Stress urinary incontinence (SUI) is defined as the involuntary loss of urine during physical exertion or when sneezing or coughing [1]. SUI can diminish patients’ quality of life and imposes an increased economic burden [2,3]. The prevalence of urinary incontinence varies according to the study population and inclusion criteria, but it has been reported to be as high as 49%, and it increases with age [4]. Approximately 50% of patients with urinary incontinence complain of SUI symptoms [5].

Both nonsurgical and surgical options can be used to treat SUI. Weight loss and exercise to strengthen pelvic-floor muscles, such as Kegel exercises, can improve SUI symptoms [6,7]. Electrostimulation may also strengthen pelvic-floor muscles, but there is insufficient evidence of its efficacy. Furthermore, there are no U.S. Food and Drug Administration-approved medications specifically for SUI [8]. Thus, physicians usually suggest surgery to improve symptoms when other treatments are unsuccessful.

In the United States, the annual number of urinary incontinence operations increased from 37,953 in 1998 to 94,910 in 2007, concomitantly with the introduction of minimally invasive operative techniques. By 2007, midurethral sling surgery increased to 75.2% of incontinency operations, while the retropubic suspension technique decreased to 13.8% [9]. In Korea, the proportion of transvaginal operations that used a midurethral sling also continued to increase [10]. Consequently, postoperative complications associated with incontinence surgery have also been increasing. Postoperative complications can have a devastating effect on individual patients and pose substantial additional cost burdens to the health care system.

Therefore, in order to reduce postoperative complications, it is important to identify preoperative clinical risk factors and to improve intraoperative procedures [11], which requires high levels of physician expertise. To this end, the aim of this study was to analyze the efficacy and to estimate the complication rate of incontinence surgery according to the surgeon’s specialty using a nationally representative dataset. In addition, we also examined the impact of performing a preoperative urodynamic study (UDS) on the efficacy and safety of incontinence surgery according to the surgeon’s specialty.

Several studies have evaluated practice patterns in incontinence surgery according to surgeon specialty [13,14], but few have compared the postoperative outcomes of incontinence surgery across specialties, and the results have varied. Anger et al. [15] suggested that patients who underwent surgery for SUI or pelvic organ prolapse (POP) performed by urologists were more likely to develop urinary tract infections and to undergo reoperation than those who underwent surgery performed by gynecologists based on medical claims data in the USA. In contrast, Slopnick et al. [16] suggested that complication rates were higher among gynecologists than urologists using the American College of Surgeons National Surgical Quality Improvement Program database. Furthermore, it is somewhat difficult to analyze the reoperation rate as a clinical outcome, because patients with recurrent SUI sometimes choose different surgeons for the subsequent reoperation. Elliott et al. [17] reported that approximately one-third of patients with recurrent SUI or POP did not notify their primary surgeon when undergoing a reoperation performed by another surgeon. Therefore, to effectively analyze the reoperation rate, population-based studies are necessary.

In this study of data from the NHIS database, which covers the entire Korean population, we found that the postoperative outcomes of incontinence surgery were dependent upon the surgeon’s specialty. The HR for reoperation was significantly lower for urologists than for non-urologists. The following reasons may at least partially explain this finding. First, greater experience can improve clinical outcomes [18]. Most incontinence surgery in teaching hospitals in Korea is performed in the urology department. Therefore, urologists may gain more surgical experience during residency. Furthermore, in the case of postoperative complications such as voiding dysfunction, additional tests such as cystoscopy and UDS that are particularly strongly associated with the specialty of urology may be required. Second, several studies have reported that there were differences in the preferred types of incontinence surgery across specialties, along with variation in the complication rate according to the type of surgery. James et al. found that urologists performed more autologous fascial sling procedures than gynecologists [13]. Wu et al. [14] reported that gynecologists performed more Kelly plications, whereas urologists performed more traditional pubovaginal sling procedures.

In addition, we estimated the impact of preoperative UDS on the efficacy and safety of incontinence surgery according to the surgeon’s specialty. UDS is the most effective test to check for preoperative clinical risk factors before incontinence surgery. In this study, we demonstrated that personalized information about patient incontinence may help physicians select the most appropriate treatment option. An online survey reported that 66%–96% of urogynecologists and urologists in the United Kingdom regarded preoperative UDS as essential in various clinical scenarios [19]. However, debate continues about whether preoperative UDS should be performed before incontinence surgery, because UDS involves invasive tests that require catheterization, such as cystometric analysis, urethral-pressure profiling, and leak-point pressure measurement.

Nager et al. [20] compared the outcomes between an officeevaluation-only group and a preoperative UDS group in a multicenter, non-inferiority, randomized trial (315 women per group). The office-evaluation-only group did not have significant inferiority in treatment success 1 year after surgery, and they also did not experience greater rates of other secondary outcomes, such as postoperative voiding dysfunction and impaired quality of life. van Leijsen et al. [21] also confirmed noninferiority in an office-evaluation-only group using the Urogenital Distress Inventory Urinary Incontinence Subscale at 1 year after surgery; however, the number of participants was small (n=59 women). Linder et al. [22] reported associations between preoperative urodynamic parameters and the risk of surgical sling release, leading to a postoperative voiding problem. None of the urodynamic parameters measured in this study, including peak flow rate, postvoiding residual volume, and detrusor pressure at maximal flow, showed a significant association with sling-release risk, although procedures involving the surgical-sling method were rare in this cohort.

In contrast, several studies have recommended preoperative UDS for incontinence surgery. Midurethral sling tension can be adjusted to be less obstructive if the Valsalva leak-point pressure is too low or if detrusor underactivity or intrinsic sphincter deficiency is diagnosed during a preoperative UDS [23]. In a study comparing the risk factors for failure between a retropubic tape group and a transobturator tape group, the odds ratios for mixed urinary incontinence and detrusor overactivity were 3.7 and 8.6, respectively, in the retropubic tape group compared with the transobturator tape group, showing that preoperative UDS could guide physicians in choosing the best surgical option [24]. Furthermore, preoperative UDS findings can help physicians and patients to predict postoperative urgency and urgency urinary incontinence. Lee et al. [25] reported that urodynamic parameters, such as detrusor overactivity, preoperative urgency symptom severity, and types of incontinence surgery, were significant risk factors for postoperative urgency or urgency urinary incontinence. These findings indicate that physicians can provide their patients more accurate counseling about potential voiding problems if they are guided by preoperative UDS data.

In our study, we could not investigate the efficacy of preoperative UDS as in previous studies, because most cases in which preoperative UDS was not performed in this cohort occurred in 2006 and early 2007. Thus, selection bias was likely to affect clinical outcome comparisons according to whether a preoperative UDS was performed. Instead, when stratified by performance of a preoperative UDS, the HRs for reoperation according to the surgeon’s specialty varied based on whether a preoperative UDS was performed. While the reoperation rate was significantly lower in the procedures performed by urologists when a preoperative UDS was performed, there was no significant difference in the HRs for reoperation between the 2 groups when a preoperative UDS was not performed. This indicates that urologists consider adjusting midurethral tape tension or postoperative treatment such as anticholinergic therapy based on UDS findings, because they typically have more experience treating patients with voiding disorders beyond SUI according to UDS results.

Our study has several limitations. First, we did not investigate the distribution of different incontinence surgery types according to surgeon specialty. Furthermore, the surgical experience and learning curve of each surgeon also could not be considered in this population-based study. The reoperation and complication rates might have varied according to surgery type and the surgeon’s experience. Thus, detection biases could have been a problem in this study. Second, this cohort did not include some important patient information, such as hysterectomy status, POP, and other comorbid diseases, with the exception of hypertension, diabetes, and dyslipidemia. These factors could affect the postoperative complication rate. Third, a matched case-control study estimating the value of preoperative UDS using a population-based cohort would be useful.

In conclusion, this population-based study showed that the postoperative outcomes of incontinence surgery depended on the surgeon’s specialty and that the reoperation rate was significantly lower in procedures performed by urologists than in procedures performed by nonurologists. The reoperation rate according to the surgeon’s specialty varied based on whether a preoperative UDS was performed. However, further research that accounts for the abovementioned limitations is needed to clarify the clinical risk factors and to improve the outcomes of incontinence surgery.