Results
Between February 2017 and November 2019, 2309 women underwent a surgical repair for pelvic organ prolapse by 110 surgeons in 19 centres, agreed to participate in the registry and were included in the analysis. We estimate that the surgical procedures included in the analysis represent about 76.6% of eligible procedures for POP repair performed during the study period in the 19 centres (Table S1).
The vaginal repair group included 504 women, the transvaginal mesh group 692, and the laparoscopic sacropexy group 1113 (including 128 with robotic assistance; Table S2). Eight women in the laparoscopic sacropexy group needed a conversion (0.7%): three times to a laparotomic sacropexy, twice to laparoscopic lateral attachment, twice to transvaginal mesh, and once to vaginal repair. One or more other surgical procedures were associated with prolapse surgery, including midurethral sling placement or hysterectomy (Table S2).
The surgical groups differed in terms of age, body mass index, diabetes, menopausal status, smoking, previous hysterectomy, previous surgery for stress urinary incontinence or for pelvic organ prolapse, and anatomical defect (Table S3). Seven of the 12 covariates in the planned propensity score had weighted maximum standardized differences below 10%, while 5 (age, menopausal status, history of POP surgery, history of hysterectomy, and anatomical anterior defect) exceeded the threshold by a maximum of 4% (Figure S1).
Mean follow-up was 16.6 months (0.4 to 33.8). During this period, 52 women had Clavien-Dindo grade III or higher complications (Table S4). During surgery or in the first 48 hours after: 7 women had an intraoperative injury, 4 a postoperative haemorrhage or hematoma, and 1 a cardiac infarct. From 2 days to 2 months 18 women required a surgical treatment of complications (some women with more than one type of complication): 1 had a peritonitis related to ileal injury, 1 an appendicitis, 1 an incisional eventration with intestinal inclusion after laparoscopy, 1 a bladder retention related to transvaginal mesh, 7 a haemorrhage or hematoma, 4 an ureteral obstruction, 4 a pelvic abscess, 1 a severe postoperative pain after sacrospinofixation, and 2 a vaginal mesh exposure. Between 2 and 12 postoperative months, 20 women required surgical treatment for a complication (5 women had 2 complications each): 11 had a vaginal mesh exposure, 1 a bladder mesh exposure after sacropexy, 3 a severe chronic pain, 2 a ureteral obstruction after transvaginal mesh, 2 an incisional hernia after laparoscopy, and 3 a vaginal granuloma. Two women returned to the operating room more than a year after the prolapse surgery: 1 for an incisional hernia after laparoscopy, and 1 for toxin injection for de novo overactive bladder. Complications necessitated 16 interventions to remove the mesh totally or partially (0.9%).
The cumulative weighted incidence of serious complications (Figure 1, Table 1) showed a difference between the groups in favour of vaginal repair: 1.8% for vaginal repair (95% CI 0-3.9), 3.9% for transvaginal mesh (2.0-5.9), and 2.2% for sacropexy (1.1-2.6) at 12 months. Compared with the sacropexy group, the risk of serious complications was higher among women in the transvaginal mesh group (weighted-adjusted HR 1.55, 1.03 to 2.34). A concomitant total hysterectomy was associated with a higher risk of complications (Table 1).
Due to a recurrence of the prolapse, a second intervention was required for 32 women (1.4%): 14 after vaginal repair (2.8%), 6 after transvaginal mesh (0.9%), and 12 after sacropexy (1.1%). The cumulative weighted incidence of reoperation for prolapse recurrence at 12 months was 1.5% for vaginal repair (95%CI 0.4-2.5), 0.7% for transvaginal mesh (0-1.4), and 1.1% for sacropexy (0.3-1.9). Compared with the vaginal repair group, the risk of reoperation for prolapse recurrence was reduced in the transvaginal mesh and sacropexy groups (Figure 2; Table 2), with no significant differences between the latter two groups.
Postoperative stress incontinence resulted in reoperation for 61 women (2.6%): 6 after vaginal repair (1.2%), 28 after transvaginal mesh (4.0%), and 27 after sacropexy (2.4%).
In accordance with the design of the registry, we audited a sample of women included in 2017 and 2018 (61% of 1,575 women responded). The responses showed that 96.3% of the serious complications were already listed in the registry as were 94.1% of the reoperations for prolapse recurrence.