JMIR Publications

ABSTRACT

Background:

Operative opportunity is the central but finite resource of surgical education, and departments must serve junior residents, senior residents, fellows, reduced-FTE trainees and clinician-scientists from the same theater time. Whether scheduling can expand education rather than ration it has rarely been tested in practice.

Objective:

To determine whether clinician-led hierarchical AI-informed operating room scheduling could scale surgical education while preserving resident access, equitable proportionality and departmental capacity.

Methods:

We studied 150 matched calendar weeks in a single academic plastic surgery department, 75 before and 75 after a hierarchical educational scheduling strategy took effect in January 2025. The strategy integrated trainee role, training year, contractual full-time equivalent (FTE), recent exposure and weekly capacity, and deliberately paired fellows with residents in complex procedures; final rosters were constructed and approved by the senior surgical leadership. Residents and fellows formed the educational population with time-varying postgraduate year and eligible FTE person-time as the denominator. Analyses used generalized estimating equations clustered by trainee with sensitivity analyses across alternative eligibility thresholds.

Results:

Total trainee operating days rose from 807 to 962 (+19.2%) while total departmental volume remained essentially stable (1503 and 1543 surgeon-days). The educational conversion rate, the share of departmental operative activity carrying a trainee, rose from 54% to 62%. Fellow operating days increased from 283 to 431 (52.3%) and absolute resident operating days were preserved at 524 and 531, giving a fellow-to-resident exposure ratio of 1.36 (95% CI 1.09-1.69) before and 1.71 (1.42-2.07) after implementation. Temporal regularity (residual variance ratio 0.93, 0.83-1.05) and weekly capacity (ratio 1.03, 0.97-1.10) were maintained. Operative exposure among trainees recorded as women remained closely proportional to their eligible FTE share, at 0.97 and 0.95, despite women holding the reduced-FTE trainee position in both periods and two of three reduced-FTE contracts departmentwide. As the cohort grew from 14 to 17, eligible person-time expanded faster than opportunity, so exposure per eligible resident week fell (adjusted rate ratio 0.81, 0.69-0.96; 0.90, 0.77-1.06 after standardization for training-year composition) and progression across resident training years attenuated from 1.17 (1.11-1.22) to 1.10 (0.99-1.23) per year, partly reflecting transition of senior trainees into fellowship.

Conclusions:

Clinician-led hierarchical AI-informed scheduling scaled operative education without sacrificing its principal equity and capacity safeguards. Advanced fellow exposure increased by more than 50% while absolute resident operating volume was maintained, and operative access among women remained proportional to eligible FTE despite their greater representation in reduced-FTE employment. Deliberate fellow-resident pairing in complex surgery provides a scalable multilevel training model. Future iterations should ensure that operative opportunity grows proportionally with an expanding resident workforce and should protect progressive exposure across training years.