Going bananas – a reliable low fidelity model for the assessment and training of suturing skills ​

Purpose Competency-based training requires reliable models with objective technical skill assessments. We present a novel scoring system, using a low-fidelity banana model that affords haptic and visual feedback.Method Participants were asked complete twelve radial stitches around a 2 cm hole in banana skin. Three surgeons and two trainees reviewed ninety-eight photographs, generated during a surgical trainees’ skills competition. Each photo was evaluated according to six criteria: “mushing”, suture count, radial and equidistant placement, and “skids/ tears”. A cumulative score was derived reflecting key needle handling principles.Results Average Intraclass Correlation Coefficient (ICC) values were reported, according to established guidelines, to reflect the reliability of the mean ratings across all raters, as these are more relevant to the study’s scoring procedure. For each ICC, 95% confidence intervals (CIs) and statistical significances were presented. All analyses were performed in IBM SPSS Statistics (Version 27). The overall ICC score demonstrated excellent reliability (ICCavg = 0.970, 95% CI: 0.960–0.979, p < 0.001). Among the individual domains, mush (ICCavg = 0.974, 95% CI: 0.964–0.981, p < 0.001) and number of suture (ICCavg = 0.976, 95% CI: 0.967–0.983, p < 0.001) also showed excellent reliability. RADIAN (ICCavg = 0.810, 95% CI: 0.742–0.865, p < 0.001), EQUIDISTANT (ICCavg =0.799, 95% CI: 0.726–0.857, p < 0.001), TEAR (ICCavg = 0.829, 95% CI: 0.767–0.879, p < 0.001), skid and tear (ICCavg = 0.848, 95% CI: 0.793–0.892, p < 0.001) demonstrated good reliability.Conclusion Our low fidelity simulation model objectively assesses fundamental suturing skills with excellent reliability, enabling monitoring and tracking of the acquisition of this essential surgical skill.