CMS Penalties and 30-Day Postdischarge Metrics: Next Chapter

October 04, 2018

Declines in 30-day readmission rates did not come at the cost of higher 30-day mortality after hospitalizations for acute MI, heart failure, or pneumonia following introduction of the Hospital Readmissions Reduction Program (HRRP), concludes a new report.

The analysis of Centers for Medicare and Medicaid Services (CMS) data extending to before the advent of HRRP, part of the game-changing Affordable Care Act (ACA), is the latest chapter in an ongoing debate about its consequences.

The report follows high-profile claims that although the introduction of HRRP was followed by substantial nationwide declines in the controversial 30-day readmission metric, 30-day mortality after HF hospitalizations climbed at the same time, possibly as a result.

Central to the new report's argument: its assertion that two critical events in the process of introducing HRRP had no significant effects on the direction of 30-day postdischarge mortality trends.

Those events, as described in the report published September 28 in JAMA Network Open, were the April 2010 announcement that CMS reimbursement penalties based on 30-day readmissions and other metrics would be forthcoming, and the October 2012 implementation of those penalties. The ACA became law in March 2010.

The CMS data included in the analyses covered the much longer period from January 1, 2006, to December 31, 2014 encompassing 1.7 million hospitalizations for acute MI, 4.0 million for heart failure, and 3.5 million for pneumonia by fee-for-service Medicare beneficiaries. The analyses included both raw and risk-standardized numbers, which tended to tell similar stories.

Indeed, "readmissions declined substantially across the three conditions" after the program was announce, Rohan Khera, MD, told | Medscape Cardiology.

"And we found that mortality fell for acute MI both in-hospital and postdischarge. In-hospital mortality declined for heart failure and pneumonia as well," said Khera, from University of Texas Southwestern Medical Center, Dallas, who is lead author of the report.

Mortality within 30 days after discharge from an HF or pneumonia hospitalization did rise over the period included in the analysis, but the trend did not seem related to the announcement or implementation of reimbursement penalties, Khera said.

"We know that this mortality rise predated the announcement of the HRRP, starting in 2007 and 2008, and we just found that there were no changes in the trend that had already started."

Bottom line is that their analysis showed no evidence that either in-hospital or 30-day postdischarge mortality went up as a result of hospitals getting wind of impending reimbursement penalties or their implementation, conclude the report's authors.

They predominantly include members of the Yale University team that had received funding to develop readmission and mortality metrics for CMS.

The HRRP has had both supporters and detractors, at least for some of its major elements, but has been especially contentious for heart-failure specialists. Many of them have seen 30-day readmissions as a poor and counterproductive HF-care performance metric, and at least one sees support for that view in the current analysis.


A Matter of Interpretation

In contrast to the authors' conclusions, the new analysis "adds to the growing body of evidence" that risk-standardized readmission rates used for HRRP-prescribed reimbursement penalties did not benefit and may have compromised the care of patients hospitalized with heart failure, Gregg C. Fonarow, MD, University of California, Los Angeles, told | Medscape Cardiology.

The current analysis points out that unadjusted 30-day postdischarge mortality for HF patients rose from 7.4% to 9.2% (P < .001 for trend) throughout the 2006 to 2014 data period; the pneumonia data showed a similar "concerning signal," he said.

The risk-adjusted data showed the same pattern, "suggesting that the strategies hospitals adopted to avoid public blaming for excess readmission and financial penalties were independently associated the unintended consequences of increased mortality for heart failure patients," Fonarow said.

Rising mortality throughout the time hospitals publically reported the outcomes data, with no directional change in the slope of the mortality curves (a criterion Khera and colleagues used to show consistent trends), is quite different from concluding there was no mortality increase, he said.

"A steady increase in heart failure patients dying in the first 30-day post-hospital-discharge is not reassuring nor evidence that a health care policy has been a success."

Fonarow, not associated with the current study, was senior author on a recent analysis that saw the vaunted post-HRRP declines in 30-day HF readmissions accompanied by a jump in 30-day mortality. The data came from hospitals participating in the Get With The Guidelines (GWTG) Heart Failure care quality improvement initiative.

Referring to that specific report, Khera said it included "a small fraction" of US hospitals, a much smaller sample than the cohort he and his colleagues studied. The two studies used similar analytical methods, he said, but the current study makes a better case due to its broad Medicare population numbering in the millions.

The GWTG analysis involved just over 100,000 patients, and the participating hospitals were "self-selected," he said. "It's a good hypothesis-generating study, and it's an important question to ask, and they saw the signal in heart failure that was concerning. But we applied it to the entire population of patients, and we did not find any signals."

Analysis of Risk-Adjusted CMS Data

For acute MI, mortality fell consistently before the HRRP was unveiled, was flat after the program was announced, then then continued to fall during the period after the penalties were implemented. But 30-day postdischarge mortality held at a consistent level throughout all three segments of the 9 years of data.

For patients with heart failure, in-hospital mortality consistently fell before there was wind of the HRRP, levelled out in the period right after the program was announced, and resumed a downward trend once the penalties kicked in. But postdischarge 30-day mortality rose both before and after the HRRP announcement and continued to climb once the program was implemented.

For those with pneumonia, in-hospital mortality declined continuously throughout the entire nine years. Postdischarge 30-day mortality showed an upward trend before the HRRP was unveiled, which somewhat flattened out after HRRP was announced and after penalties took effect.

All three conditions displayed the same time-trend patterns for risk-adjusted 30-day readmission rates: no significant changes prior to the HRRP announcement, followed by substantial, continuous declines once details of the HRRP became known and after it took effect.

What Has HRRP Accomplished?

An accompanying editorial takes a stand against what it sees as the HRRP's disproportionate focus on 30-day readmissions in the reimbursement penalties it allows.

CMS "needs to adjust the penalties for the program relative to other programs. Right now, a high-readmission, low-mortality hospital will be penalized at six to 10 times the rate of a low-readmission, high-mortality hospital. The signal from policy makers is clear — readmissions matter a lot more than mortality — and this signal needs to stop," writes Ashish K. Jha, MD, MPH, who directs the Harvard Global Health Institute, Cambridge, Massachusetts.

"Second, we need a more comprehensive examination of the impact of the HRRP and both its beneficial and deleterious effects," given the many unanswered questions about its consequences. "It is particularly important that some of these evaluations be conducted by independent experts who have no vested interest in the results," he says.

"The HRRP was a well-intentioned effort to get hospitals to focus on transitions of care, an area in desperate need of improvement. However, 8 years after the passage of this program, we still do not understand whether it has met its primary goal and at what costs."

Fonarow pointed to previous studies (which include at least one of his) that "have suggested that most of the reduction in readmissions have not been the result of hospitals investing in better transitions of care, early follow-up, and disease management."

Rather, "the reductions in 30-day readmissions represent largely gaming efforts, such as upcoding in the medical record, preventing patients presenting to the ER within 30 days from being admitted, even if their condition warranted admission, and diverting patients to 'outpatient' observation status, potentially exposing patients to financial toxicity."

Khera reported no conflicts. Disclosures for the other authors are in the report. Jha reported no conflicts. Fonarow is a member of the GWTG steering committee and has disclosed consulting for Amgen, Bayer, Janssen, Medtronic, Novartis, and St.  Jude Medical.

JAMA Network Open. Published online September 28, 2018. Article, Editorial

Follow Steve Stiles on Twitter: @SteveStiles2. For more from, follow us on Twitter and Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: