Researchers have divergent views about the significance of the new study.


August 9—A new study in Health Affairs may appear to validate a recent decision by the Centers for Medicare & Medicaid Services (CMS) to not consider patients’ socioeconomic status (SES) when determining hospital readmission penalties. But rather than closing the book on this subject, some say the study could just be an opening chapter.

Researchers, led by physicians at Yale-New Haven (Conn.) Hospital, found that hospitals that cared for large proportions of patients from low-income ZIP codes had similar readmission rates to other hospitals and that adjusting for patients’ SES “does not change hospital rates in meaningful ways.”

In the CMS-funded study, researchers examined more than 2.9 million records of heart attack, heart failure, and pneumonia patients from July 1, 2007 to June 30, 2010. They identified patients from ZIP codes where the median two-person household income was below $43,710 (300 percent of the federal poverty level in 2010).

But adjusting for patients’ SES resulted in a mere 0.1 percent change in readmission rates and would only result in only about 3 to 4 percent fewer hospitals being penalized under Medicare’s hospital readmissions reduction program (HRRP), according to the study.

The CMS estimated that about 2,588 hospitals will have their base operating DRG payments cut by a total of $528 million in FY17 compared to 2,666 hospitals experiencing cuts totaling $420 million in the current fiscal year.

The CMS used the recent inpatient prospective payment system final rule to announce that, at least for now, SES would not be factored into its payment formulas.

“We continue to have concerns about holding hospitals to different standards for the outcomes of their patients of diverse [SES] because we do not want to mask potential disparities or minimize incentives to improve the outcomes of disadvantaged populations,” CMS stated in the final rule.

NQF Studying Issue

CMS also noted the National Quality Forum (NQF) is in the midst of a two-year trial reviewing the appropriateness of socioeconomic risk adjustment of performance measures. When the trial is over, CMS will use NQF recommendations and other research to consider whether to adjust programs for socioeconomic factors.

America’s Essential Hospitals (AEH) has been advocating for adding socioeconomic risk adjustment to CMS performance measures and disagreed with the CMS decision to refrain from doing so.

In an interview, Beth Feldpush, AEH, senior vice president of policy and advocacy said the Health Affairs study shows the “academic world” has been struggling to find an accurate way to measure the impact of SES.

“A hospital should not be penalized because of who their patients are,” Feldpush said. “I think the dialogue will continue and we’re glad that CMS is at least paying attention to this.”

Legislation Advances

Congress also has joined the debate, and a bill introduced by Rep. Pat Tiberi (R-Ohio) seeks to create a transitional risk adjustment based on a hospital’s proportion of patients who are dually eligible for Medicare and Medicaid.

Tiberi’s bill, the Helping Hospitals Improve Patient Care Act of 2016, was approved by the House of Representatives by voice vote June 7 and is awaiting action in the Senate.

The connection between a patient’s geographic location and health outcomes has been documented in research efforts such as the Commonwealth Fund’s annual Scorecard on Local Health System Performance, which assesses regional healthcare access, costs, quality, and outcomes.

“We find a strong correlation between the [SES] in a healthcare market and health outcomes,” David Radley, a senior scientist with the Commonwealth Fund and the Westat statistical services company, said in an interview. “They’re intrinsically tied. We’ve known that in health care for a long time.”

Radley praised the study but had mild criticism of its methodology and its downplaying the effect of an SES risk adjustment in preventing penalties for about 65 hospitals.

“It may not seem like a huge deal, but it’s a huge deal to those 65 hospitals,” Radley said. “For people trying to keep the lights on, it makes a difference.”

Better Tools Needed

Using a ZIP code’s median income to calculate SES “diluted” the study’s findings, Feldpush said. And Radley noted that ZIP code data may work best for densely populated and geographically compact postal zones, but there is wide income variation in the large ZIP code where he resides in a suburb of Syracuse, N.Y.

Yale’s Susannah Bernheim MD, an author of the study, disagreed. She said the study’s researchers continued to study the issue after the Health Affairs paper was completed and the results didn’t change when smaller geographical units were studied.

“Some people think that results would change if we just had the perfect way to assess [SES], but that doesn’t seem to be true,” Bernheim wrote in an e-mail.

Bernheim and her colleagues acknowledged in their report that SES “is not a singular patient-level characteristic that can be easily measured.”

Radley agreed.

“It certainly highlights that we don’t have good measures of [SES] at the person level,” Radley said. “It highlights how blunt the tools are that we as researchers or the people who set payment rates have to use.”

Radley said that the questions of “should we adjust for SES?” and, if so “how do we do it?” are just beginning to be asked. The debate will continue and future research will go beyond the three conditions examined in this particular study, Radley predicted.

“It will be interesting to see how generalizable these findings will be as the body of evidence grows,” Radley said. "This is a well-done study and I’m glad it’s coming out. This it will certainly inform this debate.”

Key Takeaways

Elizabeth Wiley, MD, former president of the American Medical Student Association and now a preventive medicine resident with the Johns Hopkins Bloomberg School of Public Health, was concerned the study’s findings could be misconstrued to mean that SES and social determinants of health don’t play an important role in health outcomes.

“The results speak to the complexity of the social determinants of health and the difficulty of how to measure and adjust for their influence,” Wiley said in an interview.

As a family medicine resident training in Baltimore, Wiley said she saw how low-income patients faced barriers to improving their health such as lacking access to healthy food, lacking transportation to healthcare services, and being exposed to violence, polluted air and heavy tobacco advertising.

The study’s findings show the need for more granular and complex research, Wiley said.

Bernheim said the need to find a more appropriate way to integrate SES into hospital performance evaluations “is the most important thing to take from this paper.”

“When hospitals and researchers noticed that safety-net hospitals are somewhat more likely to be penalized in the [HRRP] most people thought the measures should be changed, but adjusting the readmissions measures for [SES] is not the right solution,” Bernheim said. “First, as we show, it doesn’t change how hospitals do on the measure, and second it probably isn’t the right policy or scientific solution.”

Instead, Bernheim suggested changing the penalty threshold for safety-net hospitals.

Future research should focus on what safety-net hospitals can do to ensure positive outcomes while avoiding readmissions.

“A really important finding from our paper is that many safety-net hospitals have great results on the quality measures and have low readmission rates,” Bernheim wrote. “We need to understand the features of those great safety-net hospitals—what it is about hospital leadership and the care that nurses and doctors and social workers provide and the community supports that help make a great safety-net hospital.”


Andis Robeznieks is a freelance writer based in Chicago. Follow Andis on Twitter at @AndisRobeznieks.

Publication Date: Tuesday, August 09, 2016