Penn Experts Urge Focus on Reducing Preventable Hospital Readmissions

Experts from the Perelman School of Medicine at the University of Pennsylvania and Vanderbilt University suggest that the reporting of hospital readmission rates should be based exclusively on preventable or potentially preventable readmissions, in a review published online in the Journal of Hospital Medicine. Currently the Centers for Medicare and Medicaid Services (CMS) does not take into account whether a readmission is preventable when assessing hospitals’ performance on this quality metric.

Under the Affordable Care Act, hospitals must report readmission rates for heart attack, heart failure, and pneumonia to CMS. CMS then imposes financial penalties on institutions having an excessive number of readmissions that take place within 30 days following patient discharge. Additional medical conditions will be added in 2015. Many patients with these conditions suffer from additional illnesses that are complex and come with many co-morbid conditions.

“Reducing hospital readmissions is clearly important on many levels,” says lead author Julia G. Lavenberg, PhD, RN, a research analyst at Penn’s Center for Evidence-based Practice. “Patients prefer to remain at home, payers save money, and hospitals avoid financial penalties for having high readmission rates. But while current policy assumes that a significant proportion of readmissions are preventable, research tells us that this is simply not so. Moreover, there is no consensus in the medical and policy communities on how to define preventable readmissions, which is essential for taking action to reduce them.”

Other payers, such as private insurers, are likely to follow the federal lead and withhold funding for high readmission rates. As a result, hospitals and health systems nationwide are devoting significant time, effort, and money to reducing readmissions. Steps include increasing patient education before discharge, introducing or expanding home health visits, and working more closely with nursing homes and rehabilitation centers.

“Current research tells us that only about 25 percent of hospital readmissions are preventable,” says senior author Sunil Kripalani, MD, MSc, chief of the Section of Hospital Medicine at Vanderbilt University. “We urge a focus on preventing these readmissions, so that hospitals can enhance efforts in areas where it will have the greatest effect as well as ensure fair and equitable reporting of hospital performance.”

Until a validated measure of preventability is developed, the authors recommend several steps. First, the readmission time horizon should be reduced from the current 30 days to seven or 15 days, as research suggests that early readmissions – those within seven to 15 days of discharge -- are more likely preventable than those occurring later. Second, policymakers should take the socioeconomic status of patients into account by only comparing hospitals serving similar patient communities when determining penalties for excess readmission rates. Finally, adjusting for other community factors such as practice patterns and access to care is necessary to more accurately reflect factors under a hospital’s control.

“We’re encouraging a major shift in perspective,” adds co-author Craig A Umscheid, MD, MSCE, a hospitalist and director of the Center for Evidence-based Practice at Penn. “We want policymakers to acknowledge that hospitals should not be penalized for factors beyond their control which may play an important role in readmissions, such as inadequate community health resources or severity of the patient’s illness.”

Such a shift would require agreement among healthcare researchers and policymakers on how to identify and measure preventable or potentially preventable readmissions. While there are existing methods for doing so, there is no consensus on which is best. Furthermore, some of these methods are proprietary and thus unavailable for evaluation by others. These methods use such techniques as identifying readmission for conditions closely related to the original diagnosis or complications arising from the original admission.

To rectify this gap, the authors urge healthcare researchers and policymakers to come to agreement on a transparent, universal method for defining preventable or potentially preventable readmissions. This could include algorithms -- based on insurance claims data -- that recognize patterned relationships between original and readmission diagnoses for a variety of medical conditions. Then, clinicians at individual hospitals could use these standards to evaluate whether individual readmissions were potentially preventable or not. In time, researchers could codify these individual decisions into best-practice standards to serve as guides for subsequent assessments.

In addition to Lavenberg and Umscheid, Penn co-authors include Brian Leas; Kendal Williams, MD; and David R. Goldmann, MD.

Umscheid was supported in part by the National Center for Research Resources and the National Center for Advancing Translational Sciences (UL1TR000003). Kripalani receives funding from the National Heart, Lung and Blood Institute (R01HL109388), and from CMS (1C1CMS331006-01 and 1C1CMS330979-01).

Source: Perelman School of Medicine at the University of Pennsylvania