Tsung-Chien Lu1,2,3, Dean-An Ling1, Chu-Lin Tsai1,2, Fuh-Yuan Shih1,2, Cheng-Chung Fang1,2. 1. Department of Emergency Medicine, National Taiwan University Hospital. 2. Department of Emergency Medicine, College of Medicine, National Taiwan University, Taipei, Taiwan. 3. Division of Biomedical and Health Informatics, Department of Biomedical Informatics and Medical Education, School of Medicine, University of Washington, Seattle, Washington, USA.
Abstract
OBJECTIVES: Emergency department (ED) revisits had traditionally been used as a quality indicator, but focused more on the same hospital revisit (SHRV). Our study investigated the 72-hour ED revisits on SHRV and different hospital revisits (DHRV), and explored the predictors of DHRV. METHODS: 72-hour ED revisits were analyzed using Taiwan's National Health Insurance Research Database that contained one-third patient records from 2012 to 2013. Rates of SHRV and DHRV were calculated and compared among levels of hospital accreditation. Linear regression analyses were used to measure the correlation between revisit rates and average monthly volumes of the index ED. Multilevel logistic analyses were performed to evaluate the predictors of DHRV. RESULTS: There were 4 065 215 index ED visits. Of them, 234 826 (5.8%) were associated with 72-hour revisits and 36.7% of them occurred at different institutions. The revisit rates showed differences across distinct hospital levels. DHRV rates had significant inverse correlation with the average monthly volume of the index ED. Independent predictors associated with the increased odds of DHRV were: ED visits on weekend or holiday, or the index ED being at a local hospital. CONCLUSIONS: About one in three ED revisits occurred in another hospital. SHRV rate alone would inevitably underestimate the true revisit rate. DHRV is associated with the hospital level and annual ED volume, and there is increased likelihood of DHRV if patient's index ED visit being a local hospital.
OBJECTIVES: Emergency department (ED) revisits had traditionally been used as a quality indicator, but focused more on the same hospital revisit (SHRV). Our study investigated the 72-hour ED revisits on SHRV and different hospital revisits (DHRV), and explored the predictors of DHRV. METHODS: 72-hour ED revisits were analyzed using Taiwan's National Health Insurance Research Database that contained one-third patient records from 2012 to 2013. Rates of SHRV and DHRV were calculated and compared among levels of hospital accreditation. Linear regression analyses were used to measure the correlation between revisit rates and average monthly volumes of the index ED. Multilevel logistic analyses were performed to evaluate the predictors of DHRV. RESULTS: There were 4 065 215 index ED visits. Of them, 234 826 (5.8%) were associated with 72-hour revisits and 36.7% of them occurred at different institutions. The revisit rates showed differences across distinct hospital levels. DHRV rates had significant inverse correlation with the average monthly volume of the index ED. Independent predictors associated with the increased odds of DHRV were: ED visits on weekend or holiday, or the index ED being at a local hospital. CONCLUSIONS: About one in three ED revisits occurred in another hospital. SHRV rate alone would inevitably underestimate the true revisit rate. DHRV is associated with the hospital level and annual ED volume, and there is increased likelihood of DHRV if patient's index ED visit being a local hospital.