Electronic Health Record Use Patterns Among Well-Being Survey Responders and Nonresponders: Longitudinal...

Area of Science:

• Medical Informatics
• Physician Well-being
• Health Services Research

Background:

• Physician surveys are crucial for understanding physician experience, but nonresponse bias can limit findings.
• Electronic health record (EHR) data offers a potential method to assess experiences of nonresponders.
• Understanding nonresponder characteristics is key to improving survey representativeness.

Purpose of the Study:

• To identify EHR use measures that correlate with physician survey responses.
• To develop methods for estimating population-level survey results among physicians.
• To assess the utility of EHR data in understanding physician burnout and survey nonresponse.

Main Methods:

• Longitudinal observational study of primary care physicians (2019-2020).
• Quantified EHR use and burnout symptoms (emotional exhaustion, interpersonal disengagement).
• Used penalized linear regressions and propensity weighting to model burnout and response prediction.

Main Results:

• Responders and nonresponders had similar demographics but differing clinical workloads and EHR efficiency.
• EHR use measures effectively discriminated between survey responders and nonresponders (AUC=0.88).
• Predicted burnout prevalence in nonresponders (52%) was higher than in responders (28%), estimating population prevalence at 31%.

Conclusions:

• EHR use data can distinguish survey responders from nonresponders.
• While limited in predicting burnout directly, EHR data aids in understanding nonresponder impact.
• EHR metrics may inform strategies to increase survey response rates and improve physician well-being assessments.