The literature search identified 1,707 unique references, and 1,409 studies were excluded during screening (Figure 1). Of the 298 full-text articles assessed for eligibility, 243 were excluded. Overall, 55 studies (n = 32,507) were included in the network meta-analysis. Summaries of all included studies are shown in Supplemental Table S4.
Pairwise meta-analyses result for different endpoints
Pairwise meta-analyses results were shown in Supplemental Figure S2. The results suggested that all treatments were more efficacious than placebo in reducing micturition frequency, urinary incontinence episodes, urgency episodes, and urgency urinary incontinence episodes. Patients receiving Antimuscarinics and Mirabegron showed no significant difference in reducing micturition frequency, urinary incontinence episodes, urgency episodes, and urgency urinary incontinence episodes. There was no difference in reducing micturition frequency and urgency urinary incontinence episodes between patients treated with Antimuscarinics and PTNS, and there was no difference in reducing urgency urinary incontinence episodes between patients treated with SNM and OnabotulinumtoxinA. In terms of 100% and ≥50% reductions from baseline in urinary incontinence episodes /day, there was no significant difference in the efficacy of placebo, Antimuscarinics and Mirabegron. Other direct comparisons were of limited significance due to the insufficient number of studies.
Network Meta-Analysis on the Outcomes of Interests
The primary network including all studies (regardless of RoB) for all outcomes were presented in Figure 2.
The network meta-analysis (Figure 3) indicated that treatment with OnabotulinumtoxinA or SNM resulted in obviously greater mean reductions in micturition frequency, urinary incontinence episodes, urgency episodes, and urgency urinary incontinence episodes compared with all of the other interventions included in the network and the efficacy of Antimuscarinics, Mirabegron and PTNS were similar and were better than that of placebo. In addition, the results suggested that patients receiving OnabotulinumtoxinA have the highest odds of achieving reduction of 100% and ≥50% in the number of urinary incontinence episodes /day [odds ratios relative to placebo: 5.92 (95% CrI 3.53–10.29) and 3.57 (95% CrI 2.56–5.00)]. Also, Antimuscarinics or Mirabegron was superior to placebo in 100% and ≥50% reductions from baseline in urinary incontinence episodes /day.
The Bayesian ranking probabilities of comparable treatments in different populations are shown in Figure 4. In overall, SNM was most likely to be ranked first for reducing micturition frequency, urgency episodes and urgency urinary incontinence episodes, and placebo ranked lowest. The ranking results for urinary frequency reduction and urgency urinary incontinence episodes reduction was as follows: SNM ranked first, OnabotulinumtoxinA ranked second, PTNS ranked third, one of Mirabegron and Antimuscarinics ranked fourth and the other ranked fifth, and placebo ranked sixth. The ranking results for urgency episodes reduction was as follows: SNM ranked first, OnabotulinumtoxinA ranked second, Antimuscarinics ranked third, Mirabegron ranked fourth, and placebo ranked fifth. The ranking results for urinary incontinence episodes reduction was as follows: SNM, OnabotulinumtoxinA and PTNS were the top three in no particular order, Antimuscarinics ranked fourth, Mirabegron ranked fifth, and placebo ranked sixth. The ranking results for achieving reduction of 100% and ≥50% in the number of urinary incontinence episodes /day were as follows: OnabotulinumtoxinA ranked first, SNM ranked second, and placebo ranked lowest.
Consistency and inconsistency assessment
The fit of the consistency model in all comparisons was similar or better than the fit of the inconsistency model (Supplemental Table S3). Node splitting analysis was performed to evaluate consistencies by comparing differences between the direct and indirect evidence and the result was shown in Supplemental Table S5, which showed no significant differences in most comparisons except for placebo vs OnabotulinumtoxinA, Anticholinergics vs SNM, OnabotulinumtoxinA vs SNM in the comparison of micturition, urinary incontinence episodes and 100% reductions from baseline in urinary incontinence episodes /day.
The results of sensitivity analysis (Supplemental Figure S3 and S4) excluding 12 studies considered to have a high RoB showed little impact on the results of the network meta-analysis. The main change of sensitivity analysis was summarized in Supplemental Table S6.