Abstract
Introduction: Physicians dedicate substantial time to documentation. Scribes are sometimes used to improve efficiency by performing documentation tasks, although their impacts have not been prospectively evaluated. Our objective was to assess a scribe program's impact on emergency department (ED) throughput, physician time utilization, and job satisfaction in a large academic emergency medicine practice. Methods: We evaluated the intervention using pre-and post-intervention surveys and administrative data. All site physicians were included. Pre-and post-intervention data were collected in fourmonth periods one year apart. Primary outcomes included changes in monthly average ED length of stay (LOS), provider-specific average relative value units (RVUs) per hour (raw and normalized to volume), self-reported estimates of time spent teaching, self-reported estimates of time spent documenting, and job satisfaction. We analyzed data using descriptive statistics and appropriate tests for paired pre-post differences in continuous, categorical, and ranked variables. Results: Pre-and post-survey response rates were 76.1% and 69.0%, respectively. Most responded positively to the intervention, although 9.5% reported negative impressions. There was a 36% reduction (25%-50%; p<0.01) in time spent documenting and a 30% increase (11%-46%, p<0.01) in time spent in direct patient contact. No statistically significant changes were seen in job satisfaction or perception of time spent teaching. ED volume increased by 88 patients per day (32-146, p=0.04) pre-to post-and LOS was unchanged; rates of patients leaving against medical advice dropped, and rates of patients leaving without being seen increased. RVUs per hour increased 5.5% and per patient 5.3%; both were statistically significant. No statistically significant changes were seen in patients seen per hour. There was moderate correlation between changes in ED volume and changes in productivity metrics. Conclusion: Scribes were well received in our practice. Documentation time was substantially reduced and redirected primarily to patient care. Despite an ED volume increase, LOS was maintained, with fewer patients leaving against medical advice but more leaving without being seen. RVUs per hour and per patient both increased. © 2015 by the article author(s).