Construction and Validation of Risk-adjusted Rates of Emergency Department Visits for Long-stay Nursing Home Residents

BACKGROUND: Long-stay nursing home (NH) residents are at high risk of having emergency department (ED) visits, but current knowledge regarding risk-adjusted ED rates is limited. OBJECTIVES: To construct and validate 3 quarterly risk-adjusted rates of long-stay residents' ED use: any ED visit, ED visits without hospitalization or observation stay (outpatient ED), and potentially avoidable ED visits (PAED). RESEARCH DESIGN: The authors calculated quarterly NH risk-adjusted ED rates from 2011 Q2 to 2013 Q3 national Medicare claims and Minimum Data Set data. Using random-effect linear regressions, the authors validated these rates against Nursing Home Compare overall 5-star quality ratings and examined their associations with hospitalization rates to provide a quality context. SUBJECTS: Resident-quarter observations (7.3 million) from 15,235 unique NHs. RESULTS: Risk-adjusted rates of any ED, outpatient ED, and PAED averaged 9.7%, 3.4%, and 3.2%, respectively. Compared with NHs with 1 or 2 stars overall rating, NHs with ≥3 stars were significantly associated with lower rates of any ED visit, outpatient ED, and PAED (β, -0.23%, -0.16%, and -0.11%, respectively; all P<0.01). Pearson Correlation coefficients between hospitalization rates and rates of any ED visit, outpatient ED, and PAED were 0.74, 0.31, and 0.46, respectively. CONCLUSIONS: The moderately negative associations of 5-star ratings with ED rates provide supportive evidence to their validity. Outpatient ED and PAED were moderately correlated to hospitalizations suggesting they provided more information about quality than any ED.

https://pubmed.ncbi.nlm.nih.gov/31764481