FINAL REPORT STUDY OF THE IMPACT OF THE ACA IMPLEMENTATION IN KENTUCKY Prepared for: Foundation for a Healthy Kentucy Prepared by: State Health Access Data Assistance Center (SHADAC) University of Minnesota 2221 University Ave SE, Suite 345 Minneapolis, MN 55414 February 2017 This page intentionally left blank FINAL REPORT STUDY OF THE IMPACT OF THE ACA IMPLEMENTATION IN KENTUCKY February 2017 TABLE OF CONTENTS I. Introduction_______________________________________________ 1 Overview_________________________________________________ 1 Purpose and Layout of Current Report__________________________ 1 II. Study Findings: Data Update__________________________________ 3 Domain #1: Health Insurance Coverage__________________________ 3 Domain #2: Access_________________________________________ 10 Domain #3: Cost___________________________________________ 18 Domain #4: Quality________________________________________ 23 Domain #5: Health Outcomes________________________________ 29 III. Study Findings: Medicaid Enrollment and Services________________ 36 IV. Study Findings: Kentucky Health Reform Trend Analysis____________45 V. Study Conclusions_________________________________________ 50 VI. Appendix: Data Sources, Methods and Indicators________________53 VII. Endnotes________________________________________________ 57 I. INTRODUCTION Overview This report was produced by the State Health Access Data Assistance Center (SHADAC) at the University of Minnesota as part of our Study of the Impact of the Implementation of the Affordable Care Act (ACA) in Kentucky, funded by the Foundation for a Healthy Kentucky (Foundation). The study evaluates Kentucky’s performance in five domains: coverage, access, cost, quality, and health outcomes. This is the Final Report for the study and provides a comprehensive presentation of study findings, includ- ing both new study findings based on analysis of recently available data and a review of key findings from prior reports. The study’s initial duration was planned to last 34-months (March 2015 through January 2018); however, the duration of the study has been shortened to conclude in February 2017. Due to this, the study did not include previously planned qualitative components — focus groups with Medicaid bene- ficiaries and interviews with key stakeholders in Kentucky. As part of this project, SHADAC has used semi-annual and annual reports to document the impact of the ACA in Kentucky using a set of indicators selected in consultation with the Foundation and its ACA Impact Study Oversight Committee. These reports have tracked changes in the indicators through the duration of the study, and in certain cases they include comparisons of Kentucky metrics with the U.S. and other states. This report includes data obtained from analysis of a variety of federal and state data resources, including both survey and administrative data. A new section of this report presents an analysis of Medicaid administrative data previously reported in study quarterly snapshots. This section examines trends in enrollment of non-elderly adults in Kentucky’s traditional Medicaid program and ACA Medicaid expansion from 2014-2016. This report also presents new findings from the Kentucky Health Reform Survey (K-HRS), which was conducted in spring 2016 by SHADAC and the University of Cincinnati’s Institute for Policy Research. 1 1 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky INTRODUCTION Purpose and Layout of Current Report The main purpose of this report is to provide an overview of study findings through: 1) updates on our analysis of key study indicators in Kentucky, 2) a new analysis of trends in Medicaid enrollment and services in Kentucky from 2014-2016, 3) additional findings from our 2016 survey, with a trend comparison of pre-ACA estimates from the Kentucky Health Issues Poll versus post-ACA estimates from the K-HRS, and 4) a conclusion section that discusses overall findings on the impact of implementation of the ACA in Kentucky from this and other reports. 1) Data Update 3) Kentucky Health Reform Survey (K-HRS) Trend This section provides a data update to the key Analysis indicators that were introduced in the study’s The following section presents findings from a new baseline report and revisited in further annual analysis of the K-HRS, which compares pre-ACA and semi-annual reports. With the exception of estimates from the Kentucky Health Issues Poll certain indicators that were discontinued because (KHIP) against our study’s K-HRS, which was updated data were unavailable after the base- conducted in 2016. The K-HRS was designed in line report, this section presents on all indicators consultation with the Foundation and the study’s observed throughout the study, as well as a few Oversight Committee to address key study ques- indicators that were added later in the study. tions about the impacts of ACA implementation Many of the indicators in this report have been in Kentucky. Additionally, the K-HRS was based on updated with new data since the most-recent the methodology of the existing KHIP, allowing us semi-annual report; however, we also present to compare estimates across the two surveys. This data that haven’t been updated, to provide a com- analysis focuses on areas in which we designed prehensive review of our study findings. These the K-HRS to match the KHIP, to support a pre- and data include indicators from all five study domains post-ACA comparison: uninsurance and coverage (coverage, access, cost, quality, and health out- type, usual source of care and type of care facili- comes). All of the updated data in this section ty, dental coverage and time since last dental visit, include the time period since implementation of and self-reported health status. the ACA; some of the updates were available for 4) Study Conclusions 2015, while others were only available for 2014 The final section of this report includes a dis- at this time. As a baseline comparison, we use cussion of the conclusions of our study on the calendar year 2012 data for most indicators impacts of ACA implementation in Kentucky. We because it pre-dated the first ACA enrollment will revisit the key findings from this and other period; however, for certain indicators in which study reports—including prior semi-annual and 2012 data were not available, we use 2013 as a annual reports; quarterly snapshots; and special baseline. For selected indicators, we also compare reports on kids’ coverage, high-deductible health Kentucky to U.S. estimates and neighboring states insurance, and substance use—to present final for comparison (Arkansas, Indiana, Illinois, Ohio, study conclusions on changes to coverage, access, Missouri, Tennessee, Virginia and West Virginia).1 cost, quality and health outcomes in Kentucky 2) Kentucky Medicaid Enrollment and Services since implementation of the ACA. The next section of the report presents an anal- ysis of Medicaid enrollment and service utiliza- tion for non-elderly adults, using administrative data provided by the Kentucky Cabinet for Health and Family Services. The analysis examines quar- terly trends since the Commonwealth expanded its Medicaid program in January 2014 through the third quarter of 2016 (July-Sept.), which are the most recently available data. The indicators assessed in this section include enrollment in traditional and ACA-expansion Medicaid, and several services covered by Medicaid, including hepatitis C screenings, newborn births, dental services, breast and colorectal cancer screenings, substance use treatment services, and diabetes screenings. FEBRUARY 2017 | WWW.SHADAC.ORG 2 2 II. STUDY FINDINGS: DATA UPDATE DOMAIN #1: HEALTH INSURANCE COVERAGE Health insurance coverage is a critical component 2015. During this same time, Kentucky has seen of access to health care services. Having health stable rates of employer-sponsored insurance and insurance is associated with increased access increases in coverage through Medicare, Med- to needed medical care, better health care out- icaid/CHIP and the individual-market coverage. comes and improved health status.2 In this study, Despite these improvements, however, some the metrics used to monitor health insurance groups continue to experience higher rates of un- coverage within Kentucky over time include the insurance. For example, the uninsurance rate for distribution of health insurance coverage by type Kentucky’s Hispanic/Latino population is nearly (public, private and uninsured); rates of underin- quadruple the state’s overall uninsurance rate, surance; and the percentage of employers that the rate for young adults (ages 19-25) is almost offer health insurance coverage. Our data sources double the overall rate and the rate for the low- in this domain include federal surveys that provide income population is more than one and a half state-level estimates of health insurance coverage times the overall uninsurance rate. including the American Community Survey (ACS), the Medical Expenditure Panel Survey-Insurance Component (MEPS-IC) and the Current Population Since 2012, Kentucky’s Survey (CPS). Overall, health insurance coverage rates in Ken- uninsurance rate has tucky have improved substantially since 2012. dropped by more than half. The Commonwealth’s uninsurance rate has been cut by more than half, from 13.6% to 6.1% in COVERAGE MEASURES Uninsurance Declined, Medicaid/CHIP and Since 2012, Medicaid/CHIP coverage increased Individual-market Increased Significantly 6.4 percentage points to 19.8%, which likely re- Figure 1.1 presents the distribution of the pop- flects the Commonwealth’s implementation of the ulation by type of health insurance coverage ACA’s Medicaid expansion. As discussed later in (employer, individual, Medicaid/CHIP, Medicare this report, since Kentucky expanded its Medicaid and uninsured), for 2012 and 2015. In Figure 1.1 program in 2014, it has experienced increased (and in all figures in this report), statistically signif- enrollment in traditional Medicaid and enrollment icant differences are marked with asterisks. by non-elderly adults in the Medicaid expansion Since 2012, Kentucky’s uninsurance rate declined grew to more than 500,000 people by the third a statistically significant 7.5 percentage points, quarter of 2016. Kentucky also has experienced from 13.6% to 6.1% in 2015. Employer-sponsored a smaller, but still statistically significant, increase insurance (ESI) remains the largest source of cov- of 0.9 percentage points in individual-market cov- erage in Kentucky (50.0%), which has remained erage, from 4.4% in 2012 to 5.3% in 2015. This statistically unchanged since 2012. This stability in also is consistent with ACA provisions to support Kentucky’s ESI coverage rate represents departure individual-market coverage, such as the creation from the long-term trend of declining ESI coverage of marketplaces (e.g., kynect) where individuals in Kentucky and nationally.3 could shop for and purchase health insurance coverage, and financial assistance (i.e., advanced The remaining three types of insurance coverage premium tax credits) to make health insurance —Medicare, Medicaid/Children’s Health Insurance more affordable for people with moderate incomes Program (CHIP), and individual-market coverage— (139-400% of Federal Poverty Guidelines). each increased significantly from 2012 to 2015. During this time, Medicare coverage in Kentucky We also examine Kentucky’s uninsurance rate increased 1.5 percentage points to 18.9% in 2015, compared to the U.S. rate and to that of nearby which is consistent with the aging of Kentucky’s states. Figure 1.2 shows uninsurance rates for population.4 Kentucky’s bordering states, plus Arkansas, and the U.S. rate.5 3 3 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #1: HEALTH INSURANCE COVERAGE Employer Individual Medicaid/CHIP Medicare Uninsured FIGURE 1.1: 100% 6.1%* 9.4%* Insurance Coverage by 13.6% 14.7% Type for Kentucky and 18.9%* 16.5%* 80% 17.4% 15.4% the U.S., 2012 & 2015 (all ages) 19.8%* 15.7%* 13.4% 13.4% 60% 4.4% 5.3%* 5.2% 7.1%* 40% 51.2% 50.0% 51.4% 51.4% 20% 0% 2012 2015 2012 2015 KY U.S. *Difference is sta�s�cally significant across years (e.g., 2012 Kentucky vs. 2014 Kentucky) at the 95% level. Source: SHADAC analysis of the non-ins�tu�onal popula�on in the 2012 and 2015 ACS using the Public Use Microdata Sample Files. Insurance types are mutually exclusive. Since some people have mul�ple sources of coverage, a primary coverage hierarchy was used. Overall, Kentucky’s uninsurance rate is significantly OH, WV) have lower uninsurance rates (7.0% lower than the U.S. and neighboring states, or lower), while the three states that haven’t with the exceptions of Ohio and West Virginia. expanded their Medicaid programs (MO, TN, VA) Although it should be interpreted with caution have higher uninsurance rates (9.1% or higher). due to the small number of states, the uninsur- Additionally, the two states (AR, IN) that expanded ance rates of our group of comparison states their Medicaid programs through an “alternative” suggest a pattern related to whether and how approach, based on a Section 1115 waiver, have they implemented Medicaid expansions. The four uninsurance rates similar to the non-expansion states that implemented “traditional” Medicaid states (9.4% and 9.8%, respectively). expansions as intended by the ACA (KY, IL, 15% Kentucky United States Medicaid Expansion States Non-expansion States FIGURE 1.2: Uninsurance, Kentucky Compared to Neighboring States and 9.8%* 10.2%* U.S. Rate, 2015 (all ages) 10% 9.4%* 9.4%* 9.6%* 9.1%* 7.0%* 6.5% 6.1% 5.5% 5% 0% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant across states (e.g. Kentucky vs. Arkansas) at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2015 ACS using the Public Use Microdata Sample Files. Note: While Indiana is a Medicaid expansion state, the state did not expand its Medicaid program un�l 2015. FEBRUARY 2017 | WWW.SHADAC.ORG 4 4 Figure 1.3 shows both 2012 and 2015 uninsurance varied. Since 2012, Kentucky had the second- rates for Kentucky, the U.S. and comparison states. largest decline in its uninsurance rate (7.5 per- While the U.S. and all states experienced declines centage points), after only West Virginia (8.8 per- in their uninsurance rates, the size of those drops centage points). FIGURE 1.3: 20% 2012 2015 Uninsurance, Kentucky Compared to 16.3% Neighboring States and 14.7% 14.4% 14.2% 15% 13.6% 13.8% 13.8% U.S. Rate, 2012 & 2015 12.9% 12.3% (all ages) 11.5% 10% 9.8%* 10.2%* 9.4%* 9.4%* 9.6%* 9.1%* 7.0%* 5% 6.1%* 6.5%* 5.5%* 0% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant within the state (e.g. Arkansas 2012 es�mate vs. Arkansas 2015 es�mate) at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 ACS using the Public Use Microdata Sample Files. The following four figures present uninsured Additionally, uninsurance rates continue to differ rates by race/ethnicity, age, income category and by age in 2015. Despite a relatively large decline gender for Kentucky. of 15.3 percentage points, young adults (ages 19-25) continued to have the highest uninsurance Uninsurance Dropped Among Whites, rate (11.2%), while children continued to have African Americans the lowest rate (4.5%), also tied with adults ages Of the five race and ethnicity categories we 55-64. examined, three experienced statistically signifi- cant declines in uninsurance between 2012-2015, while two did not experience significant changes (see Figure 1.4). Since 2012, uninsurance rates Since 2012, Kentucky had dropped a statistically significant 11.8 percent- the second-largest decline age points among African Americans, to 5.5% in 2015; 7.4 percentage points among people of in its uninsurance rate other or multiple races, to 8.2%; and 7.3 percent- age points among whites, to 5.3%. The 24.2% (7.5 percentage points), uninsurance rate for the Hispanic/Latino pop- ulation was statistically unchanged since 2012, as after only West Virginia was the 12.5% uninsurance rate among Asians. (8.8 percentage points). Uninsurance Declined Significantly Among All Ages Among the Commonwealth’s non-elderly pop- ulation, all age groups experienced statistically significant declines in uninsurance. From 2012- 2015, the sizes of these declines varied from a decline of 1.9 percentage points for children (ages 0-18) to 12.6 percentage points for adults ages 26-44 (see Figure 1.5). 5 5 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #1: HEALTH INSURANCE COVERAGE 30% 28.7% 2012 2015 FIGURE 1.4: Uninsured Rates by 25% 24.2% Race/Ethnicity for Kentucky, 2012-2015 (all ages) 20% 17.3% 16.9% 15.5% 15% 12.6% 12.5% 10% 8.2%* 5.3%* 5.5%* 5% 0% Hispanic/La�no White African Asian Other/Mul�ple American/Black *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 ACS using the Public Use Microdata Sample Files. The race categories reported are mutually exclusive. Hispanic includes all people repor�ng this ethnicity; all other Race/Ethnicity categories exclude Hispanic. People repor�ng more than one race are included in Other/Mul�ple. 30% 2012 2015 FIGURE 1.5: 26.5% Uninsured Rates by Age 25% Category for Kentucky, 22.6% 2012-2015 (ages 0-64) 20% 16.9% 15% 11.2%* 11.4% 10.0%* 10% 6.4% 6.5%* 5% 4.5%* 4.5%* 0% 0-18 19-25 26-44 45-54 55-64 *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 ACS using the Public Use Microdata Sample Files. All Income Categories Saw Declines In of 15.5 percentage points. This decline in unin- Uninsurance surance is likely due to the Commonwealth’s Figure 1.6 presents uninsured rates by income Medicaid expansion, which expanded eligibility to category. We use the Federal Poverty Guidelines adults with incomes up to 138% of FPG. (FPG) and ACA income eligibility levels for our People with incomes between 139-400% of FPG income categories (see End Notes for 2015 FPG also experienced significant declines in uninsur- levels in dollars).6,7 Figure 1.6 shows the rela- ance between 2012-2015. Those with incomes tionship between income and uninsurance: as from 139-200% of FPG—a group eligible for incomes rise, uninsurance rates decline. People financial assistance to reduce the cost of premi- with incomes below 138% of FPG had the highest ums and cost-sharing subsidies to reduce out-of- uninsurance rates (9.4% in 2015). However, this pocket costs, such as deductibles—experienced group also had the largest declines in uninsurance a 9.2 percentage point decline in uninsurance, to from 2012-2015, experiencing a significant drop 8.3% in 2015. FEBRUARY 2017 | WWW.SHADAC.ORG 6 6 People with incomes from 201-400% of FPG—a contributed to this decline, as well. For example, group eligible for financial assistance to reduce although they are not eligible for financial premium costs but not eligible for cost-sharing assistance, higher income Kentuckians may still subsidies—experienced a smaller decline of 3.0 purchase coverage through the state health insur- percentage points, to 5.1% in 2015. ance marketplace (formerly called “kynect”). Additionally, Kentuckians with incomes of 401% Additionally, the ACA’s shared responsibility pro- of FPG or higher also saw a relatively small vision, also known as the “individual mandate,” but statistically significant decline in uninsur- requires individuals to maintain health insurance ance of 1.1 percentage points, to 1.9% in 2015 coverage or pay a tax penalty, which may have (from about 32,000 to 21,000). While this income prompted some higher income Kentuckians to group was not eligible for Medicaid expansion obtain health insurance. Figure 1.7 presents unin- or financial assistance for purchasing private sured rates by gender. health insurance, other ACA provisions may have FIGURE 1.6: 30% 2012 2015 Uninsured Rates by Income as Percent 25.0% 25% of Federal Poverty Guidelines for Kentucky, 20% 2012-2015 (all ages) 17.5% 15% 10% 9.4%* 8.3%* 8.1% 5.1%* 5% 2.9% 1.9%* 0% 0-138% 139-200% 201-400% 401%+ *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 ACS using the Public Use Microdata Sample Files. The family income uses the Health Insurance Unite (HIU), which may differ from the Census defini�on of a family. The HIU defines a family based on those individuals who would most likely be considered a “family unit” in determining eligibility for public or private coverage. This defini�on of a family is narrower than the one used by the Census Bureau. FIGURE 1.7: 20% 2012 2015 Uninsured Rates by Gender for Kentucky, 2012-2015 (all ages) 15% 14.8% 12.4% 10% 7.2%* 5.1%* 5% 0% Males Females *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 ACS using the Public Use Microdata Sample Files. 7 7 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #1: HEALTH INSURANCE COVERAGE Prior to implementation of the ACA in Ken- percentage points in the percentage of employ- tucky, males had higher rates of uninsurance ers that offer health insurance between 2012- than females. Since then, both groups have 2015. However, the availability of ESI differs based experienced similar declines in uninsurance (7.6 on the size of employers. Since 2012, there was percentage points among males and 7.4 points no significant change in the percentage of large among females), with males continuing to have employers (50 or more workers) offering health a higher uninsurance rate in 2015 (7.2% versus insurance, at 98.3% in 2015 (see Figure 1.8). 5.1%). In contrast, the percentage of small employers (less than 50 workers) offering coverage declined Drop In Employers Offering Coverage Driven 9.8 percentage points, from 36.4% in 2012 to By Small Firms 26.6% in 2015. Despite the decline in the percent- In the U.S., employer-sponsored insurance (ESI) age of employers offering health insurance, it is is the largest source of coverage for individuals. important to note that Kentucky has not experi- This is true in Kentucky as well, where 50.0% of enced a significant decline in the percentage of the population has employer-sponsored insur- individuals with coverage through an employ- ance (see Figure 1.1). Because of this, whether er. This is likely because large employers—which employers offer health insurance to their workers employ more Kentuckians than small employers— is an important factor in the coverage land- have continued to offer health insurance at rates scape. Looking at employer offer rates, there similar to before the ACA. has been a statistically significant decline of 6.6 100% 2012 2015 98.7% 98.3% FIGURE 1.8: Employer Offer Rates by Private Sector Employers 80% for Kentucky, 2012-2015 60% 54.4% 47.8%* 40% 36.4% 26.6%* 20% 0% All private sector Firms with less than Firms with 50 employers 50 employees or more employees *Difference is sta�s�cally significant at the 95% level. Source: 2012 and 2015 MEPS-IC. Es�mates are for percent of private sector establishments that offer coverage by firm size. Nearly 1 in 4 Kentuckians Remained SHADAC analysis of data from the CPS found that Underinsured 23.1% of Kentuckians were underinsured in 2015, Underinsurance is a measure of the affordability of which was not significantly different from 2013 health insurance and its effectiveness at insulating (see Figure 1.9).13 Although these data do not people from high out-of-pocket costs if they need support the concern raised by some policymak- health care. While there are various ways to define ers that health insurance and health care have underinsurance, for this study we consider fami- become less affordable since implementation of lies spending 10% or more of annual household the ACA, they do suggest that many Kentuckians income on health care (premiums, deductibles, continue to face high health-related costs relative and out-of-pocket expenses) during any given year to their incomes. to be underinsured. 9,10,11,12 FEBRUARY 2017 | WWW.SHADAC.ORG 8 8 FIGURE 1.9: 25% 2013 2015 23.4% 23.1% Underinsured Rate, Kentucky and the U.S., 20.8%* 19.8% 2013-2015 (all ages) 20% 15% 10% 5% 0% Kentucky United States Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2013 CPS. Underinsured is defined as the percentage of people whose family has spent 10% or more of their income in health care in the past year. 9 9 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #2: ACCESS DOMAIN #2: ACCESS The U.S. Institute of Medicine defines health Overall, the indicators paint a nuanced portrait care access as “the timely use of personal health of health care access since implementation of services to achieve the best health outcomes.”14 the ACA in Kentucky. The Commonwealth has Even among those with health insurance cov- seen some improvements, including significant erage, financial and non-financial access barriers increases in Kentuckians reporting a usual source can persist.15 We use 11 indicators to monitor of care and having a provider visit in the past health care access in this study—more indicators year, as well as a reduction in elderly Kentuckians than in any other study domain.16 For the access making changes to their medications due to cost. domain, we obtained data from the National In other cases, measures have remained stable. Health Interview Survey (NHIS), the National In some cases, this stability may be positive. For Survey on Drug Use and Health (NSDUH) and example, more than nine in ten Kentuckians con- the Behavioral Risk Factor Surveillance System tinue to find a doctor when needed. However, in (BRFSS). We include data for children under age other cases, gaps persist in the Commonwealth. 19 as well as non-elderly and elderly adults where For example, more than one in ten young adult data are available. Kentuckians have an unmet need for alcohol abuse treatment. ACCESS MEASURES Significantly More Kentuckians Reported a Between 2012-2015, the percentage of Kentuck- Usual Source of Care ians of all ages reporting a usual source of care Having a usual source of care is “a summary increased 7.4 percentage points, a statistically measure of adequate access to primary care”17 significant change from 82.3% to 89.7% (see and some studies have found it to be even more Figure 2.1). By breaking out age groups, we find important for health outcomes than having health this measure did not change significantly for insurance.18 The measure we use is from the NHIS, children (96.6% in 2015), but it did increase which asks, “Is there a place you usually go when significantly for non-elderly adults, 8.8 percent- you are sick or need advice about your health?” age points from 75.6% to 84.4% (data were not We also use responses to the follow up question: available for elderly adults). This pattern of an “what kind of place is it?” to make sure that emer- increase in usual source of care for non-elderly gency department visits were not considered to adults while children remained stable suggests be a usual source of care. the improvement in this indicator resulted from the ACA’s coverage expansions, which mostly targeted non-elderly adults. 100% 2012 2015 FIGURE 2.1: 93.8% 96.6% Usual Source of Care by 89.7%* Age Category, Kentucky, 82.3% 84.4%* 2012-2015 80% 75.6% 60% 40% 20% 0% All 0-18 19-64 *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. FEBRUARY 2017 | WWW.SHADAC.ORG 1010 Fewer Elderly Kentuckians Responded to Prior to the ACA, most Medicare Part D prescrip- Drug-cost Barriers tion drug plans had a coverage gap, also known Another indicator of access is changes in prescrip- as a “donut hole,” in which beneficiaries had to tion drug usage due to cost. This is a summary pay 100% of the cost of their medications out of measure that includes: asking the doctor for pocket.19 The ACA will gradually close that gap by cheaper medications, delaying refills, taking less 2020.20 Consistent with our findings, data from medication than prescribed, skipping dosages, the U.S. Centers for Medicare & Medicaid Services using alternative therapies and/or buying medi- found that Kentucky Medicare beneficiaries saved cations out of the country. This measure indicates an average of $1,108 in 2015 because of the ACA, whether people are making decisions based on which was slightly higher than the U.S. average of cost that may negatively affect their health. For $1,054.21 this indicator, estimates were not available for Kentuckians of all ages or children, but they were available for non-elderly and elderly adults. Since 2012, the percentage of non-elderly adults The ACA’s closing of the reporting making changes to prescription drugs due to cost did not change significantly (27.2% in Medicare Part D “donut 2015) (see Figure 2.2). However, the percentage of hole” may have improved elderly Kentuckians who reported changes in pre- scription drugs due to cost declined a statistically affordability of medications significant 12.6 percentage points, from 26.2% in 2012 to 13.6% in 2015. That improvement among for elderly Kentuckians. elderly Kentuckians may be a result of the ACA’s provisions to improve the affordability of prescrip- tion drugs specifically for Medicare beneficiaries. FIGURE 2.2: 40% 2012 2015 Skipping, Delaying, or Altering Prescription 34.1% Drug Use Due to Cost, Kentucky, 2012-2015 30% (ages 19-64 & 65+) 27.2% 26.2% 20% 13.6%* 10% 0% 19-64 65+ *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. More Kentuckians Visited Health in Kentuckians of all ages who reported visiting a Provider in the Past Year health care provider in the past year, from 73.8% Having a visit with a health care provider during to 78.7% (see Figure 2.3). We did not find signifi- the past year is another way to gauge access to cant changes among non-elderly or elderly adults, health care. For this measure, we include visits but there was a significant 7.0 percentage point to a general provider in the 12 months preceding increase in children who had a provider visit, from the survey. Between 2012-2015, there was a sta- 85.8% in 2012 to 92.8% in 2015. tistically significant 4.9 percentage point increase 1111 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #2: ACCESS 100% 2012 2015 FIGURE 2.3: 92.8%* 91.4% 91.3% Provider Visit in Past 85.8% Year by Age Category, 80% 78.7%* Kentucky, 2012-2015 73.8% 69.8% 65.6% 60% 40% 20% 0% All 0-18 19-64 65+ *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. One in Four Kentuckians Used Emergency elsewhere or that it could increase ED use by Department reducing the cost-barrier of uninsurance, we did We also examined the prevalence of visits to not find any significant changes. Figure 2.4 shows an emergency department (ED) within the past no statistically significant changes in the percent- year. According to the Agency for Healthcare age of Kentuckians of all ages reporting that they Research and Quality (AHRQ), “ED utilization re- visited an ED in the past year (25.5% in 2015). flects the greater health needs of the surrounding Similarly, we found no significant changes for any community and may provide the only readily avail- age subgroups—children, non-elderly adults and able care for individuals who cannot obtain care elderly adults. elsewhere.” For this measure, we also present comparisons Despite competing arguments that the ACA could between Kentucky and neighboring states. In 2015, reduce ED use by allowing people to obtain care Kentucky’s ED use rate of 25.5% was significantly 40% 2012 2015 FIGURE 2.4: Emergency Department 35.3% Visits in the Past Year by Age Category, Kentucky, 30.4% 30.2% 30% 2012-2015 27.8% 27.2% 25.5% 26.0% 23.3% 20% 10% 0% All 0-18 19-64 65+ *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. FEBRUARY 2017 | WWW.SHADAC.ORG 1212 higher than the U.S. rate of 18.3% and the rate were able to find a doctor when needed, which of Illinois (18.9%), but it was not statistically diff- was not statistically different from 2012 (see erent from the other seven comparison states (see Figure 2.7). We did not find statistically significant Figure 2.5). Additionally, like Kentucky, neither changes for children, non-elderly or elderly adults, the U.S. nor any comparison states experienced either. Although there was no significant increase, significant changes in the percentage of people the fact that more than nine in ten Kentuckians who used an ED in the prior year (see Figure 2.6). continue to find a provider when needed is positive. That stability is notable because it Over 9 in 10 Kentuckians Remained Able to suggests that concerns the ACA could worsen Find a Provider provider shortages may have not been realized in Being able to find a doctor when needed is an Kentucky. important component of health access. In 2015, 94.9% of Kentuckians of all ages said that they FIGURE 2.5: 30% Kentucky United States Medicaid Expansion States Non-expansion States Emergency Department Visits in the Past Year, 25.5% 25.1% Kentucky Compared to 25% 24.3% 24.0% Neighboring States and 22.3% 21.1% 20.5% U.S. Rate, 2015 (all ages) 20% 18.9%* 19.1% 18.3%* 15% 10% 5% 0% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant across states (e.g. Kentucky vs. Arkansas) at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. Note: While Indiana is a Medicaid expansion state, the state did not expand its Medicaid program un�l 2015. FIGURE 2.6: 40% 2012 2015 Emergency Department Visits in the Past Year, Kentucky Compared to 30.4% Neighboring States and 30% U.S. Rate, 2012 & 2015 26.5% (all ages) 24.3% 24.0% 22.5% 22.3% 25.5% 25.1% 20.5% 24.3% 19.0% 18.9% 19.1% 20% 21.5% 21.1% 20.0% 19.6% 18.3% 18.7% 17.6% 10% 0% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant within the state (e.g. Arkansas 2012 es�mate vs. Arkansas 2015 es�mate) at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. 1313 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #2: ACCESS 98.0% 2012 2015 99.6% FIGURE 2.7: 100% 94.8% 94.9% 97.6% 97.0% Found Doctor When 93.1% 92.7% Needed by Age Category, Kentucky, 2012-2015 80% 60% 40% 20% 0% All 0-18 19-64 65+ *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. Over 9 in 10 Kentuckians Found a Doctor Who that individuals may face trouble finding providers Accepts Their Insurance who accept their insurance as health insurers rely When seeking medical care, some people face more on narrow-network plans to contain costs; barriers with providers not accepting their so far, we have not found evidence that Kentuck- insurance coverage. From 2012-2015, there was ians are facing more problems finding providers in no significant change in Kentucky for the rate of their insurance networks. patients reporting that providers would accept their coverage, with 97.5% of Kentuckians of all Mental Health and Substance Use ages reporting they found a doctor who accepted People with mental illness and/or substance their insurance (see Figure 2.8). There were also use disorders often require specialty health no significant changes for children or non-elderly care services and may face unique barriers to adults (data were not available for elderly adults). treatment.22,23 This stability is particularly relevant to concerns 2012 2015 FIGURE 2.8: 100% 96.5% 97.5% 98.7% 96.7% 96.5% Told Provider Accepts 95.0% Insurance by Age Category, Kentucky, 80% 2012-2015 60% 40% 20% 0% All 0-18 19-64 *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. FEBRUARY 2017 | WWW.SHADAC.ORG 1414 In this section we present state-level data on law’s coverage expansions and provisions requir- prevalence of mental illness and unmet need for ing health insurance to cover treatment of mental treatment of substance use disorders. The U.S. illness were designed to enhance access to health Substance Abuse and Mental Health Services care for people with these conditions. Administration conducts an annual survey, the For more than a decade, Kentucky has experi- National Survey on Drug Use and Health (NSDUH), enced an increase in the number of drug overdose that collects information about the prevalence deaths, with many of these related to prescrip- of mental health conditions and substance use tion opioid painkillers and their chemical cousin, disorders, along with key indicators related to heroin.25 Similar to how it addresses mental access to services for these conditions. Because illness, the ACA was designed to address substance the sample size is limited, data from this survey use by increasing health insurance coverage along are pooled across two years to produce state-level with provisions to require health insurance to estimates (i.e., the 2012 estimate is actually cover treatment of substance use disorders. This pooled 2011-2012 data, and the 2014 estimate is section examines the percentage of Kentuckians actually pooled 2013-2014 data). who needed but did not receive treatment 1 in 5 Kentucky Adults Reported Having a for alcohol abuse and illicit drug abuse, which Mental Illness includes both illegal drugs (e.g., marijuana, The NSDUH provides estimates of the preva- cocaine, heroin) and misuse of prescription lence of any mental illness and serious mental medications (e.g., painkillers, stimulants). illness. Any mental illness is defined as “having any mental, behavioral, or emotional disorder in Young Adults Reported Greatest Unmet Need for Substance Use Treatment the past year that met DSM-IV criteria (excluding From 2012-2014, Kentucky did not experience developmental and substance use disorders).” statistically significant changes in the percent- Serious mental illness is defined as “any mental, age of people (ages 12+) who needed but did behavioral, or emotional disorder that substan- not receive treatment for alcohol abuse or illicit tially interfered with or limited one or more major drug abuse. Other research has shown increas- life activities.” In 2015, 5.1% of adult Kentuckians es in treatment of substance use disorders since (ages 18+) reported a serious mental illness, which Kentucky implemented the ACA, but it is possible was not statistically different from 2012. About that these increases have not been large enough one in five Kentuckians reported any mental illness to effect a large reduction in unmet need for (20.1%) in 2015, although this also was statistical- treatment.26 Additionally, it is likely that by using ly unchanged since 2012 (see Figure 2.9). While pooled 2013/2014 data, any potential effects in it would not be expected for the ACA to reduce 2014 may have been diluted in these estimates. the prevalence of mental illness in Kentucky, the FIGURE 2.9: 25% 2012 2014 Serious and Any Self- Reported Mental Illness, Kentucky, 2012-2014 20.1% 20% 19.5% (ages 18+) 15% 10% 4.7% 5.1% 5% 0% Serious Mental Illness Any Mental Illness *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the 2011/2012 and 2013/2014 Na�onal Survey on Drug Use and Health. 1515 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #2: ACCESS Because of the limitations of these 2013/2014 treatment, compared to the overall rate of 2.3% NSDUH data, future studies of substance use in (see Figure 2.10). Kentucky should examine these indicators using Also in 2015, 11.6% of young adults had an unmet data collected only since the implementation of need for alcohol abuse treatment, compared to the ACA (e.g., 2014/2015 pooled data). the overall rate of 5.5% (see Figure 2.11). When Despite the limitations of these data for under- these data are considered along with health standing the impacts of ACA implementation in insurance coverage rates, this suggests that many Kentucky, they provide important context around young adult Kentuckians who need treatment for the need for treatment of substance use disor- substance use disorders likely gained health insur- ders. For both indicators, the rates for young ance that would cover it; however, young adults adults (ages 19-26) are more than double the still have the highest rates of uninsurance in the overall rates (ages 12+): In 2015, 5.7% of young Commonwealth, posing a barrier to obtaining adults had an unmet need for illicit drug abuse needed treatment. 10% 2012 2014 FIGURE 2.10: Needed but Did Not Receive Illicit Drug 8% Abuse Treatment by Age Category, Kentucky, 7.0% 2012-2014 (ages 12+) 6% 5.7% 4% 3.3% 2.5% 2.7% 2.3% 2% 1.6% 1.6% 0% All (12+) 12-17 18-25 26+ *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the 2011/2012 and 2013/2014 Na�onal Survey on Drug Use and Health. 15% 2012 2014 FIGURE 2.11: Needed but Did Not Receive Alcohol Abuse 11.8% 11.6% Treatment by Age Category, Kentucky, 2012-2014 (ages 12+) 10% 5.1% 5.5% 5% 4.8% 4.3% 3.2% 2.4% 0% All (12+) 12-17 18-25 26+ *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the 2011/2012 and 2013/2014 Na�onal Survey on Drug Use and Health. FEBRUARY 2017 | WWW.SHADAC.ORG 1616 Dental Treatment Nearly 4 in 10 Kentuckians Reported Access to dental care is a concern because research No Dental Visits In Past Year has found that poor oral health is associated with From 2012-2014, there was no statistically signifi- other medical conditions, such as cardiovascular cant change in the percentage of Kentucky adults disease, diabetes and microbial infections.27 In this reporting they hadn’t visited a dentist in the past section, we present data from the BRFSS to track year (39.0%) (see Figure 2.12). In contrast with the percentage of adults who had no dental visit Kentucky’s stability in this measure, the U.S. rate in the past year. Recommendations on frequency worsened during the same time, with a statistical- of preventive dental visits vary, but some research ly significant increase in the percentage of adults has suggested annual visits for people at low risk reporting no dental visits in the past year. of dental disease and more frequent visits for Although dental health was not a key focus of the those at higher risk. ACA, it did include provisions that could support access to dental care, such as allowing dental policies to be sold through health insurance marketplaces, and allowing states to cover dental Research has found that services through Medicaid expansion benefits poor oral health is (an option that Kentucky adopted). Because data for this indicator are not currently available past associated with other 2014, it is difficult to determine whether the ACA has affected use of dental services; because this medical conditions. indicator examines use of dental services over the past year, any changes from 2014 may not appear until later estimates are available. FIGURE 2.12: 45% 2012 2014 No Dental Visit in the Last Year, Kentucky and 40% 39.7% 39.0% the U.S., 2012-2014 35.6%* (ages 18+) 35% 34.6% 30% 25% 20% 15% 10% 5% 0% Kentucky United States Source: SHADAC analysis of the 2012 & 2014 BRFSS. 1717 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #3: COST DOMAIN #3: COST Health care costs are a topic of concern for many NHIS, the MEPS-IC and the CPS. Our estimates stakeholders in Kentucky. Our study focuses in the cost domain cover all ages, except where primarily on issues of health care costs for fam- noted. ilies, such as the out-of-pocket costs they spend Overall, we found the Commonwealth experi- for health care and whether they have difficulties enced improvements in most of our measures paying medical bills. While families throughout of cost. Since implementation of the ACA, fewer the U.S. experience pressures from health care Kentuckians report trouble paying medical bills, costs, these are particularly a concern in Kentucky, and fewer report delaying or going without need- which in 2015 had a significantly lower median ed health care due to cost. Additionally, Kentucky household income than the U.S. ($45,215 versus hospitals have seen a decline in charity care and $55,775) and a higher rate of people in poverty self-pay charges for the uninsured—most likely (18.5% versus 14.7%).29 We also include a measure due to the declining uninsurance rate. However, of the impact of the ACA on Kentucky hospitals: Kentuckians’ median out-of-pocket spending for uncompensated care. Additionally, because of health care has remained stable, and we found the large role that employers play in Kentucky’s evidence that premiums for employer-sponsored health insurance landscape—covering half of Ken- insurance may be continuing their pre-ACA trends tuckians—we include measures of ESI premiums of growth. to examine whether and how these have grown. Data sources for the cost measures include the COST MEASURES Fewer Kentuckians Reported Trouble Paying In 2012, nearly half of Kentuckians of all ages Medical Bills (49.1%) reported that their families had trouble To measure the burden of health care costs on in- paying medical bills. By 2015, this dropped a statis- dividuals and families, we track the percentage of tically significant 11.5 percentage points, to 37.6% Kentuckians reporting trouble paying medical bills. of Kentuckians (see Figure 3.1). We also found This finding comes from SHADAC analysis of the statistically significant declines in trouble paying NHIS, which asks, “In the past 12 months did [you/ medical bills for children, dropping from 52.3% anyone in the family] have problems paying or to 40.9%, and non-elderly adults, dropping from were unable to pay any medical bills? Include bills 52.7% to 39.8%. However, we did not find a sig- for doctors, dentists, hospitals, therapists, medica- nificant decline among elderly adults, with 21.0% tion, equipment, nursing home, or home care.” reporting trouble paying medical bills in 2015. 60% 2012 2015 FIGURE 3.1: Trouble Paying Medical 52.3% 52.7% Bills by Age Category, 50% 49.1% Kentucky, 2012-2015 40.9%* 39.8%* 40% 37.6%* 30% 22.9% 21.0% 20% 10% 0% All 0-18 19-64 65+ *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. This es�mate reports the percentage of people who had trouble paying off medical bills in the last year or were currently paying off medical bills. FEBRUARY 2017 | WWW.SHADAC.ORG 1818 We also compare this metric to the U.S. and Ken- Kentucky’s rate was not significantly different from tucky’s neighboring states. Despite its statistically our other comparison states. However, between significant reduction in trouble paying medical 2012-2015, only Kentucky, the U.S. and Ohio saw bills, Kentucky’s rate of 37.6% remained signifi- significant declines in trouble paying medical bills; cantly higher than the U.S. rate of 27.9% in 2015, none of the other comparison states experienced as well as three neighboring states (IL, OH, VA) significant changes (see Figure 3.3). (see Figure 3.2). FIGURE 3.2: 50% Kentucky United States Medicaid Expansion States Non-expansion States Trouble Paying Medical Bills, Kentucky Compared 41.0% to Neighboring States and 40% 39.1% U.S. Rate, 2015 (all ages) 37.6% 36.0% 35.6% 31.4% 30.9%* 30% 27.9%* 28.7%* 27.6%* 20% 10% 0% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant across states (e.g. Kentucky vs. Arkansas) at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. This es�mate reports the percentage of people who had trouble paying off medical bills in the last year or were currently paying off medical bills. Note: While Indiana is a Medicaid expansion state, the state did not expand its Medicaid program un�l 2015. FIGURE 3.3: 60% 2012 2015 Trouble Paying Medical Bills, Kentucky Compared 49.1% to Neighboring States and 50% U.S. Rate, 2012 & 2015 41.2% 41.0% (all ages) 40.0% 39.1% 40% 38.5% 36.5% 40.3% 34.7% 37.6%* 31.3% 30.8% 36.0% 35.5% 35.6% 30% 31.4% 30.9% 27.9%* 28.7% 27.6%* 20% 10% 0% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant within the state (e.g. Arkansas 2012 es�mate vs. Arkansas 2015 es�mate) at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. This es�mate reports the percentage of people who had trouble paying off medical bills in the last year or were currently paying off medical bills. 1919 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #3: COST Delayed and Forgone Care Declined We also found significant declines in Kentuckians Significantly going without needed care due to cost. Between Delaying or not getting needed medical care can 2012-2015, the percentage of Kentuckians of all be a major impediment to good health outcomes, ages reporting forgone care dropped by approxi- and it can sometimes cause serious conditions mately half, from 10.0% to 4.9% — a statistically to go undetected or to get worse by being left significant decline of 5.1 percentage points (see untreated — resulting in worse health status and Figure 3.4). While estimates for children and higher treatment costs. Cost is a reason frequently elderly adults were not available, we found that cited for delaying or going without medical care. non-elderly adults also experienced a significant decline in forgone care, from 14.4% in 2012 to Between 2012-2015, the percentage of Kentuck- 7.3% in 2015 (a 7.1 percentage point decline). ians of all ages who reported delaying needed care due to cost dropped a statistically significant 5.2 percentage points, from 11.7% to 6.5% (see Figure 3.4). While estimates were not available for children, non-elderly adults experienced a signifi- cant 7.9 percentage point decline (from 16.9% to 9.0%), and elderly adults experienced a significant 3.0 percentage point decline (from 5.1% to 2.1%). 20% 2012 2015 FIGURE 3.4: Delayed or Went Without 16.9% Needed Care Due to Cost by Age Category, 15% 14.4% Kentucky, 2012-2015 11.7% 10.0% 10% 9.0%* 7.3%* 6.5%* 5% 4.9%* 0% Delayed care Went without care Delayed care Went without care All 19-64 *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 NHIS using the SHADAC Data Center. Hospital Charity Care and Self-Pay Between 2012 and the first year of ACA imple- Charges Declined 67% mentation in Kentucky, 2014, charity care and Before the ACA, hospitals often provided care self-pay charges dropped by more than half— to patients without insurance for which they from nearly $2.4 billion to $942 million. The received no payment or only partial payment, decline continued into 2015, dropping to $786 commonly called “uncompensated care.” By re- million.30 Overall, between our baseline year of ducing the number of people without health 2012 and 2015, these uncompensated charges insurance, the ACA was expected also to reduce dropped 67%. Because these data come from hospitals’ uncompensated care burden. As a proxy hospital data and not a statistical sample, no for uncompensated care, we use data on hospital significance testing was performed. charges for charity care or self-pay bills (see Figure 3.5). It is important to note that these data do not include bad debt from people with insurance, such as if a person with coverage does not pay cost sharing (e.g., deductible) owed to the hospital. FEBRUARY 2017 | WWW.SHADAC.ORG 2020 FIGURE 3.5: 3,000 Hospital Charity Care and $2,570 Self-Pay Charges in Dollars $2,396 (millions), Kentucky, 2,500 2012-2015 2,000 1,500 $942 1,000 $786 500 0 2012 2013 2014 2015 Source: SHADAC analysis of 2012 to 2015 data from the Kentucky Cabinet for Health and Family Services’ Kentucky Hospital Administra�ve Claims Data. Premiums for Employer-sponsored Single over the long-term prior to implementation of the Coverage Increased, Family Coverage ACA.31 Although ESI premiums for family coverage Statistically Unchanged have not changed significantly since 2012, the Figure 3.6 provides estimates of spending on significant increase in single-coverage premiums health insurance premiums. In 2015, the average suggests that ESI coverage may be continuing its annual single premium for private-sector em- pre-ACA trend of increasing costs. ployer-sponsored insurance was $5,984, a sta- tistically significant increase of $587 from 2012. To measure the impact of health care costs on The average family premium for employer-based individuals, we use a measure of median out-of- coverage was $16,622, but this was not signifi- pocket health care costs. This includes health cantly different from 2012 (see Figure 3.6). To insurance premiums and other money that indi- better understand these findings, it is important viduals spend on health care, such as deductibles, to consider that ESI premiums were increasing co-pays and co-insurance. FIGURE 3.6: 20,000 2012 2015 Average Premium per Private Sector Employee $16,622 in Dollars, Kentucky, $15,734 2012-2015 15,000 10,000 $5,984* $5,397 5,000 $0 Family Premiums Single Premiums *Difference is sta�s�cally significant at the 95% level. Source: 2012 and 2015 MEPS-IC. These es�mates represent the total annual premium cost. 2121 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #3: COST From 2012-2015, there was no statistically sig- money Kentuckians are paying for health care on nificant increase in annual median out-of-pocket average. However, these findings also run counter costs for Kentuckians (all ages) at $1,270 in 2015 to concerns raised by some stakeholders that indi- (see Figure 3.7). The stability in out-of-pocket costs viduals’ out-of-pocket spending on health care has since implementation of the ACA in Kentucky sug- increased substantially since implementation of gests that the expansions of coverage seen in the the ACA through high deductibles or other forms Commonwealth have not reduced the amount of of cost-sharing.32 2,000 2012 2015 FIGURE 3.7: Median Out-of-Pocket Spending in Dollars, Kentucky and U.S., $1,500* 2012-2015 1,500 $1,270 $1,200 $1,100 1,000 500 $0 Kentucky United States *Difference is sta�s�cally significant across years (e.g., 2012 Kentucky vs. 2015 Kentucky) at the 95% level. Source: Es�mates were based on SHADAC’s analysis of the civilian non-ins�tu�onal popula�on in the 2012 and 2015 CPS. Includes spending on premiums. FEBRUARY 2017 | WWW.SHADAC.ORG 2222 DOMAIN #4: QUALITY Improving the quality of health care in the U.S. While some of the indicators are available through was a key goal of the ACA. There are a number 2015, some are only available through 2014. of ways in which the law is focused on improving Additionally, 2012 data were not available for all the quality of care, including avoiding prevent- of the measures; in these cases, we use 2013 as able hospitalizations, increasing the utilization of our baseline. preventive care, and encouraging recommended While we found some improvements in mea- health practices, such as breastfeeding for infants. sures of quality, these have been more limited We include several metrics that relate to quality than in the domains of coverage, access and cost. of care, focusing both on hospital quality and For example, rates of newborn breastfeeding in aggregate measures of preventive care utilization. Kentucky have increased since 2012, and more For the quality domain, our data sources include Kentuckians are reporting receiving recom- the BRFSS, the Youth Risk Behavior Surveillance mended colorectal cancer screenings. However, System (YRBSS), the Healthcare Cost and Utiliza- most of our measures remained stable—such tion Project (HCUP) and vital statistics systems. as low birth weight, cholesterol awareness and Data in this domain cover all ages except where unprotected sex among high school students— noted. and one measure worsened (diabetes short-term admissions). QUALITY MEASURES Potentially Preventable Hospital Admissions Death Rate in Low Mortality Admissions According to the Agency for Healthcare Research Stable and Quality (AHRQ), “one area where higher Figure 4.2 shows the number of deaths per 1,000 quality and lower costs coincide is potentially patients of all ages who were hospitalized for con- preventable hospital admissions—inpatient stays ditions that typically do not result in mortality. All that could be prevented with high-quality primary cases treated in hospitals are classified according and preventive care. High rates of these potential- to groups called diagnosis-related groups (DRGs). ly preventable hospital admissions identify areas DRGs are used to help determine how much a where possible improvements in the health care hospital gets paid for its services, adjusted for delivery system could be made to enhance patient severity and other factors.35 Many DRGs (e.g., outcomes and decrease costs.”33 In this study, we eye disorders, childbirth, knee procedures) are look at potentially avoidable hospitalizations for associated with low mortality rates and are used three chronic conditions: diabetes, hypertension, as one indicator of hospital quality; hospitals and asthma. The data for these come from AHRQ’s with high mortality rates associated with these HCUP dataset.34 low mortality DRGs may provide lower quality care.36 Figure 4.1 presents data on potentially prevent- able hospitalizations as the number of hospitaliza- The mortality rate presented here is risk-adjusted tions per 100,000 adults. For diabetes short-term to take into account patients’ prior health status. complications, approximately 93 out of 100,000 Figure 4.2 shows that in 2014, Kentucky’s mortal- adults were admitted in 2014, an increase from ity rate for “low-mortality DRGs” was 0.325 per 84 in 2012. In contrast, both hypertension and 1,000, only slightly lower than the state’s 2012 asthma-related admissions decreased over the baseline rate of 0.330 per 1,000. However, these same period, from approximately 68 to 58 per rates may vary from year to year (the 2013 rate 100,000 for hypertension and 58 to 44 per was 0.233 per 1,000), so future research may be 100,000 for asthma. Although these data suggest needed to follow these mortality trends in follow- Kentucky may be experiencing some improve- ing years. Because these data come from hospital ments in quality of health care, future research admissions records and not a statistical sample, may be needed to determine whether these no significance testing was performed. continue past the first year of ACA implemen- tation. Because these data come from hospital admissions records and not a statistical sample, no significance testing was performed. 2323 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #4: QUALITY 100 2012 2014 FIGURE 4.1: 92.8 Diabetes (ages 18+), 84.3 Hypertension (ages 18+) 80 and Asthma (ages 18-39) Hospital Admissions 67.9 (per 100,000), Kentucky 58.4 2012-2014 60 57.9 44.0 40 20 0 Diabetes short-term Hypertension Asthma admissions admissions admissions Source: SHADAC analysis of 2012 and 2014 HCUP data. These es�mates report the Diabetes Short-term Complica�ons Admission Rate for adults (PQI 1), the Hypertension Admission Rate for adults (PQI 7), and the Asthma in Younger Adults Admission Rate (PQI 15). 0.34 FIGURE 4.2: Mortality Rate in Low Mortality DRGs (per 1,000 cases), Kentucky, 0.335 2012-2014 (all ages) 0.33 0.33 0.325 0.325 0.32 2012 2014 SHADAC analysis of 2012 and 2014 HCUP data. The es�mate reports the “Dying in the Hospital while Ge�ng Care for a Condi�on that Rarely Results in Death Rate” cases. Breastfeeding Grew to More Than Due to changes in the availability of state-level Two-thirds of Births breastfeeding rates, we have revised this indicator Because of the positive effects of breastfeeding to track the percentage of infants breast-fed upon on the health of the mother and baby,37 the U.S. discharge from the hospital. government has set national goals to increase Figure 4.3 shows that the percentage of Kentucky the proportion of infants who are breastfed, with infants who were reported as being breastfed at a goal (by 2020) of 81.9% ever being breastfed, discharge from the hospital has increased 5.9 per- 60.6% being breastfed at 6 months, and 34.1% centage points to 68.7% in 2015. Because these being breastfed at 1 year of age.38 The ACA also data come from birth records and not a statistical included provisions aimed at supporting mothers sample, no significance testing was performed. in efforts to breast feed, including requiring that health insurance plans cover lactation counseling and the cost of a breast pump. FEBRUARY 2017 | WWW.SHADAC.ORG 2424 FIGURE 4.3: 75% Breastfeeding Initiation Rates, Kentucky, 2012-2015 70% 68.7% (newborn infants) 66.4% 63.9% 65% 62.8% 60% 55% 50% 2012 2013 2014 2015 Source: Data provided by the Kentucky Department for Public Health. Note: Data are s�ll preliminary for 2014 and 2015. Racial Disparities Continued in Low Birth likely that any effects on low birth weight would Weight lag other improvements, such as reduced uninsur- According to the Centers for Disease Control and ance rates. Because these data come from birth Prevention (CDC), low birth weight (defined as less records and not a statistical sample, no signifi- than 5 pounds, 8 ounces) is “the single most im- cance testing was performed. portant factor affecting neonatal mortality and a significant determinant of post-neonatal mortal- Colorectal Screenings Increased, Cholesterol ity. Low birth weight infants who survive are at Awareness Stayed Stable increased risk for health problems ranging from Preventive care utilization for adults also is import- neurodevelopmental disabilities to respiratory ant because early, lower-cost health interventions disorders.”39 The U.S. Department of Health and may prevent or reduce the severity of higher-cost, Human Services has set a national target to reduce severe health problems. Our study tracks two low birth weight to 7.8% of live births by 2020, examples of preventive care: cholesterol aware- (the national rate was 8.0% in 2014).40 ness and colorectal cancer screening. Although low birth weight in Kentucky has varied The cholesterol awareness metric reports the per- slightly from year to year—beginning at 8.7% in centage of adults (ages 18+) who had their blood 2012, increasing to 9.0% in 2013, and dropping cholesterol checked within the past five years. to 8.8% in 2014—it returned in 2015 to the same In 2015, 76.5% of Kentucky adults reported having rate as 2012, of 8.7% (see Figure 4.4). Despite the had this test, which was not significantly different relative steadiness in low birth weight, the data from 2013 (see Figure 4.5). show consistent disparities by race/ethnicity. In 2015, non-Hispanic whites had a rate of 8.2% low birth weight. By comparison, non-Hispanic blacks had higher rate of 13.9%, while Hispanics had a Despite the relative lower rate of 6.5%. steadiness in low birth The ACA included provisions that could help to address the issue of low birth weight, such as weight, the data show requirements for individual-market health insur- consistent disparities by ance to cover pregnancy-related care, which is intended to improve access to prenatal care by race/ethnicity. making it more affordable for pregnant women. However, because prenatal care occurs over a period of several months during gestation, it’s 2525 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #4: QUALITY 15% 2012 2015 FIGURE 4.4: 13.8% 13.9% Low Birth Weight for Births by Race/Ethnicity, Kentucky, 2012-2015 (all births) 10% 8.7% 8.7% 8.2% 8.2% 6.8% 6.5% 5% 0% All Non-Hispanic Non-Hispanic Hispanic White Black Source 2012 and 2015 Na�onal Vital Sta�s�cs Reports, Supplemental Tables (Table 1-9). Percent of low birth weight births (<5 pounds 8 ounces). 80% 2013 2015 FIGURE 4.5: Cholesterol Awareness, Kentucky and U.S., 78% 77.9% 2013-2015 (ages 18+) 77.0% 77.1% 76.5% 76% 74% 72% 70% Kentucky United States *Difference is sta�s�cally significant across years (e.g., 2013 Kentucky vs. 2015 Kentucky) at the 95% level. Source: Es�mates are based on SHADAC analysis of the BRFSS survey data of the percentage of adults who have had their blood cholesterol checked within the last 5 years. Figure 4.6 shows Kentucky’s performance on this Figure 4.7 shows Kentucky, the U.S. and compar- indicator compared to the U.S. and comparison ison states from 2013-2015. Of these, only the states. In 2015, Kentucky’s rate of cholesterol U.S. and three states saw significant increases in awareness was not significantly different from cholesterol awareness (AR, IL, WV). the U.S. rate or half of our comparison states, but The colorectal cancer screening metric reports the it was significantly higher than one (IN) and lower percentage of adults ages 50 to 75 who have met than three (WV, TN, VA). guidelines for receiving colorectal cancer screen- ing within certain time periods.41 FEBRUARY 2017 | WWW.SHADAC.ORG 2626 FIGURE 4.6: 90% Kentucky United States Medicaid Expansion States Non-expansion States Cholesterol Awareness, Kentucky Compared to Neighboring States and U.S. Rate, 2015 (ages 18+) 85% 82.3%* 81.1%* 80.2%* 80% 77.9% 78.2% 77.9% 76.5% 76.8% 75.6% 75% 73.5%* 70% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant across states (e.g. Kentucky vs. Arkansas) at the 95% level. Source: Es�mates are based on SHADAC analysis of the 2015 BRFSS survey data of the percentage of adults who have had their blood cholesterol checked within the last 5 years. Note: While Indiana is a Medicaid expansion state, the state did not expand its Medicaid program un�l 2015. FIGURE 4.7: 90% 2013 2015 Cholesterol Awareness, Kentucky Compared to Neighboring States and U.S. Rate, 2013 & 2015 85% (ages 18+) 82.3%* 81.1% 81.1% 80% 77.9%* 78.2%* 78.2% 80.2% 80.2% 77.0% 76.8%* 78.7% 77.9% 75.6% 77.1% 74.5% 75% 76.5% 74.0% 73.9% 73.5% 72.3% 70% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant within the state (e.g. Arkansas 2013 es�mate vs. Arkansas 2015 es�mate) at the 95% level. Source: Es�mates based on SHADAC analysis of the 2013 and 2015 BRFSS survey data of the percentage of adults who have had their blood cholesterol checked within the last 5 years. In 2014, 66.8% of respondents reported having U.S. and neighboring states, finding that only had a colorectal cancer screening, a statistically Kentucky experienced a statistically significant im- significant increase of 4.4 percentage points since provement, and that this improvement brought 2012 (see Figure 4.8). While this increase contin- Kentucky on par with the U.S.44 ues a longer-term trend of improving colorectal cancer screening rates in Kentucky, other data Unprotected Sex Among High School Students showing increased colorectal cancer screenings in Remained Statistically Unchanged Kentucky’s Medicaid program since 2014 suggest The 2015 YRBSS provides estimates of unprotect- that the Commonwealth’s ACA Medicaid expan- ed sex (i.e., no use of any birth control) among sion has also played a role.42,43 high school students who reported that they were sexually active. This indicator was identified by the In a prior report, we presented data comparing Foundation as an important part of the study’s Kentucky’s performance on this indicator to the population health and prevention measures. 2727 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #4: QUALITY 70% 2012 2014 FIGURE 4.8: Colorectal Cancer 66.8%* Screenings, Kentucky 65.5% 65.9% and U.S., 2012-2014 65% (ages 50-75) 62.4% 60% 55% 50% Kentucky United States *Difference is sta�s�cally significant across years (e.g., 2012 Kentucky vs. 2014 Kentucky) at the 95% level. Source: Es�mates are based on SHADAC analysis of the 2012 and 2014 BRFSS survey data of the percentage of adults who met U.S. Preven�ve Services Task Force colorectal cancer screening recommenda�ons. Although the ACA includes certain provisions Figure 4.9 provides a snapshot of the 2013 base- designed to increase access to contraception— line data and updated 2015 data for Kentucky. such as requiring private health insurance plans Among high school students, 14.5% reported en- to cover birth control prescribed by a health gaging in unprotected sex during their last sexual care provider with no cost-sharing—there are intercourse in 2015, which was not statistically many factors that influence adolescents’ use of different from 2013. Female high school students contraception,45 so the law is not expected to reported higher rates of unprotected sex (17.5%) have a strong effect on use of birth control by high compared to males (11.6%) in 2015, although school students. neither of these were statistically different than in 2013. 20% 2013 2015 FIGURE 4.9: 18.6% Unprotected Sex Among 17.5% High School Students, Kentucky, 2013-2015 15% (grades 9-12) 11.2% 11.6% 10% 5% 0% Female Male *Difference is sta�s�cally significant at the 95% level. Source: SHADAC analysis of 2013 and 2015 Youth Risk Behavior Surveillance System data. The es�mate reports the percentage of high school students who did not use any method to prevent pregnancy during their last sexual intercourse. FEBRUARY 2017 | WWW.SHADAC.ORG 2828 DOMAIN #5: HEALTH OUTCOMES An ultimate goal of the improvements in the In this study, we use five measures of health out- prior study domains—coverage, access, cost comes: obesity rates, cigarette use, self-reported and quality—is improved health for Kentucky’s health status, prevalence of chronic disease and population. Health outcomes are determined by a premature death. These measures are based on combination of factors including genetics, behav- data from the BRFSS, YRBSS and CDC vital statistics. iors, environmental exposures, social factors and Our analysis of early impacts of the ACA on health health care services and policies.46 Although these outcomes did not find large changes. For example, determinants are complex, the outcome measures although adult cigarette smoking rates declined, included in this report are at least partially influ- adolescent rates remained stable; and while adult enced by access to high quality care. While health obesity rates increased, adolescent rates again outcomes are slow to change at a state or national remained stable. Overall, indicators remained level, monitoring them is key to understanding the mostly unchanged since 2012. impacts of efforts to improve health in Kentucky. HEALTH OUTCOMES MEASURES Adult Obesity Grew Significantly, 18.5% in 2015 was not significantly different from Adolescent Stable 2013 (the baseline year for this measure, since Obesity is associated with a range of chronic 2012 data weren’t available); however, this could conditions, including heart disease, high blood be due in part to a shorter comparison timeframe pressure, and diabetes.47 Obesity is prevalent (i.e., using 2013 rather than 2012 data). Despite among adults and children in the U.S., though the ACA’s aims to improve people’s health, it is rates among children have stabilized in recent not unexpected that Kentucky’s obesity rates years.48 would remain stable or increase since implemen- Figure 5.1 shows estimates of the prevalence tation of the law. While access to health care of obesity among adult Kentuckians (ages 18+) services may serve an important role in stopping from 2012-2015. During this time period, obesity and reversing the rise of obesity in Kentucky, it among adult Kentuckians increased a statistically is a complex problem that has taken decades to significant 3.3 percentage points, to 34.6% in reach today’s levels of prevalence,49 and halting or 2015. Figure 5.2 shows obesity among adolescent reversing that trend may take years. Kentuckians. Kentucky’s adolescent obesity rate of FIGURE 5.1: 40% 2012 2015 Self-Reported Obesity, Kentucky and U.S., 2012-2015 (ages 18+) 35% 34.6%* 32.5%* 31.3% 31.3% 30% 25% 20% Kentucky United States *Difference is sta�s�cally significant across years (e.g., 2012 Kentucky vs. 2015 Kentucky) at the 95% level. Source: The Kentucky es�mates are based on SHADAC analysis of 2012 and 2015 BRFSS survey data. The es�mates report the percentage of adults with a Body Mass Index of over 30. 2929 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #5: HEALTH OUTCOMES 20% 2013 2015 FIGURE 5.2: 18.5% Self-Reported Obesity, 18.0% Kentucky and U.S., 2013-2015 (high school 15% students, grades 9-12) 13.7% 13.9% 10% 5% 0% Kentucky United States *Difference is sta�s�cally significant across years (e.g., 2012 Kentucky vs. 2015 Kentucky) at the 95% level. Source: SHADAC analysis of 2013 and 2015 Youth Risk Behavior Surveillance System data. Adult Cigarette Use Declined, Adolescent the law allows insurers to charge higher premiums Stable to people who use tobacco, and it also requires According to the CDC, smoking is associated with health insurance to cover certain recommended numerous conditions, including cancer, heart preventive health care services, including tobacco- disease, and birth defects, and it causes nearly one cessation benefits, with no cost-sharing.53 in five deaths in the U.S. each year.50 Kentuckians Figure 5.3 shows estimates of the prevalence of are particularly at risk because of the Common- cigarette use among adults in Kentucky. Since wealth’s high smoking rates. In 2015, Kentucky 2012, Kentucky’s adult smoking rate declined a had the highest adult smoking rate in the U.S.,51 statistically significant 2.3 percentage points, to and the second-highest adolescent smoking 26.0% in 2015. rate, after West Virginia.52 The ACA incorporated certain policies to discourage tobacco use and to provide people resources to quit. For example, 30% 2012 2015 FIGURE 5.3: 28.3% Cigarette Use, Kentucky 26.0%* and the U.S., 2012-2015, 25% (ages 18+) 20% 18.1% 15.1%* 15% 10% 5% 0% Kentucky United States *Difference is sta�s�cally significant across years (e.g., 2012 Kentucky vs. 2015 Kentucky) at the 95% level. Source: SHADAC analysis of 2012 and 2015 Behavioral Risk Factor Surveillance System data. The es�mate reports the percentage of adults who have smoked 100 or more cigare�es in their life�me and who currently smoke some days or every day. FEBRUARY 2017 | WWW.SHADAC.ORG 3030 Figure 5.4 shows cigarette use among Kentucky To better understand the role of the ACA on adolescents. Between 2013 and 2015, Kentucky’s tobacco use in Kentucky, additional research may adolescent smoking rate remained statistically be needed into the reasons people quit smoking, unchanged, at 16.9% in 2015; however, like and if people are quitting tobacco altogether or the obesity indicator, this could be due in part switching from cigarettes to e-cigarettes. to a shorter comparison timeframe (i.e., using 2013 rather than 2012 data). While it may be that the ACA’s tobacco policies played a role in reduced cigarette use among Kentucky adults, it To better understand the is important to consider other circumstances that role of the ACA on tobacco could also have contributed, such as the rise of electronic cigarettes. The 2016 KHIP found that use in Kentucky, additional 25% of adult Kentuckians have used electronic cigarettes,54 and other national research has found research may be needed that use of e-cigarettes has risen during the past several years, and this may be contributing to into use of e-cigarettes. declines in rates of cigarette smoking.55,56 FIGURE 5.4: 20% 2013 2015 Cigarette Use, Kentucky 17.9% and the U.S., 2013-2015, 16.9% (high school students, 15.7% grades 9-12) 15% 10.8%* 10% 5% 0% Kentucky United States *Difference is sta�s�cally significant across years (e.g., 2012 Kentucky vs. 2015 Kentucky) at the 95% level. Source: SHADAC analysis of 2013 and 2015 Youth Risk Behavior Surveillance System data. The es�mate reports the percentage of high school students who currently smoked cigare�es, on at least 1 day during the 30 days before the survey. Nearly 1 in 4 Kentucky Adults Report We also compare Kentucky’s self-reported health Fair or Poor Health status to the U.S. and neighboring states, pre- Research has consistently shown self-reported sented in Figure 5.5. In 2015, the percentage of health status from surveys to be a valid predictor Kentucky adults reporting poor or fair health was of mortality.57 The BRFSS survey asks, “Would you significantly higher than the U.S. rate of 17.5%. say that in general your health is excellent, very Kentucky’s rate of poor or fair health also was good, good, fair, or poor?” In 2015, 22.2% of higher than most comparison states (IL, IN, OH, adults surveyed in Kentucky reported poor or fair MO, VA); only West Virginia had a significantly health, which was not significantly different from higher rate of poor or fair health than Kentucky 2012. (see Figure 5.6). 3131 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #5: HEALTH OUTCOMES 25% 23.9% 2012 2015 FIGURE 5.5: 22.2% Poor/Fair Health, Kentucky and the U.S., 20% 2012-2015 (ages 18+) 17.9% 17.5% 15% 10% 5% 0% Kentucky United States *Difference is sta�s�cally significant across years (e.g., 2012 Kentucky vs. 2015 Kentucky) at the 95% level. Source: SHADAC analysis of 2012 and 2015 Behavioral Risk Factor Surveillance System data. The es�mate reports the percentage of adults who report being in poor or fair health. 30% Kentucky United States Medicaid Expansion States Non-expansion States FIGURE 5.6: Poor/Fair Health, 25.9%* Kentucky Compared to 25% 23.8% Neighboring States and 22.2% U.S. Rate, 2015 (ages 18+) 21.1% 20% 18.8%* 17.5%* 17.8%* 16.4%* 16.5%* 15.2%* 15% 10% 5% 0% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant across states (e.g. Kentucky vs. Arkansas) at the 95% level. Source: SHADAC analysis of 2012 and 2015 Behavioral Risk Factor Surveillance System data. The es�mate reports the percentage of adults who report being in poor or fair health. Figure 5.7 shows the rates of poor or fair health The stability of Kentucky’s rate of poor or fair in 2012 and 2015 for Kentucky, the U.S. and health—and those of most of its neighbors— comparison states. Although the U.S. experienced suggests that the ACA’s coverage expansions have a relatively small but statistically significant 0.4 not yet driven significant improvements in overall percentage point decline in poor or fair health, health status; however, changes in individuals’ only two of eight comparison states saw signifi- overall health could take years to develop, so this cant declines (OH and VA). should be a key measure for future research into the impacts of the ACA. FEBRUARY 2017 | WWW.SHADAC.ORG 3232 FIGURE 5.7: 20% 2012 2015 Poor/Fair Health, 25.9% Kentucky Compared to 23.8% 23.9% Neighboring States and 25.2% 21.1% U.S. Rate, 2012-2015 15% 23.7% (ages 18+) 20.0% 22.2% 18.7% 17.9% 18.3% 21.1% 17.4% 17.5% 18.8% 17.5%* 17.8% 10% 16.4% 16.5%* 15.2%* 5% 0% KY U.S. AR IL IN OH WV MO TN VA *Difference is sta�s�cally significant within the state (e.g. Arkansas 2012 es�mate vs. Arkansas 2015 es�mate) at the 95% level. Source: SHADAC analysis of 2012 and 2015 Behavioral Risk Factor Surveillance System data. The es�mate reports the percentage of adults who report being in poor or fair health. More than 1 in 4 Kentucky Kentucky reported having one or more of these Adults Reported Chronic Conditions conditions, which was not significantly different Chronic diseases result in large cost and social bur- from 2012. However, in an earlier report, we found dens. The CDC estimates that chronic conditions that the 2014 rate of 29.1% was significantly are the cause of seven of every 10 deaths in the higher than the 2012 rate of 26.8%.59 Because U.S., and that the cost of treating these conditions the difference between 2014 and 2015 was not consumes 86% of U.S. health expenditures each statistically significant, this may be due to the year.58 In this study, we estimate the burden of inherent level of uncertainty in survey estimates. chronic disease using BRFSS data; our estimates When they become available, 2016 estimates may include the percentage of adults reporting one or help to clarify whether self-reported prevalence more of the following conditions: diabetes, cardio- of chronic disease has increased significantly vascular disease, heart attack, stroke, and asthma. since implementation of the ACA. Figure 5.8 shows that in 2015, 28.7% of adults in FIGURE 5.8: 35% 2012 2015 Chronic Disease Prevalence, Kentucky and 30% 28.7% the U.S., 2012-2015 26.8% (ages 18+) 25% 21.0% 20.8% 20% 15% 10% 5% 0% Kentucky United States *Difference is sta�s�cally significant across years (e.g., 2012 Kentucky vs. 2015 Kentucky) at the 95% level. Source: The Kentucky es�mates are based on SHADAC analysis of 2012 and 2015 BRFSS survey data. The es�mates show the percentage of adults who report having one or more of the following chronic condi�ons: diabetes, cardiovascular disease, heart a�ack, stroke and asthma. 3333 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky DOMAIN #5: HEALTH OUTCOMES Although the ACA aimed to eventually improve In other words, if life expectancy is 75 years, and a the health of Americans by expanding access to person dies at age 50, she loses 25 potential years. health care, improvements in health outcomes By adding all the years of life lost to early death, would likely take years. It would not necessarily we estimate the number of YPLL for Kentucky. be surprising for self-reported rates of chronic In 2015, there were a total of 9,206 YPLL due to disease to increase in the years soon after imple- premature death per 100,000 people in Kentucky, mentation of the ACA, as more people gain health slightly higher than the rate of 8,865 in 2012 (see insurance and may learn of previously undiag- Figure 5.9). During that same time period, the nosed health conditions. For example, a person U.S. rate of YPLL also increased, from 6,407 to who was uninsured in 2013 could have learned 6,583. Although the increases in YPLL were similar he was diabetic in 2014 after obtaining health for Kentucky and the U.S. (3.8% and 2.7%, respec- insurance and visiting a health care provider. tively), Kentucky’s rate remained more than a Early Death Remained Relatively third higher than the U.S. (39.8% higher in 2015). Stable in Kentucky, But Higher Than U.S. Because these data come from death records and This study also uses a measure of premature not a statistical sample, no significance testing death (defined in this study as before age 75), was performed. sometimes called the Years of Potential Life Lost (YPLL), which is calculated from vital statistics data. The National Center for Health Statistics (NCHS) describes YPLL this way: “YPLL is a summary measure of premature mortality (early death). It represents the total number of years not lived by people who die before reaching a given age.”60 10,000 2012 2015 FIGURE 5.9: 9,206 Years of Potential Life Lost 8,865 Due to Premature Deaths, 8,000 Kentucky and the U.S., 2012-2015 (ages 75 and 6,407 6,583 younger) 6,000 4,000 2,000 0 Kentucky United States Source: Web-based Injury Sta�s�cs Query and Repor�ng System (WISQARS) database, Na�onal Center for Injury Preven�on and Control and the CDC. Es�mates report the YPLL before age 75, using the YPLL Age-Adjusted Rate and 2000 as the standard year. FEBRUARY 2017 | WWW.SHADAC.ORG 3434 Discussion for charity care and self-pay charges for patients Since implementation of the ACA, Kentucky has without insurance. Despite these improvements, seen broad improvements in measures of health though, evidence suggests that premiums for coverage. The Commonwealth’s uninsurance rate employer-sponsored insurance have continued has dropped by more than half, and coverage their pre-ACA increases, and median out-of-pock- through the individual market and Medicaid/CHIP et spending for Kentuckians has remained steady have increased significantly. In 2015, Kentucky’s rather than declining. uninsurance rate was lower than the U.S. and six The potential impacts of the ACA on health care of eight neighboring states. However, Kentucky quality in Kentucky have been less clear. While continues to experience disparities in health insur- newborn breastfeeding and colorectal cancer ance coverage. The Hispanic/Latino population’s screening rates have increased since implemen- uninsurance rate is nearly four times the overall tation of the law, the prevalence of low birth rate, the young adult uninsurance rate is nearly weight infants and cholesterol awareness rates, double the overall rate, and the low-income popu- and unprotected sex among high school students lation is about 1.5 times the overall rate. However, have remained largely unchanged. Additionally, the share of Kentuckians who are underinsured— measures of preventable health complications spending 10% or more of their family income on have been mixed, with death rates for low-risk health expenses—remained steady at nearly one hospitalizations remaining relatively steady, while in four Kentuckians. While fewer small employers admissions for diabetes short-term complications offer coverage since implementation of the ACA, increased, and admissions for hypertension and the share of Kentucky’s population with employ- asthma decreased. er- sponsored insurance has remained stable, likely because large firms continue to offer health Our study did not find clear improvements in insurance. health outcomes during these early years since implementation of the ACA in Kentucky. Obesity Kentucky also has begun to experience some rates for adult Kentuckians continued to climb, improvements in access to health care. More although adolescent rates stayed steady. Cigarette Kentuckians report having a usual place where smoking rates for adolescents remained stable they go for health care, as well as more reporting while adult rates declined; however, whether that they have visited a health care provider in the ACA played a large role in the decline among the past year. Since 2012, elderly Kentuckians are adults—or whether that was driven by other also less likely to report making changes to their factors, such as the rise of e-cigarettes—remains medications due to cost, likely a result of the ACA’s uncertain. The share of Kentucky adults report- provisions to gradually close the Medicare Part ing poor or fair health remained stable at more D “doughnut hole.” Over nine in ten Kentuckians than one in five, and the share reporting a chronic continue to report they can find a doctor when disease also stayed stable at more than one in they need one and that they can find a doctor four. Compared to the U.S., Kentucky continues to who takes their insurance. Kentuckians continue have a higher rate of poor or fair health, and only to report using the Emergency Department (ED) one other neighboring state (West Virginia) had a in rates similar to before the ACA. Although the higher rate in 2015. Additionally, years of life lost Commonwealth’s rate of ED use is higher than the due to premature death remains higher than the U.S., it is no longer significantly higher than most U.S. of its neighboring states. Additionally, some gaps in access have continued. For example, more than Overall, our analysis of these indicators suggests one in ten young adult Kentuckians have an unmet that the ACA has improved health insurance cov- need for treatment of alcohol abuse, and nearly erage in Kentucky, and it may have played a role in four in ten Kentucky adults haven’t seen a dentist improving some measures of access to health care. in the past year. The Commonwealth has also experienced signifi- cant improvements in reducing the financial strain The Commonwealth has seen improvements in of health care on families and reduced the role of most of the measures of cost that we have tracked cost as a barrier to care. However, disparities and through our study. Fewer Kentuckians report gaps remain in these three domains. Any effects trouble paying medical bills compared to before of the ACA on quality and health outcomes remain the ACA. While the U.S. also experienced a decline uncertain based on early data from the first years in trouble paying medical bills, only one other of implementation of the law in Kentucky. Future neighboring state (Ohio) also saw a decline. Addi- study is needed into health quality and outcomes tionally, fewer Kentuckians say they have delayed to better understand whether and to what extent or gone without needed medical care due to the the ACA has impacted these domains. cost. Hospitals also have seen reductions in costs 3535 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky III. STUDY FINDINGS: MEDICAID ENROLLMENT AND SERVICES One of the Affordable Care Act’s key provisions to Using data provided by the Kentucky Cabinet for increase health insurance coverage was the law’s Health and Family Services, this section of our expansion of Medicaid eligibility to adults with report examines enrollment of non-elderly adults incomes up to 138% of Federal Poverty Guidelines in Kentucky’s Medicaid expansion and in the (FPG). Although the law was intended to expand Commonwealth’s traditional income-based Med- Medicaid throughout the U.S., a 2012 ruling by icaid program. Additionally, this chapter examines the U.S. Supreme Court effectively made the utilization of selected health care services by ACA’s Medicaid expansion optional for states. non-elderly adults enrolled in traditional and Since then, 31 states and the District of Columbia expanded Medicaid programs. have expanded their Medicaid programs, includ- ing Kentucky, which expanded under executive authority of then-Governor Steve Beshear in 2014. MEDICAID ENROLLMENT Enrollment Increased In Expansion and included people who were likely ineligible to en- Traditional Medicaid roll in Medicaid before 2014. From the first quarter of Medicaid expansion in Kentucky to Quarter 3 of 2016, enrollment In comparison, the traditional Medicaid group of non-elderly adults (ages 19-64) in Medicaid was an established Medicaid eligibility category increased by 72.7%, from 376,956 to 650,867 before the ACA, so most of this growth was likely (see Figure 6.1). With the exception of one drop in the result of increased awareness of the program, Quarter 3 of 2015, enrollment has increased each commonly known as the “welcome mat” effect.61 quarter since Kentucky expanded its Medicaid program. Enrollment grew to a greater extent in the Medicaid expansion than in traditional Med- By Quarter 3 of 2016, icaid. Between Quarter 1 of 2014 and Quarter 3 of 2016, enrollment in Medicaid expansion more than 500,000 nearly doubled, from 260,535 to 506,317, while enrollment in traditional Medicaid increased a Kentuckians were enrolled more-modest 24.2%, from 116,421 to 144,550. in Medicaid expansion. The larger growth in Medicaid expansion reflects the fact that this was a new eligibility group that 600,000 Expansion Tradi�onal income-based FIGURE 6.1: Quarterly Medicaid 506,317 Enrollment, 2014-2016 500,000 400,000 300,000 260,535 200,000 144,550 116,421 100,000 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2014 2015 2016 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Services are based on claims data with dates of service from 1/1/14-9/30/16. FEBRUARY 2017 | WWW.SHADAC.ORG 3636 Enrollment Largely Reflected Regional of enrollment in Greater Lexington and Greater Population Levels Louisville, and two-thirds of enrollment in Eastern, The Commonwealth’s Medicaid expansion shows Western and Northern Kentucky (see Figures 6.2, a statewide reach, with approximately one-third 6.4 and 6.5). FIGURE 6.2: Kentucky Regions Northern Kentucky Greater Louisville Greater Lexington Eastern Western Kentucky Kentucky Enrollment proportions across regions were FIGURE 6.3: similar in both Medicaid expansion and traditional Kentucky Regional Medicaid, with Eastern Kentucky accounting Populations of Non-elderly Adults, 26.5% 21.6% for the largest share of enrollment for both Western Eastern enrollment categories (see Figures 6.4 and 6.5). 2015 Kentucky Kentucky Those enrollment ratios equate roughly to the Commonwealth’s regional populations of non- elderly adults, although with somewhat higher enrollment for Eastern Kentucky (see Figure 6.3). Between Quarter 1 of 2014 and Quarter 3 of 10.3% Northern 18.8% Greater 2016, regional enrollment remained relatively Kentucky Lexington consistent. 22.7% Greater Louisville Source: SHADAC analysis of American Community Survey data. 3737 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky Western Kentucky Northern Kentucky Greater Louisville Greater Lexington Eastern Kentucky FIGURE 6.4: 100% Medicaid Expansion 90% Enrollment by Region, 33.6% 30.7% 2014 & 2016 80% 70% 60% 16.4% 16.0% 50% 19.6% 40% 18.5% 30% 7.3% 7.5% 20% 24.6% 25.8% 10% 0% 2014 Q1 2016 Q3 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Western Kentucky Northern Kentucky Greater Louisville Greater Lexington Eastern Kentucky FIGURE 6.5: 100% Traditional Income-based Medicaid Enrollment by 90% Region, 2014 & 2016 32.2% 80% 35.8% 70% 60% 15.1% 15.0% 50% 17.8% 19.2% 40% 30% 7.2% 7.7% 20% 24.1% 25.8% 10% 0% 2014 Q1 2016 Q3 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Younger Adults Lead Medicaid Enrollment While Medicaid expansion represents the largest Enrollment in the Kentucky’s Medicaid program share of enrollees for each group, the size of also varies by age, with adults ages 26-34 repre- enrollment in traditional Medicaid varies by age. senting the largest group (104,945 enrollees) in For example, in Quarter 3 of 2016 traditional Quarter 1 of 2014, followed by young adults ages Medicaid accounted for 26.4% of Medicaid 19-25 (85,767 enrollees), ages 35-44 (82,958 enrollees ages 19-25 but only 2.3% of Medicaid enrollees), ages 45-54 (65,621 enrollees) and ages enrollees ages 55-64. The lower enrollment of 55-64 (37,665 enrollees) (see Figure 6.6). While older adults in traditional Medicaid is likely enrollment has increased, the age distribution of because this category of Medicaid covers largely beneficiaries has remained relatively consistent low-income parents and pregnant women, who through Quarter 3 of 2016 (see Figure 6.7). are more likely to be younger. FEBRUARY 2017 | WWW.SHADAC.ORG 3838 FIGURE 6.6: Expansion Tradi�onal income-based Total Enrollment by Age, Quarter 1 of 2014 19-25 49,764 36,003 26-34 60,881 44,064 35-44 55,750 27,208 45-54 57,561 8,060 55-64 36,579 1,086 0 50,000 100,000 150,000 200,000 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). FIGURE 6.7: Expansion Tradi�onal income-based Total Enrollment by Age, Quarter 3 of 2016 19-25 108,839 39,086 26-34 128,221 53,787 35-44 111,284 38,539 45-54 96,388 11,683 55-64 61,585 1,455 0 50,000 100,000 150,000 200,000 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). SERVICES Number of Colorectal Screenings Increased Most of these screening are provided to Medicaid With Medicaid Expansion expansion enrollees. That is consistent with the Early in Kentucky’s Medicaid expansion, the Com- fact that colorectal cancer screening guidelines monwealth saw a large increase in colorectal recommend that most people begin screening cancer screening services, increasing from 3,762 at age 50,62 and most beneficiaries of that age to 6,458 (71.7%) between Quarter 1 and Quarter are enrolled in expansion rather than traditional 2 of 2014 (see Figure 6.8). Medicaid. However, despite quarter-to-quarter fluctuations, colorectal cancer screenings have remained rela- tively steady since that initial increase. 3939 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky 7,000 Expansion Tradi�onal income-based FIGURE 6.8: Quarterly Colorectal 6,000 Screenings, 2014-2016 5,550 5,000 4,000 3,000 3,152 2,000 610 769 1,000 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2014 2015 2016 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Services are based on claims data with dates of service from 1/1/14-9/30/16. Number of Dental Services Increased for time period, services to expansion beneficiaries Medicaid Expansion Enrollees increased 74.7%—from 19,340 to 33,782. Dental services provided through Kentucky’s In addition to preventive dental services, tra- Medicaid program have continued to grow since ditional and Medicaid expansion also provide Medicaid expansion. From Quarter 1 of 2014 certain other dental services, such as fillings to Quarter 3 of 2016, the number of preven- to treat tooth decay. These services also have tive dental services increased from 30,088 to increased since the Commonwealth expanded its 44,065 (46.5%) (see Figure 6.9). The increase in Medicaid program, accounting for approximately preventive dental services was driven largely by half of the dental services provided to both ex- the Medicaid expansion. While services provided pansion and traditional Medicaid beneficiaries to traditional Medicaid beneficiaries remained (see Figure 6.10). essentially flat at about 10,000 throughout that 40,000 Expansion Tradi�onal income-based FIGURE 6.9: 33,782 Quarterly Preventive 35,000 Dental Services, 2014-2016 30,000 25,000 20,000 19,340 15,000 10,283 10,000 10,748 5,000 0% Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2014 2015 2016 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Services are based on claims data with dates of service from 1/1/14-9/30/16. FEBRUARY 2017 | WWW.SHADAC.ORG 4040 FIGURE 6.10: Preven�ve Services Other Services 100% Quarterly Total Dental Services, 2014 & 2016 90% 80% 46% 49% 70% 60% 50% 40% 30% 54% 51% 20% 10% 0% 2014 Q1 2016 Q3 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Services are based on claims data with dates of service from 1/1/14-9/30/16. Expansion Increased the Number of Breast increase — were due to expansion enrollees, which Cancer Screenings like colorectal cancer screenings is consistent with Similar to colorectal cancer, screenings for breast recommendations that these screenings begin cancer increased early in Medicaid expansion for women in their 40s or 50s.63,64 — and primarily for Medicaid expansion en- rollees. From Quarter 1 of 2014 to Quarter 2 of Traditional Medicaid Covered Fewer Births 2014, breast cancer screenings increased 45.1%, and Expansion Covered More from 6,535 to 9,485 (see Figure 6.11). Since then, Since the beginning of 2014, the number of births breast cancer screenings have varied from quar- covered by traditional and expanded Medicaid ter-to-quarter but only increased slightly since combined has varied substantially—from a high of Quarter 2 of 2014 to 10,143 in Quarter 3 of 2016. 7,884 in Quarter 3 of 2014 to a low of 5,007 in Most of those screenings — and most of the Quarter 2 of 2015—but overall it has not followed FIGURE 6.11: 10,000 Expansion Tradi�onal income-based Quarterly Breast Cancer 8,920 9,000 Screenings, 2014-2016 8,000 7,000 6,000 5,000 5,453 4,000 3,000 2,000 1,223 1,082 1,000 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2014 2015 2016 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Services are based on claims data with dates of service from 1/1/14-9/30/16. 4141 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky a clear increasing or decreasing trend (see Figure (from 7,009 to 5,488) but the number of births 6.12). Between Quarter 1 of 2014 and Quarter 3 covered by Medicaid expansion has increased of 2016, the number of births covered by Med- (from 122 to 1,551). This shift has likely occurred icaid differed by only about 100 (7,131 to 7,039). because women already enrolled in the state’s However, traditional and expansion Medicaid Medicaid expansion have become pregnant, with followed different trends: The number of births their births then covered by Medicaid expansion covered by traditional Medicaid has declined rather than under traditional Medicaid. 10,000 Expansion Tradi�onal income-based FIGURE 6.12: 7,009 Quarterly Births, 9,000 2014-2016 8,000 5,488 7,000 6,000 5,000 4,000 3,000 1,551 2,000 1,000 122 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2014 2015 2016 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Services are based on claims data with dates of service from 1/1/14-9/30/16. Hepatitis C Screenings Increased Since 2015 C screenings increased somewhat during 2014, Screenings for hepatitis C have increased since the largest quarterly increase (34.0%) occurred Medicaid expansion, but unlike those for breast from Quarter 4 of 2014 to Quarter 1 of 2015, and colorectal cancer, the largest increases in when screening rose from 3,422 to 4,586 (see hepatitis screenings occurred later. While hepatitis Figure 6.13). 4,500 Expansion Tradi�onal income-based FIGURE 6.13: 4,011 Quarterly Hepatitis C 4,000 Screenings, 2014-2016 3,500 3,000 2,500 2,073 2,000 2,148 1,500 1,000 1,208 500 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2014 2015 2016 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Services are based on claims data with dates of service from 1/1/14-9/30/16. FEBRUARY 2017 | WWW.SHADAC.ORG 4242 Since then screenings have continued to increase, treatments compared to utilization of other Med- though at a steadier pace. Additionally, the earlier icaid services. For example, prior to implemen- trends in hepatitis C screening differ somewhat tation of the ACA, Kentucky’s Medicaid program between traditional Medicaid and expansion: typically didn’t cover substance use treatment, From Quarter 1 to Quarter 2 of 2014, hepatitis so it could have taken time for beneficiaries and C screenings increased in expansion, while they providers to learn that treatments were now decreased in traditional Medicaid. Since Quarter covered under Medicaid. Additionally, unlike 1 of 2015, however, screenings have increased in many screenings that are performed infrequently both expansion and traditional Medicaid. (e.g., a colonoscopy every 10 years), treatment for substance use disorders often requires a series It is not clear why hepatitis screenings remained of ongoing services for a period of time, such as relatively flat in 2014, but the increases in 2015 regular behavioral therapy visits or clinic visits for and 2016 may be related to enhanced awareness medication-assisted therapy. of substance use in the Commonwealth. Because hepatitis C is a blood-borne disease, screening is Diabetes Screening Increased More recommended for people who currently inject or For Expansion Enrollees who have ever injected drugs.65 From Quarter 1 of 2014 to Quarter 3 of 2015, screenings for diabetes increased modestly, from Substance Use Treatment Services about 780 to 1,180 in traditional and expansion Grew More than Five Times Medicaid combined (see Figure 6.15). But in Since Quarter 1 of 2014, substance use disorder Quarter 4 of 2015, diabetes screenings increased treatments covered by Medicaid have increased 220.5% to 3,782—and they have continued to by more than five times (505.7%) in traditional and increase since then to 4,495 in Quarter 3 of 2016. expanded Medicaid (see Figure 6.14). These data Overall, diabetes screenings increased nearly five represent the number of treatments provided times (476.3%) since Quarter 1 of 2014, although under Medicaid coverage, and not necessarily the the increase was larger among expansion bene- number of individuals receiving treatment, as one ficiaries (709.7%) than traditional beneficiaries person can receive multiple treatments. (171.0%). While treatments have increased each quarter, the largest quarterly increase (55.1%) occurred relatively recently, between Quarter 4 of 2015 and Quarter 1 of 2016—from 8,276 to 12,837. There may be numerous factors that contributed to the higher growth in substance use disorder FIGURE 6.14: 14,000 Expansion Tradi�onal income-based Quarterly Substance 12,319 Use Services, 12,000 2014-2016 10,000 8,000 6,000 4,472 4,000 1,504 2,000 1,268 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2014 2015 2016 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Services are based on claims data with dates of service from 1/1/14-9/30/16. 4343 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky The difference may be related to greater risk for Assembly has taken steps to address the growing diabetes among expansion beneficiaries, who prevalence of diabetes, directing the Medicaid tend to be older than traditional beneficiaries.66 department and other agencies in 2011 to take Because the increase in diabetes screenings far efforts to reduce the prevalence and improve outpaces enrollment growth, this suggests that treatment of diabetes, and appropriating $2.6 other factors also probably contributed to this million for diabetes prevention and control efforts increase. For example, the Kentucky General in 2014.67 4,000 Expansion Tradi�onal income-based FIGURE 6.15: 3,579 Quarterly Diabetes 3,500 Screenings, 2014-2016 3,000 2,500 2,000 1,500 916 1,000 442 500 0 338 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2014 2015 2016 Source: SHADAC analysis of data provided by the Kentucky Cabinet for Health and Family Services (CHFS). Services are based on claims data with dates of service from 1/1/14-9/30/16. CONCLUSION Since Kentucky expanded its Medicaid program For example, in the same time period that Med- in 2014, enrollment of non-elderly adults in icaid expansion enrollment increased about Kentucky’s combined traditional and expansion 95%, utilization of colorectal cancer screenings Medicaid programs has increased by 72.7%. While increased, breast cancer screenings increased, enrollment increased by nearly a quarter in tra- preventive dental services increased and other ditional Medicaid, most of the growth occurred dental services increased between about 60- in the Commonwealth’s Medicaid expansion, 105%. In other cases, the increases in services which almost doubled from the first quarter of were much larger than the increase in enrollment expansion in 2014 to the third quarter of 2016. during this time. For example, hepatitis C screen- Nearly all of the health care services we exam- ings increased more than 200%, substance use ined also experienced increases when traditional treatment services increased more than 700% and expansion Medicaid are combined, with the and diabetes screenings also increased more than exception of births covered by Medicaid, which 700%. Additionally, births covered by Medicaid remained mostly unchanged. expansion increased more than 1,000%, but this was offset by a larger decline in births in tradi- In traditional Medicaid, utilization of services did tional Medicaid.68 not follow a consistent trend. Some services in- creased (e.g., diabetes screenings and substance Overall, these data show that both traditional and use treatment), while others remained largely expansion Medicaid have covered more Kentuck- stable (e.g., dental services and hepatitis C screen- ians since the Commonwealth implemented the ings), and births covered by Medicaid declined. ACA in 2014, and these expansions in coverage However, Medicaid expansion saw increases in through Kentucky’s Medicaid program have also utilization across the board. In several cases, these allowed beneficiaries to access preventive ser- increases roughly mirrored the size of enrollment vices, such as cancer screenings, and other types increases. of care, such as substance use treatment. FEBRUARY 2017 | WWW.SHADAC.ORG 4444 IV. STUDY FINDINGS: KENTUCKY HEALTH REFORM TREND ANALYSIS KENTUCKY HEALTH REFORM SURVEY (K-HRS) SHADAC conducted a one-time telephone survey The findings in this section of our report are of non-elderly adult Kentuckians in the spring of limited to non-elderly adults (ages 18-64); for ease 2016 as part of our study of the impacts of the of presentation we refer to them as Kentuckians. ACA in Kentucky. The Kentucky Health Reform Where we describe 2016 estimates, these refer to Survey (K-HRS) was designed by SHADAC, in con- the SHADAC K-HRS, and where we refer to other sultation with the Institute for Policy Research at estimates (e.g., 2012, 2014), these are from the the University of Cincinnati and the Foundation KHIP. for a Healthy Kentucky. The dual-frame (landline and cell-phone) survey was conducted between Uninsurance in Kentucky Dropped March 31 and May 3, 2016, and asked respondents Significantly questions related to their health status, insurance Between 2012-2016, Kentucky experienced a coverage, and experiences accessing health care. statistically significant 19.0 percentage point drop in uninsurance among non-elderly adults, To support a trend analysis examining changes from 27.9% to 8.9% (see Figure 7.1). Evidence pre- and post-ACA implementation, SHADAC used shows the decline in uninsurance was driven, in the same methodology and many of the same part, by the expansion of Kentucky’s Medicaid questions as the annual Kentucky Health Issues program and other ACA reforms designed to Poll (KHIP). The KHIP is an existing survey jointly increase private coverage, such as the creation funded by the Foundation for a Healthy Kentucky of health insurance marketplaces and financial and Interact for Health and conducted by the assistance to help people with moderate incomes Institute for Policy Research at the University of afford health insurance. The drop in uninsurance Cincinnati. Despite using the same methodology found by the K-HRS is consistent with findings and many of the same questions as the KHIP, from other surveys (see Section II), providing the K-HRS deviated from KHIP in several notable additional support to the conclusion that the ways. While the KHIP surveys adults of all ages in ACA has reduced uninsurance among non-elderly Kentucky, the K-HRS was restricted to non-elderly adults, who were specifically targeted by the law’s adults, reflecting our study’s focus on implemen- coverage expansions. tation of the ACA. In order to compare K-HRS findings with the KHIP, the Institute of Policy Research provided SHADAC with KHIP estimates that were restricted to the non-elderly adult sample. FIGURE 7.1: 100% Insured Uninsured Insurance Coverage, 91.1%* 2012-2016 80% 72.1% 60% 40% 27.9% 20% 8.9%* 0% 2012 2013 2014 2015 2016 *Difference from 2012 es�mate is sta�s�cally significant at the 95% level. Source: 2016 Kentucky Health Reform Survey and 2012-2015 Kentucky Health Issues Polls. 4545 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky Both Public and Private Health Insurance We also found significant growth in public health Coverage Increased Significantly coverage between 2012-2016, increasing 12.4 Corresponding with the drop in Kentucky’s un- percentage points from 28.6% to 41.0% of Ken- insurance rate, the Commonwealth has seen tuckians (see Figure 7.2). In 2014, Kentucky statistically significant growth in both public and implemented the ACA’s provision allowing states private health coverage between 2012-2016. to expand their Medicaid programs to adults Private coverage (i.e., employer-sponsored insur- with incomes up to 138% of the FPG. We believe ance and individual-market insurance) grew by the growth in public health coverage was largely 8.0 percentage points during this period, from driven by this policy, through offering coverage 40.1% in 2012 to 48.1% in 2016 (see Figure 7.2). to people who weren’t previously eligible and Because other findings in this report show that attracting people who were already eligible be- employer-sponsored insurance remained stable fore the ACA but not enrolled. These findings are from 2012 to 2015 (see Section II), this suggests consistent with other survey data that show an the increase in private coverage was driven mainly increase in Medicaid/CHIP coverage from 2012- by an increase in coverage through the individual 2015 (see Section II) and administrative data that market. The ACA included several provisions show an increase in enrollment by non-elderly aimed at improving access to individual-market adults in the Commonwealth’s Medicaid program health insurance, including creating health insur- since 2014 (see Section III). In the K-HRS and KHIP, ance marketplaces where individuals could shop public coverage also includes those with insurance for and purchase private health insurance, and through military plans and non-elderly Medicare financial assistance to help people with moderate beneficiaries, but those populations are relatively incomes (139-400% of FPG) afford health insur- small in Kentucky and were unlikely to have a large ance premiums. effect public coverage rates. 50% 48.1%* 2012 2016 FIGURE 7.2: Insurance Coverage by 41.0%* Type, 2012 & 2016 40.1% 40% 30% 28.6% 27.9% 20% 10% 8.9%* 0% Private coverage Public coverage Uninsured *Difference is sta�s�cally significant at the 95% level. Source: 2016 Kentucky Health Reform Survey and 2012 Kentucky Health Issues Poll. Dental Coverage Increased to Two-thirds of The ACA did include some limited provisions Kentuckians, But Dental Visits Stayed Stable that could support private dental coverage— Although dental insurance was not a key focus such as allowing dental coverage to be included of the ACA, we found a statistically significant in- in health insurance plans sold through market- crease in dental coverage in Kentucky. Between places, as well as allowing stand-alone dental 2012-2016, dental coverage increased 13.1 per- plans to be sold through marketplaces—but the centage points for nonelderly adults, from 52.9% law did not require private health insurance plans to 66.0% (see Figure 7.3). Evidence suggests that to cover dental care. In Kentucky, some market- this increase in dental coverage was due mostly place health insurance plans offer optional dental to Kentucky’s Medicaid expansion. benefits, but not all plans included these.69 FEBRUARY 2017 | WWW.SHADAC.ORG 4646 FIGURE 7.3: 75% Dental Coverage, 66.0%* 2012 & 2016 52.9% 50% 25% 0% 2012 2016 Source: 2016 Kentucky Health Reform Survey and 2012 Kentucky Health Issues Poll. Additionally, stand-alone dental plans have been Despite increased dental coverage, we did not offered through Kentucky’s marketplace, but find changes in self-reported use of dental care. enrollment in these has been relatively small Among non-elderly adult Kentuckians in 2016, in Kentucky and the U.S. compared to health 59.9% reported having a dental visit in the past insurance plans.70,71 The ACA did not require year, while 17.0% reported they hadn’t visited a that states expanding their Medicaid programs dentist within the past 5 years, which were not include dental coverage, but it allowed states significantly different than 2012 (see Figure 7.4). the choice to provide dental coverage—a policy The K-HRS did not ask respondents why they had option that Kentucky adopted. Supporting the forgone dental care, so we do not know why the idea that Kentucky’s increase in dental coverage increase in dental coverage hasn’t resulted in was the result of Medicaid expansion, the size of increased use of dental services; however, some the increase in dental coverage (13.1 percentage research suggests there are other barriers to care points) was similar to the size of the increase in in Kentucky, such as shortages of dental providers, public coverage (12.4 percentage points). particularly in rural areas of the state.72 FIGURE 7.4: Self-Reported Health Status Remained Unchanged Time Since Last Dental Consistent with similar findings in Section II of Visit, 2016 17.0% this report, we did not find statistically significant 5 or more changes in Kentuckians’ self-reported health years ago status between 2012-2016. In 2016, most non- elderly adults reported their health was “very 10.0% Within the good” (29.3%) or “good” (28.3%) (see Figure 7.5). Only 13.7% reported “excellent” health, and the past 5 years remainder said their health was “fair” (17.5%) or 59.9% Within the “poor” (10.7%). None of these estimates were significantly different from 2012. As discussed 13.0% Within the past year earlier in this report, because there are numerous past 2 years factors that influence overall health—including genetics, lifestyle, environment and other factors —and because health improvements may take years to occur, it is not surprising that Kentuck- ians’ health status has not changed significantly within the first few years of implementation of Source: 2016 Kentucky Health Reform Survey. the ACA. To gauge the potential impacts of the 4747 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky FIGURE 7.5: Health Status, 2016 Excellent 13.7% Very Good 29.8% Good 28.3% Fair 17.5% Poor 10.7% 0% 5% 10% 15% 20% 25% 30% Source: 2016 Kentucky Health Reform Survey. 100% FIGURE 7.6: Had Usual Source of 86.5%* Care 77.1% 75% 50% 25% 0% 2014 2016 Source: 2016 Kentucky Health Reform Survey and 2014 Kentucky Health Issues Poll. ACA on Kentuckians’ health, it will be important of this report. Additionally, it is important to note for future research to study changes in health that although the K-HRS measure of “usual source status over a longer time period. of care” is similar to the “usual source of care” indicator obtained from the NHIS (reported in Since 2014, More Kentuckians Reported Section II of this report), there is a key difference. Usual Source of Care Both surveys ask whether respondents have a The K-HRS also asked questions about access place they usually visit when they need health to health care. Comparable pre-ACA data were care, but the NHIS estimates (found in Section II) not available from the 2012 KHIP, but data were exclude people who say an Emergency Depart- available from 2014,73 the first year of ACA ment is their usual source of care. In contrast, the implementation in Kentucky. It is important to K-HRS estimates include people who report ED note that the following findings examine changes as their usual source of care, which we discuss in from within the first years of ACA implementation, further detail below. not a pre-ACA/post-ACA analysis as found in most FEBRUARY 2017 | WWW.SHADAC.ORG 4848 We found that from 2014-2016, there was a These were also the top three places reported statically significant increase of 9.4 percentage in 2012, and the estimates did not change points in the share of Kentuckians who reported significantly (see Figure 7.7). In 2016, 3.9% of having a usual source of care, from 77.1% to non-elderly adult Kentuckians reported using a 86.5% (see Figure 7.6). This is consistent with a hospital ED as their usual source of care, which similar measure found in Section II. Among those was similar to 3.6% in 2014. While these esti- who report having a usual source of care, the top mates should be considered with caution due to three places where people usually sought care their unreliability, it is worth acknowledging that in 2016 were a private doctor’s office (62.9%), a EDs were not named as one of the top three community-based health center (16.0%) and a usual sources of care in 2014 or 2016.74 hospital out-patient department (5.8%). FIGURE 7.7: Top Three Usual LOCATION PERCENT Sources of Care 1. Private doctor's office 62.9% 2. Community-based health center 16.0% 3. Hospital outpatient department 5.8% Source: 2016 Kentucky Health Reform Survey. CONCLUSION The findings from our trend analysis of 2016 Although pre-ACA comparison data were not K-HRS estimates and baseline 2012 KHIP esti- available, we also found improvements in the mates are largely consistent with findings from percentage of Kentuckians who report having a other components of our study. Since 2012, the usual source of care, which suggests the observed Commonwealth has seen a statistically significant increases in coverage may be having the ACA’s decline in uninsurance among non-elderly adults, intended effect of improving access to health from more than one in four (27.9%), to less than care. However, our analysis of K-HRS and KHIP one in 10 (8.9%) by early 2016. data did not find changes in Kentuckians' self-reported health status, so improvements in That decline in uninsurance appears to have been health insurance coverage and access to care driven by gains in both private and public cover- don’t yet appear to have resulted in significant age, as a result of Kentucky’s adoption of the ACA’s health improvements. coverage expansion provisions — the creation of a state-based health insurance marketplace with federal subsidies provided to income-eligible individuals in the form of tax credits and the Medicaid expansion for lower-income adults. In addition to gains in health insurance cover- age, we also documented statistically significant increases in dental coverage, from just over half (52.9%) of non-elderly adults reporting dental coverage in 2012 to approximately two-thirds (66.0%) in 2016. Despite those gains in dental coverage, however, the Commonwealth hasn’t seen increases in Kentuckians reporting visiting a dentist, suggesting other barriers to receiving dental care. 4949 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky V. STUDY CONCLUSIONS Overall, our study of the impacts of implementation of the ACA in Kentucky has found the Commonwealth has expe- rienced some clear improvements since our baseline year of 2012, but there are other areas in which Kentucky has not yet seen substantial progress. In the domain of coverage, the Commonwealth’s uninsurance rate has declined significantly, with more Kentuckians covered by both private and public health insurance. Kentuckians have also seen some improvements in access, particularly increases in Kentuckians who report having a usual source of care and visiting a doctor in the past year. The Commonwealth also has experienced improvements in cost. Since implemen- tation of the ACA, fewer Kentuckians report trouble paying medical bills and delaying or going without needed care due to cost. Additionally, Kentucky hospitals have seen charity care and self-pay charges for the uninsured drop by two-thirds since 2012. While we found evidence of some improvements in quality indicators tied to ACA provisions, such as increases in breastfeeding initiation rates for newborns, quality overall has not shown clear progress like the domains of coverage, access and cost. Similarly, health outcomes largely appeared to remain stable, which is not surprising because any improvements in the health of Kentuckians may take years, and we currently only have data from the first few years of ACA implementation. Ultimately, we believe it will be important to continue to track the performance of Kentucky in each of the five domains of this study to determine whether the improvements we found are sustained, and whether those domains that remained relatively stable through our study see improvements over a longer period of time. Coverage It is noteworthy that certain groups of Kentuckians Our study has found that Kentucky’s uninsurance rate specifically targeted by the ACA—those with lower and has dropped substantially since implementation of the moderate incomes, and non-elderly adults—experi- Affordable Care Act. Between 2012-2015, Kentucky’s enced the largest declines in uninsurance. However, uninsurance rate dropped by more than half—from other Kentuckians have also seen significant reductions 13.6% to 6.1%—based on an analysis of estimates from in uninsurance. the American Community Survey. While other states Since implementation of the ACA, even Kentuckians also have seen declines in uninsurance since the ACA, with incomes too high to qualify for Medicaid or market- Kentucky’s decline has been larger than the U.S. and place tax credits experienced a decline in uninsurance. most neighboring states. Additionally, fewer Kentucky children are uninsured, We also found evidence that Kentucky’s decline in most likely because their parents enrolled them in uninsurance was driven by ACA policies, particularly the coverage at the same time they enrolled themselves in Commonwealth’s Medicaid expansion and creation of a health insurance.76 health insurance marketplace, with financial assistance Despite increased health insurance coverage in Ken- for people with moderate incomes. Analysis of federal tucky, there are still disparities and gaps. In 2015, the survey data found increases in Medicaid coverage and Hispanic/Latino population’s uninsurance rate was individual-market coverage, which was supported nearly four times the Commonwealth’s average, the by similar findings from our Kentucky Health Reform rate for young adults was almost twice the average, and Survey. This is consistent with administrative data from the rate for low-income Kentuckians was one and a half the Kentucky Cabinet for Health and Family Services, times the Commonwealth’s average. Although Kentucky which show that enrollment in the Commonwealth’s has seen substantial reductions in uninsurance, approx- Medicaid expansion reached more than 500,000 in imately 320,000 Kentuckians remained uninsured in 2016, and enrollment of non-elderly adults in Ken- early 2016—nearly 90% of whom had lower or moder- tucky’s traditional Medicaid program grew nearly 25% ate incomes that could make them eligible for Medicaid to almost 145,000. In another study report, we found coverage or financial assistance to buy private insur- that low-income Kentuckians were less likely to be ance.77 Our 2016 K-HRS survey of non-elderly adults also uninsured than low-income residents of neighboring found that more than one in ten insured Kentuckians states that haven’t expanded their Medicaid programs said they were concerned about losing their coverage or expanded them later.75 in the next year, and that increased to more than one in Kentucky’s coverage gains have had a broad impact on four among lower-income people.78 Additionally, while the Commonwealth’s population. We found reductions the percentage of Kentuckians with employer-sponsored in uninsurance for both males and females, across ages insurance has remained steady since 2012, the per- and income levels, and among most racial and ethnic centage of small employers that offer health insurance groups. Only the Hispanic/Latino and Asian populations to their workers has declined significantly since 2012, did not experience significant declines in uninsurance. continuing a pre-ACA trend. FEBRUARY 2017 | WWW.SHADAC.ORG 5050 Access policies to be sold through health insurance marketplac- The ACA’s coverage expansions were designed in part es, and giving states the option of covering dental care to improve access to health care. Although data are through their Medicaid expansions. Our survey found limited at this point in implementation of the ACA, we the percentage of non-elderly adult Kentuckians with have found evidence of some early improvements in dental coverage increased significantly since implemen- access to health care. tation of the ACA, but we found no improvement in the percentage of Kentuckians who visited a dentist in the Through analysis of federal survey data and our own prior year. survey, we found increases since 2012 in the percent- age of Kentuckians who said they have a usual source Cost of health care. Federal survey data also showed an By expanding coverage, the ACA intended to both increase in the percentage of Kentuckians who reported reduce cost as a barrier to obtaining health care and visiting a health care provider in the past year. We also reduce the financial strain of health care costs on fam- found that fewer elderly Kentuckians are taking steps ilies. Despite concerns that have been raised about the such as delaying refills, skipping doses and taking less growth of premiums for individual-market coverage and medication than prescribed in response to high drug about the affordability of cost-sharing, our study has costs—likely an impact of the ACA’s provision to gradu- found some early improvements in measures of cost. ally close the Medicare Part D “donut hole.” Additionally, Additionally, we found that Kentucky hospitals have we did not find evidence that the increased number of experienced a dramatic decline in costs associated with Kentuckians with health insurance has made it harder to uninsured patients. Since 2012, Kentucky hospitals have get care when needed; the percentages of Kentuckians experienced a 67% drop in charity care and charges to saying they could find a doctor when needed and found self-pay patients—from $2.4 billion in 2012 to $786 a provider who accepted their insurance remained million in 2015. statistically stable and above 90%. Since 2012, the percentage of Kentuckians reporting Since Kentucky expanded its Medicaid program in 2014, trouble paying medical bills dropped significantly, from the program has seen continued growth in services almost half to slightly more than one-third in 2015. provided to beneficiaries, mirroring the growth in en- While Kentucky’s rate remained higher than the U.S. rollment. The increases in breast and colorectal cancer rate in 2015, the Commonwealth’s rate is now similar screenings, diabetes screenings, hepatitis C screenings, to most of its neighboring states. Additionally, the per- and other services suggests Medicaid expansion is centage of Kentuckians who reported delaying or going helping to provide access to care for new beneficiaries. without needed care due to cost dropped by approxi- In addition to the ACA’s broader coverage expansions, mately half from 2012-2015. However, our 2016 survey the law included some provisions specifically targeting found that about one in five non-elderly adults reported substance use disorders, such as requiring that indi- they delayed or went without care due to cost. Those vidual-market plans and Medicaid expansion cover the cost barriers were higher for lower-income Kentuckians, treatment of substance use for enrollees. In a special with more than one in four reporting delayed or forgone report examining the impacts of the ACA on substance care due to cost. Additionally, cost was a greater barrier use in Kentucky, we found indications that the ACA may for Kentuckians who may have more need for health be contributing to increases in treatment, especially care, with almost two-thirds of non-elderly adults re- through Medicaid.79 Since the Commonwealth imple- porting “poor” health saying they went without care due mented Medicaid expansion in 2014—offering coverage to cost, compared to a rate of about one-third overall. of substance use treatment to expansion enrollees and In examining costs of coverage through states’ insurance enhancing substance use treatment benefits for tradi- marketplaces, we found evidence that Kentucky had tional Medicaid enrollees—the number of substance lower premium costs than comparison states. For 2015 use treatments covered by the program has grown by coverage, one in ten people in the U.S. who bought more than 500%. Some early data from 2014 suggest health insurance through marketplaces chose gold- or that unmet need for substance use services continues to platinum-level plans, which tend to have higher premi- persist in Kentucky, but understanding any impacts will ums than bronze- and silver-level plans.80 By comparison, require future research as later years of data become one in four Kentuckians enrolled in gold- or platinum- available. level plans, suggesting that these plans may have In other areas—particularly dental—the study has not been more affordable in Kentucky. While Kentuckians found that coverage expansions have resulted in clear purchased more lower-cost bronze- and silver-level improvements in access. Although dental health was plans for 2016, our analysis found Kentucky had the not a key focus of the ACA, the law did include some pol- lowest silver-level marketplace premiums that year icies to support this, such as allowing dental insurance compared to neighboring states.81 Early data for 2017 5151 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky coverage suggest premiums in Kentucky’s market- due to diabetes short-term complications increased, place may have increased somewhat compared with while admissions related to hypertension and asthma other states, but premiums in Kentucky’s marketplace both declined. Most of the other quality indicators that remained lower than most neighboring states for the we tracked remained largely unchanged. Unprotected second-lowest cost silver-level plans—a key benchmark sex among high school students did not change since used for determining financial assistance.82 implementation of the ACA, although the ACA was not expected to have a large effect on this measure. The As in other states, concerns have been raised in Kentucky Commonwealth’s rate of low birth weight remained about the affordability of health insurance deductibles. steady at 8.7%, and racial and ethnic disparities con- We found that in 2016, most Kentuckians who bought tinued into 2015. The rate of death for hospitalizations coverage through the Commonwealth’s health insur- typically considered low-risk also remained mostly ance marketplace enrolled in silver-level plans (60%), steady between 2012-2014. which had median deductibles of $3,500 for single coverage, while about a quarter enrolled in bronze- Health Outcomes level plans with median deductibles of $6,000 for single Ultimately, the goal of health care—and by extension coverage. However, other data suggest that cost- the ACA’s coverage expansions—is to improve people’s sharing has remained relatively stable since before the health. Although there are numerous factors that influ- ACA. For Kentuckians overall, we found no statistically ence health, and improving the health of Kentuckians significant change in out-of-pocket spending since 2012, may take years, we tracked several measures of health and we found no significant increase in underinsurance. outcomes. The Commonwealth has not made improve- ments in most of the measures, but it will be important Quality to monitor these outcomes in the future to determine Beyond the ACA’s focus on expanding health insurance the long-term impacts of the ACA. coverage, the law also included provisions to improve the quality of care people receive. Although we did find Of the health outcome indicators in our study, Ken- some improvements, most of the indicators of health tucky’s performance only improved in one: The Com- care quality that we tracked did not show clear changes monwealth’s adult cigarette smoking rate declined from in either a positive or negative direction. Because any 28.3% in 2012 to 26.0% in 2015. However, the smoking measureable improvements may take time to accumu- rate among high school students did not change since late, understanding the impact of the ACA on quality our study baseline. The ACA included certain provisions of health care will likely require continued monitoring that may have contributed to the decline in cigarette research in the future. smoking—such as requiring health insurance plans to cover smoking cessation treatment and allowing health Among the ACA’s efforts to improve quality of health insurers to charge higher premiums for tobacco users. care were policies to reduce financial barriers to preven- However, additional research is needed to determine tive health services and infant breastfeeding. Between whether the decline in Kentucky’s smoking rate was 2012-2014, Kentucky experienced an increase in the driven by the ACA or other reasons, such as increased percentage of people obtaining recommended colorec- use of e-cigarettes. tal cancer screenings.83 Although there may be other factors that contributed to this increase, the ACA may Since our study baseline, Kentucky’s obesity rate for also have played a role by requiring that most health adults increased from 31.3% to 34.6% in 2015. The insurance plans cover recommended preventive screen- Commonwealth’s high school obesity rate did not ings without cost-sharing for individuals. However, change, however. Another measure of health outcomes we did not find improvements in another preventive showed ambiguous results, with the percentage of measure, cholesterol screening. Since 2012, Kentucky adult Kentuckians reporting a chronic disease increas- also has experienced improvements in the percentage ing significantly from 26.8% in 2012 to 29.1% in 2014. of newborns who were breastfed at discharge from the However, the rate in 2015 (28.7%) was not significantly hospital, from 62.8% to 68.7% in 2015 — a practice that different from 2012. Because of this, future research was supported by ACA requirements for most private may be needed to determine whether reported chronic health insurance to cover breastfeeding counseling and disease prevalence has in fact increased, and if so, the equipment, such as breast pumps. reasons it has increased. For example, it could reflect an actual increase in the number of people with diseases By improving individuals’ access to health care, as well such as diabetes, or it could mean that people are now as encouraging payment reforms to reward hospitals’ more aware of their health conditions after obtaining and providers’ improvements in quality, the ACA also health insurance and accessing care. Other measures attempted to reduce unnecessary hospital admissions. of health outcomes have not shown changes since Since 2012, Kentucky has seen mixed results in indica- implementation of the ACA. tors related to preventable hospitalizations—admissions FEBRUARY 2017 | WWW.SHADAC.ORG 5252 VI. APPENDIX: DATA SOURCES, METHODS, & INDICATORS In this Appendix, we describe our data collection procedures and methods for the study. The Appendix is organized by data source, and it includes a brief data source description, a discussion on how the estimates were obtained, and some notes about specific indicators where relevant. American Community Survey (2012, 2013, 2014, of February and April, asks about health insurance 2015) coverage for the prior calendar year and is combined The American Community Survey (ACS) is a federal with information from the main CPS survey on deter- survey conducted by the U.S. Census Bureau. The ACS minants of health insurance coverage such as employer asks about demographic and socioeconomic character- size, household spending, and other demographic and istics, and it includes a question on current health insur- socioeconomic characteristics. The sample size is about ance coverage. Despite the availability of other sources 200,000 people nationally, with over 2,300 in Kentucky to estimate health insurance coverage, we consider the in 2015. The CPS is available as a public use data file ACS the best source for annual state-level estimates, which allows for the creation of custom variables. particularly for states that have relatively low popu- The CPS income and health insurance questions were lation sizes, like Kentucky. The reason is that it has a recently redesigned to improve the quality of data large sample size relative to other federal surveys (more reported. Consequently, estimates of income and than 3.5 million people nationally and nearly 52,000 in health insurance from 2012 and before should not be Kentucky in 2015). This allows us to provide estimates by compared with more recent estimates. That is why subpopulations at higher levels of precision than would SHADAC uses baseline estimates from 2013 for our be possible using other federal surveys. An addition- underinsurance indicator (the question changes didn’t al advantage is that we are able to use the ACS public affect our other CPS indicator). In fact, 2013 was a tran- use file to create custom variables that are specific to sition year for the set of income questions, as both the analyzing the impact of the ACA. new and old questions were concurrently asked. The In this report, we use data from the ACS to estimate estimates we use in this study are based on the portion insurance coverage by type and to estimate the percent of the 2013 sample that used the new questions. uninsured by five different characteristics. When re- SHADAC used data from the CPS to estimate percent porting the distribution of insurance coverage, SHADAC underinsured and median out-of-pocket spending. The uses a mutually exclusive variable based on the concept definition for underinsurance used in this report is an of primary coverage; a hierarchy is imposed to avoid individual living in a family that has spent over 10% of its double counting people with multiple sources of cov- total income on healthcare expenses. erage. For adults, priority is given to Medicare cover- age, followed by employer based insurance (or military Medical Expenditure Panel Survey – Insurance coverage), Medicaid, and directly purchased coverage, Component (2012, 2013, 2014, 2015) respectively. For children, priority is assigned to ESI, The Medical Expenditure Panel Survey – Insurance followed by Medicaid/CHIP, individual coverage, and Component (MEPS-IC) is a federal survey spon- Medicare, respectively. For example, someone with sored by the U.S. Department of Health and Human coverage through their employer who also has direct- Services, Agency for Healthcare Research and Quality. ly purchased supplemental private coverage, would be The MEPS-IC collects information from public and considered as having employer coverage. private employers about the health insurance plans For analysis purposes, the definition of a family is they offer to employees, including benefits, costs, and important because eligibility for health insurance cov- other characteristics. The sample size in 2015 was erage is often based on family relationships and size. over 39,000 businesses at the national level. Summary SHADAC suggests defining a family using the concept of reports with detailed state-level tables for private sector a Health Insurance Unit (available here). This is partic- employers are released in July of each year following the ularly important for defining different income eligibility survey year. Unlike with the ACS and CPS, a public use categories. data file is not available from the MEPS-IC. For this report, SHADAC used data from the MEPS-IC to Current Population Survey (2013, 2014, 2015) estimate private-sector employer offer rates and premi- The Current Population Survey (CPS) is a federal survey ums. We accessed these estimates from the MEPS-IC conducted by the U.S. Census Bureau, sponsored jointly web site. with the U.S. Department of Labor/Bureau of Labor Statistics. The CPS Annual Social and Economic Supple- ment (ASEC), collected annually between the months 5353 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky National Health Interview Survey (2012, 2013, are available from the CDC for all states. The Kentucky 2014, 2015) BRFSS has an average sample size of more than 9,000 The National Health Interview Survey (NHIS) is a federal adults (ages 18+). survey sponsored by the Centers for Disease Control & SHADAC has changed the way we obtained the BRFSS Prevention (CDC) and the National Center for Health since the baseline report, opting to access and analyze Statistics (NCHS). The NHIS asks about health insurance the public use data for all estimates. To maintain con- coverage, health care utilization and access, health sistency and comparability, we have updated our base- conditions and behaviors, and general health status, as line estimates, as well. Since the initial study baseline well as many demographic and socioeconomic char- report, we added a new BRFSS indicator with the es- acteristics. It has a total sample of more than 103,000 timate of cigarette use in adults, added to the health in 2015 (the NHIS does not release state-level sample outcomes domain. This estimate reports the percentage sizes). of adults who have smoked 100 or more cigarettes in Summary reports, with state estimates for the 43 largest their lifetime and who currently smoke some days or states of types of coverage (including Kentucky) are every day. released six months after data collection. Data files with state-level and other geographic identifiers can National Survey on Drug Use and Health (2011, be accessed only through a Census Research Data 2012, 2013, 2014) Center (RDC). Access to data in Research Data Centers The National Survey on Drug Use and Health (NSDUH) is is only allowed after a proposal has been submitted and sponsored by the U.S. Department of Health and Human approved by NCHS and only to researchers who have Services’ Substance Abuse and Mental Health Services Special Sworn Status. SHADAC has an approved project Administration. The NSDUH collects information on the for accessing this restricted data in the RDC for the prevalence of tobacco, alcohol, and drug use, as well as purpose of posting estimates on our Data Center. mental health and treatment-related indicators among SHADAC used data from the NHIS to estimate nine differ- Americans ages 12 years and older. ent measures in the cost and access domains. Measures The Substance Abuse and Mental Health Services within the cost domain include trouble paying medical administration creates the estimates by pooling two bills, delayed needed care due to cost, and went without years of data. The estimates in this report are from the needed care due to cost. For the access domain, the time period 2011/2012, 2012/2013, and 2013/2014. measures include: usual source of care, provider visit Because 2014/2015 data we not available in time for in the last year, emergency department visit in the last this report, we used 2013/2014 as our time period for year, found doctor when needed, told provider accepts estimating the impacts of ACA implementation; insurance, and changes to medical drug use due to cost. however, it is important to note that because these For usual source of care, data estimates for 2012-2015 data are pooled pre- and post-ACA implementation, were updated since the last semi-annual report using a they may underestimate any effects of the law. For the revised methodology, so data may be differ from prior baseline report, we did not test for significance because reports. Data for ages 65+ were no longer available for the necessary data were not available; however, those NHIS indicators usual source of care and told provider data are now available and we included statistical tests accepts insurance, so these estimates were excluded in this report. The four measures included here under from this report. the access domain are: serious mental illness, any The changes to drugs due to cost measure includes mental illness, needed but did not receive illicit drug asking the doctor for cheaper medications, delaying abuse treatment and needed but did not receive alcohol refills, taking less medication than prescribed, skipping abuse treatment. Estimates on the prevalence of mental dosages, using alternative therapies, or buying medica- illness are based on people aged 18 or older. Estimates tions out of the country within the past year. The trouble on treatment of substance abuse provide information paying off medical bills measure includes people who for people aged 12 or older. are paying off medical bills within the past year. Healthcare Cost and Utilization Project (2012, Behavioral Risk Factor Surveillance System (2012, 2013, 2014) 2013, 2014, 2015) The Healthcare Cost and Utilization Project (HCUP), The Behavioral Risk Factor Surveillance System (BRFSS) is sponsored by the U.S. Department of Health and is a state-based survey sponsored by the CDC and Human Services’ Agency for Healthcare Research and the Kentucky Cabinet for Health and Family Services. Quality (AHRQ) and provides data on health statistics The BRFSS survey asks about health conditions, risk and information on hospital inpatient and emergency behaviors, preventive health practices, access to health department utilization. care, and health insurance coverage. State-level results FEBRUARY 2017 | WWW.SHADAC.ORG 5454 We use HCUP data for estimates in the quality domain, the National Vital Statistics System, the National Elec- including diabetes short-term admissions, hyperten- tronic Injury Surveillance System, the Census Bureau, sion admissions, asthma admissions, and death rate in and other sources. Users can create custom reports, low mortality DRGs. These indicators were previously charts, and maps using the built-in tools on the site, reported with data from a different source and due to and breakouts are available by state, gender, race, and potential differences in the methodology, these data age. The tool does not provide information on standard may not match similar data in prior reports. The diabe- errors and statistical testing of the differences between tes admission estimate reports the diabetes short-term estimates for Kentucky and the U.S. was not possible. complications admission rate for adults. The hyper- We use WISQARS to obtain information on premature tension estimate reports the hypertension admission deaths, which is an indicator that reports the years of rate for adults. The asthma estimate reports asthma in potential life lost (YPLL) before age 75, using the YPLL younger adults’ admission rate for adults ages 18 to 39. Age-Adjusted Rate and 2000 as the standard year. The death rate estimate reports the dying in the hospi- tal while getting care for a condition that rarely results National Vital Statistics Reports (2012, 2013, 2014, in death rate cases. Because these data are not based 2015) on a sample, there was no need for statistical testing of The National Vital Statistics Report, disseminated by the differences. CDC, contains data on low birth weight births, by race and Hispanic origin of the mother in each U.S. state. Low Youth Risk Behavior Surveillance System (2013, birth weight is categorized as weighing less than 2,500 2015) grams (5 lb. 8 oz.). Because these data are not based The Youth Risk Behavior Surveillance System (YRBSS) on a sample, (the system records all known occurrences survey asks students in grades 9-12 about tobacco use, of low birth weight, and reports are released annually), sexual behaviors, alcohol and drug use, diet and exer- there was no need for statistical testing of differences. cise, obesity, asthma, and behaviors related to violence and injury. Kentucky also administers a middle-school Kentucky Outpatient & Inpatient Hospital Adminis- version for grades 6-8. The YRBSS is given to a sample trative Claims Data (2012, 2013, 2014, 2015) of students, and is a bi-annual survey conducted in The Kentucky Outpatient & Inpatient Hospital Admin- odd-numbered years, with results released the year istrative Claims Data were provided by the Kentucky following the survey. In 2015, the Kentucky sample Cabinet for Health and Family Services. For our study, from the YRBSS included more than 2,500 students. we use charges for self-pay and charity care as a proxy The source for the indicators obtained for this source is for uncompensated care. In these data we are not able online data from the CDC. to discern between paid and unpaid charges. Since hos- We include the following three measures from the pitals are likely to receive some payment for at least survey: unprotected sex among high school students of portion of self-pay charges, we acknowledge that in the quality domain, as well as obesity rates and ciga- not all self-pay charges become “uncompensated”. For rette use in the health outcomes domain. The estimate the purposes of estimating uncompensated care, we on unprotected sex reports the percentage of sexu- assume that the majority of the self-pay charges are not ally active high school students who did not use any paid in full. Unlike for our baseline report, we obtained method to prevent pregnancy during their last sexual these data directly from the Cabinet, and the data may intercourse. The obesity measure reports the percent- differ from the baseline due to different methodology. age of students who were above the 95th percentile Additionally, the Cabinet has revised its 2015 data, so for Body Mass Index based on gender and age specific those charity care and self-pay charges in this report reference data from the 2000 CDC growth charts. The differ from earlier reports. cigarette measure reports the percentage of high school students who currently smoked cigarettes, on at least Kentucky Breastfeeding at Hospital Discharge Data one day during the 30 days before the survey. Like (2012, 2013, 2014, 2015) the adult cigarette use measure, this indicator did not The Kentucky Cabinet for Health and Family Services appear in our baseline report because it was added later provided hospital data on initiation of breastfeeding in the study. prior to hospital discharge for 2012 through 2015. The 2014 and 2015 data are preliminary and have not yet Web-based Injury Statistics Query and Reporting been finalized as of February 2017. The source for this System (2012, 2013, 2014, 2015) indicator has changed since the baseline report due to The Web-based Injury Statistics Query and Reporting changes in the availability of the prior source. Because System (WISQARS™) is the CDC’s public-use database of these data are not based on a sample, (the Common- information on injury, violent death, and cost of injury wealth records all known births), there was no need for in the United States. The database pulls in data from statistical testing of differences. 5555 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky Kentucky Medicaid Enrollment and Services Data Reporting and analysis of data (2014, 2015, 2016) Suppression rules depended on the source of the data This report also contains data provided by the Cabinet and the availability of measures of uncertainty and/or for Medicaid Enrollment and Services from the first sample sizes. In the ACS and CPS where we used public quarter of 2014 through the third quarter of 2016. use files, we suppressed data when the relative stan- These data include only traditional income-based dard error was greater than 30%. Estimates from the Medicaid and ACA expansion Medicaid enrollees ages NHIS are suppressed if either the number of sample 19-64. We exclude special enrollee categories: Medi- cases was too small or the relative standard error was care-Medicaid dual eligible; foster, former foster, and greater than 30%. In cases where standard errors were kinship care; intermediate care facility, nursing home, not available, we did not suppress any estimates. Lastly, and hospice populations; Medicare savings and special we did not include some trend estimates due to recent populations; SSI recipients; waiver populations, or changes in the questions of some federal surveys that incomplete claims that do not show enrollee category. made it difficult to compare data points over time (e.g., Dental services represent preventive dental visits only; the CPS). other dental visits are excluded. Because this is not a It should be noted that we lacked the necessary survey, but rather an inventory of this occurrences, information to perform an “overlap adjustment” to there is no sampling or sample size and no need for our statistical tests. Since we are comparing Kentucky’s statistical testing of differences. estimates to national estimates (which include Kentuck- Kentucky Health Reform Survey (2016) ians), the proportion of Kentuckians in the population The Kentucky Health Reform Survey (K-HRS) was con- considered in the estimate should be taken into account. ducted by SHADAC and the University of Cincinnati However, this specific information was not available for Institute for Policy Research from March-May 2016. most estimates. By not conducting an overlap adjust- The methodology and a substantial part of the survey ment we are slightly less likely to report that a difference instrument were based on the existing Kentucky Health is statistically significant. Issues Poll (KHIP), allowing for comparisons of the estimates from the K-HRS to prior KHIP estimates and potentially future KHIP estimates yet more depth in several policy-relevant areas than possible in the KHIP. Survey questions were selected in consultation with the Foundation and study Oversight Committee, with over- arching goals of maintaining consistency with the KHIP to allow trend analyses and investigating key compo- nents of ACA implementation in Kentucky, such as the Commonwealth’s kynect state-based marketplace. The dual-frame (landline and cell phone) survey sampled non-elderly adult Kentuckians for a total of 1,639 interviews. The measures in this report include data on uninsurance and coverage types, concern about losing coverage, forgone or delayed care due to cost, dental coverage and care, and emergency department use. Kentucky Health Issues Poll (2012, 2014) The Kentucky Health Issues Poll (KHIP) is an annual tele- phone opinion poll of Kentucky adults commissioned jointly by the Foundation for a Healthy Kentucky and Interact for Health, and conducted by the Institute for Policy Research at the University of Cincinnati (UC-IPR). The KHIP has been conducted annually each fall since 2008 and provides a snapshot of Kentuckians’ views on various health topics. In this report, KHIP data for 2012 and 2014 are used for comparison with 2016 K-HRS data from questions aligned with the KHIP instrument. For the K-HRS/KHIP analysis, UC-IPR provided SHADAC with KHIP estimates based on the non-elderly adult sample, to ensure comparability of the estimates. FEBRUARY 2017 | WWW.SHADAC.ORG 5656 VII. ENDNOTES 1 Though Arkansas is not technically a border state, we include it because it is often compared to Kentucky due to similarities in health status, demographics, and state policies. 2 Institute of Medicine. (2009). America’s uninsured crisis: Consequences for health and health care. (Report Brief). Available at: http://iom.nationalacademies.org/~/media/Files/Report%20Files/2009/Americas-Uninsured-Crisis-Consequences-for-Health- and-Health-Care/Americas%20Uninsured%20Crisis%202009%20Report%20Brief.pdf 3 State Heath Access Data Assistance Center (SHADAC). (2013) State-level Trends in Employer-sponsored Health Insurance: A State-by-state Analysis. Available at: http://www.shadac.org/sites/default/files/Old_files/shadac/publications/ESI_Report_2013.pdf 4 United States Census Bureau. (2012). American FactFinder. Available at: https://factfinder.census.gov/ 5 The state-comparison charts in this report indicate which of Kentucky’s neighboring states have opted to expand their Medicaid programs as part of ACA (Arkansas, Illinois, Indiana, Ohio and West Virginia). The only one of these states that did not implement its expansion in 2014 was Indiana, which expanded its Medicaid program via a Section 1115 waiver in 2015. 6 In 2015, for a 1-person household, 100% of the FPG was $11,770 and 138% of FPG was $16,243. The categories are adjusted for family size; in 2014, 100% of the FPG for a 4-person household was $24,250 and 138% of FPG was $33,465. 7 U.S. Department of Health and Human Services, Assistant Secretary for Planning and Evaluation. (2015). 2015 HHS Poverty Guidelines: One Version of the U.S. Federal Poverty Measure. Available at: https://aspe.hhs.gov/2015-poverty-guidelines 8 Because of rounding, the percentage point differences described in the report may not match those calculated from the estimates found in the figures. 9 SHADAC. (Undated). Measuring the Adequacy of Coverage or Underinsurance. Available at: http://www.shadac.org/files/ MeasureUnderinsurance.pdf 10 Blewett, L., Ward, A., & Beebe, T. (2006). How much health insurance is enough? Revisiting the concept of underinsurance. Medical Care Research and Review, 63(6). Available at: http://journals.sagepub.com/doi/abs/10.1177/1077558706293634 11 SHADAC uses 10% of annual household income spent on health care based on the definition used by the National Center for Health Statistics (NCHS) 12 Agency for Healthcare Research and Quality. (2010). National Health care Quality Report. Chapter 9: Access to Care. Available at: http://archive.ahrq.gov/research/findings/nhqrdr/nhqr10/Chap9.html 13 The underinsurance indicator uses 2013 as the baseline year instead 2012. This is because the U.S. Census Bureau implemented new income questions starting with the 2013 CPS, so 2012 underinsurance estimates would not be comparable to later years. Institute of Medicine, Committee on Monitoring Access to Personal Health Care Services. (1993). Access to Health Care in America. 14 Washington, DC. National Academy Press. Available at: https://www.nap.edu/catalog/2009/access-to-health-care-in-america 15 Call, K., McAlpine, D., Garcia, C., Shippee, N., Beebe, T., Adeniyi, T., & Shippee, T. (2014). Barriers to Care in an Ethnically Diverse Publicly Insured Population: Is Health Care Reform Enough? Medical Care, 52(8). Available at: http://journals.lww.com/lww-medicalcare/ Abstract/2014/08000/Barriers_to_Care_in_an_Ethnically_Diverse_Publicly.8.aspx 16 Prior reports noted that SHADAC was investigating the feasibility of adding another measure: wait time to see a primary care provider. However, adding this measure was not possible. 17 U.S. Department of Health and Human Services, Assistant Secretary for Planning and Evaluation. (Undated). Health System Measurement Project: Percentage of People Who Have a Specific Source of Ongoing Medical Care. 18 DeVoe, J., Tillotson, C., Wallace, L., Lesko, S., & Pandhi, N. (2012). Is health insurance enough? A usual source of care may be more important to ensure a child receives preventive health counseling. Maternal and Child Health Journal, 16(2), 306-315. Available at: http://link.springer.com/article/10.1007/s10995-011-0762-4 19 Centers for Medicare and Medicaid Services. (2016). More than 10 million people with Medicare have saved over $20 billion on prescription drugs since 2010. Available at: https://www.cms.gov/Newsroom/MediaReleaseDatabase/Press-releases/2016-Press-releases- items/2016-02-08.html 20 Blum, Jonathan. (2010). What is the donut hole? The Medicare Blog. Available at: https://blog.medicare.gov/2010/08/09/ what-is-the-donut%C2%A0hole/ 21 Centers for Medicare and Medicaid Services. (2015). Part D donut hole savings by state 2015 YTD. Available at: https://downloads.cms. gov/files/Part%20D%20dount%20hole%20savings%20by%20state%20YTD%202015.pdf 22 U.S. Centers for Disease Control & Prevention. (2009). The Power of Prevention: Chronic disease…the public health challenge of the 21st century. Available at http://www.cdc.gov/chronicdisease/pdf/2009-Power-of-Prevention.pdf 5757 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky 23 National Association of State Mental Health Program Directors Medical Directors Council. (2006.) Morbidity and Mortality in People with Serious Mental Illness. Available at http://www.nasmhpd.org/sites/default/files/Mortality%20and%20Morbidity%20Final%20Report%20 8.18.08.pdf 24 Substance Abuse and Mental Health Services Administration. (2014). Substance Use and Mental Health Estimates from the 2013 National Survey on Drug Use and Health: Overview of Findings. Available at: http://www.samhsa.gov/data/sites/default/files/NSDUH-SR200- RecoveryMonth-2014/NSDUH-SR200-RecoveryMonth-2014.htm 25 SHADAC. (2016). Substance Use and the ACA in Kentucky. Available at: https://www.healthy-ky.org/res/images/resources/Full-Substance- Use-Brief-Final_12_28.pdf 26 SHADAC. (2016). Substance Use and the ACA in Kentucky. Available at: https://www.healthy-ky.org/res/images/resources/Full-Substance- Use-Brief-Final_12_28.pdf 27 Institute of Medicine, Committee on an Oral Health Initiative. (2011). Advancing Oral Health in America. The National Academies Press. Available at: http://www.hrsa.gov/publichealth/clinical/oralhealth/advancingoralhealth.pdf 28 Giannobile, W., Braun, T., Caplis, A., Doucette-Stamm, L., Duff, G. & Kornman, K. (2013). Patient Stratification for Preventive Care in Dentistry. Journal of Dental Research, 92(8), 694-701. Available at: http://journals.sagepub.com/doi/abs/10.1177/0022034513492336 29 SHADAC analysis of 2015 American Community Survey estimates. 30 The 2015 charity care and self-pay charges have been revised by the Kentucky Cabinet for Health and Family Services. An earlier report found a drop of 77% to $552 million, but this has been revised as a 67% drop to $786 million in 2015. 31 Lukanen, E., Schwehr, N., & Hest, R. (2017). State-Level Trends in Employer-Sponsored Health Insurance, 2011-2015. Minneapolis, MN: State Health Access Data Assistance Center. Available at: www.shadac.org/MEPSESIReport2016 32 SHADAC. (2016). High-deductible Health Insurance in Kentucky. Available at: https://www.healthy-ky.org/res/images/resources/KY-high- deductible-brief-Final-Combined.pdf 33 Torio, C., Elixhauser, A., & Andrews, R. (2013). Trends in Potentially Preventable Hospital Admissions among Adults and Children, 2005– 2010. Agency for Healthcare Research and Quality. Healthcare Cost and Utilization Project. Statistical Brief #151. Available at: https://www.hcup-us.ahrq.gov/reports/statbriefs/sb151.pdf 34 Commonwealth of Kentucky, Cabinet for Health and Family Services. (2012). Prevention Quality Indicators. Available at: http://chfs.ky.gov/ohp/healthdata/pqis.htm 35 Centers for Medicare and Medicaid Services. (2014). Acute Inpatient PPS. Available at: https://www.cms.gov/Medicare/Medicare-Fee-for- Service-Payment/AcuteInpatientPPS/index.html?redirect=/AcuteInpatientPPS 36 Agency for Healthcare Research and Quality. (2009). Patient Safety Indicators Technical Specifications PSI #2 Death in Low-mortality DRGs. Available at: https://www.qualityindicators.ahrq.gov/Downloads/Modules/PSI/V41/TechSpecs/PSI%2002%20Death%20in%20Low- mortality%20DRGs.pdf 37 U.S. Department of Health and Human Services, Office on Women’s Health. (2014). Why Breastfeeding is Important. Available at: https://www.womenshealth.gov/breastfeeding/breastfeeding-benefits.html 38 U.S. Centers for Disease Control and Prevention. (2010). Healthy People 2020 Breastfeeding Objectives for the Nation. Available at: http://www.cdc.gov/breastfeeding/policy/hp2010.htm?topicId=26 39 U.S. Centers for Disease Control and Prevention. (2010). Is Low Birthweight a Health Problem? 40 U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion. (2015). Healthy People 2020: Maternal, Infant, and Child Health Objectives. Available at: https://www.healthypeople.gov/2020/topics-objectives/topic/ maternal-infant-and-child-health/objectives 41 The U.S. Preventive Services Task Force recommends men and women ages 50 to 75 have a fecal blood test annually; a sigmoidoscopy every five years, plus a fecal blood test every three years; or a colonoscopy every 10 years. For this indicator, we provide an approximation of whether individuals have met this recommendation. For more information on this, see the section on the Behavioral Risk Factor Surveillance System in Appendix I. 42 Redmond K., Kanotra S., Siameh S., Jones J., Thompson B., Thomas-Cox S. (2015). Understanding Barriers to Colorectal Cancer Screening in Kentucky. Accessible at: http://www.cdc.gov/pcd/issues/2015/14_0586.htm 43 The state and partners have made efforts to increase colorectal cancer screening through programs including the Kentucky Colon Cancer Screening Program, created by the Kentucky General Assembly in 2008, and the 2012 creation of a public-private partnership between the state and the Kentucky Cancer Foundation to fund screenings for low-income uninsured Kentuckians. 44 SHADAC. (2016). Annual Report: Study of the Impact of the ACA Implementation in Kentucky. Available at: https://www.healthy-ky.org/ res/images/resources/FINAL-FULL-Annual-Report-2.29-1-.pdf 45 Klein, J., Barratt, M., Blythe, M., Braverman, P., Diaz, A., Rosen, D., & Wibbelsman, C. (2007). Contraception and Adolescents. Pediatrics, 120(5). Available at: http://pediatrics.aappublications.org/content/120/5/1135 FEBRUARY 2017 | WWW.SHADAC.ORG 5858 46 U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion. (2015). Healthy People 2020: Foundational Health Measures: Determinants of Health. Available at: http://www.healthypeople.gov/2020/about/foundation-health-measures/Determinants-of-Health 47 Boban, M., Francis, L., Kayalar, A., & Cone, J. (2008). Obesity: Effects of Cardiovascular Disease and its Diagnosis. Journal of the American Board of Family Medicine, 21(6). Available at http://www.jabfm.org/content/21/6/562.full 48 National Institutes of Health, National Heart, Lung, and Blood Institute. (2013). Why Obesity is a Health Problem. Available at: https://www.nhlbi.nih.gov/health/educational/wecan/healthy-weight-basics/obesity.htm 49 Robert Wood Johnson Foundation. (2014). The State of Obesity: Better Policies for a Healthier America, 2014. Available at: http://www.rwjf.org/content/dam/farm/reports/reports/2014/rwjf414829 50 The Centers for Disease Control and Prevention. (Undated). Health Effects of Cigarette Smoking. Available at: https://www.cdc.gov/ tobacco/data_statistics/fact_sheets/health_effects/effects_cig_smoking/ 51 Robert Wood Johnson Foundation’s National DataHub. Indicator: Percent of adults who have smoked 100 or more cigarettes in their lifetime and who currently smoke some days or every day. Available at: http://www.rwjf.org/en/how-we-work/rel/research-features/rwjf- datahub/national.html#q/scope/national/ind/59/dist/0/char/0/time/33/viz/rankings/cmp/brkdwn 52 Robert Wood Johnson Foundation’s National DataHub. Indicator: Percent of high school students who have smoked at least one cigarette in the past 30 days. Available at: http://www.rwjf.org/en/how-we-work/rel/research-features/rwjf-datahub/national.html#q/scope/national/ ind/28/dist/0/char/0/time/33/viz/rankings/cmp/stcmp 53 The Centers for Disease Control and Prevention. (Undated) Coverage for Tobacco Use Cessation Treatments. Available at: https://www.cdc.gov/tobacco/quit_smoking/cessation/pdfs/coverage.pdf 54 Foundation for a Healthy Kentucky and Interact for Health. (2017) “Nearly 4 in 10 Kentucky Young Adults Have Tried and E-cigarette.” Available at: http://files.constantcontact.com/1b4946a6001/20ffc83c-fd08-428e-b475-d45165ca2b40.pdf 55 Arrazola, R., Singh, T., Corey, C., Husten, C., Neff, L., Apelberg, B., Bunnell, R., Choiniere, C., King, B., Cox, S., McAfee T., Caraballo, R. (2015) Tobacco Use Among Middle and High School Students — United States, 2011–2014. Available at: https://www.cdc.gov/mmwr/preview/ mmwrhtml/mm6414a3.htm?s_cid=mm6414a3_w 56 Longo, D. (2016) “The Health Effects of Electronic Cigarettes.” The New England Journal of Medicine. Available at: http://www.nejm.org/doi/full/10.1056/NEJMra1502466 57 Idler, E. & Benyamini, Y. (1997). Self-Rated Health and Mortality: A Review of Twenty-Seven Community Studies. Journal of Health and Social Behavior, 21-37. Available at: http://www.jstor.org/stable/2955359?seq=1#page_scan_tab_contents 58 The Centers for Disease Control and Prevention. (2015). Chronic Disease Prevention and Health Promotion. Available at: http://www.cdc.gov/chronicdisease/ 59 SHADAC. (2016). Annual Report: Study of the Impact of the ACA Implementation in Kentucky. Available at: https://www.healthy-ky.org/ res/images/resources/FINAL-FULL-Annual-Report-2.29-1-.pdf 60 National Center for Health Statistics, Health Indicators Warehouse. (Undated). Years of potential life lost before age 75 (per 100,000). Available at: http://www.healthindicators.gov/Indicators/Years-of-potential-life-lost-before-age-75-per-100000_3/Profile 61 Sonier, J., Boudreaux, M., & Blewett, L. (2013). Medicaid ‘Welcome Mat’ Effect Of Affordable Care Act Implementation Could Be Substantial. Health Affairs, 32(7), 1319-1325. Available at: http://content.healthaffairs.org/content/32/7/1319.abstract 62 The Centers for Disease Control and Prevention. (2013). Colorectal Cancer Screening Guidelines. Available at: https://www.cdc.gov/ cancer/colorectal/basic_info/screening/guidelines.htm U.S. Preventive Services Task Force. (2016). Breast Cancer Screening. Available at: https://www.uspreventiveservicestaskforce.org/Page/ 63 Document/UpdateSummaryFinal/breast-cancer-screening 64 The American Cancer Society medical and editorial content team. (2016). American Cancer Society Guidelines for the Early Detection of Cancer. Available at: http://www.cancer.org/healthy/findcancerearly/cancerscreeningguidelines/ american-cancer-society-guidelines-for-the-early-detection-of-cancer 65 The Centers for Disease Control and Prevention. (2015). Testing Recommendations for Hepatitis C Virus Infection. Available at: https://www.cdc.gov/hepatitis/hcv/guidelinesc.htm 66 The Centers for Disease Control and Prevention. (2015). Distribution of Age at Diagnosis of Diabetes Among Adult Incident Cases Aged 18-79 Years, United States, 2011. Available at: https://www.cdc.gov/diabetes/statistics/age/fig1.htm 67 Kentucky Cabinet for Health and Family Services. (2015). 2015 Kentucky Diabetes Report. Available at: http://chfs.ky.gov/NR/ rdonlyres/7D367886-671C-435E-BCF4-B2A740438699/0/2015DiabetesReportFinal.pdf 68 In the case of births covered by Medicaid expansion, these increased by more than 1,400 — from 122 in Quarter 1 of 2014 to 1,551 in Quarter 3 of 2016 — although births in traditional Medicaid dropped by more than 1,500. This was likely due in part to changes in how pregnant women enroll in Medicaid. Prior to the ACA, many women would enroll in traditional Medicaid after learning they were pregnant. 5959 STATE HEALTH ACCESS DATA ASSISTANCE CENTER FINAL REPORT Study of the Impact of the ACA Implementation in Kentucky But since 2014, more women are covered by Medicaid before becoming pregnant, so their births would mostly be covered by expansion Medicaid rather than traditional Medicaid. 69 Kentucky Health Benefit Exchange. (2017). Summary of Benefits and Coverage (SBC). Available at: http://healthbenefitexchange.ky.gov/ Pages/Summary-of-Benefits-and-Coverage-(SBC).aspx 70 U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation. (2016). Health Insurance Marketplace 2016 Open Enrollment Period: Final Enrollment Report. Available at: https://aspe.hhs.gov/sites/default/files/pdf/187866/ Finalenrollment2016.pdf 71 American Academy of Actuaries. (2014). State of Exchanges. Available at: https://www.actuary.org/files/ LunchtimeWebinarSeriesStateExchanges_May29.pdf 72 Center for Health Workforce Studies. (2016). Oral Health in Kentucky. Available at: http://www.oralhealthworkforce.org/wp-content/ uploads/2016/02/Oral_Health_Kentucky_Technical_Report_2016.pdf 73 The 2014 KHIP was conducted from October 8 – November 6, 2014. 74 While we have suppressed the estimates of other usual sources of care (e.g., urgent care center, retail store clinic, other) based on National Center for Health Statistics criteria due to concerns about their reliability, we report the estimate for Emergency Departments because of concern in Kentucky about possible overuse of hospital EDs. 75 SHADAC. (2017).Quarterly Snapshot: July-September 2016. Available at: https://www.healthy-ky.org/res/images/resources/FINAL-7th- Quarterly-Snapshot-7.pdf 76 SHADAC. (2015). Issue Brief: ACA Improves Health Insurance Coverage for Kentucky Children. Available at: https://www.healthy-ky.org/ res/images/resources/SHADAC_KY-Children-Issue-Brief_Final-10.16.2015.pdf 77 SHADAC. (2017).Quarterly Snapshot: July - September 2016. Available at: https://www.healthy-ky.org/res/images/resources/FINAL-7th- Quarterly-Snapshot-7.pdf 78 SHADAC. (2016). Semi Annual Report 2016. Available at: https://www.healthy-ky.org/res/images/resources/FINAL-Sept-2016-Semi- Annual-report.pdf 79 SHADAC. (2016). Substance Use and the ACA in Kentucky. Available at: https://www.healthy-ky.org/res/images/resources/Full-Substance- Use-Brief-Final_12_28.pdf 80 SHADAC. (2015). Quarterly Snapshot: April – June 2015. Available at: https://www.healthy-ky.org/res/images/resources/SHADAC_ACA- Impact-Study_Quarterly-Snapshot-Q22015_0.pdf 81 SHADAC. (2016). Quarterly Snapshot: July – September 2015. Available at: https://www.healthy-ky.org/res/images/resources/FINAL- Quarterly-Snapshot-January-2016.pdf 82 SHADAC. (2017).Quarterly Snapshot: July-September 2016. Available at: https://www.healthy-ky.org/res/images/resources/FINAL-7th- Quarterly-Snapshot-7.pdf 83 SHADAC. (2016). Annual Report: Study of the Impact of the ACA Implementation in Kentucky. Available at: https://www.healthy-ky.org/ res/images/resources/FINAL-FULL-Annual-Report-2.29-1-.pdf FEBRUARY 2017 | WWW.SHADAC.ORG 6060 6161 STATE HEALTH ACCESS DATA ASSISTANCE CENTER