Nearly 1 in 4 patients hospitalized with heart failure and 1 in 5 patients hospitalized with acute myocardial infarction (AMI) are readmitted within 30 days of discharge (1- 3). These rates have been fairly stable or have increased slightly in recent years (4- 5). Higher readmissions rates have been associated with lower patient satisfaction (6) and are estimated to cost Medicare more than $17 billion per year in hospital payments (7).

Randomized controlled trials have demonstrated successful efforts to reduce readmissions in a variety of patient populations (8- 11), and a recent review suggested a substantial proportion of readmissions might be avoidable (12). Strategies commonly recommended for reducing readmissions include improved patient education about their medications, patient-centered discharge instructions, follow-up telephone calls, home visits, and increased coordination with outpatient providers (13- 16). Despite the national focus on readmission rates, contemporary data on these hospital practices aimed at reducing readmissions are lacking.

Accordingly, we conducted a descriptive study to determine the range and prevalence of practices being implemented by hospitals to reduce 30-day readmissions of patients. We surveyed hospitals that were enrolled in Hospital to Home (H2H), a quality campaign sponsored by the American College of Cardiology and the Institute of Healthcare Improvement with the goal of reducing readmission rates by 20% by the end of 2012. More than 1,000 hospitals have enrolled in this national effort, supported by multiple professional associations and partners. Given its national spread and size, the H2H campaign (17) provides an ideal opportunity to examine changes over time in hospital practices; the present study reports baseline data on these practices.

We conducted a cross-sectional study using a Web-based survey (Online Appendix) of hospitals to examine their reported use of specific hospital practices to reduce readmissions for patients with heart failure or AMI. We contacted all hospitals that enrolled in H2H during its first 8 months (October 1, 2009 to July 1, 2010) (n = 594). Of the 594 hospitals, 537 completed the survey, for a response rate of 90.4%. We sent a letter of invitation to the contact person registered with H2H. The roles reported by respondents varied, and many respondents reported having >1 role; nearly 60% were from quality management departments, 25% were from cardiology departments, 25% were from other clinical departments, 16% were from case management or care coordination, and 8% reported working in nonclinical roles. Respondents were instructed to coordinate with other relevant staff to complete a single survey reflecting hospital practices. All research procedures were approved by the Institutional Review Board at the Yale School of Medicine.

We examined hospital practices in 3 areas: quality improvement efforts and performance monitoring regarding readmission, medication management, and discharge and follow-up procedures. In addition, to summarize the data, we created a summary count variable of 10 specific practices across each of the 3 areas: 1) having a quality improvement team for reducing readmissions for heart failure, for AMI, or both; monitoring the percent of patients with follow-up appointments within 7 days of discharge; and monitoring 30-day readmission rates; 2) having medication management efforts, including providing patient education about the purpose of each medication and any alterations to the medication list; having a pharmacist usually responsible for conducting medication reconciliation at discharge; and having a pharmacy technician primarily responsible for obtaining medication history as part of medication reconciliation process; and 3) having discharge processes in which patients or their caregivers receive an emergency plan; having patients leave the hospital with an outpatient follow-up appointment already arranged; having a process in place to ensure the outpatient physicians are alerted to the patient's discharge status within 48 h of discharge; and contacting patients after discharge to follow-up on post-discharge needs or to provide additional patient education.

We assessed the internal consistency of the 10-item summary score using the Cronbach alpha coefficient, which was acceptable at 0.61. Practices included in the survey were selected based on existing literature and recommendations of the H2H campaign, taking into account practices that hospital staff might be expected to be able to address. We field tested the survey items using cognitive interviews (18) with hospital quality improvement directors to assess clarity and comprehensiveness. The summary score ranged from 0 to 10, and was supported by a number of studies (8,10,15- 16,19- 33), although definitive evidence on their effectiveness is lacking.

We also ascertained hospital characteristics, including number of staffed hospital beds, teaching status (member of the Association of American Medical Colleges Council of Teaching Hospitals [COTH] vs. nonteaching), multihospital affiliation (yes/no), and ownership (for-profit, nonprofit, and government) using data from the Annual Survey of the American Hospital Association (AHA) from 2009. We determined census regions from the U.S. Census Bureau and urban/suburban/rural location from the 2003 Urban Influence Codes.

We used standard frequency analysis to describe the sample of hospitals, the prevalence of each hospital practice, and the distribution of summary variables. We also used independent samples t-tests and chi-square tests as appropriate to compare our sample of H2H hospitals to all the other adult medical/surgical hospitals in the 2009 Annual Survey of the AHA. To examine variations in summary scores by hospital characteristics, we conducted unadjusted analysis with correlation coefficients (for number of staffed beds), with analysis of variance (for geographic location, ownership type, and census region), and with t-tests (for teaching status and multihospital affiliation). The research was funded by the Commonwealth Fund, which had no influence on the methodology, findings, or interpretation. All analyses were conducted in SAS, version 9.2 (SAS Institute, Cary, North Carolina).

The sample of 537 hospitals (response rate 90.4%) had a mean size of 316 beds, with 11% having ≥600 beds (Table 1). Compared with all other adult medical/surgical hospitals in the United States, hospitals in our sample had more beds, were more likely to be COTH hospitals and part of a multihospital system/chain, were less likely to be government-owned, and more likely to be urban; they also varied significantly in census region, with greater numbers in the South Atlantic region (p < 0.05).

Nearly 90% of hospitals agreed or strongly agreed that they had a written objective of reducing preventable readmission for patients with heart failure or AMI (Table 2). Most hospitals reported having a reliable process to identify patients with heart failure at the time of admission; most hospitals reported having a quality improvement team in place to reduce preventable readmission for these patients with heart failure. Members of that team were typically staff nurses, quality improvement staff, social workers, physicians, and senior management. Pharmacists were included on teams in two-thirds of the hospitals. Hospitals had less focus on readmission after hospitalization for AMI, with 54% of hospitals reporting they had a quality improvement team in place to reduce preventable readmission after hospitalization for AMI. Slightly more than one-half of hospitals had a multidisciplinary care team to manage patients at high risk for readmissions, more than two-thirds had partnered with home care or skilled nursing facilities to reduce readmission rates, less than one-half had partnered with community physicians, and one-quarter had partnered with local hospitals to reduce readmissions.

In terms of monitoring performance data, nearly all hospitals reported tracking the 30-day readmission rate; two-thirds had a designated person or group to review unplanned readmissions that occurred within 30 days of discharge. Other indicators were monitored by hospitals less frequently, such as the proportion of discharge summaries sent to the primary physician (22.7% of hospitals monitored this proportion), percent of patients with follow-up appointment within 7 days (32.1%), and proportion of patients readmitted to another hospital (11.4%). Approximately one-fifth (22.3%) of hospitals reported that they formally estimated the risk of readmission and used it in clinical care during patient hospitalization.

In about 14% of hospitals, the responsibility for medication reconciliation was sometimes not formally assigned (Table 3). Nevertheless, nearly three-quarters of hospitals reported having some electronic medical record or Web-based form to facilitate medication reconciliation. Typically, emergency medicine staff or the admitting medical team obtained the medical history, with less common involvement of pharmacy staff. For nearly one-half of the hospitals, a pharmacist or pharmacy technician was never involved with obtaining the medication history, and a small proportion of hospitals reported always making contact with an outside pharmacy or with the primary physician as part of the medication reconciliation process (in 3.2% and 13.9% of hospitals, respectively). The majority of hospitals reported that all patients or their caregivers received discharge instructions and names, doses, and frequency of discharge medications when they left the hospital, and the majority reported using “teach-back” techniques for patient and family education.

Almost 60% of hospitals reported providing patients (or their caregivers) some type of emergency plan and providing an action plan for patients with heart failure if symptoms changed (Table 4). Less than one-half responded that patients with home health services were provided direct contact for a specific inpatient physician in case of questions. Fewer hospitals reported that there was a process in place to ensure that outpatient physicians were alerted to the discharge within 48 h, and about 30% of hospitals did not routinely make discharge summaries available for viewing within 7 days of discharge. About one-third of hospitals assigned someone to follow-up on test results that were received after the patient was discharged, nearly two-thirds reported that they regularly called patients after discharge, and less than one-quarter reported arranging home visits for most or all patients with heart failure or AMI after discharge. In the case of patients discharged to skilled nursing facilities, less than two-thirds of hospitals reported that nurse-to-nurse report was always conducted before transfer.

Less than 3% of hospitals had in place all the practices that comprised our summary score (Figure 1). The average number of 10 key practices reported to be in place was 4.8. More than 30% of hospitals had all quality improvement and performance monitoring practices in place, and approximately 15% had all the discharge and follow-up practices in place (Table 5). A minority (5%) of hospitals had all the medication management practices in place. Census region was significantly associated with summary scores (with Pacific region having the highest scores and the Mountain region having the lowest scores). Hospital teaching status, urban/suburban/rural location, number of beds, chain affiliation, and ownership type were not significantly associated with summary scores for heart failure or for AMI.

We found that the majority of hospitals reported having written objectives to reduce readmission, quality improvement teams focused on readmissions, and ongoing monitoring of 30-day readmission rates, whereas many of the specific practices considered to be important for preventing readmissions were implemented by fewer hospitals. We also found wide variation among this set of hospitals, although all had enrolled in a national campaign to reduce readmissions. Although 12% of hospitals had implemented ≤2 of 10 key practices, another 12% had implemented ≥8 of these 10 practices, indicating substantial heterogeneity among hospitals in strategies to reduce readmissions of patients with heart failure and AMI.

Practices were particularly variable within the area of medication management, and the findings suggested that medication reconciliation processes were nonstandardized at most hospitals. Although the majority of hospitals reported that physicians, physician assistants, or nurse practitioners were always responsible for conducting medication reconciliation, nearly one-third of hospitals reported pharmacists were sometimes responsible, and 14% of hospitals suggested that the responsibility was at least sometimes not formally assigned to anyone. Additionally, for more than one-half of the hospitals, making contact with outside pharmacies or the primary physician was sometimes, but not always, part of the medication reconciliation process, again underscoring the variability in this process. Although patient education about medications was apparently robust, with 70% of hospitals reporting the use of “teach-back” techniques and 77% reporting that all medication details were given to patients at discharge, lack of standard processes for both the reconciliation and patient education regarding medications is potentially problematic.

Several of the discharge and follow-up practices, which were shown to be associated with reduced readmissions (16), were practiced by less than one-half of hospitals. Central to effective continuity of care is the linking of inpatient and post-discharge (e.g., outpatient, home care, or skilled nursing facility) providers and information. Nevertheless, a process to alert outpatient physicians within 48 h of the patient's discharge and a process to follow-up on test results that were returned after a patient's discharge were present in 37% and 36% of hospitals, respectively. In the cases of patients discharged with home health services or to skilled nursing facilities, direct contact information for an inpatient physician in case of questions was reported to be provided in 47% and 34% of hospitals, respectively. The limited use of some evidence-based practices found in the present study was consistent with recent qualitative data suggesting that hospital and professional cultures tend to focus on the inpatient part of the patient's care and are less endorsing of responsibilities post-discharge (34).

Why might these practices be lacking in so many hospitals? One reason might be because they require added resources. Particularly in the case of producing timely discharge summaries and ensuring adequate patient and caregiver education about complex medication and other issues, it is possible that constraints on staff time is a major rate-limiting step to implementing some of these potentially best practices. Perhaps a more important challenge, however, is the management challenge of coordinating efforts to ensure timely and proper discharge. Especially under the pressure of morning discharges, coordination among various physicians, pharmacists, nurses, and the many ancillary staff to achieve a smooth discharge might be extremely complex. Without standardized systems and across a diverse care team balancing multiple priorities, ensuring these practices requires substantial coordination, which might be difficult to accomplish. Lastly, clinicians and administrators might be uncertain about the efficacy of various strategies as we lack definitive studies demonstrating their impact on readmission. As a result, adoption of these strategies might be slower, particularly in the absence of definitive evidence supporting their effectiveness for reducing readmission rates.

This is the first national study we know of that documented specific hospital practices undertaken to reduce readmission rates. Nevertheless, the results should be interpreted in light of the study's limitations. First, the study was descriptive and could not evaluate the effect of these practices on outcomes, including readmission rates. Readmission rate data will be linked to these practices in future reports from this study as the data become available. Second, implementation of these practices was self-reported by the primary hospital contact for the H2H campaign, often by the quality improvement director. We did request, however, that respondents consult with other key staff to complete the survey to ensure a more comprehensive and informed view of practices implemented. Furthermore, such practices are complex and cannot be fully characterized using quantitative methods, although items were field tested before the survey. Lastly, our findings might have overestimated the use of these practices nationally, because hospitals enrolled with H2H might be more motivated to reduce readmissions, or might have underestimated their use if hospitals with particular concerns might have been more likely to enroll. Nevertheless, among this large sample of hospitals with a high response rate, we found limited use of several practices that have been widely recommended.

We examined the reported use of key practices to reduce readmissions for patients with heart failure and AMI. Although some practices were implemented by many hospitals, most hospitals did not report having a comprehensive set of recommended practices in place. The lack of implementation of key practices was most apparent in the areas of medication management and discharge and follow-up processes. Particularly striking was the substantial variability in hospital practices to reduce readmission rates. Given the diversity of efforts to reduce readmission rates, establishing more definitive evidence about the effective hospital practices in this area is warranted. Nevertheless, our findings suggest opportunities for continued improvement in communication and care coordination, which may assist in hospital efforts to reduce readmission rates.

For the Hospital-to-Home (H2H) Survey, please see the online version of this article.