Anger, Jennifer T. MD, MPH1; Litwin, Mark S. MD, MPH1,2; Wang, Qin MA3; Pashos, Chris L. PhD3; Rodríguez, Larissa V. MD1
Slings are an effective treatment for stress urinary incontinence (SUI) caused by intrinsic urethral sphincter deficiency and are associated with high efficacy and low morbidity.1–3 Many variations in sling material, location, and anchoring technique have been reported.1,4–6 Although slings, along with the Burch colposuspension, are considered the gold standard in management of stress urinary incontinence, few rigorous health services research studies have been published that evaluate patient outcomes after sling surgery and other procedures for stress urinary incontinence.4
Data on complication rates after sling surgery have traditionally been derived from retrospective studies of clinical subjects or limited single-surgeon or single-institution prospective case series.7,8 The majority of these reports have focused on the sign of SUI as a postoperative outcome, while some have reported on the symptoms of SUI. Nevertheless, most published series have not studied in detail the complications associated with these procedures. Although reports of mortality9 or major complications after sling procedures are rare, significant morbidity has been reported, with complication rates reported to be as high as 20–36%.4,10–12
Unfortunately, most of the literature reporting results of slings does not meet scientific standards for outcomes analysis. In 1996, the Update of the Agency for Health Care Policy and Research Urinary Incontinence Clinical Practice Guidelines concluded that the surgical literature is deficient in standards to describe the outcome of surgical procedures for pelvic floor disorders.13 In addition, a systematic review of the literature identified a number of methodological flaws in the published series, including a lack of comprehensive assessment of postoperative complications.14 This prompted the National Institutes of Health (NIH) in 1999 to sponsor a workshop to address the state of research addressing female pelvic floor disorders. At that time, an expert panel was gathered and agreed on a series of standards for research in female pelvic floor disorders.15
One of the recommendations made by the NIH for outcomes research on SUI was that outcomes should not only focus on SUI symptoms but should also include associated symptoms and unwanted effects resulting from any intervention, including new urinary symptoms such as urge incontinence, frequency, and urinary urgency; changes in sexual function; onset of urinary tract infections; adverse effect on bowel function; surgical complications; and the development or worsening of pelvic organ prolapse.15 Because stress urinary incontinence frequently occurs in conjunction with pelvic organ prolapse, new-onset symptomatic prolapse after sling surgery may result, at least in part, from a surgeon’s decision to address only stress incontinence in a patient with both incontinence and prolapse. Therefore, the decision to perform concomitant prolapse surgery is likely influenced by surgical experience and expertise. In this study we analyze Medicare claims data to determine short-term complications after sling surgery among female beneficiaries aged 65 years and over, with a focus on the adverse effects deemed important by the NIH.15 Such outcomes may serve as a means of measuring the quality of care for patients with stress incontinence.
MATERIALS AND METHODS
Our study qualified for exemption by the University of California, Los Angeles, Institutional Review Board. Women undergoing sling procedures were identified from a 5% national random sample of Medicare beneficiaries using 1999–2002 claims data from the Public Use Files provided by the Centers for Medicare and Medicaid Services. Subjects were selected to be in the 5% beneficiary sample on the basis of the last two digits of their health insurance claim number, which is virtually synonymous with the Social Security number. Women who underwent such procedures between January 1, 1999, and July 31, 2000, (the index period) were identified on the basis of the presence of Physicians Current Procedural Terminology Coding System (4th edition, CPT-4) code 57288 (sling operation for stress incontinence, fascia or synthetic) and tracked from 6 months before the procedure through 12 months after surgery. Individual deidentified subjects were longitudinally tracked for 6 months before surgery to assess for comorbidities and preoperative testing (urodynamics and cystoscopy). We used International Classification of Diseases (9th revision) (ICD-9) codes to calculate Charlson scores to measure comorbidity. The Charlson Comorbidity Index Score represents the sum of weighted diagnosis codes for each comorbid condition. A higher score indicates greater comorbidity.16 Subjects were tracked for 12 months after surgery to assess for short-term complications.
Demographic characteristics (including subject age group and ethnicity), clinical characteristics (including comorbidity16,17), and complications were obtained from an analytical file created by linking encrypted beneficiary identification numbers from three Medicare Standard Analytic Files representing hospital inpatient, hospital outpatient, and physician-supplier part B care. Previous studies provide internal validation for similar research using Medicare claims data.18–20 Perioperative complications were assessed in the hospital setting using the Centers for Medicare and Medicaid Services hospital inpatient file. With the inpatient, outpatient, and physician-supplier (carrier) files, we identified key postoperative outcomes using all available Current Procedural Terminology and ICD-9 codes for relevant procedures and diagnoses among female Medicare beneficiaries (see the box, “ICD-9 and CPT-4 Codes Used for Analyses”). We used multivariate analysis to simultaneously control for subject age, race, and comorbid diagnoses.
In these data representing the claims of a random 5% of Medicare beneficiaries, 1,356 sling procedures were performed by 988 physicians from January 1, 1999, through July 31, 2000. This extrapolates to approximately 27,120 slings performed on the female Medicare population aged 65 years and over during the 18-month index period. The majority of women undergoing sling surgery were aged 65–74 years. Based on the ICD-9 coding data from the 6-month period before surgery, 69% of subjects had a Charlson score of zero, indicating low comorbidity. White women underwent 95.0% of the slings performed during the index period (Table 1).
In the first 3 months after surgery, surgical complications were identified in 12.5% of subjects. Also during that period, 33.6% of subjects were diagnosed with a urinary tract infection. By the end of the first postoperative year, 49.7% of subjects had been diagnosed with a urinary tract infection; 9.4% of subjects had a new diagnosis of pelvic pain (Table 2).
In the 6 months before sling surgery, 24.8% of subjects underwent cystoscopy, and 27.4% underwent urodynamic testing. In the first year after surgery, 32.4% of subjects underwent cystoscopy, and 30.5% underwent urodynamics. Bivariate analysis revealed that subjects who underwent preoperative testing (cystoscopy and urodynamics) were significantly less likely to undergo such testing postoperatively (P<.001, data not shown). Within 1 year of the procedure, 6.9% of subjects had a new diagnosis of outlet obstruction. Eight percent of subjects, not necessarily the same group of subjects diagnosed with outlet obstruction, underwent treatments to manage outlet obstruction, including urethral dilations and suprapubic tube placements; 15.2% of subjects were given a new diagnosis of urge incontinence within one year after sling surgery (Table 2).
Overall, there was a high incidence of new diagnoses of pelvic prolapse (19.0%) and prolapse repairs (23.2%) occurring within 1 year after the sling (Table 2). Of the 34.4% of subjects who underwent a prolapse repair at the time of sling surgery, 13.7% underwent subsequent prolapse surgery within 1 year after the sling.
Multivariate analysis revealed that subject race, age, and comorbidity each had a significant influence on outcomes. Specifically, nonwhite subjects were more likely to experience urologic and nonurologic postoperative complications, the development of prolapse or outlet obstruction in the first year after sling surgery (Table 3). Compared with subjects older than age 75, subjects aged 65–69 years were significantly less likely to experience nonurologic complications. Younger subjects were also less likely to undergo treatments to relieve outlet obstruction (odds ratio [OR] 0.58, 95% confidence interval [CI] 0.36–0.94), to have a new diagnosis of urge incontinence (OR 0.44, 95% CI 0.29–0.65), or to experience treatment failure as indicated by a repeat incontinence procedure within 1 year after sling surgery (OR 0.58, 95% CI 0.32–0.87). Subjects with a Charlson index of one or greater were 1.5 times more likely to experience nonurologic complications than subjects with a Charlson index of zero (95% CI 1.17–1.90).
Most of the literature reporting results of sling surgery does not meet scientific standards for outcomes analysis in that it lacks a comprehensive assessment of postoperative complications and focuses primarily on the degree of improvement of stress incontinence. By only focusing analyses on the symptoms of stress incontinence, several key postoperative outcomes will be missed. Based on the NIH’s recommendations that outcomes should not only focus on cure of stress incontinence but should also include associated symptoms and unwanted effects resulting from any intervention,15 we chose to analyze several outcome categories. By doing so, we identified a broad spectrum of unexpected negative outcomes.
Our study had several principal findings. The surgical complication rate in the first 3 months after surgery was 12.5%. This finding is similar to that of Taub et al,4 which also used a nationally representative data set. Using the Nationwide Inpatient Sample of 147,473 women who underwent surgery for stress incontinence, Taub et al reported a 13.0% morbidity rate for the pubovaginal sling. Clinical series have reported complication rates after sling surgery to be as low as 4%.21 The fact that a higher complication rate was found in claims-based studies suggests that most surgeons who perform sling surgeries may not have the same results as experts who report outcomes in the literature and that the actual overall complication rate may be much higher.
We found a surprisingly high frequency of diagnosis of urinary tract infections occurring after sling surgery, with 49.7% of subjects diagnosed with urinary tract infections in the first postoperative year, including 33.6% in the first 3 months after the sling surgery. Such a high infection diagnosis rate may be due to any of several factors. It is possible that patients develop urinary tract infections because of insufficient treatment with perioperative antibiotics at the time of sling surgery. Alternatively, female Medicare beneficiaries, who represent an older population of sling recipients, may experience urinary tract infections at a high frequency, regardless of a recent history of sling surgery. It is possible, however, that patients who develop either temporary postoperative irritative symptoms or de novo urgency or urge incontinence are initially diagnosed and treated presumptively for possible urinary tract infections. Whether the diagnosis of such urinary tract infections is based on documented urine cultures is not known. Regardless of cause, the high incidence of diagnoses of urinary tract infections within 1 year after sling surgery warrants further investigation.
We found that 7% of subjects were given a new diagnosis of outlet obstruction, and 8% underwent treatments to manage outlet obstruction in the first postoperative year, including urethral dilations and suprapubic tube placements. Prolonged urinary retention from minimally invasive synthetic slings has previously been reported to occur in 2–13% of patients,12 with up to 4% requiring urethrolysis (sling take-down) procedures.22 Classic pubovaginal slings placed at the bladder neck may have even higher rates of outlet obstruction.6,22–24 Unfortunately, the Current Procedural Terminology code for urethrolysis (53500) was unavailable during the index period of our analysis. Therefore, we could not ascertain which cases of outlet obstruction were severe enough to require a formal urethrolysis.
Fifteen percent of our subjects were diagnosed with new-onset urge incontinence after the sling procedure. Subjects with a new ICD-9 diagnosis of urge incontinence may include women with de novo urge incontinence, those with urge symptoms secondary to an obstructing sling, those with a urinary tract infection, those who previously had mixed stress and urge incontinence before the sling was placed, and those who had pure urge incontinence preoperatively but were misdiagnosed with stress incontinence. Regardless of etiology, the frequent diagnosis of new-onset urge incontinence signifies that voiding dysfunction occurs in a large proportion of women undergoing sling surgery.
Pelvic pain was also diagnosed frequently during the first postoperative year (9.4%). Further studies are needed to determine if such pain develops as a result of the sling or is attributable to other causes. Nonetheless, our findings suggest the need to counsel patients preoperatively about pelvic pain as a possible complication of sling surgery.
Although urodynamic and cystoscopic procedures performed in the first year after sling surgery are not considered postoperative complications per se, they may serve as indicators of voiding dysfunction. These procedures are performed as part of a routine preoperative work-up by a minority of practitioners. In fact, in this database, only 24.8% of subjects underwent cystoscopy, and 27.4% underwent urodynamic testing in the 6 months before sling surgery. Cystoscopy and urodynamics in the postoperative setting are not routinely performed by most providers, and therefore, such postoperative testing is suggestive of either voiding dysfunction or another complication. The fact that 32.4% of subjects underwent cystoscopy and 30.5% underwent urodynamics implies that the incidence of postoperative complications from sling surgery in this population is higher than would be expected based on previous studies in the literature.
We found a high incidence of prolapse diagnoses occurring in the year after the sling operation. We also found that subjects who underwent concomitant prolapse surgery at the time of the sling procedure were less likely to undergo subsequent prolapse repair than those who underwent a sling surgery alone. If not addressed at the time of sling surgery, concomitant pelvic organ prolapse may become symptomatic after a sling is placed, often requiring surgical management. Alternatively, surgeons without experience in prolapse surgery may choose to address incontinence and then refer the patient to another surgeon with more prolapse experience. Such practice patterns could result in a second surgery within 1 year of the sling procedure.
Other factors that had a significant influence on outcomes included subject age, comorbidity, and race. The age of the subjects in our sample of Medicare beneficiaries is older than that of many reported studies, allowing for the detection of such age-related differences in outcomes. Compared with subjects older than age 75, subjects aged 65–69 years were significantly less likely to experience outlet obstruction, a new diagnosis of urge incontinence, or treatment failure as indicated by a repeat incontinence procedure within 1 year after sling surgery. Because elderly patients already have inherently greater prevalence of voiding dysfunction,25 it is not surprising that older subjects had worse outcomes after sling surgery.
Subjects with a Charlson index of one or greater were 1.5 times more likely to experience nonurologic complications than subjects with a Charlson index of zero (P<.05). The finding of a higher nonurologic complication rate among subjects with more comorbidities is an expected finding. Nonwhite subjects were more likely to experience nonurologic complications from the sling procedure, the development of prolapse, and a new diagnosis of outlet obstruction in the first year after sling surgery. Given that the vast majority of subjects who underwent sling surgery were white women (95% of subjects), further analysis is needed to determine the basis for these racial differences in outcomes.
Although Medicare claims provide data from a nationwide population-based sample of subjects from various clinical settings, the use of Medicare claims data has limitations.17 First, claims files are designed primarily to provide billing information, not detailed clinical information or use of health care resources not covered by Medicare. Therefore, outcomes such as improvement of incontinence or decreased pad usage could not be measured. Second, although numerous studies have used these data to understand real-world clinical practice and outcomes, some subjects with complications may not have been identified because severity was low (and therefore the complication was not coded) or the individual chose not to seek care (for whatever reason, which again resulted in no claim being submitted). Third, all types of slings are grouped together under one procedure code. Therefore, we could not make comparisons between classic pubovaginal slings placed at the bladder neck and minimally invasive midurethral slings, which have been shown in large case series to have substantially different rates of postoperative complications and outlet obstruction.6,21–24
Finally, findings in the Medicare population may be different from those in younger women undergoing sling procedures. In any case, these data reflect the real-world experience of a randomly selected, geographically diverse, older American population and the actual care delivered by the medical community.
1. Tash J, Staskin DR. Artificial graft slings at the midurethra: physiology of continence. Curr Urol Rep 2003;4:367–70.
2. The American Urological Association Female Stress Urinary Incontinence Clinical Guidelines Panel. Report on: The surgical management of female stress urinary incontinence. Clinical Practice Guidelines. Baltimore (MD): American Urological Association Inc; 1997.
3. Leach GE, Dmochowski RR, Appell RA, Blaivas JG, Hadley HR, Luber KM, et al. Female Stress Urinary Incontinence Clinical Guidelines Panel summary report on surgical management of female stress urinary incontinence. The American Urological Association. J Urol 1997;158:875–80.
4. Taub DA, Hollenbeck BK, Wei JT, Dunn RL, McGuire EJ, Latini JM. Complications following surgical intervention for stress urinary incontinence: a national perspective. Neurourol Urodyn 2005;24:659–65.
5. Moran PA, Ward KL, Johnson D, Smirni WE, Hilton P, Bibby J. Tension-free vaginal tape for primary genuine stress incontinence: a two-centre follow-up study. BJU Int 2000;86:39–42.
6. Dunn JS Jr, Bent AE, Ellerkman RM, Nihira MA, Melick CF. Voiding dysfunction after surgery for stress incontinence: literature review and survey results. Int Urogynecol J Pelvic Floor Dysfunct 2004;15:25–31.
7. Tsivian A, Kessler O, Mogutin B, Rosenthal J, Korczak D, Levin S, et al. Tape related complications of the tension-free vaginal tape procedure. J Urol 2004;171:762–4.
8. Levin I, Groutz A, Gold R, Pauzner D, Lessing JB, Gordon D. Surgical complications and medium-term outcome results of tension-free vaginal tape: a prospective study of 313 consecutive patients. Neurourol Urodyn 2004;23:7–9.
9. Comiter V, Vasavada S, Kavaler E, Carbone J, Raz S. The surgical treatment of female SUI: making an intelligent choice. Contemp Urol 2000;12:62–87.
10. Cross C, Cespedes R, McGuire E. Our experience with pubovaginal slings in patients with stress urinary incontinence. J Urol 1998;159:1195–8.
11. Kuuva N, Nilsson CG. A nationwide analysis of complications associated with the tension-free vaginal tape (TVT) procedure. Acta Obstet Gynecol Scand 2002;81:72–7.
12. Cetinel B, Demirkesen O, Onal B, Akkus E, Alan C, Can G. Are there any factors predicting the cure and complication rates of tension-free vaginal tape? Int Urogynecol J Pelvic Floor Dysfunct 2004;15:188–93.
13. Fantl JA, Newman DK, Colling J. Urinary incontinence in adults: acute and chronic management. AHCPR Publication No. 96-0682. Rockville (MD): Agency for Health Care Policy and Research Publication; 1996.
14. Black NA, Downs SH. The effectiveness of surgery for stress incontinence in women: a systematic review. Br J Urol 1996;78:497–510.
15. Weber AM, Abrams P, Brubaker L, Cundiff G, Davis G, Dmochowski RR, et al. The standardization of terminology for researchers in female pelvic floor disorders. Int Urogynecol J Pelvic Floor Dysfunct 2001;12:178–86.
16. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–83.
17. Hu JC, Gold KF, Pashos CL, Mehta SS, Litwin MS. Role of surgeon volume in radical prostatectomy outcomes. J Clin Oncol 2003;21:401–5.
18. Brandeis J, Pashos CL, Henning JM, Litwin MS. A nationwide charge comparison of the principal treatments for early stage prostate carcinoma. Cancer 2000;89:1792–9.
19. Saigal CS, Pashos CL, Henning JM, Litwin MS. Variations in use of imaging in a national sample of men with early-stage prostate cancer. Urology 2002;59:400–4.
20. Baron JA, Lu-Yao G, Barrett J, McLerran D, Fisher ES. Internal validation of Medicare claims data. Epidemiology 1994;5:541–4.
21. Morgan TO Jr, Westney OL, McGuire EJ. Pubovaginal sling: 4-YEAR outcome analysis and quality of life assessment. J Urol 2000;163:1845–8.
22. Deval B, Jeffry L, Al Najjar F, Soriano D, Darai E. Determinants of patient dissatisfaction after a tension-free vaginal tape procedure for urinary incontinence. J Urol 2002;167:2093–7.
23. Rackley RR, Abdelmalak JB, Tchetgen MB, Madjar S, Jones S, Noble M. Tension-free vaginal tape and percutaneous vaginal tape sling procedures. Tech Urol 2001;7:90–100.
24. Anger JT, Amundsen CL, Webster GD. Obstruction after Burch colposuspension: a return to retropubic urethrolysis. Int Urogynecol J Pelvic Floor Dysfunct 2006;17:455–9.
25. Carr LK, Walsh PJ, Abraham VE, Webster GD. Favorable outcome of pubovaginal slings for geriatric women with stress incontinence. J Urol 1997;157:125–8.