For data extraction, we searched through Scopus, PubMed, Web of Science, Embase, and Google Scholar databases and used PRISMA 2020 template for the study design. The initial search identified 3060 articles (Fig. 1). After de-duplication, 2922 articles were screened for further analysis, out of which 2797 were excluded upon the initial screening, leaving 117 studies for the full review. Of the 117 studies, 45 were excluded due to not reporting an outcome of interest (n = 45), or the data were not separated between women and men (n = 23), or because of being a review (n = 18), a case report (n = 6), a pilot study (n = 5), or there was not enough data (n = 3).

PRISMA study selection flow chart

Seventeen studies were included, comprising 857 female patients with SUI caused by ISD, using AUS. Of these, 736 were using AMS 800 (American Medical Systems, Inc., Minnetonka, Minnesota) AUS (n = 12), and 121 were using other types of AUS (n = 5). About 188 had undergone minimally invasive surgery (n = 8) and 669 had undergone open surgery (n = 9). The mean age ranged from 56.7 to 70.5 years for patients treated through minimally invasive surgery and 51 to 77 years for patients treated through open surgery. The cuff size ranged from 4.5 to 9 cm for patients undergoing minimally invasive surgery and 5 to 10 cm for patients undergoing open surgery (Table 1).

We divided the follow-up studies into (1) short-term follow-up (mean follow-up < 5 years), including 469 women (n = 12), of which 188 had undergone minimally invasive procedures (n = 8) and 281 of them had undergone open procedures (n = 4), and (2) long term follow-up (mean follow-up > 5 years) includes 388 women (n = 5) that all of them had undergone open procedure (Table 1).

In this study, we compared the complications of AUS among the population undergone minimally invasive and open surgeries. Complications comprised some necrosis, atrophy, erosion, infection, mechanical failure, revision needing, and leak (sphincter insufficiency) (Table 1).

The results of the quality assessment are summarized in Table 2. Using the criteria that a study with 2 stars in the selection domain AND 1 or 2 stars in the comparability domain AND 2 or 3 stars in the outcome/exposure domain of the Newcastle–Ottawa quality assessment for cohort studies, was considered to be fair quality. Eight studies had good qualities, eight studies had fair qualities and one study had poor quality(Table 2). Most of the studies were designed as retrospective cohort studies and hence were subject to the lacking of data that could be affecting the results. Lacking data in studies included the selection of the non-exposed variables in cohorts, lack of result assessment description, lack of data related to cuff size and type of sphincter in some of the studies, and in case competing interests and source of funding, were not reported.

Out of 857 patients under study, 2 cases of atrophy were reported (0.05 [95% CI, 0–0.19], P = 1.00, I2 = 0%). Atrophy occurred in 1 of 188 (0.53%) patients treated with minimally invasive surgery and in 1 of 669 (0.15%) patients treated with open surgery, in the 17 included studies (Table 1). The type of surgery was not associated with a statistically significant increase in atrophy number (p value = 0.713). The pooled effect size was 0.14% (0%, 0.69%) in patients treated with minimally invasive surgery and 0.04% (0%, 0.19%) in patients treated with open surgery (Table 3). In addition, the pooled effect size was 0.06% (0.0%, 0.27%) among the population with a follow-up of < 5 years and 0.04% (0%, 0.23%) among the population with a follow-up of > 5 years (Table 4).

None of the 17 included studies reported the occurrence of necrosis in the patients under study (Table 1).

A total of 17 studies reported 50 erogenous events number (p-value = < 0.001) (Fig. 2). Erosion occurred in 9 of 188 (4.78%) patients treated with minimally invasive surgery and in 41 of 669 (6.12%) patients treated with open surgery (Table 1). The pooled effect size was 1.65% (0%, 3.43%) in patients treated with minimally invasive surgery and 4.36% (1.79%, 6.92%) in patients treated with open surgery (Table 3). Also, the pooled effect size was 1.52% (0.43%, 2.61%) among the population with a follow-up < 5 years and 5.55% (3.29%, 7.81%) among the population with a follow-up > 5 years (Table 4).

Meta-analysis for number of erosion in female patients with urinary incontinence due to intrinsic sphincter insufficiency

Infection occurred in 34 of 857 (3.9%) patients, among them 12 out of 188 (6.38%) patients were treated with minimally invasive surgery and 22 out of 669 (3.2%) patients treated with open surgery, in the 17 included studies (p-value = 0.01) (Table 1). The mean follow-up period was associated with a statistically significant increase in infection numbers (p value = 0.21) (Fig. 3). The pooled effect size was 3.96% (1.22%, 6.71%) in patients treated with minimally invasive surgery and 1.21% (0%, 2.49%) in patients treated with open surgery (Table 3). In open surgery, the chance of infection is lower than in minimally invasive surgery(p-value = 0.021)(Table 5). Also, the pooled effect size was 0.90% (0.06%, 1.74%) among the population with a follow-up of < 5 years and 2.58% (0%, 5.34%) among the population with a follow-up of > 5 years (Table 4).

Meta-analysis for number of infection in female patients with urinary incontinence due to intrinsic sphincter insufficiency

A mechanical failure occurred in 56 of 857 (6.5%) patients, including 1 of 188 (0.53%) patients treated with minimally invasive surgery and 55 of 669 (8.22%) patients treated with open surgery, in the 17 included studies (p value = 0.01) (Table 1). The mean follow-up period was associated with a statistically significant increase in mechanical failure numbers. The pooled effect size was 0.28% (0%, 1.04%) in patients treated with minimally invasive surgery and 4.51% (1.96%, 7.06%) in patients treated with open surgery (Table 3). Also, the pooled effect size was 1.78% (0.05%, 3.50%) among the population with a follow-up of < 5 years and 4.03% (0.77%, 7.28%) among the population with a follow-up of > 5 years (Table 4).

Reconstructive surgery occurred in 102 of 857 (11.9%) patients, including 7 of 188 (3.72%) patients treated with minimally invasive surgery and in 95 of 669 (14.2%) patients treated with open surgery, in the 17 included studies (p value = 0 < 001) (Table 1). The mean follow-up period was associated with a statistically significant increase in reconstructive surgery numbers (Fig. 4). The pooled effect size was 0.98% (0%, 2.69%) in patients treated with minimally invasive surgery and 17.73% (9.55%, 25.92%) in patients treated with open surgery (Table 3).In Comparing open surgery methods with minimally invasive surgery methods, the probability of reconstructive surgery in patients treated with minimally invasive surgery methods is lower than open surgery methods(p value = 0.049)(Table 5). In addition, the pooled effect size was 1.29% (0%, 2.73%) among the population with a follow-up of < 5 years and 26.53% (13.15%, 39.91%) among the population with a follow-up of > 5 years (Table 4).

Meta-analysis for number of revision needing in female patients with urinary incontinence due to intrinsic sphincter insufficiency

Leaks occurred in 10 of 857 (1.1%) patients, including 4 of 188 (2.12%) patients treated with laparoscopic surgery and in 6 of 669 (0.89%) patients treated with open surgery, in the 17 included studies (p value = 0.33) (Table 1). The type of surgery was associated with a statistically significant increase in leak number (p value = 0.788). The pooled effect size was 0.21% (0%, 0.90%) in patients treated with minimally invasive surgery and 0.11% (0%, 0.37%) in patients treated with open surgery (Table 3). Also, the pooled effect size was 0.39% (0%, 0.96%) among the population with a follow-up of < 5 years and 0.06% (0%, 0.33%) among the population with a follow-up of > 5 years (Table 4).