Discussion
We report an analysis of data collected from routine care to compare the medium-term efficacy and safety of the 3 most common surgical procedures for pelvic organ prolapse repair in a population of 2309 women. The events were uncommon: the rate of serious complications ranged from 1.4% after native tissue vaginal repair to 3.8% after transvaginal mesh placement, and the rate of reoperations for recurrence from 2.8% after vaginal repair to 0.9% after transvaginal mesh. The women in the vaginal repair group were exposed to the lowest risk of serious complications and the highest risk of reoperation for recurrence, the women in the transvaginal mesh group to the highest risk of serious complications and the lowest risk of reoperation for recurrence, and the women in the laparoscopic sacropexy group to an intermediate risk of serious complications and a low risk of reoperation for recurrence, like in the transvaginal mesh group.
The strengths of our study are its large prospective registry including 19 centres and numerous surgeons; these features enable the detection of rare events. The analysis covers operations performed in real-life situations: complex clinical situations were not excluded. Indeed, a high proportion of our population would not have been included in a randomized trial, because they were too old, because of prolapse recurrence after previous surgery, or various comorbidities. The regular and routine verification of the information from the hospitals databases and from a sample of the participants is evidence of validity.
Our primary outcome was based on a robust criterion, as the modified Clavien–Dindo classification has been found to be a valid and reproducible classification of complications in various surgical domains.11,12 It is easy to verify serious complications and reoperations by consulting the hospital’s data department and the women who had surgery.
Our results must be interpreted considering the absence of randomization. The surgical groups had different characteristics. However, we took the women’s preoperative characteristics into account with the propensity score to emulate a comparative trial.14 The data about the construction of the propensity score appear reassuring regarding potential unmeasured confounders that may bias our results. In addition, within each surgical group, there were multiple and different surgical procedures. These limitations are the price of a ”real world” study design, which can introduce variance and limit statistical power.
Another limitation is the medium-term follow-up averaging 16 months. However, a Finnish cohort found that most complications of POP surgery occurred within the two first postoperative months.11Wihersaari O, Karjalainen P, Tolppanen A-M, Mattsson N, Jalkanen J, Nieminen K. Complications of Pelvic Organ Prolapse Surgery in the 2015 Finnish Pelvic Organ Prolapse Surgery Survey Study. Obstet Gynecol 2020;136:1135-44. Five year after transvaginal mesh placement, another study found that 79% of the mesh exposures occurred during the first postoperative year.22Jacquetin B, Hinoul P, Gauld J, Fatton B, Rosenthal C, Clavé H, Garbin O, Berrocal J, Villet R, Salet-Lizée D, Debodinance P, Cosson M. Total transvaginal mesh (TVM) technique for treatment of pelvic organ prolapse: a 5-year prospective follow-up study. Int Urogynecol J 2013;24:1679–86.
Our results about the relative risk of serious complications are similar to earlier comparative studies, reporting fewer complications with vaginal surgery without mesh.33Siddiqui NY, Cara L. Grimes CL, Casiano ER, Abed HT, Jeppson PC, Olivera CK, Sanses TV, Steinberg AC, South MM, Balk EM, Sung VW. Mesh Sacrocolpopexy Compared With Native Tissue Vaginal Repair, A Systematic Review and Meta-analysis. Obstet Gynecol 2015;125:44–55.,44Maher C, Feiner B, Baessler K, Christmann-Schmid C, Haya N, Brown J. Surgery for women with anterior compartment prolapse. Cochrane Database of Systematic Reviews 2016, Issue 11. Art. No.: CD004014. The incidence of complications in our registry was lower than that reported in several trials. This difference may be explained by the definition used for complications, which considered only serious Clavien-Dindo complications. The risk of reporting failure seems marginal, as we systematically and regularly verified the information with the hospitals’ data departments and with the women. Most participating hospitals are teaching hospital centres specialized in management of recurrent prolapse. This point may explain the low complication rate. Verification of this hypothesis might also mean that transvaginal mesh use by experienced centres might be at low risk of complications.
Today as concerns about mesh safety abound, our results offer real-world information for women that can enable them to participate in the choice of technique most appropriate for them. It suggests that in the medium term, laparoscopic sacropexy may have the best risk profile. Vaginal repair has a minimal risk of serious complications, but a higher risk of reoperation for recurrence.
Because reporting the specific functional results of prolapse surgery for each type of surgical procedure appears useful, we have planned a questionnaire to send to all the women in our registry. Identifying the technical surgical details that are at greater risk of complications than others (such as the type of mesh or the sutures used) should help improve surgical procedures. Assessing the effect of the experience of both surgeons and centres could be useful for determining if some procedures should be reserved to expert centres.