Underreporting of falls and pressure ulcers in nursing homes was frequent and “widespread” in a new study of such incidents and whether they were accurately reflected on Care Compare.
Over six years, 40% of hospitalizations related to major fall injuries that were found in hospital claims were not reported in facility-level Minimum Data Sets, researchers at the University of Chicago found. For pressure ulcer hospitalizations, 33.3% went unreported from 2011 through 2017.
Overall, the researchers examined records from 13,179 nursing homes and found 131,000 long-stay residents had to be hospitalized due to either condition. Underreporting of those hospitalizations wasn’t limited to a few bad actors. Rather, the study showed about 70% of nursing homes reported less than 80% of their patients’ fall and pressure ulcer hospitalizations. Lower reporting rates also were associated with certain racial and ethnic patient populations.
“Underreporting of clinical events to CMS was widespread during the study period,” wrote lead author Prachi Sanghavi, PhD, an assistant professor at University of Chicago. “Although CMS appears to recognize this situation, as evidenced by the lower weighting of quality measures in the ratings, the measures remain on public display. In other words, the quality measures are in limbo, but the public may be using these measures to make important life decisions.”
Full findings were published in JAMA Network Open Tuesday.
Sanghavi and co-author Zihan Chen, also of the University of Chicago, noted that CMS has tried in recent years to improve accuracy by using more objective data sources such as claims data.
“Even with alternative data sources, the study found that interpretation of quality measures needs to consider the racial and ethnic composition of nursing homes, which may affect both the prevalence of major injury falls and pressure ulcers and the accuracy of reporting,” they added.
Nursing homes with a higher, or more accurate, fall reporting rate were more likely to be in rural areas, hospital-based and had more RN hours per 100 resident days. They were also more likely to have a high percentage of White residents and a smaller share of dual-eligible residents, the study found.
Body of proof grows
While other studies have shown providers underreport data that feeds some quality metrics, this study is larger than its predecessors and also was the first to identify facility characteristics related to underreporting. It also aimed to compare reporting differences among long-stay and short-stay patients and whether failure to fully report one metric might influence reporting of the other metric.
Reporting rates of pressure ulcers among long-stay residents increased with higher reporting of pressure ulcers among short-stay residents, whereas reporting rates of pressure ulcers among long-stay residents decreased with higher reporting rates of major injury falls among long-stay residents.
Mean fall rates among White residents were between 2.2 and 2.5 times higher than the fall rates among Black residents, whereas pressure ulcer rates among Black residents were between 1.8 and 2.5 times higher than the pressure ulcer rates among White residents, the researchers reported. As the facility-level percentage of White residents decreased, overall fall rates and fall rates among White and Black residents decreased, whereas the pressure ulcer rates increased across all groups.
Hospitalizations in which the patient did not return to the same nursing home were excluded from the study, Sanghavi told McKnight’s Long-Term Care News. That was to try to limit the research to cases for which nursing home staff would be award of the resident’s hospitalization and condition.
In the case of pressure ulcers, low reporting rates were associated with higher overall, quality, and survey ratings as well as a lower publicly reported pressure ulcer rate. Low reporting rates were associated with a higher quality rating for both types of events.
At the state level, as many nursing homes reported 33% of their major injury falls as reported 94.3%, and a similar spread was found for pressure ulcers. The researchers said that showed that “accurate reporting is achievable, although the circumstances that allow it remain unclear, and underreporting is common.”
“Enforcement through investigations and inspections is challenging and costly to perform on a large scale,” Sanghavi added in an email. “I think CMS should lean heavily on the claims data that it already owns and make better use of it for validation of nursing home-reported data and estimation of nursing home quality. The current measures that are based on nursing home-reported data should be replaced or at least supplemented with measures from more objective data sources, such as CMS claims from other healthcare providers.”
