Nursing homes house vulnerable people who may have physical and psychological limitations and also often lack the financial, political, and social means required for advocacy. To ensure people in nursing homes are safe and receiving high quality of care, public monitoring of these facilities is crucial. A core part of current efforts to monitor and improve patient safety in nursing homes rely on the Minimum Data Set 3.0 (MDS), which contains patient-level health assessments that are self-reported by nursing homes. The Centers for Medicare and Medicaid Services quality measures and five-star ratings based on these assessments suggest progress has been made over the past decade. However, in light of inconsistencies between the MDS and other data sources, and investigations and audits that suggest MDS reporting may be unreliable, it is reasonable to question the integrity of the data and thereby the current federal approach to monitoring nursing home patient safety. We propose to use Medicare and Medicaid claims of nursing home patients to identify health events that led to hospital visits. Our focus will be on four patient safety sections of the MDS, namely falls, pressure ulcers, urinary tract infections (UTIs), and pneumonia, all of which are generally preventable but without proper care can lead to serious physical and psychological morbidity and mortality. This will allow assessment of the quality of reporting on the MDS, our first aim. The guiding principle here is that while only a portion of the cases that occur in nursing homes would be serious enough to warrant a hospital visit, all cases that result in a hospital visit should be appropriately recorded by the MDS. We will also assess variations in MDS reporting accuracy by patient and nursing home characteristics, with a particular focus on identifying disparities by race and Medicaid enrollment. In our second aim, we will estimate changes in MDS reporting behavior after major injury falls were included in the five-star ratings calculations. In the third aim, we will use the claims to develop alternative measures of patient safety in nursing homes, that is the incidence rates of the more severe cases of these clinical conditions. We will assess the associations between these claim-based rates and other measures of patient safety. In addition to studying these rates and their trends, we will assess variations in incidence rates by nursing home characteristics, with a particular focus on identifying disparities by race mix and proportion Medicaid. Finally, we will describe how current rankings of nursing homes may shift under the claims-based measures. This work will advance safety and quality research for the nursing home care system. It will be a crucial step towards evaluating the effectiveness of current programs and improving future efforts by the public. We will pay specific attention to race and Medicaid payer status in MDS reporting rates and our claims-based patient safety measures to identify disparities, as minorities and the poor may be especially at risk of receiving low-quality care. We expect to provide specific, evidence-based insights that policymakers can use to substantially improve patient safety tools in these settings.
To ensure people in nursing homes are safe and receiving high quality of care, public monitoring of these facilities is crucial. A core part of current efforts to monitor and improve patient safety in nursing homes rely on patient-level health assessments that are self-reported by nursing homes, however recent investigations and audits suggest these data may be unreliable. We propose to use Medicare and Medicaid claims of nursing home patients in a validation exercise to assess the quality of self-reporting by nursing homes, estimate changes in reporting behavior following the inclusion of major injury falls in the five-star-rating system, and develop new claims-based patient safety measures for nursing home residents.
