Prospective Evaluation of a Scribe Program's Impacts on Provider Experience, Patient Flow, Productivity, And Teaching in an Academic Emergency Medicine Practice

Abstract

Physician charting burdens are linked to reduced job satisfaction and, in academic settings, can impinge on teaching. Some practices are using scribes to address these issues but the impacts have not been prospectively evaluated. Our goal was to prospectively assess scribe program impacts on provider job satisfaction, teaching, and productivity in an academic emergency medicine (EM) practice. We prospectively evaluated the impacts of a new scribe program in a large, urban, academic EM practice in the Southeast. We administered surveys three months before and six months after the intervention to physicians in the group and compared results pre- to postsurvey. We also collected and compared seasonally matched indicators of physician productivity and ED flow. Productivity indicators included relative value units (RVUs) per patient, RVUs per hour, and patients per hour. Flow metrics included total ED volume, ED length of stay (LOS), and percentage of patients that left without being seen (WBS) and against medical advice (AMA). We analyzed the surveys using descriptive statistics, chi-square tests, and Wilcoxon signed rank tests. We analyzed productivity and flow data using paired-sample t tests. Response rates were 76.1% and 69.0%; 91.8% of respondents reported never or almost never dictating charts postimplementation. Most liked working with scribes, though 9.5% had negative impressions of the scribe intervention. Job satisfaction scores increased but changes were not statistically significant. There was an absolute reduction of 16% (11%-22%) in time spent charting and a relative reduction of 36% (25%-50%; P<.01). Respondents reported reductions in time spent charting in and outside the ED after shifts but these were not statistically significant. Time spent with patients showed an absolute increase of 11% (4%-17%) and relative increase of 30% (11%-46%; P<.01). Respondents perceived that the intervention substantially increased time for teaching but relevant pre-to-post trends were not statistically significant. Volume increased by 88 patients per day (32-146; P=.04) during the study. Absolute percentage of patients leaving AMA dropped by 0.29 (−0.52 to −0.06; P=.02), a 20% relative decrease, and absolute percentage of patients leaving WBS increased by 1.47 (0.83-2.11; P<.01), a 50% relative increase. LOS was unchanged. Average RVUs/hour increased by 5.5% and average RVUs/patient by approximately 5.3% (multiple confidence intervals and P values for each week, all significant). Patients/hour showed a nonsignificant increase of 5.2% (same, but none significant). In this prospective cohort study scribes were well received in an academic EM practice and substantially reduced provider charting burdens while increasing revenues. Job satisfaction increased marginally but changes were not statistically significant. Providers reported that their time was redirected to patient care and teaching. Pre-to-post assessments supported the former but not the latter. During the study period ED volume increased, ED LOS held constant, and there was a drop in the percentage of patients leaving AMA but an increase in patients leaving WBS. Operationally and statistically significant increases in provider productivity were primarily linked to increases in RVUs per hour and per patient.