Nursing homes underreport antipsychotic use but overreport diagnoses qualifying for appropriate use

Abstract

Abstract Antipsychotic drug use in US nursing homes remains a priority concern, but less is understood about the characteristics associated with reporting. Using linked Medicare claims and Minimum Data Set (MDS) assessments for long-stay nursing home residents from January 2018 to December 2019, we assessed the consistency of antipsychotic drug reporting and diagnosis of conditions (schizophrenia, Tourette's syndrome, and Huntington's disease) that qualify as appropriate drug use across data sources by calculating reporting rates in facility-reported MDS and Medicare claims. The antipsychotic reporting outcome is conditional on claims reporting while the condition reporting outcomes are conditional on MDS reporting. We found underreporting (87% reporting rate) in facility-reported antipsychotic use versus Medicare claims. In contrast, we found overreporting of the qualifying conditions with several facility-reported diagnoses unsupported by a corresponding claims diagnosis. Only 54.8% of schizophrenia, 46.5% of Tourette's syndrome, and 72.4% of Huntington's disease diagnoses reported in the MDS had a claims diagnosis. There was also variation in reporting odds for antipsychotic drug use by dual-eligibility status and race, with higher odds for dual-eligible and lower odds for Black residents. These findings suggest that CMS should continue investigating the source of reporting discrepancies in antipsychotic drug use and qualifying diagnoses.