Study: MIPS hospital readmission metric likely ineffective

May 20, 2019



As part of its ongoing efforts to reduce hospital readmissions among Medicare beneficiaries, CMS made readmissions one of the quality metrics it uses to penalize or provide bonuses to primary care physicians (PCPs) under the Merit-based Incentive Payment System (MIPS).

But a new study suggests that the measure will be ineffective, because PCPs vary little in the readmission rates of their Medicare patients.

The study examines 30-day post-discharge readmission rates among about 566,000 Medicare beneficiaries cared for by approximately 4,2000 PCPs in Texas in the years 2012-2015. The results show virtually no significant variations by PCP, with the adjusted readmission rate at the 99th percentile differing from the mean rate by only 1.1%.

The results have implications for MIPS, because one of the program’s mandatory quality measures is 30-day hospital readmission rates for physician groups of 16 more who collectively have more than 200 admissions annually. The assumption underlying that metric, the authors note, is that “readmission rates vary by PCP independent of patient characteristics,” and thus financial incentives will lead to doctors doing more to prevent readmissions.

The study also looked at how often the PCPs saw patients within seven days of the patient’s hospital discharge. It found that only about 20 percent did so, despite the financial incentives for prompt post-discharge follow-up under the transitional care management codes introduced in 2013.

Patients were more likely to be seen in the seven-day post-discharge period, the study found, if they were older, Hispanic, or male; had an emergency admission, or lived in an area with a high proportion of high school graduates. Those less likely to be seen were black, belonged to a more highly-weighted diagnosis-related group, or been frequently hospitalized during the prior year.

The study, “Variation Among Primary Care Physicians in 30-Day Readmissions,” appears in the May 21, issue of Annals of Internal Medicine.

x