To characterize the number of treat-and-release ED visits in the US from 2006 to 2018 by their physician services evaluation and management codes (level of service), and to characterize how risk-adjusted likelihood of high-intensity billing in 2018 compares to 2006. This is an observational analysis of the Nationwide Emergency Department Sample from the AHRQ Healthcare Cost and Utilization Project in 2006 and 2018. We included all treat-and-release ED visits (not resulting in admission to the same hospital) for adult patients, excluding visits with disposition against medical advice, left without being seen, or death in ED. Current Procedural Technology (CPT) codes for these visits were available for evaluation and management codes level of service. Consistent with prior work, we defined ED visits with level of service 5 (CPT 99285) or critical care (CPT 99291/99292) as high-intensity billing. We defined diagnostic categories using Clinical Classification Software, developed by AHRQ to provide meaningful groupings of diagnoses. The top ten modal diagnosis categories, as well as primary payer (Medicare, Medicaid, private insurance, self-pay), sex, and age were utilized in subsequent modeling. The outcome was high-intensity visit rate, utilizing survey weights for nationally-representative estimates, in 2006 and 2018. We report both the unadjusted ratio of high-intensity billing rate for 2018 as compared to 2006 and the risk-adjusted observed:expected ratio for high-intensity billing in 2018. We constructed a logistic regression model using 2006 data with the following covariates: disposition (discharged, transfer, home healthcare), age, sex, primary payer, and diagnosis category amongst the ten modal categories. In order to generate risk-adjusted observed:expected ratio in 2018, we divided the observed rate with that predicted for high- intensity billing using the above logistic regression model and inputted 2018 data. In 2006 there were an estimated 79,515,535 treat-and-release ED visits nationwide meeting sample criteria, of which 4,773,046 ED visits (6.0%) were classified as high-intensity. By 2018, there were 96, 739, 780 such visits nationwide of which 17,172,884 (17.7%) were classified as high-intensity. The unadjusted ratio of the high intensity billing rate in 2018 as compared to 2006 was 2.95. In logistical regression analysis using data from 2006, increasing age, transfer to a short-term hospital, Medicare, Medicaid, abdominal pain, and lower respiratory diseases were associated with greater likelihood of high-intensity visitation. The observed high-intensity billing rate in 2018, 17.7%, was slightly higher than that expected based on the model, 15.3%, yielding an observed:expected ratio of 1.15. High-intensity billing for treat-and-release ED visits had increased in the period from 2006 to 2018. Several patient-level factors predict high-intensity billing, and accounting for these factors given observed trends in ED case-mix such as increased visitation by older adults, those insured by public programs, evolution of the ED as an acute diagnostic certain, results in fairly modest differences in risk-adjusted billing rates. Policymakers must consider the increasingly complex needs of patients receiving outpatient services when designing payment policies.
Read full abstract