Electronic Health Record Clinical Workload Metrics among Pediatric Gastroenterologists

ABSTRACT

BACKGROUND: Electronic health record systems (EHRs) represent one of the most widely adopted digital healthcare technologies in the past decade. Among the potential benefits of EHRs has been the quantification of individual physician time spent performing key components of clinical workload. Epic EHR is a global system with the majority market share in North American acute care and ambulatory arenas and may offer a means to quantify the clinical workload of pediatric gastroenterology, as a subspecialty field of medicine. OBJECTIVE(S) To quantify clinical workload of pediatric gastroenterology across Epic EHR systems. METHOD(S) From January 2020 through April 2022, we evaluated Signal EHR data captured in Epic for all pediatric gastroenterologists (PGI), defined as physicians (MDs) with an Epic specified PGI profile. Signal data provides detailed data on clinician time spent daily (defined by days where a MD was clinically active or logged into the EHR) interfacing with the EHR, including clinical work process data in 4 key areas In-Basket (including communications with patients and other healthcare providers), Orders, Notes and Letters, and Clinical Review. For our study purposes, clinical workload was characterized by 4 monthly metrics days with appointments;appointments per scheduled day (data from April-July 2020 during COVID-19 lockdown were not included to accurately reflect current practice);pajama EHR time (530 PM to 7 AM);and EHR time outside templated clinic hours. Proportional time spent in different clinical arenas was reported for April 2022 only. Monthly process metrics captured in each of the 4 key areas focused on work volume and time spent. Outcome metrics were reported as average+/-standard deviation (SD) and median (interquartile range (IQR)). All metrics were evaluated for change over time using regression modeling. Statistical significance was set at p<0.05. RESULT(S) Signal data from 993 PGI at 213 institutions were analyzed. 95.8% (n=204) institutions were located in the US. Clinical workload Over the reporting period, PGI had clinical appointments an average of 43+/-3% [median (IQR) = 46% (35%, 57%)] days per month or about 3 days per week. PGI had 7.6+/-0.3 [7.0 (5.8, 8.9)] clinical appointments per scheduled day. On average, PGI spent an additional 23.7+/-1.6 [14.4 (4.6, 30.2)] pajama time minutes and 36.1+/-1.9 [30.3 (15.8, 43.3)] minutes outside scheduled hours interacting with the EHR each day. Clinical workload metrics remained stable over the study period. On average, PGI spent 60% time in the ambulatory arena, 9.7% in inpatient, 0.3% in the emergency department and 30% in other. In-Basket The average time spent in In-Basket by PGI was 23.0+/-1.3 [20.4 (13.2, 26.5)] minutes per day. Average time in In-Basket increased significantly over the study period (p<0.0001). Primary drivers for this change included increases in certain types of In-Basket messages, including results (p=0.01), patient medical advice (p<0.0001), hospital chart completion requests (p<0.0001), prescription authorization requests (p=0.003), and staff messages (p<0.0001). Orders On average, PGI prescribed 1 medication every other appointment, or 0.5+/-0.02 [0.4 (0.3, 0.6)] medications per visit. PGI ordered 2.2+/-0.3 [2 (1.4, 2.8)] tests/evaluations per appointment. Notes and Letters The average note length was 6392+/-193 [6072 (4344, 7696)] characters, equivalent to over 3.5 pages of text. Time spent in notes was 10.2+/-0.4 [9.7 (6.7, 13.1)] minutes per appointment and 46.9+/-2.4 [43.6 (29.9, 56.2)] minutes per day. Length of notes increased significantly over the study period (r=0.51, p=0.01) but time spent in notes did not. Clinical Review PGI spent an average of 17.7+/-1.5 [17 (12.7, 20.3)] minutes per scheduled day in chart review, equivalent to 4+/-0.2 [3.9 (2.7, 5.3)] minutes per appointment. CONCLUSION(S) Quantification of some key components of clinical workload inherent to PGI is possible using EHRs. PGIs routinely spend time outside of work hours performing EHR work. Over the past 2 years, In-Basket time has contributed substantially to PGI workload and has trended towards increasing messages from both external (patients and pharmacies) and internal sources (staff and hospital compliance). Considerable PGI time has also been spent constructing clinical notes of lengths that appear to have increased during the same 2-year period. Limitations to the study include non-standardized, opaque metric definitions and unclear fidelity of provider categorization. We would also note that our results document increasing EHR-related workload burdens on PGIs that can contribute to physician burnout. Through identification of best outcome metrics, quantification of PGI clinical workload using EPIC Signal data may allow quality improvement activities that reduce provider burden while enabling our subspecialty field to benefit from widespread implementation of EHRs.