Surgery for stress urinary incontinence by mid urethra sling (MUS) has become a standard widely used in routine clinical practice. The retropubic procedure is the one for which there is the most long-term data. Technical variants (transobturator and single incision sling) have been developed with the intention of simplifying the surgical procedure and reducing the risk of complications.
Clinical trials are limited in size and are not designed to specify the incidence of rare complications. The registers offer the possibility of specifying the occurrence of rare events in routine clinical practice [1].
Our objective is to specify the incidence, and to describe the serious complications relating to surgery by MUS in routine clinical practice according to the type of sling (retropubic, transobturator, or single incision) using the medium-term data from the VIGI-MESH register MESH [2].

All surgeons described each surgical procedure performed on a specific case report form. We checked the data collection by reviewing sling deliveries from the hospital pharmacies and surgical procedure codes recorded by each hospital's medical data department. 
We defined serious complications using Clavien-Dindo classification: sling placement cancelled due to peroperative injury (Grade III), subsequent surgical intervention related to a complication (Grade III), life-threatening complication (Grade IV), woman's death (Grade V). We collected also surgical revisions for SUI relapse. We used several sources to identify complications and revisions: the monitoring of surgical procedures by the data departments, surgeons' spontaneous reports, and a questionnaire sent to the women a year later. These annual follow-up questionnaires collected also information about perceived health and improvement (WHO, EQ5D, and PGI-I questionnaires).

Between February 2017 and November 2019, 1814 women underwent a surgical procedure with MUS in 18 centers and agree to participate. The sling procedures were performed through a retropubic route in 923 cases, a transobturator in 521 cases, and a single incision in 370 cases. 
An history of hysterectomy or previous incontinence surgery was more common (15.5%) in women who underwent a retropubic sling (Table).
The median follow-up time was 15 months. We observed serious complications in 82 women, 5.7% after retropubic sling (53/923), 4.0% after transobturative sling (21/521) and 2.2% after single incision sling (8/370).
Ten complications occurred during surgery and six in the first 48 hours: nine intraoperative injuries (six bladder injuries, two urethral injuries, one vaginal injury), two intraoperative haemorrhage, six obstructed micturition (related to five retropubic and one single-incision sling). Placement of the MUS was cancelled ten times and in six cases the sling was loosened vaginally.
There were 33 complications treated from two days to two months after the initial intervention: 27 cases of obstructed micturition (related to 20 retropubic, 6 transobturator, and 1 single-incision sling) resulted in loosening the sling in 22 women, dividing it in 3, and removing it in 2; one woman with severe groin pain after transobturator sling placement needed to return to OR on day two to remove the sling and replace it by a retropubic one; a woman needed  a laparotomy on D9 due to infected haematoma of Retzius space after transobtur sling; one woman needed to return to the operative room (OR) on day 13 to evacuate a suburethral thrombus that had resulted in obstructed micturition after transobturator sling; an early suprapubic abscess on retropubic BSU with purulent vaginal discharge led to the removal of the sling 13 days after placement; in the second month a part of the MUS had to be removed in 2 women because of vaginal exposure (associated with pain in one case).
Between 2 and 12 postoperative months, 26 complications were managed: 8 related to late urinary retention, 5 to overactive bladder, 6 to chronic pain, 15 to vaginal exposure, 2 to urethral exposure. These complications required MUS division (3 cases), partial removal of the MUS (19 cases), removal of a non-absorbable thread (1 case), and vaginal trimming without MUS resection (2 cases). 
Seven women returned to OR more than a year after applying the procedure (including 4 single incision slings) to remove part of the sling (6 cases) or the vaginal scar (one case). The reasons were pain (4 cases), vaginal exposure (1 case), overactive bladder (1 case), and obstructed micturition (2 cases). 
The survival curve without serious complication showed a significant difference between MUS types (Figure in months, logrank test p = 0.011). The estimated incidence at 18 months of serious complications was 6.3% [95% CI 4.6-8.0] in case of retropubic sling, 4.2% [2.4-6.0] in the case of a transobturator, and 1.6% [0.2- 3.1] in case of single incision. The risk of complication was 3 times lower with a single incision compared to a retropubic sling (RR 0.36 [0.17-0.75]), the risk was not significantly different for the transobturator approach compared to the retropubic (RR 0.67 [0.40-1.11 ]). 
Twenty-seven women (1.5%) benefited from a surgical revision due to a failure or a recurrence of stress urinary incontinence, 15 after retropubic MUS (1.6%), 7 after transobtur (1.4%) and 5 after single incision (1.4%; p=0.95). The revision procedure was the re-tensioning of the BSU (9 cases including 2 adjustable sling, 4 retropubic, 2 transobturator, and 1 single-incision sling), an injection of Bulkamid (2 cases), and the new placement of MUS (16 cases). 
Among the 1,167 women operated on between February 2017 and December 2018 contacted by mail, we received 692 responses to the health questionnaire sent a year or more after their surgery (59.3%). To the question "What do you think of your current state of health compared to what it was before your surgery for incontinence or prolapse?” 91.4% (608/665) felt better (much better, better, or a little better) (PGI-I). The perceived improvement was better in the absence of serious complications (p=0.008). 96.2% (630/655) rated their general health as good (very good, good, or fairly good) (Figure 3). The perceived health status was similar with or without complication. Compared to the French population of the same age, the operated women reported a better perceived state of health.

The retropubic approach was more often chosen in the event of failure of previous surgery. The main complications needing surgical revision were in order of frequency: voiding obstruction (2.4%), MUS exposure (1.1%), and chronic pain (0.8%). The retropubic route was associated with a higher risk of complication compared to the single incision. However, after 18 months the curves seem to get closer (Figure). In the medium term, the risk of surgical revision for failure on SUI is low (1.5%) and similar regardless of the type of sling applied.
Most women (over 90%) were improved by surgery and their health was better than in the general population.

Analysis of the first medium-term results in our registry shows that the single incision is a valid option for the first-line surgical treatment of stress urinary incontinence. However, we believe it is necessary to continue follow-up in the longer term.

Fritel X. Evidence About Surgical Revision for MUS Complications Will Come from Large Retrospective Cohorts and Prospective Registers. BJOG 2010.
Fritel et al. Complications after pelvic floor repair surgery (with and without mesh): short-term incidence after 1873 inclusions in the French VIGI-MESH registry. BJOG 2020.