Physician Tenure and Clinical Productivity for Internal and External Hires: A Retrospective Cohort Study

New employees generally demonstrate lower productivity than experienced colleagues in non-healthcare sectors, but there is limited evidence on how tenure affects physician productivity.

To evaluate the association between tenure and clinical productivity for attending physicians in the Veterans Health Administration (VHA) and explore whether this relationship differs by prior VHA residency or fellowship training.

Retrospective cohort evaluation.

A total of 34,878 VHA attending physicians from 27 specialties who performed outpatient encounters between October 1, 2017, and August 1, 2023, accounting for 1,518,463 physician-months.

The primary exposure variable was VHA tenure. Physicians were categorized as new hires during their initial 2 years of employment and existing employees beyond 2 years. The secondary exposure variable was whether new hires had any VHA residency or fellowship training before becoming attending physicians (internal vs. external hires).

The outcome was productivity, defined as the average number of encounters per clinic day at physician-month level.

Among 34,878 VHA attending physicians (1,518,463 physician-months), 17,274 (279,263 physician-months) were new hires. New hires in their 1st quarter of employment had 1.72 fewer encounters (95% CI - 1.79 to - 1.65) per clinic day than existing employees. The negative association between tenure and productivity decreased over the first 2 years of employment. By their 8th quarter of tenure, new hires had 0.44 fewer encounters (95% CI - 0.52 to - 0.35) per clinic day. Among new hires, internal hires had higher productivity than external hires, particularly in primary care and large surgical specialties.

New attending physicians had lower productivity than existing employees, though their productivity improved with increasing tenure. Among new hires, internal hires had higher productivity than external hires and were faster to reach equivalent levels. These findings are important for health systems calculating the extended cost and access implications of physician turnover and replacement.