Pennsylvania Patient Safety Advisory Leveraging Healthcare Policy Changes to Decrease Hospital 30-Day Readmission Rates ABSTRACT $15 billion.4 In response to rising healthcare costs, Hospitalizations account for nearly one-third of the the Obama administration’s 2010 budget proposes $2 trillion spent on healthcare in the United States a combination of incentives and penalties to reduce annually. Nearly 20% of these hospitalizations are hospital readmission rates, thereby saving approxi- rehospitalizations occurring within 30 days of dis- mately $26 billion over 10 years to help pay for charge. In 2008, there were 57,852 readmissions in healthcare reform.5 Pennsylvania, amounting to approximately $2.5 billion in charges. Thirty-eight percent of these readmissions On a national level, the Centers for Medicare & Medicaid Services (CMS) posts 30-day, all-cause, risk- were related to complications or infections. From adjusted readmission rates for three conditions on its June 2004 through August 2009, 1,791 events of Web site: (1) heart failure, (2) acute myocardial infarc- readmission to the emergency department within 48 tion, and (3) pneumonia. Participating hospitals are hours were reported to the Pennsylvania Patient Safety classified as “better than U.S. national rate,” “no dif- Authority, 8% of which were Serious Events (indicating ferent than U.S. national rate,” or “worse than U.S. harm to the patient). In June 2008, the Medicare Pay- national rate.” Exclusionary criteria include patients ment Advisory Commission calculated the annual cost readmitted for the purpose of planned cardiac treat- of readmissions to the Medicare program at $15 bil- ment, patients who leave the hospital against medical lion. The Obama administration’s 2010 budget aims advice, and hospitals with fewer than 25 cases. These to reduce Medicare readmissions in order to fund measures are updated quarterly.6 Historically, hos- healthcare reform. The Centers for Medicare & Medic- pitals could only track readmissions back to their aid Services posts hospital readmission rates for three own facilities; collecting and sharing multihospital conditions on its Web site. National readmission rates aggregate data may shed new light on the readmission show a wide variance across states as well as vari- issue. MedPAC has recommended that CMS confi- ance between facilities within the same state. This high dentially report readmission rates and resource use variance rate suggests that significant financial sav- around hospitalization episodes (30-day periods) to ings could be realized if best practices for preventing hospitals and physicians for two years. Beginning in unnecessary readmissions were adopted. This article the third year, providers’ relative resource use should reviews both national policy related to readmissions be publicly disclosed. To encourage providers to col- and best practices that could help hospitals reduce laborate and better coordinate care, MedPAC believes readmission rates while simultaneously improving that payments should be reduced for those hospitals patient-centered care and patient safety. (Pa Patient with relatively high readmission rates for select condi- Saf Advis 2010 Mar;7(1):1-8.) tions and favors shared financial accountability (gain sharing) between physicians and hospitals.4 Jencks et al. conducted a retrospective review of Background Policy Medicare fee-for-service claims data from October 2003 to September 2004 to analyze Medicare 30-day Hospitalizations account for nearly one-third of the readmission rates in an effort to describe patterns of $2 trillion annual cost of healthcare in the United readmissions and the relation of rehospitalizations to States.1,2 In the majority of cases, hospitalization is demographic characteristics of the patients and of the necessary and appropriate. However, experts estimate hospitals. Their findings revealed that nearly 20% of that as many as 20% of hospitalizations are rehos- hospitalized Medicare beneficiaries were readmitted pitalizations within 30 days of discharge.1,2 These to the hospital within 30 days and 34% were read- rehospitalizations are costly, potentially harmful, mitted within 90 days. Additionally, they found that and often preventable. The Pennsylvania Patient nearly 69% of patients who had been admitted with Safety Authority received more than 3,500 reports a medical diagnosis and 53% of patients who had of hospital readmissions from June 2004 through been admitted with a surgical diagnosis were either August 2009. According to data from the Agency for readmitted or had died within one year following the Healthcare Research and Quality’s (AHRQ) Health- initial hospitalization. Surprisingly, less than half of the care Cost and Utilization Project (HCUP), in 2006, Medicare patients who had been readmitted to the hospital nearly 4.4 million hospital admissions, totaling nearly within 30 days had visited an outpatient physician before the $30.8 billion, could have been potentially prevent- readmission. It was estimated that only 10% of the read- able with timely and effective ambulatory care or missions were likely to have been planned, leaving 90% adequate patient self-management of the condition.3 of the readmissions potentially preventable, at a cost of Additionally, nearly one in five Medicare admissions $17.4 billion to the Medicare program in 2004.7 (18%) was for a potentially preventable condition.3 More recently, in June 2008, the Medicare Payment The Table illustrates the high variability of read- Advisory Committee (MedPAC) calculated the annual mission rates across states. This high variance rate cost of readmissions to the Medicare program at suggests that significant financial savings could be Vol. 7, No. 1—March 2010 REPRINTED ARTICLE - ©2010 Pennsylvania Patient Safety Authority Page 1 Pennsylvania Patient Safety Advisory realized if best practices for preventing unnecessary readmissions were adopted. Table. Rates of Rehospitalization In 2007, the Commonwealth Fund studied key within 30 Days after Hospital Discharge* indicators of health system performance, including PERCENTAGE NUMBER OF Medicare 30-day readmissions in 2003, and found RANGE STATES IN RANGE a two-fold variation in rates of hospital readmission 13.3% to 17.5% 13 within 30 days among Medicare beneficiaries, from 17.6% to 19.1 % 14 24% in Louisiana and Nevada to 13% in Vermont 19.2% to 20.1% 13, including and Wyoming. Pennsylvania’s Medicare 30-day read- Pennsylvania at 19.7% mission rate in 2003 was 20.1%, ranking 43rd of 50 20.2% to 23.2% 10 states. If Pennsylvania’s performance improved to the *The rates include all patients in fee-for-service Medicare programs level of the best performing state on this indicator, who were discharged between October 1, 2003, and September 30, 2004. 13,866 fewer readmissions would occur, saving the Source: Jencks SF, Williams MV, Coleman EA, et al. Medicare program nearly $164 million annually.8 Rehospitalizations among patients in the Medicare fee-for-service program. N Eng J Med 2009 Apr 2;360(14):1418-28. More recently, Friedman et al. conducted a retrospec- tive review of nearly 1.5 million adult surgery patients initially treated in 1,088 short-stay hospitals in 2004, all at risk for one of nine patient safety events (see Containment Council (PHC4).12 Rates are calcu- the box “Nine Patient Safety Events”).* Their find- lated for all-cause readmissions and readmissions for ings showed that patients who experienced one of the complications or infections. They are categorized by nine patient safety events had a higher incidence of condition into “significantly higher than the expected hospital 30-day readmissions than those who did not rate,” “not significantly different than the expected experience a patient safety event (11% versus 16%; risk rate,” and “significantly lower than the expected rate.” adjusted result for readmission within one month 1.20 Exclusionary criteria include hospitals with less than [p < 0.01]).9 The connection between patient safety five cases, nonadult cases, and missing or invalid events and hospital readmissions, while not surprising, discharge status, as well as patients who leave against further complicates the preventable 30-day hospital medical advice. readmission scenario. Furthermore, 30-day readmis- In 2008, there were 57,852 readmissions for any sion rates have been considered a marker of low reason in the categories covered by the report. These quality care and suboptimal patient safety.10 readmissions resulted in nearly $2.5 billion in charges These recent studies have helped land 30-day readmis- and 350,000 additional hospital days. Thirty-eight sions on Medicare’s program-integrity radar screen. percent (22,094) of the readmissions were for compli- In fact, CMS’ program integrity contractors (recovery cation or infection, amounting to approximately audit contractors) will continue postpayment audits $1.1 billion in charges and 157,000 additional to identify hospital readmissions within 30 days of a hospital days.13 For the 21 conditions for which hospital discharge.10 According to MedPAC’s plan,4 readmissions are calculated, the overall Pennsylvania once 30-day readmission rates are systematically calcu- readmission rate was 18.9%; respiratory failure with lated and analyzed, financial penalties and incentives mechanical ventilation was the highest at 27.6%, and to reduce 30-day readmissions will follow. vaginal hysterectomy was the lowest at 3%. (For a visual summary of the background information, see The Pennsylvania Environment “Timeline of 30-day Avoidable Readmission Informa- In 2007, Governor Rendell introduced “Prescrip- tion,” available on the Authority’s Web site.) tion for Pennsylvania,” a statewide healthcare reform agenda focused on reducing costs, providing access to Authority Data universal coverage, improving quality, and decreasing The Authority received more than 3,500 reports inefficiencies in the Pennsylvania healthcare system. related to readmissions from June 2004 through His plan identified avoidable readmissions as an August 2009. However, this is just the “tip of the ice- area ripe for both quality improvement and financial berg,” as only readmissions associated with Incidents savings.11 or Serious Events are reported in the Authority’s data- base. For example, 1,791 events of “unplanned return In Pennsylvania, rates of hospital readmission (i.e., an to emergency department (ED) in 48 hours requiring acute care hospitalization for any reason which occurs admission” were reported between June 2004 and within 30 days of the original hospitalization) are August 2009. calculated for 21 medical and surgical conditions and are published by the Pennsylvania Health Care Cost The Authority reviewed 392 events related to hospital readmissions reported from January through August 2009, 120 of which were reported as Serious Events * Patient safety events, as specified in software in the public (those events which harm patients) (31%) and 272 domain by AHRQ. The main data sources are seven state-wide databases of hospitalizations in 2004, maintained by HCUP. of which were reported as Incidents (near-misses) (Cited 2009 Sep 21; available from Internet: http://www. (69%). Common themes among the hospital read- qualityindicators.ahrq.gov/.) mission reports included ineffective communication Page 2 REPRINTED ARTICLE - ©2010 Pennsylvania Patient Safety Authority Vol. 7, No. 1—March 2010 Pennsylvania Patient Safety Advisory Patient discharged; readmitted one week later. Dur- Nine Patient Safety Events ing the admission assessment, it was discovered that 1. Iatrogenic pneumothorax patient had [had] no anticoagulant education [during previous admission]. 2. Selected infections due to medical care The patient had a transurethral resection of the pros- 3. Postoperative hemorrhage or hematoma tate and was ordered an antibiotic postoperatively. The 4. Postoperative physiologic and metabolic patient never took the ordered medication, which con- derangements tributed to a readmission due to back pain. The patient was found to have an UTI [urinary tract infection]. 5. Postoperative respiratory failure An example of ineffective communication between 6. Postoperative pulmonary embolism or deep providers across healthcare settings is as follows: vein thrombosis The patient met discharge criteria and was discharged 7. Postoperative sepsis to a personal care home after leg surgery. He fell at 8. Postoperative wound dehiscence after the personal care home and was sent back to the hos- abdominopelvic surgery pital [the next day]. The physician from the personal care home stated he did not think a return to the 9. Accidental puncture or laceration home should have occurred on a Saturday because the Source: Friedman B, Encinosa W, Jiang HJ, et al. home did not have licensed staff on the weekend. Do patient safety events increase readmissions? Ineffective Transitions of Care Med Care 2009 May;47(5):583-90. Examples of ineffective transitions of care within hos- pitals include the following: among providers, between providers and patients, and A patient was transferred from the medical surgical between providers across healthcare settings and inad- unit to the inpatient rehab center in the mid-after- equate transitions of care, both within hospitals and noon. The patient was sent to the ED that evening between hospitals and community settings. The report with shortness of breath and hypoxia. The patient was narratives reveal the breadth of reasons why patients readmitted to facility secondary to the respiratory con- experience potentially preventable readmissions. dition. The event was reviewed, and staff confirmed that the patient was receiving oxygen at 4L/min via Ineffective Communication nasal cannula prior to discharge. Oxygen was omitted Examples of ineffective communication among pro- on the transfer orders to the rehab facility. viders include the following: A patient was admitted to the ED with an overdose. A patient was admitted from a nursing home with a The patient was treated and admitted to the inten- four-page list of medications. The admitting diagnosis sive care unit (ICU). When stable, the patient was was dehydration and vomiting. The triage nurse transferred to the inpatient mental health unit. The listed all medications on the ED triage form. The patient was in a gown at the time of transfer. The admitting nurse completing the medication reconcili- patient’s belongings were searched. Later, the patient ation missed one page of the patient’s nursing home was found unresponsive on the floor of her room, with medications. The admitting physician listed all of shallow respirations. 911 was called, and the patient the medications in the [history and physical] but did was given Narcan® and transferred to the ED. The not add to the ED physician orders to include any patient was treated and readmitted to the ICU. cardiac medications. At discharge, a covering physi- Examples of ineffective transitions of care between hos- cian who was sending the patient back to the nursing pitals and community settings include the following: home reviewed the medication reconciliation list and A patient was transferred to long-term care from acute did not order any cardiac medications. The nursing care without oxygen; oxygen saturation was 46% home considered the medications discontinued. The on room air. The patient had been on oxygen at the patient was [subsequently] readmitted to the hospital acute care facility. Rebreather mask and respiratory in congestive heart failure. treatments were given; oxygen saturation was 87% Amylase/lipase [levels were] highly elevated, and the after one hour. The patient became confused. The patient was discharged. The patient had to return to physician determined that the patient was medically the ED; no phone call for critical value was received unstable and gave orders to transfer the patient back while the patient was registered in the ED. to acute care. The patient was readmitted there. The patient was transferred to us [long-term care] again Examples of ineffective communication between pro- without oxygen and only partial medical records . . . viders and patients include the following: A patient was seen in the ED for evaluation of The patient was admitted to the ED for an animal syncope. Labs revealed blood urea nitrogen of 85 bite. Rabies prophylaxis was initiated in the ED. The and creatinine of 5.2. . . CT [computed tomography] patient was admitted. Later, the patient was discharged scan of the head was negative. Patient sent home home without plan to continue rabies booster . . . alone [emphasis added]. The patient returned to Vol. 7, No. 1—March 2010 REPRINTED ARTICLE - ©2010 Pennsylvania Patient Safety Authority Page 3 Pennsylvania Patient Safety Advisory the ED [one day later] in acute renal failure with by the orthopedic consultant. He was discharged rhabdomyolysis following a fall at home. The patient back to the skilled nursing facility. A nursing assis- was unable to get up and was found by family on the tant helped him out of bed, and he complained floor. A large surface pressure ulcer was noted. of pain in his groin. She called the geriatrician, A patient was seen in the ED after a fall. The who sent him back to the ED of the hospital. patient complained of knee pain and had x-rays The emergency physician confirmed that the done of the right knee and lower leg. The x-rays pain was from the fracture, which remained were normal. The patient was in pain and unable stable, and sent him back to the skilled nursing to ambulate. The patient was discharged and sent facility with confirmation that weight bearing as home by ambulance. The patient returned to the tolerated was appropriate. Later, he was found to ED two days later with continued right leg pain and have a high blood sugar (about 500 mg/dl) and was x-rayed and found to have a fractured hip that was sent back to the ED, where he was noted required surgical care. to also be dehydrated. He was readmitted to Of the 392 events related to hospital readmissions the service of his cardiologist, who changed his reported from January through August 2009, four diabetes medications. He was sent back to the root-cause analyses (RCAs) were completed and for- skilled nursing facility but returned to the hos- warded to the Authority, three of which indicated pital the next day, again with high blood sugar. that “communication among staff members” was the The cardiologist had dictated a note to the root cause of the failure. If more RCA information geriatrician, but the note had not arrived, and related to readmissions were routinely submitted by the patient had been put on the same diabetes facilities, the Authority would be better able to pro- medication regime that he had been on previ- vide analysis of the causes of some of these events. ously. The patient went on to develop decubiti that took months to heal. He eventually became Barriers to Successfully Reducing Hospital a permanent resident of the skilled nursing facil- Readmissions ity within the retirement community. Clearly, hospital readmissions are costly, and both In the above example, each facility appropriately cared federal and state agencies are interested in reducing for the patient and treated his medical condition, 30-day readmission rates in an effort to save health- yet the over-arching care plan failed. Because there is care dollars. With policy makers focused on reducing no payment structure to absorb the cost of care plan healthcare costs and improving patient safety, 30-day management across care settings, this important task readmission rates are an area of improvement that no is frequently missed or poorly performed.4,7 In the Pennsylvania facility can afford to ignore. U.S. healthcare environment, few built-in safeguards One major barrier to reducing hospital readmis- identify and rectify failures spanning more than one sions is misalignment of financial incentives. While healthcare setting. In Pennsylvania, the Authority is reducing readmissions saves money for insurers and unlikely to receive reports referencing fragmented payers, there is no financial incentive for hospitals to care, because no mechanism exists to track readmis- decrease utilization. The current fee-for-service pay- sions across facilities. Nonetheless, poorly executed ment system not only encourages patient admissions, transitions in care, whether interhospital transfers or it also encourages silos among healthcare providers, transfers between healthcare settings, can negatively creating barriers to effective communication and care affect patients’ health and well-being and often result coordination across care settings. in avoidable readmissions to the hospital. An anecdotal example from Pennsylvania follows: Success Stories An elderly patient fell going up some outdoor National Success: Reducing Readmissions by concrete steps with his wife, hitting his head. Improving Transitions in Care Collaborative He complained of dizziness. He was taken to his In fall 2009, the Institute for Healthcare Improve- local hospital, where he was given a CT scan and ment began a four-year multistate initiative to admitted on the service of his primary care phy- measurably reduce hospital readmissions. The Reduc- sician, a cardiologist. The wife understood that ing Readmissions by Improving Transitions in Care he had “blood in his brain.” His primary physi- Collaborative focuses on creating an ideal transition cian discontinued his Coumadin® and started for patients from hospital to home. The aim of this aspirin. His wife did not know why he had been collaborative is to reduce 30-day readmission rates on Coumadin. He continued to complain of by 30% and increase patient and family satisfaction dizziness. He was discharged back to the skilled with optimal transitions and coordination of care. nursing facility in his retirement community. This collaborative focuses on four major areas of risk After discharge from the skilled nursing facility, reduction: (1) performing enhanced admission assess- he got dizzy and fell again. He was readmitted to ments, (2) providing effective teaching and enhanced the hospital on the service of his cardiologist with learning, (3) conducting real-time patient- and family- a “fracture of the pelvis,” according to his wife. centered handover communication, and (4) ensuring She was unaware of the treatment recommended posthospital care follow-up.14 This initiative is one of Page 4 REPRINTED ARTICLE - ©2010 Pennsylvania Patient Safety Authority Vol. 7, No. 1—March 2010 Pennsylvania Patient Safety Advisory several successful care models designed to reduce hos- a decrease in the system’s readmission rate by 44% as pital 30-day readmission rates. well as the decline of overall treatment costs.16,17 National Success: Project Reengineered Planning for the Future Discharge (RED) A 2009 Cochrane systematic review to determine A randomized controlled trial in a general medical the effectiveness of in-hospital discharge planning of service at an urban, academic, safety-net hospital to patients moving from hospitals to outpatient settings test the effects of interventions designed to minimize failed to show an associated reduction in readmission hospital utilization after discharge showed that partici- rates. Specifically, the review pooled data from seven pants in the discharge intervention group (n = 370) randomized controlled trials that recruited elderly had a lower rate of hospital utilization than those patients with a medical condition and reported read- receiving usual care (n = 368) (0.314 versus 0.451 visit mission rates at up to three months of discharge from per person per month; IRR 0.695 [95% CI, 0.515 to the hospital. The review failed to detect a difference 0.937]; p = 0.009).15 Interventions included a nurse between those allocated to discharge planning and the discharge advocate (DA) who worked with patients control group, with respect to hospital readmission in the hospital to arrange follow-up appointments, rates (OR 0.91, 95% CI to 0.67 to 1.23).18 However, confirm medication reconciliation, and conduct as the above examples illustrate, other studies have patient education using “teach-back” methodology for shown significant reductions in 30-day readmission patient-centered education. The nurse DAs also used rates, as well as cost savings, associated with a variety an individualized instruction booklet (an after hospi- of enhanced discharge processes, most of which used a tal care plan), a copy of which was sent directly to the combination of enhanced in-hospital communication primary care provider at discharge. A clinical pharma- plus improved discharge processes, postdischarge care cist was an integral part of the discharge team, as well, coordination, and restructured financial incentives. and called the patient two to four days after discharge to reinforce the discharge plan and to review medi- The State Action on Avoidable Rehospitalizations cations with the patient. Key success factors in the (STAAR)* Initiative identified several potential handoff between hospital and home were (1) using reasons for high hospital readmission rates, includ- a plan that the patient understood, (2) putting it in ing the following: quality of care issues in the initial writing, and (3) bridging gaps between the hospital hospitalization, lack of access to physicians to receive doctors and the patient’s doctor in the community. follow-up care following the initial hospitalization, Project RED showed that bundled interventions hospital admission norms that discourage treatment including patient-centered education, comprehensive in other care settings, home healthcare access and discharge planning, and postdischarge reinforcement quality, effective discharge planning, breakdowns worked to decrease postdischarge hospital utiliza- in transitions of care between settings, and nursing tion (combination emergency room admissions and home access and quality.19 Hospitals can assess the hospital readmissions) within 30 days of discharge by characteristics of their readmission population to approximately 30%.15 determine which of these factors may be influencing their readmission rate and to determine how many of Local Success: Geisinger Health System their readmissions are potentially preventable. Geisinger Health System (Danville, Pennsylvania) has realigned financial incentives for care, thereby mini- Strategies to Reduce 30-Day Hospital mizing variance and reducing costs by implementing Readmission Rates a medical home concept. The medical home concept The STAAR Initiative reviewed the medical literature focuses on personal care coordination by shifting from and identified five promising, evidence-based strate- episodic acute care to a continuous, comprehensive gies to reduce readmissions:19 team approach to care, called ProvenHealth Navigator, which uses financial incentives to alter the care model. 1. Comprehensive discharge planning with timely Payments are made to physicians for a variety of communication. Thorough preparation of the actions that contribute to a more cohesive treatment patient and family for discharge is important. process, including seeing patients more often, seeing Having a strong transition plan, prompt postdis- them during off-hours, and playing a more direct role charge communication, and follow-up care can in coordinating care throughout the system. Internists, significantly reduce rehospitalizations.20 surgeons, and specialists are paid for adherence to 2. Postdischarge support. Early, post-acute follow- evidence-based medical guidelines in the treatment up care by transition coordinators, coaches, of chronic diseases and other illnesses. Additionally, telephone nurses, or clinicians has been shown physicians are rewarded for collecting and managing to reduce readmissions.21-23 patient data, which allows trends to be identified and analyzed. Simultaneously, Geisinger has changed the 3. Multidisciplinary, team-based management. way it charges payers. For example, for a number of Multidisciplinary heart failure management surgeries, costs are bundled into a single flat fee. If the programs have shown a decrease in hospital patient experiences complications or needs additional treatment within 90 days, the system covers the costs. * An initiative of the Commonwealth Fund and the Institute for This innovative financial architecture has resulted in Healthcare Improvement, launched May 1, 2009. Vol. 7, No. 1—March 2010 REPRINTED ARTICLE - ©2010 Pennsylvania Patient Safety Authority Page 5 Pennsylvania Patient Safety Advisory admissions.24 For example, the Program for All- the past 30 days, ask questions to determine the rea- Inclusive Care for the Elderly (PACE) provides son for the readmission. Did the patient: comprehensive, interdisciplinary care through an — Understand discharge instructions? adult day-care center coupled with PACE teams — Take medications correctly? that provide care in the hospital, nursing home, or home, as needed.25 — Have adequate home resources? 4. Patient education and self-management support. — Follow self-care instructions? Developing a commonly understood care plan that — Understand the signs of clinical deterioration contains instructions for medications, diet, activity to report to the primary physician? level, and identification of signs of disease pro- — Seek medical follow-up after discharge from gression is a critical part of the discharge process. the hospital? Providing the patient with a nurse educator for one hour as an adjunct to the normal discharge ■ Consider a dedicated transitional coach to perform enhanced admission assessments, focusing on post- process can reduce the risk of rehospitalizations discharge needs as soon as possible.15 or death.26 ■ Include the patient and family in the discharge pro- 5. Remote monitoring. Remote monitoring uses cess, and be vigilant in assessment of the support a variety of modalities to track patients’ health systems available in the postacute care setting. and well-being in order to identify early signs of ■ Perform a thorough physical and cognitive func- clinical deterioration. Used in conjunction with tional health status assessment to identify the other support systems, remote monitoring can appropriate postacute care setting for the patient. help patients remain in their homes and avoid rehospitalizations.19 ■ Refer the patient to appropriate community resources (e.g., home care, assisted living, In light of impending national- and state-level policy long-term care). changes, Pennsylvania hospitals can and should evalu- ■ Provide evidence-based and error-free care for the ate their 30-day readmission rates and formulate both patient in the hospital. short- and long-term plans to reduce these rates while simultaneously working toward improving integrated, In-Hospital Assessment: Effective Teaching and patient-centered care. Following is a list of potential Enhanced Learning14 strategies that hospitals can implement now, and into ■ Identify the “learners” on admission by asking, the future, depending upon available financial and “Who will be helping you when you leave the hos- human resources. pital?” Realize that the patient’s visitors may not be Immediate the designated “learners.” ■ Use customized, individualized discharge instruc- Environmental Scan4,10,14,18 tions that incorporate health literacy principles, ■ Collect monthly data related to readmission rates to written at a literacy level that does not exceed track organizational performance, and compare per- patient comprehension.10,14 Health literacy prin- formance data with national and state benchmarks ciples include using simple one-to-two syllable words available online from http://www.hospitalcompare. written in a font size of 14 points, short four-to-six hhs.gov and http://www.phc4.org. word sentences, and short two-to-three sentence paragraphs without medical jargon and with abun- ■ Develop a plan related to the proposed or potential dant white space. financial impact of the alternatives being discussed ■ Use a “teach-back” method to ensure patient under- for Medicare readmissions (e.g., financial incentives, standing of discharge and follow-up care instructions. disincentives, bundling). Ask patients in a nonjudgmental way to discuss what ■ Survey community healthcare resources including they have learned, identify gaps in understanding, primary care physicians, home healthcare services, and offer additional instruction as needed. assisted living, and nursing home or long-term care ■ Develop a plan of care that follows the patient facilities. Does each of these facilities send patients home and/or to the next care setting. to the hospital? Are they associated with a portion In-Hospital Assessment: Real-Time Patient and of the readmissions? Is there a way to collaborate Family Centered Handover Communication4,10,14,15,23 with these entities to improve care transitions across healthcare settings? ■ Reconcile the patient’s medication on admission to the hospital and at each transition of care (in- In-Hospital Assessment: Enhanced Admission hospital and across care settings). Assessments10,14,20 — If the patient’s prescription medications have ■ Ask patients about previous admissions; document changed, clearly document and instruct the any admission occurring within 30 days of a previous patient about the changes, identifying those hospital discharge (from your facility or from another medications and doses that the patient should facility). If the patient was previously admitted within take now. Page 6 REPRINTED ARTICLE - ©2010 Pennsylvania Patient Safety Authority Vol. 7, No. 1—March 2010 Pennsylvania Patient Safety Advisory — If the patient’s medications have been held physician settings to facilitate transfers of care and during the hospital admission, clarify if and information. when those medications should be continued. ■ Investigate integrated electronic health records and — Assess whether a home care nurse or transi- remote monitoring technology to share real-time tional care nurse or coach should reconcile clinically relevant patient medical information the medications during a home visit with the across the care continuum. patient after discharge. (A recently released guide from the Health Research ■ Send the patient home with a copy of the plan of and Educational Trust provides an overview of strate- care, and share the care plan with the primary phy- gies and interventions hospitals can implement during sician, home healthcare agency, or long-term care hospitalization, at discharge, and postdischarge. The facility that will be accepting the patient into care. guide is available online at http://www.hret.org/hret/ ■ For dialysis patients, send a copy of the plan of care, programs/content/Readmission_Guide.pdf.) including the reconciled medication list, to the nephrologist at the dialysis center. Conclusion All-cause readmission rates highlight the importance ■ Improve coordination of care between hospitals of understanding factors that influence rehospitaliza- and primary care physician offices, home health- tion. There is extensive literature on rehospitalization care agencies, assisted-living facilities, or other outpatient settings by faxing or e-mailing discharge related to medical conditions; less so for studies summaries directly to primary care offices, mailing analyzing the multiple diseases and processes that discharge packets, or using a community discharge contribute to hospital readmissions.7 A review of planner to facilitate the timely transfer of discharge the literature and success stories points toward two information. major processes that, if improved, can help decrease 30-day readmission rates: (1) improved communica- ■ Make the initial outpatient appointment for the pa- tion among providers within and across care settings tient before he or she leaves the hospital. A primary and (2) enhanced transitional care processes including care physician should see patients with a significant postdischarge intervention. Additionally, financial chronic disease within one week of discharge. incentives and disincentives have proven effective in ■ Speak with the “emergency contact” listed in the decreasing avoidable readmissions, and both federal patient record, and give an accurate, up-to-date and state policymakers have focused on restructuring report of the patient’s condition. hospital payments as one way to reduce avoidable read- Posthospital Care Follow-Up4,14-16,19 missions. Geisinger Health System is one example of a Pennsylvania healthcare system that has reduced hospi- ■ Consider implementing a follow-up telephone call tal readmissions by restructuring both its payment and from a pharmacist, nurse, or transitional care staff clinical care models. member one to three days after discharge from the hospital to confirm understanding of all discharge Improving healthcare delivery means eliminating barri- instructions and prescribed medications. ers between silos of service and information that have dominated healthcare to create a seamless, human- ■ Establish an emergency call number at the hospital centered, and more cost-effective delivery system.16 The to help patients until their primary care physicians risk reduction strategies in this article allow facilities to take over. begin gradually reducing readmissions with simple, cost- ■ Assess the patient’s home environment to evaluate effective strategies and move to more fiscally challenging self-reported ability to manage healthcare needs strategies as the financial incentives to do so evolve. independently, and refer supplemental services as Notes needed. 1. Fazzi R, Agoglia R, Mazza G, et al. The Briggs National Future Quality Improvement/Hospitalization Reduction Study. ■ Investigate relationships with primary care physi- Caring 2006 Feb;25(2):70-5. cians, home care agencies, or other community 2. Alliance for Health Reform. Covering health issues 2006- service providers to establish collaboration across 2007 [online]. 2007 Feb 21 [cited 2009 Jul 17]. Available the care continuum.4,14,16 from Internet: http://www.allhealth.org/sourcebooktoc. ■ Work toward establishing an integrated system asp?SBID=1. of care across multiple care settings with shared 3. Jiang HJ , Russo CA, Barrett ML. Nationwide frequency accountability for patient-centered care and the and costs of potentially preventable hospitalizations, ability to communicate, review each other’s work, 2006. HCUP Statistical Brief #72. 2009 Apr [cited 2009 and collaborate to deliver consistently high-quality, Nov 16]. Available from Internet: http://www.hcup-us. patient-centered care. ahrq.gov/reports/statbriefs/sb72.pdf. ■ Establish data collection criteria and share readmis- 4. Medicare Payment Advisory Commission. Report to sion information within the community of providers. Congress: reforming the delivery system [online]. 2008 ■ Consider establishing a common care plan used Jun 13 [cited 2009 Aug 7]. Available from Internet: across care settings, and shared patient educational http://www.medpac.gov/documents/Jun08_ materials, as well as a nurse who travels to outpatient EntireReport.pdf. Vol. 7, No. 1—March 2010 REPRINTED ARTICLE - ©2010 Pennsylvania Patient Safety Authority Page 7 Pennsylvania Patient Safety Advisory 5. Office of Management and Budget. President Obama’s 15. Jack B, Chetty V, Anthony D, et al. A reengineered hos- fiscal 2010 budget [online]. [cited 2009 Aug 7]. Avail- pital discharge program to decrease rehospitalizations: a able from Internet: http://www.whitehouse.gov/omb/ randomized trial. Ann Intern Med 2009 Feb 3;150(3):178-87. fy2010_key_healthcare. 16. Menninger B. Reinventing healthcare delivery: innova- 6. Centers for Medicare & Medicaid Services (CMS). CMS tion and improvement behind the scenes. California expands information for consumers about outcomes of Healthcare Foundation Issue Brief 2009 Sep;1-14. care in America’s hospitals [fact sheet online]. 2009 Jul 9 [cited 2009 Aug 7]. Available from Internet: http://www. 17. Steele GD, Bothe A, Paulus R. Geisinger links quality cms.hhs.gov/apps/media/press/factsheet.asp?Counter= with reimbursement: coronary artery bypass graft sur- 3478&intNumPerPage=10&checkDate=&checkKey= gery. PA PSRS Patient Saf Advis [online] 2007 Dec &srchType=1&numDays=3500&srchOpt=0&srchData= [cited 2010 Jan 5]. Available from Internet: http:// &keywordType=All&chkNewsType= 6&intPage= patientsafetyauthority.org/ADVISORIES/ &showAll=&pYear=&year=&desc=&cboOrder=date. AdvisoryLibrary/2007/dec4(4)/Pages/109a.aspx. 7. Jencks SF, Williams MV, Coleman EA, et al. Rehospital- 18. Shepperd S, Parkes J, McClaran JJM, et al. Discharge izations among patients in the Medicare fee-for-service program. N Eng J Med 2009 Apr 2;360(14):1418-28. planning from hospital to home. Cochrane Database Syst Rev 2004; (1):CD000313. 8. Cantor JC, Schoen C, Belloff D, et al. Aiming higher: results from a state scorecard on health system perfor- 19. Hines S. State action on avoidable rehospitalizations mance [online]. 2009 Oct [cited 2010 Jan 19]. Available (STAAR): a tool for state policy makers [online]. [cited from Internet: http://www.commonwealthfund.org/ 2010 Jan 18]. Available from Internet: http://www.ihi.org/ Charts-and-Maps/State-Scorecard-2009/DataByState/ nr/rdonlyres/6900f685-6510-44a3-8f3f-b143acee7cd0/0/ State.aspx?state=PA. staar_a_tool_for_state_policy_makers.pdf. 9. Friedman B, Encinosa W, Jiang HJ, et al. Do patient 20. Phillips CO, Wright SM, Kern DE, et al. Comprehensive safety events increase readmissions? Med Care 2009 discharge planning with postdischarge support for older May;47(5):583-90. patients with congestive heart failure: a meta-analysis. 10. Atlantic Information Services. CMS targets readmission JAMA 2004;291(11):1358-67. through payment, audits; ‘coaching’ model reduces rates [online]. Rep Medicare Compliance 2008 Jun 30 [cited 21. Naylor MD, Brooten DA, Campbell RL, et al. Transi- 2009 Aug 7]. Available from Internet: http://www.cfmc. tional care of older adults hospitalized with heart failure: org/files/rmc063008.pdf. a randomized, controlled trial. J Am Geriatr Soc 2004 May;52(5):675-84. 11. Rendell EG. Governor’s Office of Health Care Reform. Prescription for Pennsylvania [strategic plan online]. 2007 22. Naylor MD, Brooten D, Campbell R, et al. Compre- Jan [cited 2009 Oct 2]. Available from Internet: http:// hensive discharge planning and home follow-up of www.gohcr.state.pa.us/prescription-for-pennsylvania/ hospitalized elders: a randomized clinical trial. JAMA Prescription-for-Pennsylvania.pdf. 1999 Feb 17;281(7):613-20. 12. Pennsylvania Healthcare Cost Containment Council. 23. Coleman EA, Parry C, Chalmers S, et al. The care transi- [Web site]. [cited 18 Sep 2009]. Harrisburg (PA). Avail- tions intervention: results of a randomized controlled able from Internet: http://www.phc4.org. trial. Arch Intern Med 2006 Sep;166(17):1822-8. 13. Pennsylvania Health Care Cost Containment Council. Hospital performance report 2007—key findings [online]. 24. Rich MW. Heart failure disease management: a critical [cited 2009 Aug 8]. Available from Internet: http://www. review. J Card Fail 1999 Mar;5(1):64-75. phc4.org/reports/hpr/08/docs/hpr2008keyfindings.pdf. 25. Boult C, Kane RL, Brown R. Managed care of chroni- 14. Nielson GA, Rutherford P, Taylor J. Transforming care at cally ill older people: the US experience. BMJ 2000 the bedside how-to guide: creating an ideal transition home Oct;321(7267):1011-4. for patients with heart failure. 2009 [cited 2009 Sep 15]. Available from Internet: http://www. ihi.org/IHI/Topics/ 26. Koelling TM, Johnson ML, Cody RJ, et al. Discharge MedicalSurgicalCare/MedicalSurgicalCareGeneral/Tools/ education improves clinical outcomes in patients with TCABHowToGuideTransitionHomeforHF.htm. chronic heart failure. Circulation 2005 Jan;111(2):179-85. Page 8 REPRINTED ARTICLE - ©2010 Pennsylvania Patient Safety Authority Vol. 7, No. 1—March 2010 PENNSYLVANIA PATIENT SAFETY ADVISORY This article is reprinted from the Pennsylvania Patient Safety Advisory, Vol. 7, No. 1—March 2010. The Advisory is a publication of the Pennsylvania Patient Safety Authority, produced by ECRI Institute and ISMP under contract to the Authority. Copyright 2010 by the Pennsylvania Patient Safety Authority. This publication may be reprinted and distributed without restriction, provided it is printed or distributed in its entirety and without alteration. Individual articles may be reprinted in their entirety and without alteration provided the source is clearly attributed. This publication is disseminated via e-mail. To subscribe, go to https://www.papsrs.state.pa.us/ Workflow/MailingListAddition.aspx. To see other articles or issues of the Advisory, visit our Web site at http://www.patientsafetyauthority.org. Click on “Patient Safety Advisories” in the left-hand menu bar. THE PENNSYLVANIA PATIENT SAFETY AUTHORITY AND ITS CONTRACTORS The Pennsylvania Patient Safety Authority is an independent state agency created by Act 13 of 2002, the Medical Care Availability and Reduction of Error (“Mcare”) Act. Consistent with Act 13, ECRI Institute, as contractor for the Authority, is issuing this publication to advise medical facilities of immediate changes that can be instituted to reduce Serious Events and Incidents. For more information about the Pennsylvania Patient Safety Authority, see the Authority’s Web An Independent Agency of the Commonwealth of Pennsylvania site at http://www.patientsafetyauthority.org. ECRI Institute, a nonprofit organization, dedicates itself to bringing the discipline of applied scientific research in healthcare to uncover the best approaches to improving patient care. As pioneers in this science for nearly 40 years, ECRI Institute marries experience and independence with the objectivity of evidence-based research. More than 5,000 healthcare organizations worldwide rely on ECRI Institute’s expertise in patient safety improvement, risk and quality management, and healthcare processes, devices, procedures and drug technology. The Institute for Safe Medication Practices (ISMP) is an independent, nonprofit organization dedicated solely to medication error prevention and safe medication use. ISMP provides recommendations for the safe use of medications to the healthcare community including healthcare professionals, government agencies, accrediting organizations, and consumers. ISMP’s efforts are built on a nonpunitive approach and systems-based solutions.