30-Day Readmission: A Lousy Quality Metric in HF?

What works great for ACS seems to be failing miserably as a quality metric in the setting of heart failure (HF). Thirty-day readmit numbers have been driven down in recent years for patients treated for acute MI; the comparable numbers for HF, meanwhile, seem largely mired in the muck.

Given that hospitalization costs are responsible for the majority of the $39 billion spent annually by Medicare for patients with HF, it's no surprise that early rehospitalization rates have become a focal point for quality improvement and cost reduction attempts in the setting of HF, too.

Over the last several years, the Centers for Medicare & Medicaid Services (CMS) have undertaken several initiatives to reduce readmissions among the Medicare fee-for-service population. These include:

  • reporting hospital readmission rates through Hospital Compare
  • funding hospital-level improvements through the Partnership for Patients program
  • changing payment policies through the Hospital Readmissions Reduction Program
  • various shared savings initiatives

In turn, many hospitals and other organizations have employed strategies to reduce readmissions, such as enhanced patient education, more post-discharge follow-up care, and increased coordination with outpatient providers.

Hospitals have responded to the intense interest. Recently, Bradley et al. reported the results of a survey to describe the practices employed by hospitals enrolled in the Hospital to Home (H2H) quality improvement initiative to reduce readmissions for HF or acute MI.1 The initiative, organized by the ACC and the Institute for Healthcare Improvement (IHI), is a rich resource of information and tools to reduce readmission. In the recent survey, the response rate of 90.4% reflects the strong motivation of the participating hospitals and 87% of responding facilities had a quality improvement team. The hospitals tracked a variety of metrics and employed various approaches for medication reconciliation and discharge planning. Clearly, hospitals are investing significant resources for quality improvement and their practices are diverse.

So, Have We Made Any Progress?

Six Strategies to Reduce 30-Day HF Readmissions

Combining several strategies may be the key to lower risk-standardized 30-day readmission rates (RSRRs) for patients with HF. Elizabeth H. Bradley, PhD, Yale School of Public Health, and colleagues conducted a survey of almost 600 hospitals participating in national quality initiatives to reduce readmission and then constructed a multivariable linear regression model to determine strategies independently associated with 30-day RSRRs.

Six strategies were associated with significantly lower RSRRs:

  • Partnering with community physicians or physician groups to reduce readmission
  • Partnering with local hospitals to reduce readmissions
  • Having nurses responsible for medication reconciliation
  • Arranging a follow-up appointment before discharge
  • Having a process in place to send all discharge paper or electronic summaries directly to the patient's primary physician
  • Assigning staff to follow up on test results that return after the patient is discharged
  • Many of these strategies were implemented by less than two-thirds of the hospitals.
Source: Bradley EH, Curry L, Horwitz LI, et al. Circ Cardiovasc Qual Outcomes. 2013;6:444-50.
Hospital readmission rates for Medicare fee-for-service beneficiaries significantly declined in 2012. Compared to the 19% rate for 30-day all-cause readmission seen for 2007 to 2011, the monthly readmission rate dropped to a low of 18% in October 2012 and averaged 18.4% for the year.2 An analysis from Dartmouth Atlas revealed similarly stagnant levels of readmission rates in recent years.

With so much effort, both punitive and rewarding, is a change of only a six-tenth of one percentage point worth the effort? Said Ashish Jha, MD, MPH, of Harvard School of Public Health: "Either we have no idea how to really improve readmissions, or most of the readmissions are not preventable and the efforts being put on it are not useful." Put another way, in a commentary in JACC, Drs. Butler and Kalogeropoulos believe in the current environment, policy trumps science.3 Hospitals are being compelled to adopt a variety of strategies to reduce HF readmissions with little evidence to justify the enormous resources being expended on this effort.

Well, maybe it's an issue of guideline adherence. Maybe the best centers do better. Consider recent data from the American Heart Association's Get With The Guidelines Program for HF (GWTG-HF).4 Investigators compared hospitals enrolled in GWTG-HF from 2006 to 2007 with other hospitals using data on four process measures of HF care, five noncardiac process measures, risk-adjusted 30-day mortality, and 30-day all-cause readmission after an HF hospitalization, as reported by CMS.

While hospitals enrolled in the GTWG-HF program demonstrated better processes of care than other hospitals, there were few clinically important differences in outcomes. After adjusting for hospital characteristics, 30-day mortality rates were no different (p = 0.45), although 30-day readmission was slightly lower for GWTG hospitals (-0.33%; p = 0.002).

Searching for a Better Measure

Former president of the American Heart Association, Clyde W. Yancy, MD, coauthored that paper and is now asking, "Are we targeting the right metric for heart failure?" Recent data suggest that length-of-stay for HF admissions has decreased over the past decade, with a subsequent increase in 30-day readmission rates. In a recent paper, he looks at an alternative measure: total hospital days (including the index admission and any hospital days related to readmission) within 30 days.4 This is referred to as overall episode-of-care (EOC) inpatient days. Could this be a better indicator of resource use, HF quality, and outcomes?

Dr. Yancy and colleagues evaluated data on 17,387 patients 65 years or older enrolled in GWTG-HF and treated in 149 hospitals. The median hospital-level 30-day readmission rate was 23.2% and this metric was not associated with hospital length-of-stay, better performance on quality measures, or 30-day mortality risk. There was only a modest association between hospital ranking by overall 30-day EOC and readmission rates. Surprisingly, higher readmission rates were paradoxically associated with better adherence to quality indicators, including a composite measure of 100% adherence to CMS performance measures.5 Yes, that means hospitals with the lowest 30-day readmission rates performed worse on current HF performance measures.

While 30-day rehospitalization was not associated with lower 30-day mortality, the shortest EOC quartile was associated with decreased odds of 30-day mortality.

Dr. Yancy said this is not a theoretical concern: from the initial set of observations, assigning penalties to facilities with higher-than-average 30-day readmission led to lower 30-day readmission rates—and higher 30-day mortality. "We need to continue to exercise very careful due diligence," said Dr. Yancy, "because if this is a pattern that is repeated as we develop more experience with managing this 30-day window, it's not a direction that is appropriate for best patient outcomes, even if it does save resources."

Where's the Data?

The idea that 30-day readmission rates may be a lousy quality metric for HF is no surprise to Javed Butler, MD, MPH and Andreas Kalogeropoulos, MD, PhD (both from Emory University School of Medicine), who wrote the aforementioned commentary3 to the JACC paper by Dr. Bradley and colleagues. They wrote that the overarching question is not what the hospitals are doing to prevent HF readmission, but why are they doing it and what is the evidence that these interventions are effective and cost-effective?

The answer to why is simple: because hospitals are compelled to do so. The evidence to support these measures, on the other hand, is just not there. Unlike acute MI, where the evidence base in support of the 30-day readmit metric is strong, Drs. Butler and Kalogeropoulos note that HF is a complex disease, and what should work logically does not always work in practice. As for many of the efforts to reduce readmissions (spawned by the CMS decision to not pay for early readmissions), they wrote, "Many of these activities are neither proven nor primarily based on the motivation to improve patient outcomes, but rather on the fear of punitive financial disincentives."

Dr. Yancy and others also have expressed concern that the CMS penalties are unfair because hospitals that treat the poorest patients are getting hit harder than others. Penalties can be a drain on safety net hospitals, many of which operate on slim profits or at a loss.

"The idea is right, but the implementation has been greatly flawed by penalizing hospitals that take care of the most vulnerable patients," said Atul Grover, MD, PhD, chief public policy officer at the Association of American Medical Colleges. Low-income patients are more likely to have trouble following hospital instructions for taking care of themselves after discharge. They don't always have easy access to doctors to monitor their recuperations and sometimes can't afford needed medications.

Medicare has disagreed that the readmissions penalty program needs revisions, but in a June 2013 report to Congress, the Medicare Payment Advisory Commission, or MedPAC, agreed with critics that there are "shortcomings" that "can work at cross purposes to the policy's intent." The criticisms carry extra weight because MedPAC helped devise the readmission penalties, calling for them back in 2008. "Income is still an important variable in explaining variation in readmissions," the commission said. It recommended that in future years, when determining penalties, Medicare should compare a hospital's readmission rates to those of hospitals with comparable numbers of poor patients.

Post-Hospital Syndrome

Another way of looking at the problem comes from a paper by Stephen F. Jencks, MD, MPH, and colleagues in the New England Journal of Medicine that found nearly two-thirds of hospital readmissions are not for the original condition cited at initial discharge.6 If only one-third of readmissions are for the same condition as the recent hospitalization, what's the problem? Obviously, readmission is an issue, but why all the emphasis on preventing readmissions when clearly most readmissions are for something else and not linked to the hospitalization?

Maybe yes, maybe no: according to Harlan M. Krumholz, MD, professor of cardiology and health policy at Yale University and Yale-New Haven Hospital, one of the chief architects of the government's new, punitive hospital readmission measures. In the April 2013 issue of Today's Hospitalist, he offered this scenario: if someone crashes his car less than a month after being discharged from the hospital with pneumonia, many might say the readmission resulting from the accident had nothing to do with the initial hospitalization. Hence, the reasoning would go, the new hospitalization shouldn't count against the hospital as a preventable readmission.

But, he said, what if people are "at elevated risk of accidents" because they were recently hospitalized? Then, he surmised, it is related. In an essay in the January 10th issue of the New England Journal of Medicine, Dr. Krumholz suggested that the immediate post-hospitalization period may need to be re-imagined as "phase 2" of an illness, a period marked by heightened "generalized risk" for many conditions. "I call it 'post-hospital syndrome' because a label gets the attention of doctors," Dr. Krumholz explains. "It gets you thinking that this could actually be a coherent syndrome that is acquired, transient and something patients didn't have before they came to the hospital."

While preventable readmissions should be a target, Dr. Yancy said, "it's hard to embrace a 30-day readmission window as an appropriate quality measure when the centers at risk are the centers that need the resources the most. And it is even more disquieting that the hospitals that are successful in reducing 30-day readmission rates may be doing this at the expense of 30-day mortality rates."

Dr. Yancy added, "We're entering a very different health care space and we need to be held accountable for quality, no question about that, and we should be expected to follow best practices that will not produce needless expenditures. But in our zeal to make all of these things happen concurrently and quickly, we can't lose sight of the importance of best patient outcomes."


1. Bradley EH, Curry L, Horwitz LI, et al. J Am Coll Cardiol. 2012;60:607-14. http://content.onlinejacc.org/article.aspx?articleid=1221473
2. Gerhardt G, Yemane A, Hickman P, et al. Medicare Medicaid Res Rev. 2013;3(2). http://www.cms.gov/mmrr/Briefs/B2013/mmrr-2013-003-02-b01.html
3. Butler J, Kalogeropoulos A. J Am Coll Cardiol. 2012;60:615-7. http://content.onlinejacc.org/article.aspx?articleid=1221472
4. Kociol RD, Liang L, Hernandez AF, et al. Am Heart J. 2013;165:987-94.e1.
5. Heidenreich PA, Hernandez AF, Yancy CW, et al. Circ Cardiovasc Qual Outcomes. 2012;5:37-43.
6. Jencks SF, Williams MV, Coleman EA. N Engl J Med. 2009;360:1418-28.
7. Krumholz HM. N Engl J Med. 2013;368:100-2.

Clinical Topics: Heart Failure and Cardiomyopathies, Acute Heart Failure

Keywords: Quality Improvement, Fee-for-Service Plans, Patient Readmission, Quality Indicators, Health Care, Medication Reconciliation, Heart Failure, Guideline Adherence, Inpatients, Hospitalization

< Back to Listings