Data sources and searches

We searched MEDLINE and Cochrane Central Register for Controlled Trials from Jan. 1, 1990 through April 12, 2013 ( Figure 1 ). We excluded older studies because the TVT was not available in the United States prior to this. Search terms included “urinary incontinence,” “urgency,” “sling,” “obturator,” “retropubic,” “pubovaginal,” “vaginal tape,” “urologic surgical procedures” (instrumentation or adverse effects), and related terms. The search was limited to comparative studies, cohort studies, and systematic reviews. The search was further limited to human and English-language studies. Meeting abstracts were excluded. Any review articles obtained in this search were excluded after reference lists were reviewed and articles not originally in the search were obtained. Study authors were not contacted.

Literature flow

PVS , pubovaginal slings.

^∗ These studies were potentially eligible to be included for adverse event (AE) analyses; ^† Several studies had 3 arms and provided data for multiple comparisons; ^‡ For noncomparative studies, the following minimum sample size criteria were used: minisling obturator, n ≥120; minisling retropubic, n ≥100; obturator midurethral sling (MUS), n ≥1000; pubovaginal fascial, n ≥300; pubovaginal synthetic, n ≥120; retropubic MUS, n ≥1000; ^§ Several studies reported on ≥2 slings; ^# Only from randomized controlled trials (RCTs).

Schimpf. Sling surgery for stress urinary incontinence. Am J Obstet Gynecol 2014 .

Twelve reviewers independently double-screened the abstracts using the computerized screening program Abstrackr (Tufts Medical Center, Boston, MA). To establish relevance and consensus among reviewers, all 12 screened and achieved consensus on an initial batch of 300 abstracts. Potentially relevant full-text articles were also independently double-screened by 12 reviewers.

Study selection

For the principal evaluation of outcomes, we included peer-reviewed randomized controlled trials (RCTs) with at least 12 months of follow-up ( Table 1 ). Trials were excluded from outcomes analysis for poor randomization schemes, such as alternate assignment of patients or assignment based on day of the week or birth date. We included RCTs that compared ≥2 sling procedures or a sling procedure to Burch urethropexy performed in adult women for SUI. Studies that compared Burch urethropexy to any other surgery were excluded. Bulking injections were excluded because they are not similar enough to sling surgeries regarding cure, perioperative data, or AEs. When a study included 3 arms, it was analyzed as multiple 2-arm comparisons. For the evaluation of AEs we also included trials excluded from RCT analysis, nonrandomized comparative studies, and cohort (pre-post) studies of any follow-up duration. Because of the volume of these studies, sample size limitations were placed to restrict the number of studies to only those with the most patients and therefore highest potential for identifying a complication ( Figure 1 ). Studies included for AEs had to evaluate at least 1 sling type, and information about any other comparator surgery was not collected. Sling types of interest included MUS (retropubic, obturator), pubovaginal slings at the bladder neck (biologic, synthetic, or autologous), and minislings. All studies had to report results for cohorts (or study arms) of women who all received the same sling type (or Burch urethropexy); studies that combined women who received different sling types in their analyses were excluded. Studies that examined various aspects of surgical technique, anesthesia, or surgeon training were excluded if the same type of sling was used in each arm. Data were excluded if the surgical product used was not available in the United States as of April 2013.

Outcomes of interest from RCTs fell into 6 categories: objective cure, subjective cure, perioperative outcomes, quality of life or satisfaction, sexual function, and AEs ( Table 2 ). Studies with nonrandomized designs were included only for AEs. Information on cost was not collected.

Data extraction and quality assessment

Data were extracted by 1 of 12 reviewers using a standard data extraction form and confirmed by another; discrepancies were resolved by consensus. We extracted data on study characteristics, participant characteristics, funding source, details on the interventions, length of follow-up, outcomes of interest measured, and how these outcomes were assessed. After data extraction, the lead reviewer and methodologist categorized all outcomes extracted from the RCTs into the 6 outcome categories listed above. Two reviewers also categorized all AEs into 22 categories as listed in Table 3 . The underlying data, together with additional extracted information, are accessible online at http://srdr.ahrq.gov/ in the project Sling surgery for stress urinary incontinence in women: Society of Gynecologic Surgeons 2013.

We assessed the methodological quality of each RCT using predefined criteria from a 3-category system modified from the Agency for Healthcare Research and Quality. Studies were graded as good (A), fair (B), or poor (C) quality based on the likelihood of biases and completeness of reporting. Grades for different outcomes could vary within the same study.

Data synthesis and analysis

We were able to identify comparisons for MUS vs Burch, pubovaginal slings vs Burch, pubovaginal slings vs MUS, retropubic MUS vs obturator MUS, retropubic MUS vs retropubic MUS (based on route of passage), obturator MUS vs obturator MUS (based on route of passage), and minisling vs other sling. When at least 3 RCTs compared the same surgeries for the same outcomes and provided adequate data for metaanalysis (including for AEs), we performed random effects model metaanalyses to estimate pooled odds ratios (ORs). We included data from the time point closest to 12 months’ follow-up that were reported. For objective cure, studies used cough stress test, pad test, or both methods. Across studies, we treated the different methods as equivalent (ie, we included both methods in the metaanalyses), but when a single study reported both methods, we preferentially chose stress test over pad test or a combined outcome (both pad and stress tests). When at least 3 studies (pre-post, nonrandomized comparative, or RCT) reported the same AE for the same sling type, we performed random effects model metaanalyses of the arcsine transformed proportion of women with the outcome. The arcsine transformed proportion was used to minimize bias due to the nonnormal distribution when proportions are close to 0. However, when the total number of events was <3 or metaanalysis gave an implausible summary estimate, the exact proportion and confidence interval (CI) were calculated for the total number of events and women at risk. These absolute rates of AEs are compared qualitatively between procedures, and all data are presented in Table 3 .

For each comparison of different sling types (or vs Burch), we generated an evidence profile by grading the quality of evidence for each outcome according to the Grades for Recommendation, Assessment, Development, and Evaluation system. The process considered the methodological quality, consistency of results across studies, directness of the evidence, and imprecision or sparseness of evidence to determine an overall quality of evidence. Four quality rating categories were possible: high (A), moderate (B), low (C), and very low/insufficient (D). Evidence profiles for the reviewed studies are in the Appendix .

We developed clinical practice guideline statements incorporating the balance between benefits and harms of the compared interventions when the data were sufficient to support these statements. Each guideline statement was assigned an overall level of strength of the recommendation (1 = strong, 2 = weak) based on the quality of the supporting evidence and the size of the net benefit. The strength of a recommendation indicates the extent to which one can be confident that adherence to the recommendation will do more good than harm. The wording and its implications for patients, physicians, and policymakers are detailed in Table 4 .

Table 4

Society for Gynecologic Surgeons Systematic Review Group sling surgery for stress urinary incontinence in women, clinical practice guidelines

Midurethral sling vs Burch (open or laparoscopic)

For women considering midurethral slings or Burch procedures for treatment of SUI, we suggest either intervention for objective and subjective cure and that decision be based on: (1) which adverse events are of greatest concern to patient; and (2) any other planned concomitant surgeries (vaginal vs abdominal route). (1A) • Midurethral slings may result in lower rates of perioperative adverse events such as blood loss, postoperative pain, operating room time, hospital stay, bowel injury, wound infection, and hematomas. (1C) • Burch procedures may result in lower rates of return to operating room for retention, erosion, overactive bladder symptoms, and groin pain. (1C)

 

Pubovaginal sling vs Burch
For women considering pubovaginal slings or Burch procedures for treatment of SUI, we recommend pubovaginal slings to maximize cure outcomes. (1A) • Burch procedure results in lower rates of erosion, overactive bladder symptoms, and retention requiring reoperation. (1C) • Pubovaginal slings result in lower rates of wound infection, bladder/vaginal perforation, and bowel injury. (1C)

 

Pubovaginal sling (biologic and synthetic) vs midurethral sling (only TVT was studied)

For women considering pubovaginal or midurethral sling for treatment of SUI, we recommend midurethral sling for better subjective cure outcomes. (2C) • Midurethral slings may result in lower rates of perioperative outcomes such as operating room time, blood loss, and hospital stay. (2D) • Pubovaginal slings may result in lower rates of adverse events such as urinary tract infection and vaginal perforation. (2D)

 

Retropubic vs obturator midurethral slings

For women considering retropubic or transobturator midurethral sling, we recommend either intervention for objective and subjective cure and that decision be based on which adverse events are of greatest concern to patient. (1A) • Retropubic slings result in lower rates of sling erosion, need to return to operating room for treatment of sling erosion, groin/leg pain, and vaginal perforation. (1D) • Transobturator midurethral slings result in shorter operative time, fewer bladder/urethral perforations, less perioperative pain, fewer urinary tract infections, and less overactive bladder symptoms. (1D)

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Beyond consent to sterilization: facing up to the full range of barriers to post-abortion contraception
2. Utility of cell salvage in women undergoing abdominal myomectomy
3. Probiotic Lactobacillus rhamnosusGR-1 supernatant prevents lipopolysaccharide-induced preterm birth and reduces inflammation in pregnant CD-1 mice
4. A gut-wrenching feeling

Stay updated, free articles. Join our Telegram channel

Join