or Home Health setting, excluding hospice services rendered during a hospitalization Long-Term Care (LTC) Stays Institutional fees billed by LTC facilities; the per diem rate includes supplies, drugs, equipment, and services such as therapy Home Health and Home and Community-Based Payments for services provided by a home health agency (HHA) and services provided through the home and community-based Services (HH/HCBS) services (HCBS) waiver Community-Based Adult Payments for community-based adult services and for services Services /(CBAS) rendered at an adult day health care center Federally Qualified (FQHC) and Rural Health Center (RHC) Services Payments for all services provided in an FQHC or RHC Laboratory Services Payments for laboratory services, except those provided during a hospitalization or ED visit Imaging Services Payment for imaging services, except those provided during a hospitalization, ED visit, or LTC stay UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research Health Economics and Evaluation Research Program March 2022 Category of Service Calculation of Estimated Payment Outpatient Medication Payments for outpatient drug claims, excluding prescriptions filled on the same day as an ED visit or on the day of discharge from a hospitalization Transportation Services Payments for medically required transportation, excluding transportation on the same day as an inpatient admission or an emergency department visit Primary Care Services Payments for services provided by a primary care physician Specialty Care Services Payments for services provided by a specialist, excluding services provided during an inpatient stay or an emergency department visit, and excluding facility fees Outpatient Facility Services Facility fees paid to hospital outpatient departments and ambulatory surgical centers Dialysis Services Payments for dialysis services rendered in a dialysis center Therapy Services Payments for occupational, speech, physical, and respiratory therapy services Urgent Care Services Payments for services provided in an urgent care setting Other Services Payments for services not captured above Outpatient Services Payments for all services delivered in an outpatient setting Source: UCLA Methodology. UCLA used all available Medi-Cal fee schedules and supplemented this data with other data sources as needed. Payment data sources, brief descriptions, and the related categories of services they were attributed to are provided in Exhibit 97. Exhibit 97: Payment Data Sources Source Description Applicable Service Categories Contains rates set by DHCS for all Level | procedure codes that are reimbursable ED, Hospitalizations, Hospice, LTC, HH/HCBS, CBAS, Imaging, Medi-Cal Physician Fee Schedule Annual files 2013 to 2020 inflated/ deflated to 2019 by Medi-Cal for services and procedures rendered by physicians and other providers Transportation, Primary Care, Specialty Care, Dialysis, Urgent Care, Other, and Outpatient Services Durable Medical Equipment (DME) Fee Contains rates set by CMS for Level II procedure codes for durable medical ED, Hospitalizations, Hospice, LTC, HH/HCBS, UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 Source Schedule Annual files 2017 to 2020 inflated/ deflated to 2019 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Description equipment such as hospital beds and accessories, oxygen and related respiratory equipment, and wheelchairs Applicable Service Categories CBAS, Transportation, Primary Care, Specialty Care, Dialysis, Urgent Care, and Other Medical Supplies Fee Schedules October 2019 Contains rates set by DHCS for supplies such as needles, bandages, and diabetic test strips ED, Hospitalizations, Hospice, LTC, HH/HCBS, CBAS, Transportation, Primary Care, Specialty Care, Dialysis, Urgent Care, and Other Average Sales Price Data (ASP) for Medicare Part B Drugs Annual files 2014 to 2020 inflated/ deflated to 2019 Contains rates set by CMS for procedure codes for physician-administered drugs covered by Medicare Part B ED, Hospitalizations, Hospice, LTC, Primary Care, Specialty Care, and Other CMS MS-DRG grouping software, DHCS's APR- DRG Pricing Calculator 12/1/2019 Contains Diagnostic Related Grouping (DRG) codes used for hospitalizations (CMS), base rate per DRG (DHCS) and DRG weights (CMS) Hospitalizations, LTC FQHC and RHC Rates Contains rates set by DHCS for services FQHC and RHC 12/19/2018 provided by FQHCs and RHCs inflated to 2019 Hospice per diem rates | Contains rates set by DHCS for hospice Hospice 9/28/2020 stays and services deflated to 2019 Nursing Facility Level A per diem rates Contains per diem rates set by DHCS per county for Freestanding Level A Nursing LTC, Hospice 8/1/2019 Facilities Distinct Part Nursing Contains per diem rates set by DHCS for | LTC, Hospice Facilities, Level B nursing facilities that are distinct parts 8/1/2019 of acute care hospitals UCLA Evaluation | Appendix A: Data Sources and Analytic Methods Source Home Health Services Rates 8/1/2020 deflated to 2019 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program Description Contains billing codes and reimbursement rates set by DHCS for procedure codes reimbursable by home health agencies March 2022 Applicable Service Categories Home health Home and Community- Based Services Rates 8/1/2020 deflated to 2019 Contains billing codes and reimbursement rates set by DHCS for the home and community-based services program Home and community- based services Community-Based Adult Services Rates 8/1/2020 deflated to 2019 Contains billing codes and reimbursement rates set by DHCS for community-based adult services Community-based adult services National Average Drug Acquisition Cost Contains per unit prices for drugs dispensed through an outpatient Outpatient medication (NADAC) File pharmacy setting based on the 12/30/2019 approximate price paid by pharmacies, calculated by CMS Clinical Laboratory Fee | Contains rates set by CMS for clinical lab | Laboratory Schedule services 12/30/2019 Therapy Rates Contains billing codes and Therapy 8/1/2020 reimbursement rates set by DHCS for deflated to 2019 physical, occupational, speech, and respiratory therapy Ambulatory Surgical Center (ASC) Fee Schedule January 2019 Contains billing codes and reimbursement rates set by CMS for facility fees for ASCs ED, Hospitalizations, Outpatient Facility Outpatient Prospective Payment System (OPPS) File October 2019 Contains billing codes and reimbursement rates set by CMS for facility fees for hospital outpatient departments ED, Hospitalizations, Outpatient Facility Payments were attributed based on available service and procedures codes included in each claim. A specific visit may have included a physician claim from the providers for their medical UCLA Evaluation | Appendix A: Data Sources and Analytic Methods NETSB Opy) UCLA Center for Health Policy Research Health Economics and Evaluation Research Program services and a facility claim for use of the facility and resources (e.g., medical/ surgical supplies and devices) where service was provided. The Medi-Cal Physician Fee Schedule contained monthly updated rates for all procedures that were reimbursable by Medi-Cal to providers and hospital outpatient departments. Each procedure code had multiple rates that varied based on provider type (e.g. physician, podiatrist, hospital outpatient department, ED, community clinic) and patient age. UCLA distinguished between these rates, but the paid amount for FFS still varied within the same procedure code, likely due to the directly negotiated rates between the providers and DHCS. For the purpose of HHP cost evaluation, UCLA used the procedure code with the most expensive rate when adequate information was lacking. UCLA also included a payment augmentation of 43.44% for claims for physician services provided in county and community hospital outpatient departments following DHCS guidelines. UCLA did not include any other reductions or augmentations that may have been applied by Medi-Cal due to limited information in claims data. Some procedures such as those performed by a qualified physical therapist in the home health or hospice setting did not have a fee in the Medi-Cal physician fee schedule but had fees in the Medi-Cal Provider Manual and UCLA used these fees when applicable. A number of claims lacked procedure codes but had a revenue code such as "Emergency Room- General" or "Freestanding Clinic- Clinic visit by member to RHC/FQHC". UCLA obtained documentation from DHCS that enabled identification of a price using outpatient revenue codes alone. CMS's Durable Medical Equipment (DME) Fee Schedule included billing codes that are reimbursable by Medi-Cal for DMEs such as hospital beds and accessories, oxygen and related respiratory equipment, and wheelchairs. Rates for other medical supplies such as needles, bandages, and diabetic test strips were found in DHCS's Medical Supplies Fee Schedules. FQHCs and RHCs consist of a parent organization with one or more clinic sites and are paid a bundled rate for all services during a visit. DHCS publishes FQHC and RHC Rates for each clinic within the parent organization. Payments for outpatient medication claims were calculated using the national drug acquisition cost (NADAC), which contains unit prices for drugs. UCLA calculated the drug cost by multiplying the unit price by the number of units seen on the claim. Drugs administered by physicians were priced using CMS's Average Sales Price Data (ASP) for Medicare Part B drugs. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods UCLA Center for Health Policy Research NTP Health Economics and Evaluation Research Program Facility fees were priced based on the ambulatory surgical center (ASC) fee schedule or the outpatient prospective payment system (OPPS) depending on whether the billing facility was an ASC or an outpatient department. Medi-Cal paid most LTC institutions such as nursing and intermediate care facilities for the developmentally disabled on a per-diem rate, while long-term care hospital stays were reimbursed via diagnosis related group (DRG) payments. Per diem rates for LTC facilities were obtained directly from DHCS's long-term care reimbursement webpage, and these rates varied by type of facility. Rates for hospice services were based on DHCS's hospice care site and hospice room and board rates were based on the Nursing Facility/ Intermediate Care facility fee schedule. UCLA lacked some variables in claims data that were needed to calculate some LTC and hospice payments, such as accommodation code which specifies different rates for each nursing facility depending on the type of program including the "nursing facility level B special treatment program for the mentally disordered" or "nursing facility level B rural swing bed program". In these cases, UCLA used the rates associated with accommodation code 1: "nursing facility level B regular', which were higher than other accommodation code rates. Hospitalizations are paid based on diagnosis related groups (DRGs), a bundled prospective payment methodology that is inclusive of all services provided during a hospitalization, except for physician services. Identification and pricing of DRGs varies by payers such as Medi-Cal and Medicare. In California, DHCS uses 3M's proprietary APR-DRG Core Grouping Software to assign DRGs and 3M's APR-DRG Pricing Calculator to calculate prices for Medi-Cal DRG hospitals. APR- DRGs have more specific DRGs for Medicaid populations such as pediatric patients and services such as labor and delivery, and incorporate four levels of illness severity. However, UCLA did not have access to this software and used 3M's publicly available CMS MS- DRG grouping software for the Medicare population, which includes Medicare-Severity DRGs (MS-DRGs) and their corresponding weights. MS-DRGs only include two levels of severity of illness, with complications or without complications. UCLA used this software to assign a DRG to each hospitalization based on procedure code, diagnosis, length of stay, payer type, patient discharge status, and patient age and gender. Although CMS uses the Inpatient Prospective Payment System to assign hospital prices based on the MS-DRGs, UCLA used available data and publicly available prices for DHCS's APR-DRG Pricing Calculator to calculate payments for each DRG. DHCS's APR-DRG Pricing Calculator used multiple hospital and patient-level variables to calculate the final payment for hospitals, and UCLA incorporated some of these variables into the estimated payment (such as patient age and hospital status of rural vs. urban) but could not incorporate other modifiers due to data limitations (such as other health coverage and whether or not the hospital was an NICU facility). UCLA Evaluation | Appendix A: Data Sources and Analytic Methods March 2022 UCLA Center for Health Policy Research Health Economics and Evaluation Research Program UCLA calculated the estimated payment by starting with the base rate from DHCS's APR-DRG Calculator, which was $12,832 for rural hospitals and $6,507 for urban hospitals. This base rate was multiplied by the weight assigned to each MS-DRG, which modified the base rate to account for resources needs for a given DRG. For example, more severe hospitalizations such as "Heart Transplant or Implant of Heart Assist System with major complications" had a high weight of 25.4241 but "Poisoning and Toxic Effects of Drugs without major complication" had a lower weight of 0.7502. This rate was further modified by one available policy adjuster, which increased the payment amount by patient age and was higher for those under 21 (1.25) than those 21 and older (1). Overall payment for a hospitalization was calculated by adding the estimated payments for physician specialist services that occurred during the hospitalization. When no fees were found for procedure codes in any payment data sources, UCLA used the most frequent paid amount seen in fee-for-service claims for the procedure code. These included procedures such as tattooing/ intradermal introduction of pigment to correct color defects of skin and excision of excessive skin. When outlying units of service were found on the claim, UCLA used the 90" percentile value of units for the procedure code rather than the observed units. All claims were included in a category of service and were assigned a price. For dual beneficiaries, Medi-Cal is the secondary payer (payer of last resort) and covers a portion of the costs of the service. However, UCLA lacked information on percentage of services paid for by Medi-Cal for dual managed care beneficiaries. Therefore, UCLA used Medi-Cal claims data to calculate payments for these dual beneficiaries using the same methodology as non-dual managed care beneficiaries. Dual beneficiaries made up 6% of the managed care population and 4% of the FFS population in 2019. For the purpose of evaluation, all payments were calculated using the 2019 fee schedules when available. In the absence of 2019 data, UCLA inflated or deflated payment amounts using the paid amounts for similar FFS claims in available data. Using the 2019 fees removed the impact of inflation and pricing changes in subsequent analyses. Comparison of Estimated Payments with Medi-Cal Paid Amounts UCLA examined the potential bias that may have resulted due to the methodology used to estimate payments by comparing the estimated FFS payments with Medi-Cal paid amounts in FFS claims. Exhibit 98 shows that the estimated FFS payments were 6% higher than paid amounts for all services. There was underlying variation by category of services. For example, ED payments were 9% higher while estimated payments for hospitalizations were 9% lower. UCLA Evaluation | Appendix A: Data Sources and Analytic Methods