Public Reporting of Discharge Planning and Rates of Readmissions

Transcription

1 special article Public Reporting of Discharge Planning and Rates of Readmissions Ashish K. Jha, M.D., M.P.H., E. John Orav, Ph.D., and Arnold M. Epstein, M.D. Abstract Background A reduction in hospital readmissions may improve quality and reduce costs. The Centers for Medicare and Medicaid Services has initiated a national effort to measure and publicly report on the conduct of discharge planning. We know little about how U.S. hospitals perform on the current discharge metrics, the factors that underlie better performance, and whether better performance is related to lower readmission rates. Methods We examined hospital performance on the basis of two measures of discharge planning: the adequacy of documentation in the chart that discharge instructions were provided to patients with congestive heart failure, and patient-reported experiences with discharge planning. We examined the association between performance on these measures and rates of readmission for congestive heart failure and pneumonia. From the Department of Health Policy and Management, Harvard School of Public Health (A.K.J., E.J.O., A.M.E.); the Division of General Medicine, Brigham and Women s Hospital (A.K.J., E.J.O., A.M.E.); and the Veterans Affairs Boston Healthcare System (A.K.J.) all in Boston. Address reprint requests to Dr. Jha at the Department of Health Policy and Management, Harvard School of Public Health, 677 Huntington Ave., Boston, MA 02115, or at ajha@hsph.harvard.edu. N Engl J Med 2009;361: Copyright 2009 Massachusetts Medical Society. Results We found a weak correlation in performance between the two discharge measures (r = 0.05, P<0.001). Although larger hospitals performed better on the chart-based measure, smaller hospitals and those with higher nurse-staffing levels performed better on the patient-reported measure. We found no association between performance on the chart-based measure and readmission rates among patients with congestive heart failure (readmission rates among hospitals performing in the highest quartile vs. the lowest quartile, 23.7% vs. 23.5%; P = 0.54) and only a very modest association between performance on the patient-reported measure and readmission rates for congestive heart failure (readmission rates among hospitals performing in the highest quartile vs. the lowest quartile, 22.4% vs. 24.7%; P<0.001) and pneumonia (17.5% vs. 19.5%, P<0.001). Conclusions Our findings suggest that current efforts to collect and publicly report data on discharge planning are unlikely to yield large reductions in unnecessary readmissions. n engl j med 361;27 nejm.org december 31,

2 The U.S. health care system faces challenges on two fronts: pressure to improve quality 1 and the necessity to reduce costs. 2 Unfortunately, quality-improvement efforts often increase costs even when they are costeffective, and efforts to constrain costs can lead to concerns about reductions in the quality of care. Thus, improving care in clinical areas where efforts can lead simultaneously to better outcomes for patients and lower costs represents an important step forward. Preventing readmissions is one such opportunity. Previous studies have indicated large variations in readmission rates among hospitals 3-5 and noted substantial problems with the transition of care from the hospital to the ambulatory care sector. These problems include suboptimal communication of discharge instructions, 6 failure to reconcile hospital and ambulatory care records, and failure to arrange for appropriate ambulatory care follow-up. 7 Variations in readmission rates and evidence of poor quality of care during transitions have caught the attention of policymakers and catalyzed efforts by the Centers for Medicare and Medicaid Services (CMS) to collect and publicly report data on discharge practices. The CMS Hospital Quality Alliance (HQA) program now publicly reports indicators that assess proper documentation of discharge planning in the medical records of patients with congestive heart failure. The HQA also includes the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey, which reports on patients perceptions of the adequacy of their discharge planning. 8 Despite the interest in ensuring effective discharge planning (in part because of its presumed impact on decreasing unnecessary readmissions), there are few empirical data to support this approach or guide policymakers. In this study, we sought to answer five questions. First, what is the performance of U.S. hospitals according to measures of discharge planning? Second, what hospital characteristics are associated with better performance on these measures? Third, how variable are readmission rates among health care markets? Fourth, is better performance on discharge measures associated with lower rates of readmission? And, finally, assuming an association exists, how many readmissions are potentially eliminated with better performance on discharge measures? Methods Data We used the September 2008 release of the HQA database, which provided data on the clinical care of patients discharged from U.S. hospitals during The HQA provides performance data on clinical processes (for acute myocardial infarction, congestive heart failure, pneumonia, and prevention of surgical complications) and outcomes (riskadjusted mortality associated with acute myocardial infarction, congestive heart failure, and pneumonia) and patient-reported measures for medical and surgical hospitalizations from the HCAHPS surveys. 8 We linked HQA data to the 2007 American Hospital Association annual survey, which provides information on the structural characteristics of hospitals. Discharge Planning We used two types of performance data on discharge planning from the HQA data set: a process measure that examines whether patients hospitalized with congestive heart failure received adequate discharge instructions as noted in the medical record 9 and a patient-reported measure that is a composite of two questions about discharge planning (see the Supplementary Appendix, available with the full text of this article at NEJM.org). 8 Both measures are scored on a scale of 1% to 100%, with higher scores indicating better performance. Readmission Rates We used a version of the 2007 Medicare Provider Analysis and Review that includes data on all hospitalizations of enrollees in the Medicare fee-forservice plan. We focused on all-cause 30-day readmission rates for congestive heart failure and pneumonia because they are the two most common causes of readmissions among Medicare enrollees and because previous studies indicate that better discharge planning efforts and support after discharge can reduce readmission rates for congestive heart failure. 10 To examine baseline differences in the risk of readmission based on the severity of illness, we used the approach described by Elixhauser et al., 11 with adjustment for age, sex, race or ethnic group, and the presence or absence of up to 29 coexisting conditions (see the Supplementary Appendix). We limited our analyses to pa 2638 n engl j med 361;27 nejm.org december 31, 2009

3 Discharge Planning in U.S. Hospitals tients discharged before December 1 in order to allow at least 30 days for patients to be readmitted. Statistical Analysis We first used chi-square tests and t-tests to compare the characteristics of hospitals that reported both the HQA chart-based measure of discharge planning and HCAHPS survey-based measure with those hospitals that chose not to report these data. We examined hospital-level performance on the chart-based and patient-reported measures and the correlation of performance between them. We built multivariable linear regression models to examine whether certain hospital characteristics were independently associated with discharge planning, choosing a priori to focus on nursestaffing levels (positing that greater use of nursing resources might lead to more effective discharge planning), hospital size (hypothesizing that smaller hospitals might be more attuned to patients needs after hospitalization), and teaching status (hypothesizing that teaching hospitals, where multiple providers are often involved in a patient s care, may be less adept at providing adequate discharge planning). 8 To gauge potential opportunities for decreasing readmissions, we examined geographic variations in readmission rates. Previous studies have shown variations in readmission rates across hospitals or states. 3-5 However, hospital-based analyses might show large variation due to small numbers, and state-based analyses might mask considerable intrastate variation. To examine variations in readmission rates for congestive heart failure and pneumonia across the United States, we aggregated hospitals according to health care markets (defined as 1 of 306 hospital-referral regions described in the Dartmouth Atlas of Health Care, based on patterns of travel for tertiary care among Medicare enrollees 12 ). We excluded hospital-referral regions with fewer than 100 patients admitted for pneumonia or congestive heart failure during In these patient-based analyses, we included all patients discharged from hospitals in each hospital-referral region during the first 11 months of 2007 and calculated the proportion of patients who were readmitted within 30 days. We adjusted the analyses according to the method of Elixhauser et al. and accounted for clustering at the hospital level. We examined the correlation between readmission rates for congestive heart failure and pneumonia in hospitalreferral regions, using the Pearson correlation. To further reduce the chance of exaggerating our estimate of variation because of small numbers, we calculated empirical Bayesian predictions of readmission for each hospital-referral region, using hierarchical generalized linear models. This approach shrinks the raw estimate of readmissions in a given hospital-referral region toward the national mean, with the amount of shrinkage inversely related to the size of the hospital-referral region. 13 Next, we examined the association between each discharge-planning measure and the riskadjusted 30-day rate of readmission among patients admitted to a hospital that reported both discharge measures. We built patient-level multivariable models that were adjusted for patient characteristics and accounted for clustering at the hospital level. We subsequently further adjusted the models for hospital-level factors, including size, teaching status (as defined by membership or nonmembership in the Council of Teaching Hospitals), presence or absence of an intensive care unit (as an additional marker of patient mix), census region, location (urban vs. rural), profit status, and nurse-to-census ratio (calculated by dividing the number of full-time equivalent nurses on staff by 1000 patient-days). Our approach to variable selection is further detailed in the Supplementary Appendix. All patients in a particular hospital were assigned that hospital s values for hospital-level variables such as performance on the discharge measures and size. We report readmission rates in relation to discharge planning according to quartiles of performance on the discharge-planning measure. Finally, to illustrate the potential reduction in readmission rates that could be achieved if each discharge-planning measure was improved to the 90th percentile of national performance, we used linear regression to quantify the magnitude of the association between discharge performance and rates of readmission. We then projected rates of readmission at each hospital if performance on that measure increased to the 90th percentile of performance nationally. For example, if the 90th percentile of national performance on the HCAHPS survey was 85% and a 10-point increase in the HCAHPS score was associated with a 1% reduction in the readmission rate, a given hospital that n engl j med 361;27 nejm.org december 31,

4 received an initial HCAHPS performance score of 70% and then improved the HCAHPS score to 85% should lower the readmission rate by 1.5%. We performed this projection for each hospital that had an average score below the 90th percentile of national performance and summed the potential effect of improving discharge planning. These estimates are based only on hospitals reporting both HQA and HCAHPS data, which account for nearly 80% of inpatient care for pneumonia and congestive heart failure. More information about the modeling approach and the specific models used are detailed in the Supplementary Appendix. All reported P values are two-sided and have not been adjusted for multiple testing. All analyses were conducted with the use of SAS software (version 9.0) or SUDAAN software. Results A total of 2222 hospitals reported performance on both chart-based and patient-reported discharge measures. Reporting hospitals provided 77% of all inpatient care for congestive heart failure and 78% of inpatient care for pneumonia; of the hospitals that did not report performance on both measures, 86% did not provide data on the patientreported measure, 11% did not provide data on the chart-based measure, and 3% did not provide data on either one. Reporting hospitals differed from nonreporting hospitals in several key characteristics, including size, teaching status, ownership status, and geographic location (Table 1). Performance with respect to chart-based discharge instructions for patients with congestive heart failure varied widely: about 1 of 10 hospitals documented appropriate discharge instructions for less than 50% of their patients, whereas more than 1 of 5 hospitals scored 90% or above (Fig. 1A). The distribution of performance on the HCAHPS survey was different: most U.S. hospitals reported that between 71% and 90% of their patients answered affirmatively to both questions about receiving key discharge instructions (Fig. 1B). Table 1. Characteristics of Hospitals That Reported and Those That Did Not Report Discharge Instructions on the HQA and HCAHPS Measures.* Characteristic Reporting Hospitals (N = 2222) Nonreporting Hospitals (N = 1809) P Value Size no. (%) <0.001 Small, 6 99 beds 476 (21) 1237 (68) Medium, beds 1412 (64) 498 (28) Large, 400 beds 334 (15) 74 (4) Region no. (%) <0.001 Northeast 394 (18) 170 (9) Midwest 567 (26) 615 (34) South 857 (39) 692 (38) West 404 (18) 332 (18) Ownership status no. (%) <0.001 For profit 365 (16) 249 (14) Private not for profit 1564 (70) 979 (54) Public not for profit 293 (13) 581 (32) Teaching hospital no. (%) 216 (10) 56 (3) <0.001 Urban hospital no. (%) 2073 (93) 1066 (59) <0.001 Presence of medical intensive care unit no. (%) 1873 (84) 888 (49) <0.001 Ratio of nurses to 1000 patient-days 6.2± ± day readmission rate % For congestive heart failure 25.2± ±15.5 <0.001 For pneumonia 20.6± ± * Plus minus values are means ±SD. Numbers may not add to 100 because of rounding. HCAHPS denotes Hospital Consumer Assessment of Healthcare Providers and Systems, and HQA Hospital Quality Alliance n engl j med 361;27 nejm.org december 31, 2009

5 Discharge Planning in U.S. Hospitals We found essentially no association between performance on the chart-based discharge measure and the patient-reported discharge measure (Spearman correlation coefficient, 0.05; P<0.001). Not surprisingly, we found little overlap between hospital characteristics associated with better performance on the two discharge measures: larger, for-profit hospitals located in the Midwest had modestly better performance on chart-based discharge instructions (Table 2), whereas not-forprofit, smaller institutions with higher nursestaffing levels had somewhat better performance on the patient-based measure; the latter finding was consistent with our hypotheses. Variation in Readmission Rates across Health Care Markets We found considerable variation in 30-day riskadjusted readmission rates across the United States (Table 3). The median readmission rate among patients with congestive heart failure in the hospital-referral regions was 23.3% (interquartile range, 21.3 to 25.4). Ogden, Utah, had the lowest readmission rate for patients with congestive heart failure (13.2%), whereas Oxford, Mississippi, had the highest (36.6%). Among patients with pneumonia, the median readmission rate in the hospital-referral regions was 18.3% (interquartile range, 16.0 to 20.6). Bend, Oregon, had the lowest readmission rate (10.1%), whereas Texarkana, Arkansas, had the highest (27.3%). The correlation between the readmission rate for congestive heart failure in a hospital-referral region and the readmission rate for pneumonia in the same region was 0.69 (P<0.001). Using Bayesian hierarchical models, we found that variation decreased somewhat: readmission rates across hospital-referral regions ranged from 17.5% to 29.6% among patients with congestive heart failure and from 14.1% to 25.6% among patients with pneumonia. Association between Performance on Discharge Measures and Readmission Rates We found no association between a hospital s performance on the chart-based discharge measure and the 30-day readmission rate among the hospital s patients with congestive heart failure. Unadjusted results are presented in the Supplementary Appendix. On analyses adjusted for institutional characteristics, hospitals in the highest quartile of performance on this measure had readmission rates that were nearly identical to those of hospitals in the lowest quartile (23.7% vs. 23.5%, P = 0.54). A Performance on HQA Discharge Instructions Hospitals (%) Performance Score (%) B Performance on HCAHPS Discharge Instructions Hospitals (%) Performance Score (%) Figure 1. Performance of Hospitals on Two Measures of the Adequacy of Discharge Instructions. Panel A shows the performance scores of hospitals based on data from the charts of patients with congestive heart failure as captured by the Hospital Quality Alliance (HQA) program. Panel B shows the performance scores of hospitals based on patients reports on discharge instructions as captured by the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) patient program. We found a very modest association between a hospital s performance on the patient-reported measure and its readmission rates among both patients with congestive heart failure and patients with pneumonia (Table 4). Hospitals in the highest quartile of performance on the patient-reported measure, as compared with hospitals in the lowest quartile, had readmission rates that were 2.4% lower (95% confidence interval [CI], 1.7 to 3.0) for patients with congestive heart failure and 2.0% lower (95% CI, 1.4 to 2.6) for patients with pneumonia. Potential Effect of Better Discharge Planning on Readmission Rates Because we found no association between the chart-based measure and readmission rates, we present only the results from projections associated with the improvement in the patient-reported measure. Annually, there are approximately 505,000 n engl j med 361;27 nejm.org december 31,

6 Table 2. Hospital Characteristics Independently Associated with Performance on Discharge Instruction Measures.* Characteristic Size Region No. of Hospitals (N = 2222) HQA for Congestive Heart Failure P Value HCAHPS P Value % % Small Medium Large Northeast < <0.001 Midwest South West Ownership status For profit < <0.001 Private not for profit Public Teaching status Member of the COTH Not member of the COTH Location Urban Rural Dedicated coronary care unit Yes No Ratio of nurses to 1000 patient-days Lowest quartile <0.001 Second quartile Third quartile Highest quartile * P values are for a test for heterogeneity. COTH denotes Council of Teaching Hospitals, HCAHPS Hospital Consumer Assessment of Healthcare Providers and Systems, and HQA Hospital Quality Alliance. The presence of a coronary care unit is a marker of technological capability. admissions for congestive heart failure and 476,000 admissions for pneumonia among Medicare patients discharged from hospitals that report performance on the discharge measures. We found that improving performance on the patient-reported measure in each hospital to the 90th percentile nationally would lower the readmission rate for congestive heart failure, on the basis of the magnitude of the observed association, from 24.1% to 23.3%. This decrease would result in approximately 4700 fewer readmissions among Medicare beneficiaries. For the readmission rate for pneumonia, the reduction would be from 18.3% to 17.6%, leading to 2800 fewer readmissions. Discussion The CMS has initiated a national effort to reduce readmission rates in U.S. hospitals. We found large variations in readmission rates across health care markets, underscoring the need for such a program. To date, a major focus of federal efforts has 2642 n engl j med 361;27 nejm.org december 31, 2009

7 Discharge Planning in U.S. Hospitals Table 3. Hospital-Referral Regions with the Highest and Lowest Readmission Rates for Congestive Heart Failure and Pneumonia. Region Readmission Rate for Congestive Heart Failure Region Readmission Rate for Pneumonia % % Highest readmission rates Oxford, MS 36.6 Texarkana, AR 27.3 Joplin, MO 29.8 Lake Charles, LA 27.0 Elyria, OH 29.8 Oxford, MS 26.9 Jackson, TN 29.8 Monroe, LA 26.9 Munster, IN 29.4 Chicago, IL 25.9 Lowest readmission rates Lafayette, IN 15.2 Salt Lake City, UT 11.3 Appleton, WI 15.3 Idaho Falls, ID 11.2 Salt Lake City, UT 14.5 South Bend, IN 10.9 Bend, OR 14.5 Neenah, WI 10.6 Ogden, UT 13.2 Bend, OR 10.1 been on measurement and public reporting of the conduct of discharge planning, which has engaged more than 4000 hospitals. Although public-reporting efforts are often associated with improvements in performance measures, 14,15 they also have substantial administrative costs. The very modest association that we observed between readmission rates and discharge planning measures suggests that the use of public reporting as a strategy to improve performance on these measures is unlikely to yield large reductions in unnecessary readmissions. To substantially reduce readmissions, additional efforts are needed. The National Quality Forum has endorsed the publication of hospital readmission rates for common conditions, and these data are now publicly available on the Hospital Compare Web site ( The Quality Improvement Organizations of the CMS have initiated efforts in several communities to implement a Care Transitions Program designed to reduce readmissions. 16 President Barack Obama s budget includes provisions to reduce payments for hospital readmission. This approach will probably be needed to realize a meaningful effect. However, whether these efforts will primarily reduce unnecessary readmissions or have unintended consequences will need to be tracked. 17 We were surprised to find essentially no association between performances on the two discharge measures. Presumably, hospitals that document consistent provision of written discharge instructions to patients with congestive heart failure would have higher ratings on patients reports of whether they were asked about having adequate help at home or received written discharge instructions, especially given that the latter question is nearly identical to the chart-based measure. Important differences between the two measures might explain the poor correlation. The HQA measure is only for patients with congestive heart failure, whereas the HCAHPS measure is for all hospitalized medical and surgical patients; therefore, hospitals that target improvement of discharge instructions for patients with congestive heart failure may not see substantial improvements on their HCAHPS scores. Alternatively, chart-based measures may simply capture whether hospitals document their activities, not how effectively they perform. The patients reports come closer to a standard, although the HCAHPS measure focuses on limited aspects of discharge planning. Some previous studies suggest that comprehensive discharge planning, effective case management, remote monitoring of the patient s condition from his or her home, and meticulous follow-up can reduce the frequency with which patients with congestive heart failure return to the emergency room or are rehospitalized. 10,18 However, other studies have not shown, for ex n engl j med 361;27 nejm.org december 31,

8 Table 4. Association between the Two Measures of Performance on Discharge Planning and All-Cause 30-Day Readmission Rates.* Rating Patients with Congestive Heart Failure (N = 252,266) P Value readmission rate (95% CI) Patients with Pneumonia (N = 237,935) P Value readmission rate (95% CI) Score on HQA discharge instructions Lowest quartile 23.5 ( ) 18.4 ( ) Second quartile 23.2 ( ) 18.5 ( ) Third quartile 23.6 ( ) 18.2 ( ) Highest quartile 23.7 ( ) 18.7 ( ) Score on HCAHPS discharge instructions <0.001 <0.001 Lowest quartile 24.7 ( ) 19.5 ( ) Second quartile 23.7 ( ) 18.3 ( ) Third quartile 23.1 ( ) 18.2 ( ) Highest quartile 22.4 ( ) 17.5 ( ) * Both the Hospital Quality Alliance (HQA) score and the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) score are based on a scale of 1% to 100%, with higher scores indicating better performance. The score was adjusted for patient characteristics (age, race or ethnic group, sex, and the presence or absence of 29 coexisting conditions) as well as hospital characteristics (size, location [urban vs. rural], region, ratio of nurses to patientdays, profit status, teaching status, and the presence or absence of a cardiac intensive care unit). P values are for a test of trend. ample, that the use of nurse case managers or remote monitoring is helpful, 19 suggesting that the context and specifics of any individual program to reduce readmissions are important drivers of success. 20 Our study has limitations. First, we were unable to examine performance on the discharge measures among a minority of U.S. hospitals because they have not yet begun to report HCAHPS data. The number of readmissions averted could exceed our estimates as more hospitals begin to report these data. However, the nonreporting hospitals collectively care for a small minority of patients. Second, we assessed only associations between discharge measures and risk-adjusted readmission rates. Although it is plausible that better performance on HCAHPS discharge instructions is causally related to marginally lower readmission rates, it is also possible that such performance is a measure of managerial focus on quality and that further improvement may not reduce readmission rates. Our data cannot be used to assess causality or provide guidance on whether efforts to lower readmission rates will provide benefits that are commensurable with greater costs. Finally, we used administrative data to perform risk adjustment, and although their accuracy and completeness have improved over time, these data have limitations. However, they are increasingly used, even for public reporting of performance on measures of health outcomes. In conclusion, we examined national performance on two quality measures currently used to assess hospitals discharge planning. Our findings indicate that improvements in performance on current measures of discharge planning are unlikely to have a meaningful effect. The large variation in readmission rates across health care markets suggests ample opportunity for improvement, but to substantially reduce readmission rates, we will probably need new strategies to improve transitional care in the ambulatory sector. Supported by the Commonwealth Fund. Dr. Jha reports receiving consulting fees from UpToDate. No other potential conflict of interest relevant to this article was reported n engl j med 361;27 nejm.org december 31, 2009

9 Discharge Planning in U.S. Hospitals References 1. McGlynn EA, Asch SM, Adams J, et al. The quality of health care delivered to adults in the United States. N Engl J Med 2003;348: Schoen C, Davis K, How SK, Schoenbaum SC. U.S. health system performance: a national scorecard. Health Aff (Millwood) 2006;25:w457-w Fisher ES, Wennberg JE, Stukel TA, Sharp SM. Hospital readmission rates for cohorts of Medicare beneficiaries in Boston and New Haven. N Engl J Med 1994; 331: Report to Congress: promoting greater efficiency in Medicare. Washington, DC: Medicare Payment Advisory Commission, Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med 2009;360: Coleman EA, Berenson RA. Lost in transition: challenges and opportunities for improving the quality of transitional care. Ann Intern Med 2004;141: Bodenheimer T. Coordinating care a perilous journey through the health care system. N Engl J Med 2008;358: Jha AK, Orav EJ, Zheng J, Epstein AM. Patients perception of hospital care in the United States. N Engl J Med 2008; 359: QualityNet. Specifications manual for national hospital inpatient quality measures. Baltimore: Centers for Medicare & Medicaid Services, (Accessed December 4, 2009, at QnetPublic%2FPage%2FQnetTier2&cid= ) 10. Phillips CO, Wright SM, Kern DE, Singa RM, Shepperd S, Rubin HR. Comprehensive discharge planning with postdischarge support for older patients with congestive heart failure: a meta-analysis. JAMA 2004;291: Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care 1998; 36: Wennberg JE, Cooper MM, eds. The Dartmouth atlas of health care Chicago: American Hospital Publishing, Krumholz HM, Wang Y, Mattera JA, et al. An administrative claims model suitable for profiling hospital performance based on 30-day mortality rates among patients with heart failure. Circulation 2006;113: Fung CH, Lim YW, Mattke S, Damberg C, Shekelle PG. Systematic review: the evidence that publishing patient care performance data improves quality of care. Ann Intern Med 2008;148: Jha AK, Orav EJ, Li Z, Epstein AM. The inverse relationship between mortality rates and performance in the Hospital Quality Alliance measures. Health Aff (Millwood) 2007;26: Medicare Quality Improvement Community (MedQIC) home page. (Accessed December 4, 2009, at Epstein AM. Revisiting readmissions changing the incentives for shared accountability. N Engl J Med 2009;360: Rich MW, Beckham V, Wittenberg C, Leven CL, Freedland KE, Carney RM. A multidisciplinary intervention to prevent the readmission of elderly patients with congestive heart failure. N Engl J Med 1995;333: McAlister FA, Lawson FM, Teo KK, Armstrong PW. A systematic review of randomized trials of disease management programs in heart failure. Am J Med 2001;110: Peikes D, Chen A, Schore J, Brown R. Effects of care coordination on hospitalization, quality of care, and health care expenditures among Medicare beneficiaries: 15 randomized trials. JAMA 2009; 301: Copyright 2009 Massachusetts Medical Society. powerpoint slides of journal figures and tables At the Journal s Web site, subscribers can automatically create PowerPoint slides. In a figure or table in the full-text version of any article at NEJM.org, click on Get PowerPoint Slide. A PowerPoint slide containing the image, with its title and reference citation, can then be downloaded and saved. n engl j med 361;27 nejm.org december 31,