Physician Implicit Racial Bias and Guideline-Concordant Vascular Care

Physician implicit racial bias is associated with the performance of low-value procedures for Black patients in clinical contexts lacking clear practice guidelines. Less is known about the association between implicit bias and adherence to well-defined and monitored clinical guidelines.

To investigate whether physician implicit bias is associated with the performance of evidence-based and monitored guidelines.

This cross-sectional study linked race Implicit Association Test (IAT) scores of vascular specialists from the Vascular Quality Initiative (VQI) with clinical registry data. Based on their IAT scores, specialists were categorized as having strong, moderate, or slight bias in favor of either Black or White individuals (ie, pro-Black or pro-White bias), or no bias. IAT data were collected from October 2021 to October 2022 (analyses completed through March 2025). Procedure-level registry data included peripheral revascularizations (2011-2022).

Race IAT scores, categorized as pro-White bias vs pro-Black or no bias.

The primary outcomes were preoperative prescription of antiplatelets, statins, and antihypertensives and patient smoking status. Mixed effects logistic regression models were used to assess the association of physician implicit bias with each outcome. A summary index of guideline-driven clinical practices was evaluated via a mixed effects Poisson model. Models were fit for both the full sample and an intermittent claudication (IC) subsample, with interaction terms testing effect modification by patient race.

Of 2512 eligible specialists, 218 (144 [66%] White, 74 [34%] other race and ethnicity; 160 [73%] male; mean [SD] age, 45.9 [9.3] years) completed the race IAT and were linkable to revascularization procedures for 21 632 patients across 960 US centers (4992 [23%] Black, 16 640 [77%] White; 13 613 [63%] male; mean [SD] age, 68 [11] years; 6588 in the IC subsample). No significant associations were found between physicians' implicit bias and performance of guideline-driven clinical practices. Stratifying by patient race, among all physicians, antiplatelets were prescribed preoperatively in 13 788 White (82.9%) and 4057 Black (81.3%) patients (P = .01); statins, in 13 107 White (78.8%) and 3905 Black (78.3%) patients (P = .42); and antihypertensives, in 9327 of 9924 White (94.0%) and 2690 of 2888 Black (93.1%) patients (P = .11); additionally, 10 972 White (65.9%) and 3164 Black (63.4%) patients did not smoke at the time of the procedure (P = .001). Findings were consistent in the IC subsample. Physician implicit bias was not associated with the guideline summary index (incidence rate ratio, 0.98; 95% CI, 0.96-1.00).

In this cross-sectional study of vascular specialists and the care they deliver, in contrast with findings from guideline-ambiguous clinical settings, physicians' implicit racial bias was not associated adherence to established and monitored vascular care guidelines. These results suggest that clear and consistently monitored clinical guidelines may be important to combatting disparities among Black and White patients stemming from physician implicit bias.