Approximately one in eight American women will undergo surgery to manage pelvic organ prolapse (POP) by the age of 80 years.1 After surgery there is a high recurrence rate of 39%2 and a repeat repair rate of up to 29%.3 The first synthetic, permanent vaginal mesh system for POP repair was reviewed by the U.S. Food and Drug Administration (FDA) in 2001.4 These mesh systems experienced a rapid uptake in practice,5 and an estimated 100 vaginal mesh kits have been approved by the FDA.4 They were marketed as an easy and durable option for POP with little high-quality evidence to support this claim. A 2016 Cochrane review reported lower rates of repeat surgery for POP after transvaginal mesh implantation; however, 7–18% required reoperation for recurrent POP, stress incontinence, or mesh complication.6
Significant concerns regarding the safety of vaginal mesh have been raised by the FDA and Health Canada, and thousands of women have joined class action lawsuits for complications of vaginal mesh.7,8 Reoperation for mesh complications can cause significant morbidity and emotional distress.9,10 In 2016, transvaginal mesh was classified as a high-risk device by the FDA, and some manufacturers have chosen to discontinue their products.11
In response to the concerns raised by regulatory bodies, this study was designed to address the following objectives: 1) measure the incidence of mesh removal or revision after vaginal mesh implantation for the treatment of POP, 2) evaluate the potential association between an individual surgeon's yearly volume of mesh-based POP surgery and mesh removal or revision rates, and 3) determine specific risk factors that may contribute to vaginal mesh removal or revision.
MATERIALS AND METHODS
This study was conducted through the Institute for Clinical Evaluative Sciences according to a prespecified protocol. We performed a population-based, retrospective cohort study of all adult women who underwent vaginal mesh implantation for the correction of POP from April 2002 until December 2013 in the province of Ontario (population of approximately 12 million people with unlimited access to a universal health care system). Study approval was granted through the Research Ethics Board at Sunnybrook Hospital (Toronto, Ontario, Canada).
We used four linked databases to perform this study: The Canadian Institute for Health Information's Discharge Abstract and Same Day Surgery databases (diagnostic and procedural information for patients who are admitted to the hospital or who undergo a surgical procedure), the Ontario Health Insurance Plan (claims for physician services), the Institute for Clinical Evaluative Sciences' Physician Database (physician characteristics), and the Registered Persons Database (demographic data). These data sets were linked using unique encoded identifiers and analyzed at the Institute for Clinical Evaluative Sciences. These four databases were considered complete for all study variables, except for physician specialty, which was unknown for 1.7% of the relevant physicians. Previous work has demonstrated that these databases are reliable and valid.12–14
Canadian Classification of Health Intervention codes, extracted from the hospital databases, were used to define the cohort of adult women who underwent transvaginal implantation of mesh for POP during the study period (Appendix 1, available online at http://links.lww.com/AOG/A808). After the date of the procedure, patients were observed until death, emigration, the first occurrence of an outcome of interest, or the end of the study period (March 31, 2014).
Hospital intervention codes, International Classification of Diseases, 10th Revision codes, and physician billing codes were used to detect the composite outcome of a first reoperation for transvaginal mesh-related complications. These codes included the surgical removal or revision of vaginal mesh or a foreign body, repeat prolapse repair with evidence of a complication, resection of the vaginal wall, or repair of a vaginal fistula (Appendix 2, available online at http://links.lww.com/AOG/A808). Acute surgical complications such as major vessel and adjacent organ injuries15 and subacute complications such as pelvic abscess16 have been previously characterized and were not included as outcomes.
Surgeon procedural volume was specified a priori as the primary exposure of interest. We hypothesized an inverse relationship between surgeon procedural volume and complications. Surgeon volume was defined as the number of vaginal mesh-based POP surgeries performed per year. This was determined using unique physician identifiers within the hospital databases. Quartiles were used to select “high-volume” surgeons (greater than the 75th percentile of vaginal mesh users in a given year) as has been done in other volume outcome studies.17,18 It was possible that the same surgeon could be classified differently over time depending on the operative volume.
Patient characteristics that were included as covariates were: patient age, obesity (body mass index [calculated as weight (kg)/[height (m)]2] 40 or greater), diabetes, general medical comorbidity (health care utilization was used as a proxy for overall health: the Aggregated Diagnostic Groups resource utilization bands from the validated Johns Hopkins University Adjusted Case Groups case-mix system19,20), fiscal year, geographic region of the patient, and socioeconomic status (using the Ontario Marginalization Index, a measure of regional marginalization, which was used as a proxy for individual socioeconomic status21). Surgical covariates included: previous or concurrent hysterectomy, previous or concurrent stress urinary incontinence procedure (mesh-based compared with nonmesh-based operations), previous POP procedure (mesh-based compared with native tissue repair), specialty of the surgeon performing the procedure, academic or nonacademic hospital where the procedure was performed, and need for a blood transfusion postoperatively. Coding definitions are included in Appendix 3, available online at http://links.lww.com/AOG/A808.
Frequencies (count) and medians (interquartile range) are reported. Baseline characteristics were compared between high- and low-volume surgeons using standardized differences. Standardized differences greater than 10% identify potentially meaningful differences.22
The primary analysis was a multivariable survival analysis with robust standard errors to account for patients clustered within surgeons (SAS 9.3). Results are reported as hazards ratios or adjusted hazards ratios with 95% confidence intervals and P values (P<.05 was considered significant). The proportional hazards assumption was assessed using Schoenfeld residuals. In the case of nonproportional hazards, we stratified by the affected variables to confirm the consistency of the results. The Cochrane-Armitage test for linear trends was used to assess for significant changes in the 1-year event rate over time.
Given the low volume of procedures performed by surgeons at the 75th percentile, we conducted two additional analyses: first, we created a new statistical model with a very high-volume surgeon defined as greater than the 90th percentile, and second we assessed the learning curve by comparing the risk of complications from the surgeon's initial 10 cases to all future cases. For the learning curve analysis, surgeons who were operating in the initial year of our study period (fiscal year 2002–2003) were excluded as a result of an inability to be sure when their first case was done; we used logistic regression to model the probability of a complication against a surgeon's number of prior procedures adjusting for the same covariates as the primary model.
We identified 5,448 women who underwent transvaginal mesh implantation for the surgical correction of POP between 2002 and 2013. Median follow-up was 5.4 years (interquartile range 3.0–8.0 years), 5.1% (277/5,448) were censored for death, and 1.5% (80/5,448) were censored for emigration. Women were operated on by 1 of 368 unique surgeons (85% [313/368] of whom were gynecologists). The uptake of transvaginal mesh implantation increased from 2002 to 2008 and then declined steadily through 2012 (Fig. 1). Select baseline patient characteristics of the cohort are shown in Table 1 (with full baseline characteristics shown in Appendix 4, available online at http://links.lww.com/AOG/A808). The median cutoff that defined high-volume surgeons (based on our prespecified definition of greater than the 75th percentile), in each given year of the 10-year study period, was five procedures per year (interquartile range 5–6). High-volume surgeons were more likely to operate in an academic center compared with low-volume surgeons (57.0% [2,187/3,835] compared with 33.5% [540/1,613], standardized difference 0.49). Patients who had mesh implanted by a high-volume surgeon had fewer visits (two visits [interquartile range 1–4] compared with three [interquartile range 1–4], standardized difference 0.18) in the year before their surgery and were more likely to have urodynamic studies performed preoperatively (34.6% [1,325/3,835] compared with 26.5% [427/1,613], standardized difference 0.18). Patients of high-volume surgeons were less likely to have a concurrent hysterectomy (32.5% [1,247/3,835] compared with 37.7% [608/1,613], standardized difference 0.11) and had a shorter postoperative stay in the hospital (2.0 days [interquartile range 2.0–3.0] compared with 3.0 days [interquartile range 2.0–4.0], standardized difference 0.40). Finally, there was regional variation with some provincial regions having a higher proportion of patients treated by a low- or high-volume surgeon.
The proportion of women who underwent mesh removal, revision, or repair of a mesh-associated vaginal fistula was 4.0% (218/5,448) at a median of 1.17 years (interquartile range 0.58–2.90 years) from the original vaginal mesh procedure (Table 2). The same surgeon who placed the mesh also performed the surgery for a mesh complication in 50% of cases. The cumulative incidence of patients experiencing one of the mesh-related complications was 1.9% at 1 year and increased to 5.15% by 10 years of follow-up (Appendix 5, available online at http://links.lww.com/AOG/A808). The rate of reoperation for vaginal mesh-related complications within the first year after implantation increased from 8.8 to 30 per 1,000 person-years between 2002 and 2006 and then decreased, but overall there was no significant linear trend over time (P=.62, supplemental digital content, Appendix 6, available online http://links.lww.com/AOG/A808).
High-volume surgeons performed 70% (3,835/5,448) of the vaginal mesh POP procedures, and low-volume surgeons performed the remaining 30% (1,613/5,448). In the high-volume group, 3.9% (150/3,835) of women with vaginal mesh implanted required a repeat operation for a mesh-related complication, whereas 4.2% (68/1,613) of those in the low-volume group required reoperation (Table 2). The adjusted analysis did not show a significant effect of surgeon volume when defined as greater than the 75th percentile (adjusted hazard ratio 0.95, 95% confidence interval [CI] 0.64–1.39).
Because there was a still a low number of yearly procedures defining the 75th percentile of mesh POP users, we repeated the analysis using the 90th percentile of mesh POP users as a definition of very high volume (representing less than 10 surgeons in any given year). The cutoff for very high volume with this definition was a median of 13 (interquartile range 11–14) procedures per year. Using this new definition, 2,447 of 5,448 (45%) patients were operated on by very high-volume surgeons and 3,001 of 5,448 (55%) by low-volume surgeons. With this definition for very high-volume surgeons, there was a significant decreased risk of surgical mesh complications associated with patients treated by very high-volume surgeons (Table 2; adjusted hazard ratio 0.59, 95% CI 0.40–0.86, P<.01). This represents an absolute risk reduction of 1.85% (95% CI 0.81–2.87%). To prevent one operative intervention for a mesh complication, 54 women would need to be treated by a very high-volume surgeon instead of a low-volume surgeon.
Other factors associated with transvaginal mesh removal or revision were consistent across the analysis, independent of whether the definition of high volume (75th percentile) or very high volume (90th percentile) was used. These included younger age, a concurrent hysterectomy, a blood transfusion, and a higher comorbidity score (Table 3; full results in Appendix 7, available online at http://links.lww.com/AOG/A808).
A multivariable survival analysis using the same covariates, but defining surgeon experience as a learning curve (first 10 cases compared with subsequent cases), as opposed to yearly volume, did not show that there was a significantly increased risk associated with patients being operated on by a surgeon early in their transvaginal mesh experience (adjusted hazard ratio 1.13, 95% CI 0.68–1.86, P=.65, full model not shown). Figure 2 illustrates this learning curve showing a small but continual decrease in the probability of reoperation for vaginal mesh with a surgeon's increasing number of prior transvaginal mesh-based POP procedures.
Among women who underwent transvaginal mesh implantation solely for POP, we found a cumulative incidence rate of mesh complications requiring operation of 5.15% at 10 years. A systematic review in 2009 reported a total reoperation rate of 8.5% (which were predominately for transvaginal mesh kit complications and a smaller proportion of POP recurrences).16 A retrospective cohort study of transvaginal mesh compared with native tissue repairs for the treatment of anterior vaginal wall prolapse (using health care claims data) found a 5-year cumulative risk of 15.2% for any repeat surgery and a 5.9% 5-year risk of mesh revision or removal for mesh-associated complications.23 These results are consistent with our study.
Transvaginal mesh regulatory notifications have recommended that surgeons should obtain specialized training and experience if they wish to perform transvaginal mesh surgery.7,8 We examined this utility of this recommendation by assessing the volume–outcome relationship between mesh-based POP procedures and operative mesh complications. Our original categorization of surgeons (using the 75th percentile) represented a low median number of procedures per year. However, patients of the 90th percentile of mesh implanters (very high-volume surgeons who generally performed 14 or more procedures per year) had a 41% reduced risk of mesh complications requiring surgical intervention and likely better represent a “high-volume” surgeon. Patients of surgeons who had performed less than 10 prior procedures did not have a significantly increased risk of mesh complications requiring reoperation. It is likely that a longer learning curve over a shorter time period is necessary and that the maintenance of procedure-specific operative experience (with a high yearly volume) is the more important factor. Prior research has demonstrated that surgeon experience (determined by the number of years in practice) has a significant effect on transvaginal mesh-associated complications24 and rates of vaginal mesh erosion.25 Our results support the recommendations that surgeons performing transvaginal mesh operations for POP should have advanced training and sufficient experience to minimize mesh complications requiring reoperation.
We identified four risk factors for POP transvaginal mesh complications. Younger patient age was associated an increased risk of surgical management of mesh complications, which is consistent with prior systematic reviews and likely the result of a lower tolerance for conservative management and a larger proportion of women who are sexually active.25,26 Concurrent hysterectomy at the time of mesh implantation is a known risk factor for mesh erosion potentially as a result of devascularization the vaginal cuff, bacterial contamination,25 and tension across the vaginal incisions. Blood transfusion (representing intraoperative blood loss) is also related to mesh erosion,27,28 perhaps attributable to the formation of a vaginal wall hematoma leading to poor healing, or blood loss may be a marker for a difficult dissection, problematic trocar passage, or failure to identify the proper surgical planes. Finally, increased patient comorbidity was associated with reoperation for mesh complications. Although diabetes is a reported risk factor for vaginal mesh erosion29 (and not demonstrated in our study), general medical comorbidity has not been found to be a risk factor in prior studies.30
Our study does have limitations. We used administrative data coding elements to define our cohort and our outcome. These data sources could not identify the specific complication leading to excision or revision. Also, despite including several known covariates, we could not measure parity, smoking status, the type of mesh, which vaginal compartments were repaired, the subspecialty training of the implanting surgeon, postoperative vaginal wound infections, or perioperative estrogen administration with our data sources. These factors have been associated with vaginal mesh exposure in prior studies.25,26 Finally, we relied on the surgical removal or revision of transvaginal mesh as our outcome, because we believe this is likely to represent a significant, symptomatic event. However, our study likely underestimates the true rate of mesh complications, because asymptomatic patients and those managed nonoperatively were not included.
The clinical implications of our study are significant. We found that 1 in 20 women required a second surgery for a transvaginal mesh complications after 10 years of follow-up. Patients of very high-volume surgeons had a 41% reduction in this risk. Our results underscore the recommendations that transvaginal mesh procedures be performed by surgeons with advanced surgical training and sufficient experience. Other clinically important factors associated with a second surgery for a mesh complication (younger age, concurrent hysterectomy, blood transfusion, and greater patient comorbidity) should be used to identify patients most likely to benefit from close postoperative follow-up.
1. Wu JM, Matthews CA, Conover MM, Pate V, Jonsson Funk M. Lifetime risk of stress urinary incontinence or pelvic organ prolapse surgery. Obstet Gynecol 2014;123:1201–6.
2. Fialkow MF, Newton KM, Weiss NS. Incidence of recurrent pelvic organ prolapse 10 years following primary surgical management: a retrospective cohort study. Int Urogynecol J Pelvic Floor Dysfunct 2008;19:1483–7.
3. 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.
4. Vaginal placement of synthetic mesh for pelvic organ prolapse. Committee Opinion No. 513. American College of Obstetricians and Gynecologists. Obstet Gynecol 2011;118:1459–64.
5. Rogo-Gupta L, Rodriguez LV, Litwin MS, Herzog TJ, Neugut AI, Lu YS, et al. Trends in surgical mesh use for pelvic organ prolapse from 2000 to 2010. Obstet Gynecol 2012;120:1105–15.
6. Maher C, Feiner B, Baessler K, Christmann-Schmid C, Haya N, Marjoribanks J. Transvaginal mesh or grafts compared with native tissue repair for vaginal prolapse. The Cochrane Database of Systematic Reviews 2016, Issue 2. Art. No.: CD012079. DOI: 10.1002/14651858.CD012079.
7. Urogynecologic surgical mesh: update on the safety and effectiveness of transvaginal placement for pelvic organ prolapse. Silver Spring (MD): U.S. Food and Drug Administration; 2011. Available at: http://www.fda.gov/downloads/medicaldevices/safety/alertsandnotices/UCM262760.pdf
. Retrieved August 17, 2015.
8. Surgical mesh—complications associated with transvaginal implantation for the treatment of stress urinary incontinence and pelvic organ prolapse—notice to hospitals—recalls & alerts—Healthy Canadians. Ottowa (Canada): Health Canada; 2014. Available at: http://www.healthycanadians.gc.ca/recall-alert-rappel-avis/hc-sc/2014/39475a-eng.php
. Retrieved August 17, 2015.
9. Pickett SD, Barenberg B, Quiroz LH, Shobeiri SA, O'Leary DE. The significant morbidity of removing pelvic mesh from multiple vaginal compartments. Obstet Gynecol 2015;125:1418–22.
10. Dunn GE, Hansen BL, Egger MJ, Nygaard I, Sanchez-Birkhead AC, Hsu Y, et al. Changed women: the long term impact of vaginal mesh complications. Female Pelvic Med Reconstr Surg 2014;20:131–6.
11. FDA strengthens requirements for surgical mesh for the transvaginal repair of pelvic organ prolapse to address safety risks. Silver Spring (MD): U.S. Food and Drug Administration; 2016. Available at: http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm479732.htm
. Retrieved February 27, 2016.
12. Raina P, Torrance-Rynard V, Wong M, Woodward C. Agreement between self-reported and routinely collected health-care utilization data among seniors. Health Serv Res 2002;37:751–74.
13. Williams JI, Young W. A summary of studies on the quality of health care administrative databases in Canada. In: Goel V, Williams JI, Anderson GM, Fooks C, Blackstien-Hirsch P, Naylor CD, editors. Patterns of health care in Ontario. 2nd ed. Toronto (Canada): ICES; 1996:339–47.
14. Jha P, Deboer D, Sykora K, Naylor CD. Characteristics and mortality outcomes of thrombolysis trial participants and nonparticipants: a population-based comparison. J Am Coll Cardiol 1996;27:1335–42.
15. Nguyen JN, Jakus-Waldman SM, Walter AJ, White T, Menefee SA. Perioperative complications and reoperations after incontinence and prolapse surgeries using prosthetic implants. Obstet Gynecol 2012;119:539–46.
16. Diwadkar GB, Barber MD, Feiner B, Maher C, Jelovsek JE. Complication and reoperation rates after apical vaginal prolapse surgical repair: a systematic review. Obstet Gynecol 2009;113:367–73.
17. Segal J, Sacopulos M, Sheets V, Thurston I, Brooks K, Puccia R. Online doctor reviews: do they track surgeon volume, a proxy for quality of care? J Med Internet Res 2012;14:e50.
18. McAlister FA, Youngson E, Bakal JA, Holroyd-Leduc J, Kassam N. Physician experience and outcomes among patients admitted to general internal medicine teaching wards. CMAJ 2015;187:1041–8.
19. Weiner J. The Johns Hopkins ACG system reference manual, version 8.2. 2008. Available at: http://www.ensolution.se/upl/files/14429.pdf
. Retrieved February 27, 2015.
20. Austin PC, van Walraven C, Wodchis WP, Newman A, Anderson GM. Using the Johns Hopkins Aggregated Diagnosis Groups (ADGs) to predict mortality in a general adult population cohort in Ontario, Canada. Med Care 2011;49:932–9.
21. Matheson FI, Dunn JR, Smith K, Moineddin R, Glazier RH. Development of the Canadian Marginalization Index: a new tool for the study of inequality. Can J Public Health 2012;103(suppl 2):S12–6.
22. Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput 2009;38:1228–34.
23. Jonsson Funk M, Visco AG, Weidner AC, Pate V, Wu JM. Long-term outcomes of vaginal mesh versus native tissue repair for anterior vaginal wall prolapse. Int Urogynecol J 2013;24:1279–85.
24. Barski D, Otto T, Gerullis H. Systematic review and classification of complications after anterior, posterior, apical, and total vaginal mesh implantation for prolapse repair. Surg Technol Int 2014;24:217–24.
25. Deng T, Liao B, Luo D, Shen H, Wang K. Risk factors for mesh erosion after female pelvic floor reconstructive surgery: a systematic review and meta-analysis. BJU Int 2016;117:323–43.
26. Shah HN, Badlani GH. Mesh complications in female pelvic floor reconstructive surgery and their management: a systematic review. Indian J Urol 2012;28:129–53.
27. Sirls LT, McLennan GP, Killinger KA, Boura JA, Fischer M, Nagaraju P, et al. Exploring predictors of mesh exposure after vaginal prolapse repair. Female Pelvic Med Reconstr Surg 2013;19:206–9.
28. Gold KP, Ward RM, Zimmerman CW, Biller DH, McGuinn S, Slaughter JC, et al. Factors associated with exposure of transvaginally placed polypropylene mesh for pelvic organ prolapse. Int Urogynecol J 2012;23:1461–6.
29. Frankman EA, Alperin M, Sutkin G, Meyn L, Zyczynski HM. Mesh exposure and associated risk factors in women undergoing transvaginal prolapse repair with mesh. Obstet Gynecol Int 2013;2013:926313.
30. 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.