Abstract
BACKGROUND: Apparent increase in mortality associated with being admitted to hospital on a weekend compared to weekdays has led to controversial policy changes to weekend staffing in the United Kingdom. Studies in the United States have been inconclusive and diagnosis specific, and whether to implement such changes is subject to ongoing debate.
OBJECTIVE: To compare mortality, length of stay, and cost between patients admitted on weekdays and weekends.
DESIGN: Retrospective cohort study.
SETTING: National Inpatient Sample, an administrative claims database of a 20% stratified sample of discharges from all hospitals participating in the Healthcare Cost and Utilization Project.
PATIENTS: Adult patients who were emergently admitted from 2012 to 2014.
INTERVENTION: The primary predictor was whether the admission was on a weekday or weekend.
MEASUREMENTS: The primary outcome was in-hospital mortality and secondary outcomes were length of stay and cost.
RESULTS: We included 13,505,396 patients in our study. After adjusting for demographics and disease severity, we found a small difference in inpatient mortality rates on weekends versus weekdays (odds ratio [OR] 1.029; 95% confidence interval [CI], 1.020-1.039; P < .001). There was a statistically significant but clinically small decrease in length of stay (2.24%; 95% CI, 2.16-2.33; P < .001) and cost (1.14%; 95% CI, 1.05-1.24; P < .001) of weekend admissions. A subgroup analysis of the most common weekend diagnoses showed substantial heterogeneity between diagnoses.
CONCLUSIONS: Differences in mortality of weekend admissions may be attributed to underlying differences in patient characteristics and severity of illness and is subject to large between-diagnoses heterogeneity. Increasing weekend services may not result in desired reduction in inpatient mortality rate.