Intraobserver and Interobserver Variability in Scoring Laparoscopic Diagnosis of Pelvic Adhesions

Videotapes of laparoscopic procedures are used for patient and physician education, independent consultation, and even in the courtroom as evidence. Validation of the interpretation of the videotape with respect to intraobserver and interobserver difference has been scarcely studied. As part of the protocol of a prospective, double-blind, placebo study a standardized adhesion scoring system was developed by the investigators. After laparoscopic adhesiolysis and instillation of either an active or placebo agent designed to prevent reformation of adhesions, a second look laparoscopy was performed between 6 and 10 weeks later to allow for a second adhesion score. Five reproductive surgeons evaluated via videotape pre and post adhesion scores in which 11 patients (22 procedures) were operated by one member of the group. Three months after the initial scoring of the videotapes each tape was reevaluated in order to measure intraobserver variability. Intraobserver and interobserver variability was assessed for adhesion extent as well as adhesion severity scores and also included consideration of the differences between the first and second procedures using sophisticated statistical methods. Intraobserver variability for adhesion severity was 12% versus 19% for interobserver variability. The replicate (intraobserver) variability for adhesion extent was 15.4% and interobserver variability was 21.2%. Two of the five observers consistently scored pathology as more severe than two others. The surgeon scored differences between the two procedures as greater than the other observers. Nevertheless, the variability was not so great as to preclude videotape evaluation of pathology in multicentric trials, or to allow for independent consultation of laparoscopic procedures via videotape.