Provider Experience with EHRs 

Evaluation of the Impact of Evaluation and Management Coding Changes on Physician Time Spent on Documentation and Burnout
Sponsor: The American Medical Association

To reduce EHR documentation burden, CMS changed documentation requirements for E&M encounters, with large-scale changes going into effect in 2021. The goal of this project is to conduct a comprehensive assessment of the impact of these changes on EHR time, EHR use patterns, and physician burnout. We will partner with Cerner to assess changes in a national sample and complement this with a deep dive into the experiences of UCSF physicians, including assessment of changes in validated measures of burnout.

Association between Objective Measures of EHR Use and Primary Care Clinician Burnout (STEP-PEP)

The Association between Objective Measures of EHR Use and Primary Care Clinician Burnout (STEP-PEP) study aims to explore the relationship between provider burnout and objective measures of EHR use. This study combines self-reported measures from Maslach Burnout Inventory survey of primary care clinicians across UCSF with objective EHR use measures - related to time, volume of work, and proficiency from EHR log files. As organizations increasingly rely on objective EHR measures to design and evaluate interventions to reduce burnout, our findings will point to the measures that should be targeted.    

Differences in Ambulatory EHR Use Patterns for Male vs. Female Physicians

Widespread adoption of electronic health records (EHRs) has transformed patient care. While  prior studies have linked EHR use to physician burnout, understanding different approaches to EHR use facilitates a more nuanced understanding of this relationship. Prior work has revealed that physician gender is tied to burnout as well as to patient expectations and outcomes. Therefore, in this study, we sought to understand differences in EHR use patterns — across an array of measures related to time spent on documentation, approaches to EHR documentation, and use of EHR tools — by physician gender for all ambulatory physicians (n = 1,336) at a large academic medical center. 


Algorithmic determination of level of service codes for clinic visits in a primary care setting using audit log data
Sponsor: MITRE Corporation

This research aims to apply machine learning techniques to Electronic Health Record (“EHR”) metadata found in audit logs, to automatically and accurately determine the appropriate level of service, which could support an alternative to documentation-based coding. The specific goal of this project is to develop a machine learning model that accurately predicts ambulatory level of service codes (Evaluation & Management CPT codes) from data describing physician actions in the EHR.

Characterizing Provider Styles of Clinical Note Production Using EHR Audit Logs

Clinical documentation is an integral part of patient care, but the quality of clinical documentation is variable, and documentation is a significant driver of physician burnout. In order to improve clinical documentation efficiency and quality, it is important to understand how providers currently document patient information. We characterize individual styles of note production using EHR audit log data and find that individuals have distinct styles that are variable in terms of when edits occur and how long the edits take.

Insights into the Value of Clinical Notes from Measuring Writing and Viewing Patterns Using EHR Audit Log Data

In order to develop effective tools and strategies to address documentation burden and improve documentation quality and efficiency, it is essential to understand when, why, and to whom clinical notes are useful. To address these questions, we utilize EHR audit log data to measure clinical note production and consumption behaviors for a broad set of users.