Readmission, death rates not necessarily reflected by SNF quality measure performance

Is a skilled nursing facility’s (SNF’s) performance on quality measures a reliable indicator of the likelihood that those it treats will or will not be rehospitalized or die within 30 days of being discharged from the hospital? The short answer is no, according to a study in the Oct. 15 issue of JAMA.

“Ultimately, although SNF performance measurement plays an important role in promoting transparency and accountability in the U.S. healthcare system, our findings suggest that in their current form they are unlikely to serve as a sole basis for large-scale reductions in readmissions unless accompanied by other strategies,” Mark D. Neuman, MD, MSc, of the University of Pennsylvania, Philadelphia, and colleagues write.

Studying data related to fee-for-service (FFS) Medicare beneficiaries, researchers found that better performance on various measures of quality of care was not consistently associated with a lower risk of hospital readmission or death at 30 days. Specifically, the authors studied national Medicare data on FFS beneficiaries discharged to a SNF after an acute care hospitalization and looked at these metrics: SNF staffing intensity, health deficiencies identified through site inspections, and the percentages of SNF patients with delirium, moderate to severe pain, and new or worsening pressure ulcers.

Of almost 1.54 million discharges to SNFs, 21 percent were followed by readmission within 30 days, and 4.7 percent were followed by a death within 30 days. The overall rate of 30-day readmission or death was 23.3 percent.

Although SNFs that performed better on measures such as staffing ratings and facility inspection ratings had improved outcomes in unadjusted analyses, the associations were diminished substantially after researchers adjusted for patient/resident factors, the discharging hospital and SNF facility characteristics. SNFs with better facility inspection ratings had a slightly lower adjusted risk of readmission or death. Staffing ratings and performance on clinical measures related to pain or delirium did not meaningfully affect adjusted outcomes for SNFs, however, and other measures did not predict clinically meaningful differences in the adjusted risk of readmission or death.

In an accompanying editorial, Elizabeth A. McGlynn, PhD, and John L. Adams, PhD, of the Kasier Permanente Center for Effectiveness and Safety Research, Pasadena, Calif., comment on this study and another one from the same issue of JAMA; both examine quality indicators.

“Neither study looked at other processes as potential predictors of the outcomes of interest (readmission, death, childbirth outcomes) and therefore missed an opportunity to identify areas for future measure development,” they write. “Further, careful consideration of the ultimate goal of a quality measure must be made. The information required for consumers to choose among nursing homes or hospitals may be different than the information required to improve clinical outcomes. Measures that work for one purpose and not another are still valuable.”

The editorial writers recommend that future studies of quality measures establish a “clear framework and expectations for the intended goals of quality measures.”