Construct validity of the chicken model in the simulation of laparoscopic radical prostatectomy suture

For a simulation to be valid, it has to be close to reality, correlate with the performance in a real case, and be able to discriminate between individuals with different degrees of experience. We explored the construct validity (ability to discriminate between experienced and inexperienced subjects) of a simulation for the urethrovesical anastomosis during laparoscopic radical prostatectomy (LRP).

Dead chickens were used. After partial emptying of the corporal cavity, an 18F catheter was placed through the esophagus to the stomach. In the Pelvic Trainer, a laparoscopic section of the esophago-glandular-stomach junction and a suture between the two edges were performed in the same fashion as for a urethrovesical suture in LRP. Five subjects with different levels of experience in laparoscopy, ranging from nil to more than 250 LRPs, tested two types of suture. One-way ANOVA was used to detect differences in suturing time among the operators. A multiple comparison test was used to detect specific differences between operators.

Although suturing time had a strong negative correlation with the operator's experience, significant statistical differences in suturing time were found only between the least experienced operator and the other ones. Qualitative differences in the suture were found among operators.

This simulation correlates well with the laparoscopic (LRP) subject's experience. It can discriminate between inexperienced and experienced subjects, exhibiting moderate construct validity, but failed to reflect the different levels of experience among the most experienced subjects.