Residency application publication metrics and early academic career placement in diagnostic radiology: a...

To evaluate whether residency application publication metrics are associated with early academic career placement (academic employment 3 to 4 years after residency graduation) among diagnostic radiology residents beyond characteristics visible at application, and whether more elaborate metrics improve on a simple publication count.

In this retrospective cohort study, we analyzed 551 diagnostic radiology residents from the 2020 and 2021 graduating classes, identified from public U.S. program rosters, with career outcome classified from public post-training profiles. Application-visible variables included publicly inferred sex, top-40 medical school status, and PhD attainment. Three publication metrics measured at the residency-applicant timepoint were compared: total number of residency applicant publications (TNRAP), cumulative publication value units (cPVU), and the Arms Race Control Score (ARCS). Associations and incremental discrimination were assessed using multivariable logistic regression.

Of 551 residents, 138 (25.0%) were academic. Top-40 medical school status (adjusted OR 2.64, 95% CI 1.37-5.11; p = 0.004), publicly inferred female sex (adjusted OR 1.87, 95% CI 1.23-2.85; p = 0.004), and PhD attainment (adjusted OR 1.73, 95% CI 1.07-2.79; p = 0.025) were associated with academic placement. TNRAP (adjusted OR 1.091, 95% CI 1.010-1.178; p = 0.026) and cPVU (adjusted OR 1.048, 95% CI 1.003-1.095; p = 0.036) showed small effects; ARCS was not significant (adjusted OR 1.051, 95% CI 0.998-1.107; p = 0.062). The metrics were strongly intercorrelated, and adding any metric produced minimal, nonsignificant improvement in discrimination.

TNRAP and cPVU were modestly associated with early academic placement, but residency application publication metrics added limited incremental discrimination beyond baseline applicant-visible characteristics. Greater complexity did not improve on a simple count.