JMIR Publications

ABSTRACT

Background:

Training in percutaneous arterial intervention (PAI) for peripheral artery disease requires repeated practice while maintaining patient safety. Evidence describing procedural performance across successive clinical assessments during simulation-based training remains limited.

Objective:

This study aimed to compare procedural learning trajectories and posttraining educational outcomes between training incorporating simulation and conventional clinical training in percutaneous arterial intervention.

Methods:

This prospective, single-center comparative educational study included 50 physician trainees at Peking University People’s Hospital, China, between January 2020 and January 2025. Trainees had assisted PAI but had not performed it independently and were grouped according to training received (25 per group). Both groups completed a shared 12-hour curriculum followed by six weekly 4-hour sessions of conventional clinical or simulator-based training. Procedure time was recorded during supervised patient procedures at baseline and after each session. Posttraining assessments included technical performance, theoretical knowledge, and trainee ratings of satisfaction, perceived improvement, PAI understanding, and procedural confidence. Exploratory analyses included repeated-measures analysis of variance and baseline-adjusted analysis of covariance. Holm adjustment was applied separately within the technical outcome, knowledge-domain, and questionnaire-item families.

Results:

All 50 trainees completed training and assessments. Final mean procedure time was 69.50 (SD 6.00) minutes in the simulator group versus 82.00 (SD 6.50) minutes in controls. Procedure-time trajectories differed between groups (group-by-assessment interaction: F[5.40,259.36]=14.86; Greenhouse–Geisser-corrected P<.001). The baseline-adjusted final difference (simulator minus control) was −11.67 minutes (HC3 robust 95% CI −14.75 to −8.59; P<.001). Guidewire selection time favored simulation (mean difference −3.39 minutes, 95% CI −5.51 to −1.27; Holm-adjusted P=.009), as did stent deployment stability (median 5 [IQR 4–5] vs 4 [IQR 4–4]; Holm-adjusted P=.009). Overall procedural success was 96% (24/25) versus 84% (21/25), with a risk difference of 12 percentage points (95% CI −6.3 to 30.9). No binary technical outcome showed a statistically significant difference after Holm adjustment. Total theoretical knowledge scores did not differ significantly (mean difference −2.24 points, 95% CI −6.71 to 2.23; P=.32). All four trainee-rated items favored simulation (all Holm-adjusted P<.05).

Conclusions:

Training incorporating simulation was associated with shorter procedure times during supervised PAI, better performance on selected technical measures, and more favorable trainee evaluations. These findings support integrating simulation with supervised clinical training. Effects on sustained competence and patient outcomes require further evaluation.