It has been more than 7 years since Congress enacted the Hospital Readmissions Reduction Program (HRRP) as part of the Affordable Care Act (ACA). This program created financial penalties for hospitals with higher-than-expected rates of readmissions within 30 days of discharge for Medicare patients with certain conditions, such as heart failure, based on the assumption that high readmission rates were a marker for poor quality care. Although early data suggested that HRRP was improving care, new evidence suggests that the benefits may have been much smaller than understood and that there might have been meaningful, unintended consequences.
One thing is becoming clear: HRRP is not working as well as its architects had hoped, and it is time to take a fresh look at the program.
Readmissions Are a Controversial Measure
The program was highly controversial right from the start. Although policy makers saw the metric as a valuable quality measure, clinicians were far less enthusiastic. Policy makers often used the analogy that if you have to take your car back into the auto shop within 30 days of having it serviced, the mechanic probably didn’t do a good job. The clinicians retorted that people are much more complex than cars and for many patients—especially older, sicker ones—returning to the hospital can be a natural part of the disease course, no matter the quality of care received.
Despite concerns, the HRRP was enacted, and its early successes were heralded as a bright spot during some of the most contentious days of the implementation of the ACA. Several studies showed that after years of stagnation, readmission rates were starting to fall. The timing seemed unmistakable: soon after the ACA’s passage, readmission rates fell quickly. These decreases, and the penalties that would soon follow for hospitals with higher-than-expected readmission rates, were estimated to be saving the federal government hundreds of millions, if not billions of dollars. The presumption was that patients were better off as well.
However, the evidence that the HRRP was responsible for the reduction in readmissions was based on analyses of admission rates (without control groups) before and after implementation, largely because most hospitals in the United States are subject to the HRRP penalties. More troubling, readmission rates appeared to have decreased immediately after the ACA’s passage—well before the penalties for excessive readmissions kicked in. The notion that hospitals could quickly lower readmission rates within the first few months after the passage (but prior to the enactment) of the HRRP program was surprising—and should have been the first clue that these reductions were not what they appeared.
Although there is little doubt that risk-standardized readmission rates fell after the enactment of the ACA, only recently have we begun to understand the true reason why. A critically important article by Andrew M. Ibrahim, MD, MSc, of the Institute for Healthcare Policy and Innovation, and colleagues published in JAMA Internal Medicine, finds that nearly two-thirds of the reduction in national readmission rates was due to changes in coding practice—excluding some patients from the denominator and coding substantially more comorbidities. The authors suggest that the decrease in national readmission rates were drastically overestimated. The actual reduction for the Medicare population, accounting for changes in billing practices, appears to be approximately 1.14 percentage points as opposed to the 2.41 percentage points typically attributed to the HRRP.
Relationship Between Readmissions and Mortality
Although much of the reduction in readmission was due to changes in coding, there was, nevertheless, some decrease in how frequently people were being readmitted to the hospital. This likely came from hospitals shifting their focus to preventing readmissions. But emerging evidence that this new focus might have had unintended consequences is concerning.
We knew, going into the HRRP, that for congestive heart failure (CHF), there was clear evidence of an inverse relationship between a hospital’s readmission rate and its mortality rate—low-mortality hospitals tended to have higher readmissions. This makes sense: hospitals with low mortality rates are better at keeping their sickest patients alive. However, these sicker patients are at higher risk of being readmitted than the average patient, which increases readmission rates. Therefore, it is no surprise that for some conditions, hospitals with low mortality rates have higher readmission rates.
Whether the modest readmission rate decreases that occurred during the implementation of HRRP have been associated with higher mortality is unclear, but at least 2 studies taking very different approaches have examined this question and obtained contradictory results. First, Harlan M. Krumholz, MD, SM, of Yale University School of Medicine, and colleagues (the original creators of the readmission measure) examined hospital-level “postdischarge” mortality and its correlation with postdischarge readmission rates and found no deleterious effect. Postdischarge mortality is a novel metric and presumes that the patient survives the hospitalization. This metric ignores anyone who died during the hospitalization, which is a critical component of assessing quality of care for a hospitalized patient. In contrast, a group of leading cardiology experts led by Ankur Gupta, MD, PhD, of Harvard Medical School, found that over time, as readmission rates for CHF fell, 30-day all-cause mortality rates increased across the nation.
Is it possible that HRRP led to an increase in mortality rates? It is. We know that with the program’s implementation, hospitals’ incentives for reducing readmissions were between 6 and 10 times greater than the incentives for reducing mortality. Given this signal by the national “pay for performance” programs, it isn’t surprising that hospitals shifted their attention from mortality to readmissions. Mortality, which had been decreasing for nearly a decade, seems to have plateaued or even increased after HRRP was implemented, while readmissions have fallen. Of course, it is possible that there is no relationship between HRRP implementation and the changes in mortality rates. But the evidence that HRRP caused the decrease in readmissions is of comparable quality to the evidence that HRRP caused an increase in mortality. It is hard to accept one without accepting the other.
It has been more than 7 years since HRRP was enacted and 5 years since it was implemented. The penalties to hospitals continue, even as readmission rates appear to have plateaued over the past few years. With the emerging evidence that much of the initial gains seen were likely due to changes in coding and that there may have been an unintended consequence of worsening mortality rates, it is time to take a fresh look at this program.
Congress could do several things to update the HRRP, based on the new body of evidence. First, and foremost, it is time to reweight the incentive programs so that mortality matters more than readmissions. The current weighting method makes little clinical sense and is not consistent with the values of most patients. Readmissions are not ideal but they are not a fate much worse than death.
Second, it may be worth looking beyond readmissions. The Centers for Medicare & Medicaid had made strides in this area by focusing on bundled payments. Bundling all payments within 30 or 90 days and tying this payment to quality measures has the added advantage of promoting high quality, efficiently delivered care across the entire spectrum of postdischarge care.
Finally, Congress could let patients determine the incentives hospitals receive. Ultimately, whether a readmission was appropriate or whether a death was preventable is better understood by patients and their families than through any national policy created in Washington, DC. Therefore, having patients decide whether a hospital should receive a bonus or a penalty for the care they received can substantially alter the incentives for gaming and get hospitals to focus on what matters to patients.
The Hospital Readmissions Reduction Program is a well-intentioned program crafted to make hospitals more accountable for the care patients receive after discharge. The evidence suggests that its benefits have been small and its costs potentially large. By making changes to the program, Congress and the Trump administration can signal that improving care for Medicare beneficiaries is a serious priority. Heeding the evidence would be a good place to start.
About the author: Ashish K. Jha, MD, MPH, is K. T. Li Professor of Global Health and Health Policy at the Harvard T. H. Chan School of Public Health and a practicing internist at the Veterans Affairs Boston Healthcare System. He received his doctor of medicine from Harvard Medical School and was trained in internal medicine at the University of California, San Francisco. He received his master’s in public health from Harvard School of Public Health. Dr Jha’s major research interests lie in improving the quality and costs of health care. His work has focused on 4 primary areas—public reporting, pay for performance, health information technology, and leadership—and the roles they play in fixing the US health care system. (Image: Aubrey LaMedica/Harvard T.H. Chan School of Public Health)
About The JAMA Forum: JAMA has assembled a team of leading scholars, including health economists, health policy experts, and legal scholars, to provide expert commentary and insight into news that involves the intersection of health policy and politics, economics, and the law. Each JAMA Forum entry expresses the opinions of the author but does not necessarily reflect the views or opinions of JAMA, the editorial staff, or the American Medical Association. More information is available here and here.