Nguyen, John N. MD; Jakus-Waldman, Sharon M. MD; Walter, Andrew J. MD; White, Terry MD; Menefee, Shawn A. MD
Pelvic organ prolapse is a common condition affecting 15–30% of women over the age of 50 years, and the lifetime risk of undergoing prolapse or incontinence surgery in the United States is quoted as 1 in 11.1,2 Although mesh use for abdominal sacrocolpopexy and suburethral sling procedures has been shown to be effective3,4 and is considered the standard of care, there is considerable debate over the use of permanent mesh and biologic grafts for transvaginal pelvic organ prolapse repair. Transvaginal mesh repairs for pelvic organ prolapse appear to offer superior anatomic outcomes compared with suture-based repairs in the short term.5–7 However, severe complications have been reported with the former in small studies.8–14 Reports of mesh-related complications has prompted the U.S. Food and Drug Administration (FDA) to release a Public Health Notifications in October 2008 concerning the use of mesh for pelvic organ prolapse repair and urinary incontinence and more recently a Safety Communication in July 2011 on the use of vaginal mesh for prolapse repair.15,16 The primary aim of this study was to estimate the perioperative complication and reoperation rates associated with slings and prolapse repairs using mesh and biologic grafts performed on Southern California, Northern California, and Hawaii Kaiser Permanente female members.
MATERIALS AND METHODS
This analysis was undertaken after institutional review board approval was obtained from Kaiser Permanente Southern California, Northern California, and Hawaii. Informed consent was not required for this study. Funding was provided by Southern California Kaiser Permanente Department of Research and Evaluation who were not involved in the study design or article writing. The primary aim of this study was to estimate the perioperative complication and reoperation rates associated with slings and prolapse repairs using mesh and biologic grafts. The study cohort comprised of all female members of Kaiser Permanente Southern and Northern California and Hawaii who underwent sling or pelvic organ prolapse surgeries using implanted grafts or mesh between September 1, 2008, and May 31, 2010.
An algorithm using specific search criteria was developed and checked for validity in three phases. A sample of 150 consecutive patients who had incontinence and prolapse surgery, with and without implanted prostheses, performed by the first author (J.N.N.) was first used to develop the algorithm and verify that the search criteria accurately identified the patients and captured the information of interest. The search criteria were subsequently applied to a random sample of patients from each Kaiser Permanente region. The medical records of these patients were each reviewed for accuracy of both device and patient data elements as well as differences in the Physicians' Current Procedural Terminology Coding System, 4th edition (CPT-4) and International Classification of Diseases, 9th Revision (ICD-9) coding across regions. Regionally specific modifications were made to the search criteria to ensure that the algorithm accurately captured the desired data elements from each Kaiser Permanente region. The regionally specific criteria were then applied to the respective region. The medical records of the entire study cohort were randomly sampled and checked for accuracy of the data elements.
The HealthConnect Clarity database served as a primary data source for identification and characterization of the study cohort, relevant surgical procedures, implant used, perioperative complications, and readmissions and reoperations. Potential participants were identified using CPT-4 codes 57288 (sling operation for stress incontinence), 57280 (colpopexy, abdominal approach), 57425 (laparoscopy, surgical, colpopexy), and 57267 (insertion of mesh or other prosthesis for repair of pelvic floor defect). Surgical implant logs containing the product name, product number and lot number, as well as additional CPT-4 codes were used to identify the implants used and site of placement (see the Appendix, available online at http://links.lww.com/AOG/A283). Operative reports of patients with multiple implants and those in which the implant site could not be delineated by the associated CPT-4 or ICD-9 codes were reviewed to determine the site of implant placement. We excluded male patients and those who had surgeries for stress urinary incontinence or prolapse without implants, surgeries in which implants could not be identified or were used for other indications, InterStim, or urethral bulking procedures.
Participants were characterized by demographic and clinical data, including age, parity, body mass index (BMI, calculated as weight (kg)/[height (m)]2), menopausal status (52 years old or older at the time of index surgery), race, smoking history, and presence or absence of diabetes mellitus. Sling implants were classified as retropubic, transobturator, single-incision, or other. The ‘other’ group consisted of pubovaginal slings and slings in which a reusable introducer was used to place the prosthesis. Prolapse implants were classified as apical, anterior, or posterior. Apical prolapse implants consisted of sacrocolpopexies performed laparoscopically (CPT-4 57425) or abdominally (CPT-4 57280) and did not include any transvaginal mesh procedures coded as apical suspensions. The latter were coded based on the vaginal site of mesh placement and not the marketed indication. For example, Perigee and Elevate Anterior both were classified as “anterior” although the latter is marketed as an anterior and apical prolapse repair system. If a patient had multiple implants placed during the same surgery, each implant would be counted based on the site of placement. Thus, patients who had the Prolift Total Pelvic Floor System were classified as having both an “anterior” and “posterior” implant. A similar classification system was used for posterior vaginal implants. All implants were subgrouped based on manufacturer and product code. Queries were run in the Kaiser Permanente HealthConnect Clarity database using ICD-9 and CPT-4 codes to identify suspected complications, hospital readmissions, and relevant reoperations occurring within 12 months of surgery (see the Appendix, http://links.lww.com/AOG/A283). The medical records of all patients with suspected complications were reviewed for accuracy of data elements.
Statistical analysis was performed using the SAS statistical software. Mesh-related complications such as visceral organ perforation or injury, reoperations for prosthesis erosion, or infection were calculated per unique implant as the denominator. Complications that could be attributed to concurrent procedures such as transfusion, wound seroma, and bowel obstruction were calculated per unique procedure as the denominator. Continuous variables were compared with the two-tailed t test or Wilcoxon rank-sum tests. Categorical variables were compared with the χ2 or Fisher's exact test. Multivariable logistic regression with backward elimination was performed to determine independent factors associated with vaginal mesh excision. Statistical significance was set at P<.05. All terminology conformed to the guidelines proposed by the International Continence Society unless specified otherwise.17
During the 21-month period, 4,223 women underwent gynecologic procedures in which a prosthetic implant was used. Eighty-one women were excluded from analysis because the prostheses implanted were not slings, not used for prolapse repair, or could not be identified. The study cohort consisted of 4,142 women with a mean age of 57 years (standard deviation 12.2), median parity 3 (interquartile range 1–4), and median BMI 28 (interquartile range 25–32) (Table 1). Most were menopausal (64%), had never smoked cigarettes (67%), and not diabetic (86%).
Across all regions, 3,747 (71%) and 1,508 (29%) prostheses were used for slings and prolapse repairs, respectively (Table 2). Permanent mesh was used more often than biologic grafts for slings (n=3,671, 98%) compared with prolapse repairs (n=858, 57%) (P<.001). Slings were most commonly performed retropubically (n=2,339, 63%) compared with transobturator (n=794, 21%), single-incision (n=75, 2%), or other (n=539, 14%) approaches (P<.001). Tension-free vaginal tape and Advantage slings were performed most commonly.
Prostheses were implanted more commonly in the anterior vagina (n=721, 48%) than the vaginal apex (n=498, 33%) or posterior vagina (n=289, 19%) (P<.001). Permanent mesh was used more frequently than biologic grafts in apical suspensions (n=487, 98%) compared with anterior (n=307, 42%) and posterior (n=64, 22%) vaginal prolapse repairs (P<.001). Gynemesh (n=309, 63%) was most frequently used for apical repairs. Repliform was most commonly used for both anterior (n=237, 57%) and posterior (n=131, 58%) prolapse repairs.
Surgical procedures were performed primarily by gynecologists (97%) and less frequently by urologists (3%). Across all regions, 377 different surgeons (Northern California, n=196; Southern California, n=173; Hawaii, n=8) were listed as primary surgeons and were primarily gynecologists (n=308) followed by urologists (n=46) and fellowship-trained female pelvic medicine and reconstructive surgeons (n=23). Slings were implanted in nearly equal proportions by gynecologists (n=1,806, 48%) and fellowship-trained female pelvic medicine and reconstructive surgeons (n=1,638, 44%) and less frequently by urologists (n=303, 8%). In contrast, fellowship-trained female pelvic medicine and reconstructive surgeons placed the majority (n=1,159, 77%) of the prolapse implants followed by gynecologists (n=296, 20%) and urologists (n=53, 3%).
Perioperative and postoperative complications for slings and prolapse repairs are listed in Tables 3 and 4, respectively. Trocar bladder perforations (51 of 3,747 [1.4%]) occurred more commonly than urethral perforations (2 of 3,747 [0.05%]) in sling procedures (P<.001) and occurred most often with retropubic slings (P=.005). There were no significant differences in trocar bladder perforation rates between sling manufacturers (P=.10). There were no trocar-related injuries involving the lower urinary tract or rectum for prolapse repair kit procedures. Dissection-related bladder injuries occurred in 7 of 721 (1%) anterior prolapse repairs and 8 of 498 (1.6%) apical suspensions (P=.47). Dissection-related bowel injuries occurred in 1 of 498 (0.2%) apical suspensions and 2 of 289 (0.7%) posterior prolapse repairs (P=.63). Ureteral and vascular injuries occurred infrequently.
Participants were tracked for hospital readmissions and relevant reoperations occurring up to 12 months after the index surgery. The majority (3,879 of 4,142 [94%]) of participants remained Kaiser Permanente members at least 12 months after their index surgery and were available for follow-up. Sling loosening or transection for voiding dysfunction or urinary retention occurred in 49 (1.3%) procedures at a mean of 80 days (standard deviation [SD] 79, range 7–351 days) after the index procedure. Four of 49 (8%) sling loosening or transection procedures were performed twice. Excision for vaginal mesh erosion occurred in 30 (0.8%) sling procedures at a mean of 175 days (SD 107, range 46–388 days) after the index procedure. Multiple excisions for persistent mesh erosion occurred in 2 of 31 (6%) procedures. Vaginal mesh excision rates did not differ significantly between sling approaches (P=.93) or between manufacturers (P=.73). Reoperation for urethral mesh erosion occurred in three (0.08%) sling procedures all after “bottom-up” retropubic approaches.
Reoperations after prolapse procedures were performed more often for vaginal mesh erosion (29 of 846 [3%]) than for biologic graft infection (2 of 650 [0.3%]; P=.01). The mean time to reoperation was longer for mesh-related complications (267 [SD 98] days) than for graft infections (10 [SD 2.8] days). Excision of vaginal mesh erosion was performed more commonly after anterior (19 of 307 [6%]) compared with apical (9 of 487 [2%]) or posterior vaginal implants (1 of 64 [2%]; P=.018). Vaginal mesh excision after apical prolapse repairs was performed less frequently when uncoated lightweight, macroporous polypropylene mesh (4 of 337 [1%]) was used compared with collagen-coated mesh (4 of 59 [7%]) and Marlex hernia mesh (1 of 13 [8%]; P=.01). After multivariable logistic regression, anterior vaginal implant location (odds ratio [OR] 2.8, 95% confidence interval [CI] 1.3–6.4; P=.0) was the only significant factor associated with undergoing vaginal mesh excision; concomitant hysterectomy (OR 0.97, 95% CI 0.31–3.06; P=.99), menopause (OR 1.01, 95% CI 0.21–4.81; P=.99), active or passive cigarette exposure (OR 0.89, 95% CI 0.11–7.16; P=.91), race (OR 1.24, 95% CI 0.26–5.01; P=.77), age (OR 0.98, 95% CI 0.94–1.02; P=.42), diabetes mellitus (OR 1.55, 95% CI 0.61–3.95; P=.36), BMI (OR 2.05, 95% CI 0.77–5.44; P=.23), and vaginal parity (OR 0.60, 95% CI 0.28–1.23; P=.17) were not. Mesh infection occurred in one patient after a sacrocolpopexy with uncoated lightweight, macroporous polypropylene mesh and concomitant vaginal hysterectomy, transobturator sling, and posterior vaginal prolapse repair. The pelvic abscess was treated with intravenous antibiotics and computed tomography-guided drainage.
Adverse events associated with concomitant surgeries occurred uncommonly. Operative and perioperative adverse events included transfusions (n=60), bladder injury (n=9), ureteral injury (n=6), bowel injury (n=5), vascular injury (n=1), and rectal injury (n=1). Reoperations and readmissions associated with concomitant surgeries included pelvic abscess drainage (n=4), repair of rectovaginal fistula (n=2), and vesicovaginal fistula (n=2) as well as readmission for treatment of wound infection (n=2) and small bowel obstruction (n=1).
The primary aim of this study was to estimate perioperative complication and reoperation rates associated with sling and prolapse repair procedures using implanted prostheses. Our study demonstrated low rates of perioperative adverse events as well as reoperations for mesh-related complications after sling procedures and sacrocolpopexy. Our rate of trocar-related bladder perforations was slightly lower than that of other published studies. In a meta-analysis of retropubic and transobturator midurethral slings,18 bladder perforations occurred in 6.7% of patients undergoing the former. Other complications associated with sling procedures were uncommon and also consistent with previously published studies.4,18
Reoperations for vaginally placed mesh in this study were lower than expected and highest after anterior vaginal prolapse repair. Mesh location has not been previously shown to be an independent risk factor for vaginal mesh exposure in other studies.19,20 The explanation for our finding is not entirely clear. We speculate that differences in surgical techniques, types of concomitant procedures, and patient characteristics may have contributed to this finding. Additionally, the orientation of the pelvic forces may more likely cause separation of an anterior vaginal incision compared to an apical or posterior vaginal incision and contribute to a higher rate of vaginal mesh exposure in the former group. Unlike other studies,19,20 we did not find age, smoking, or diabetes mellitus to be risk factors for vaginal mesh excision.
Among the various types of polypropylene mesh implanted, hernia mesh and collagen-coated mesh were associated with higher rates of vaginal mesh erosion requiring excision than uncoated lightweight monofilament macroporous polypropylene mesh after apical prolapse repair. However, we did not find significant differences in mesh excision rates among different uncoated lightweight macroporous monofilament polypropylene meshes. Additionally, our mesh infection rate was extremely low and likely related to the use of lightweight monofilament macroporous polypropylene meshes produced for prolapse repair.12
This study had several limitations as a result of its retrospective design. Intraoperative complications were reported at the surgeon's discretion. Thus, minor complications such as trocar injuries not requiring repair or unreported injuries or repairs could not be identified and potentially underestimated. Participants did not undergo systematic or standardized evaluations for mesh-related complications, which can also result in such complications being underestimated. Additionally, outpatient clinic data were not collected in this pilot project as a result of budgetary constraints. As a result, complications that did not require hospital readmission or reoperation (inpatient or outpatient) were not identified. Thus, we suspect that mesh-related pain and dyspareunia not requiring surgery were likely underreported. An audit of the outpatient medical records is required to determine the rates of these complications. We also realize that our follow-up time is short and continued surveillance of mesh-related complications is necessary. Lastly, we did not report quality of life, sexual function, and patient satisfaction because these subjective data were not readily available.
The principle strength of this study was that it examined the outcomes of a large racially diverse patient cohort treated by a diverse group of surgeons at multiple medical centers. Admittedly, randomized controlled trials (RCTs) are considered the “gold standard” of medical research because they assess disease states and treatments in ideal settings. However, clinicians may not find the results obtained from RCTs useful as a result of differences in practice patterns as well as patient demographics and compliance. Although we do not dismiss the results of well-conducted RCTs, we believe that our results represent “true life practices” and may be more generalizable than the aforementioned study design. Furthermore, performing RCTs on mesh-related complications in a large study cohort may not be feasible as a result of time and cost constraints.21 Postmarket surveillance of prosthetic implants in a large study cohort over long periods of time may be best suited through a registry. We believe that our system for identifying and tracking patients with pelvic floor surgeries using prosthetic implants will allow us to develop such a registry. Continued surveillance is planned to determine the long-term disposition of implanted mesh in this study cohort. We also plan to review outpatient medical records of patients with mesh-reinforced prolapse surgery to determine whether the incidence of mesh-related complications is different than the reoperation rate in this group. Lastly, we are currently developing methods of capturing patient responses to validated quality-of-life and sexual function questionnaires in our electronic medical records. Once implemented, we will also be able to collect and report such data.
The recent FDA Safety Communication in July 2011 on the use of vaginal mesh for prolapse repair16 has prompted passionate arguments on both sides of the fence. Central elements to these arguments include the safety and effectiveness of mesh over “native tissue” repairs for pelvic organ prolapse and the FDA 510(k) approval process for medical devices. The recommendations from a recent FDA Obstetrics and Gynecology Device Panel Meeting included the need for reclassification and premarket evaluation requirements as well as postmarket surveillance for vaginal mesh for prolapse repair.22 Improved postmarket studies should shed more light on the long-term effectiveness and safety of vaginal mesh, risk factors for adverse events as well as the population of women that is best suited to undergo these procedures. Until these data are available, it is imperative that health care providers not only properly counsel their patients, but also diligently monitor for adverse events associated with prosthetic implants used for incontinence and prolapse surgeries.
1. Hendrix SL, Clark A, Nygaard I, Aragaki A, Barnabei V, McTiernan A. Pelvic organ prolapse in the Women's Health Initiative: gravity and gravidity. Am J Obstet Gynecol 2002;186:1160–6.
2. Olsen AL, Smith VJ, Bergstrom JO, Colling JC, Clark AL. Epidemiology of surgically managed pelvic organ prolapse and urinary incontinence. Obstet Gynecol 1997;89:501–6.
3. Maher C, Feiner B, Baessler K, Adams EJ, Hagen S, Glazener CM. Surgical management of pelvic organ prolapse in women. The Cochrane Database of Systematic Reviews 2010, Issue 4. Art. No.: CD004014. DOI: 10.1002/14651858.CD004014.pub4.
4. Ogah J, Cody JD, Rogerson L. Minimally invasive synthetic suburethral sling operations for stress urinary incontinence in women. The Cochrane Database of Systematic Reviews 2009, Issue 4. Art. No.: CD006375. DOI: 10.1002/14651858.CD006375.pub2.
5. Webber AM, Walters MD, Piedmonte MR, Ballard LA. Anterior colporrhaphy: a randomized trial of three surgical techniques. Am J Obstet Gynecol 2001;185:1299–304.
6. Nguyen JN, Burchette RJ. Outcome after anterior vaginal prolapse repair: a randomized controlled trial. Obstet Gynecol 2008;111:891–8.
7. Altman D, Väyrynen T, Engh ME, Axelsen S, Falconer C; Nordic Transvaginal Mesh Group. Anterior colporrhaphy versus transvaginal mesh for pelvic-organ prolapse. N Engl J Med 2011;364:1826–36.
8. Sung VW, Rogers RG, Schaffer JI, Balk EM, Uhlig K, Lau J, et al.. Graft use in transvaginal pelvic organ prolapse repair: a systematic review. Obstet Gynecol 2008;112:1131–42.
9. Baessler K, Hewson AD, Tunn R, Schuessler B, Maher CF. Severe mesh complications following intravaginal slingplasty. Obtet Gynecol 2005;106:713–6.
10. Deffieux X, de Tayrac R, Huel C, Bottero J, Gervaise A, Bonnet K, et al.. Vaginal mesh erosion after transvaginal cystocele repair using Gynemesh or Gynemesh-Soft in 138 women: a comparative study. Int Urogynecol J Pelvic Floor Dysfunct 2007;18:73–9.
11. Jia X, Glazener C, Mowatt G, MacLennan G, Bain C, Fraser C, et al.. Efficacy and safety of using mesh or grafts in surgery for anterior and/or posterior vaginal wall prolapse: systematic review and meta-analysis. BJOG 2008;115:1350–61.
12. Chen CC, Ridgeway B, Paraiso MF. Biologic grafts and synthetic meshes in pelvic reconstructive surgery. Clin Obstet Gynecol 2007;50:383–411.
13. Marqulies RU, Lewicky-Gaupp C, Fenner DE, McGuire EJ, Clemens JQ, Delancey JO. Complications requiring reoperation following vaginal mesh kit procedures for prolapse. Am J Obstet Gynecol 2008;199:678.e1–4.
14. Luck AM, Steele AC, Leong FC, McLennan MT. Short-term efficacy and complications of posterior intravaginal slingplasty. Int Urogynecol J Pelvic Floor Dysfunct 2008;19:795–9.
15. U.S. Food and Drug Administration. FDA Public Health Notification: serious complications associated with transvaginal placement of surgical mesh in repair of pelvic organ prolapse and stress urinary incontinence. Available at: www.fda.gov/MedicalDevices/Safety/AlertsandNotices/ucm061976.htm
. Retrieved July 13, 2011.
17. Bump RC, Mattiasson A, Bo K, Brubaker LP, DeLancey JO, Klarskov P, et al.. The standardization of terminology of female pelvic organ prolapse and pelvic floor dysfunction. Am J Obstet Gynecol 1996;175:10–7.
18. Latthe PM, Foon R, Toozs-Hobson P. Transobturator and retropubic tape procedures in stress urinary incontinence: a systematic review and meta-analysis of effectiveness and complications [published erratum appears in BJOG 2007;114:1311]. BJOG 2007;114:522–31.
19. Kaufman Y, Singh SS, Alturki H, Lam A. Age and sexual activity are risk factors for mesh exposure following transvaginal mesh repair. Int Urogynecol J Pelvic Floor Dysfunct 2011;22:307–13.
20. Withagen MI, Vierhout ME, Hendriks JC, Kluivers KB, Milani AL. Risk factors for exposure, pain, and dyspareunia after tension-free vaginal mesh procedure. Obstet Gynecol 2011;118:629–36.
21. Begley S. The best medicine. Sci Am 2011;305: 50–5.