Maithel S, Sierra R, Korndorffer J, Neumann P, Dawson S, Callery M, Jones D, Scott D
Department of Minimally Invasive Surgery, Beth Israel Deaconess Medical Center, Boston, MA 02215, USA.
Surg Endosc. 2006 Jan;20(1):104-12. doi: 10.1007/s00464-005-0054-4. Epub 2005 Dec 7.
Video trainers may best offer visually realistic laparoscopic simulation, whereas virtual reality (VR) modules may best provide multidimensional objective measures of performance. This study compares the construct and face validity of three different laparoscopic simulators.
Subjects were voluntarily enrolled at the Learning Center during the 2004 SAGES annual meeting. Each subject completed two repetitions of a single task on each of three simulators, MIST-VR, Endotower, and CELTS; performance scores were automatically generated and recorded. Scores of individuals with various levels of experience were compared to determine construct validity for each simulator. Experience was defined according to four parameters: (a) PGY level, (b) fellowship training, (c) basic laparoscopic cases, and (d) advanced laparoscopic cases. Subjects rated each simulator regarding six face validity (realism of simulation) parameters using a 10-point Likert scale (10 = best rating) and participant scores were compared to previously established expert scores (proficiency goals for training).
Ninety-one attendees completed the study. Construct validity was demonstrated for all three simulators; significant differences in scores were detected according to one parameter for MIST-VR, two parameters for Endotower, and all four parameters for CELTS. Face validity was rated as good to excellent for all three simulators (7.0 +/- 0.3 for MIST-VR, 7.9 +/- 0.3 for Endotower [p < 0.001 vs MIST-VR], and 8.7 +/- 0.1 for CELTS [p = 0.001 vs MIST-VR, p = 0.01 vs Endotower]); 6%, 0%, and 36% of "expert" participants obtained expert scores on MIST-VR, Endotower, and CELTS, respectively.
All three simulators demonstrated significant construct and reasonable face validity. Although virtual reality holds great promise to expand the scope of laparoscopic simulation, current interfaces may limit their utility for assessment. Computer-enhanced video trainers may offer an improved interface while incorporating useful multidimensional metrics. Further work is needed to establish standards for appropriate skills assessment methods and performance levels using simulators.
视频训练器可能最适合提供视觉上逼真的腹腔镜模拟，而虚拟现实（VR）模块可能最适合提供多维的性能客观测量。本研究比较了三种不同腹腔镜模拟器的结构效度和表面效度。
在2004年SAGES年会期间，受试者在学习中心自愿报名。每个受试者在三种模拟器（MIST-VR、Endotower和CELTS）上各完成一项任务的两次重复；性能分数自动生成并记录。比较不同经验水平个体的分数，以确定每个模拟器的结构效度。经验根据四个参数定义：（a）住院医师培训阶段，（b）专科培训，（c）基本腹腔镜手术病例数，以及（d）高级腹腔镜手术病例数。受试者使用10分制李克特量表（10 = 最佳评分）对每个模拟器的六个表面效度（模拟逼真度）参数进行评分，并将参与者的分数与先前确定的专家分数（培训熟练程度目标）进行比较。
91名参与者完成了研究。所有三种模拟器均显示出结构效度；根据一个参数，MIST-VR的分数存在显著差异；根据两个参数，Endotower的分数存在显著差异；根据所有四个参数，CELTS的分数存在显著差异。所有三种模拟器的表面效度被评为良好至优秀（MIST-VR为7.0 +/- 0.3，Endotower为7.9 +/- 0.3 [与MIST-VR相比，p < 0.001]，CELTS为8.7 +/- 0.1 [与MIST-VR相比，p = 0.001，与Endotower相比，p = 0.01]）；分别有6%、0%和36%的“专家”参与者在MIST-VR、Endotower和CELTS上获得了专家分数。
所有三种模拟器均显示出显著的结构效度和合理的表面效度。尽管虚拟现实有望扩大腹腔镜模拟的范围，但当前的界面可能会限制其在评估方面的效用。计算机增强视频训练器可能会提供改进的界面，同时纳入有用的多维指标。需要进一步开展工作，以建立使用模拟器进行适当技能评估方法和性能水平的标准。
