A survey of the complications of vaginal prolapse surgery performed by members of the Society Of Gynecologic Surgeons.

This study aimed to examine the frequency and nature of complications of vaginal prolapse surgery performed by members of SGS over a year and to determine the feasibility and the problems associated with prospective, multicentered collaborative data acquisition. A survey form, which included demographics, surgical indications, colpopexy type, concomitant procedures, technique, estimated blood loss (EBL), OR time, and intra/postoperative complications, was distributed to society members. The nature, extent, and solution of the complications were examined. There were 147 members of SGS at the time of the study. Many were reproductive endocrinologists and gynecologic oncologists. Twenty-one (14%) members participated. Three hundred forty-nine (349) completed forms were received: 187 sacrospinous fixations (SSF), 92 colposacropexies (CSP), and 70 high utero sacral suspensions (HUS). There were seven (3.7%) intraoperative complications for SSF, seven (7.6%) for CSP and three (4.3%) for HUS. There were four (2.1%) postoperative complications for SSF, six (6.5%) for CSP and none for HUS (NS). OR time was significantly longer for CSP vs. HUS ( P<.003) and for SSF vs. HUS ( p=.042). The EBL was significantly higher for SSF compared with CSP for the colpopexy procedure ( p=.013) and for entire cases ( p<.003). Analysis showed that all three colpopexies had significant intraoperative and postoperative complications of less than 8%. Intraoperative visceral damage was a concern for all three procedures. With SSF and CSP there was risk of bleeding and with HUS there was a risk of ureteral obstruction. Postoperative CSP complications were bowel obstruction, bleeding or hernia; for SSF neuropathy, and for HUS none. No life-threatening intraoperative or postoperative complications were reported. OR time was significantly shorter for HUS than SSF. The highest EBL was with SSF. Only 14% of the SGS membership responded, despite multiple requests for participation, demonstrating the difficulty of multicenter data gathering.