Construct Validity Testing of a Laparoscopic Surgical Simulator

Abstract

Background: We present initial data on the construct, content, and face validity of the LAPMentor (Simbionix), virtual reality laparoscopic surgical simulator. Study design: Medical students (MS), residents and fellows (R/F), and experienced laparoscopic surgeons (ES), with < 30 laparoscopic cases per year (ES < 30) and those with > 30 laparoscopic cases per year (ES > 30), were tested on 9 basic skill tasks (SK) including manipulation of 0-degree and 30-degree cameras (SK1, SK2), eye-hand coordination (SK3), clipping (SK4), grasping and clipping (SK5), two-handed maneuvers (SK6), cutting (SK7), fulguration (SK8), and object-translocation (SK9). Results: Mean MS (n = 23), R/F (n = 24), ES < 30 (n = 26), and ES > 30 (n = 30) ages were 26 years (range 21 to 32 years), 31 years (range 27 to 39 years), 49 years (range 31 to 70 years) and 47 years (range 34 to 69 years), respectively. In the lower level skill tasks (SK3, SK4, SK5, and SK6) the ES > 30, ES < 30, and R/F had similar scores, but were all substantially better than the MS scores. In the higher level skill tasks (SK7, SK8, and SK9), the ES > 30 scores tended to be better than the R/F and ES < 30, which were similar, and these, in turn, were markedly better than the MS. The ES > 30 had notably higher SK8 scores than the R/F and ES < 30, who had similar scores, and these had notably better scores than the MS. Conclusions: The noncamera skills (SK3 to 9) of the LAPMentor surgical simulator can distinguish between laparoscopically naive and ES. SK8 showed the highest level of construct validity, by accurately differentiating among the MS, R/F, ES < 30 and ES > 30. © 2006 American College of Surgeons.