Objective
We sought to assess interrater reliability of the International Continence Society (ICS)/International Urogynecological Association (IUGA) classification system of vaginal mesh-related complications and compare this with several other available complication classification systems.
Study Design
This was a retrospective analysis of mesh-related complications in patients presenting after pelvic organ prolapse or incontinence surgery. The complications were classified by 2 independent reviewers using the ICS/IUGA classification system as well as 3 other available classification systems. Interrater reliability was assessed using percent agreement and the weighted κ statistic.
Results
The ICS/IUGA mesh complication classification system was found to have poor interrater reliability (κ = 0.15-0.78). The other systems yielded a κ that ranged from 0.18-0.60, but were too general or could only be applied to 68% of the complications.
Conclusion
The complexity of the ICS/IUGA mesh complication system, the large number of categories, and lack of clarity likely contribute to its poor interrater reliability.
Approximately 11% of women in the United States will require surgical intervention for either pelvic organ prolapse (POP) or stress urinary incontinence by age 80 years. Of these women, up to 29% will undergo repeat surgery for symptom recurrence. Increasingly, the placement of synthetic mesh during repair has been used in hopes of reducing such recurrence. Current evidence suggests that while in some circumstances the use of such mesh may reduce objective symptom recurrence when compared to native tissue repair only, complications appear to increase. Common complications include mesh erosion, chronic pelvic pain, dyspareunia, infection, intraoperative perforations, infection, and fistula formation.
In October 2008, the Food and Drug Administration issued a public health warning after receiving >1000 voluntarily reported transvaginal mesh (TVM)-related complications in the preceding 3 years. In July 2011, an update was issued based on an additional 2500 reported complications. The update stated that the complications from mesh for transvaginal repair of POP are not rare and included specific guidelines regarding a thorough discussion of potential risks before consenting patients for pelvic floor surgeries involving the placement of synthetic mesh. To provide an accurate risk/benefit assessment, more data are needed regarding the nature, management, and morbidity of mesh-related complications. In one step toward this goal, the International Urogynecological Association (IUGA) and International Continence Society (ICS) have developed a classification system for complications related to the insertion of prostheses and grafts in pelvic floor surgery that is intended to standardize terminology and stratification of complications. As this new system was empirically derived, a timely evaluation of its clinical applicability and utility is warranted. The objective of our study was to assess interrater reliability of the ICS/IUGA classification system of mesh complications and compare this with the reliability of several other complication classification systems.
Materials and Methods
This was a retrospective analysis of mesh-related complications in patients who presented to the tertiary referral center after a surgery for POP and/or stress urinary incontinence that involved the placement of synthetic mesh. The study population consisted of all patients presenting to our institution for management of a complication related to the use of synthetic mesh in the pelvis on or after Jan. 1, 2006.
The patients were identified using the electronic medical records based on Current Procedural Terminology and International Classification of Diseases, Ninth Revision codes ( Table 1 ). The medical records were then reviewed independently by 2 investigators and the subjects were selected based on the inclusion and exclusion criteria. The investigators underwent extensive training in the use of all the classification systems used in the study. Patients aged ≥21 years who received either outpatient or inpatient management for complications relating to mesh placement were included in the study. Qualifying procedures included synthetic tapes or slings for stress urinary incontinence, vaginal mesh placed via a prepackaged mesh kit, free vaginal mesh (ie, nonkit), and synthetic mesh placed during sacral colpopexy. Patients who had undergone a biologic graft placement were excluded from the study.
| CPT |
| 57267 Insertion of mesh or other prosthesis for repair of pelvic floor defect 57295 Revision or removal of prosthetic vaginal graft (vaginal approach) 57296 Revision or removal of prosthetic vaginal graft (abdominal approach) 57426 Revision or removal of prosthetic vaginal graft (laparoscopic approach) 57287 Revision or removal of sling for stress incontinence |
| ICD-9 |
| 623.2 Vaginal stricture 625.0 Dyspareunia 625.5 Pelvic pain syndrome 625.9 Pelvic pain unspecified 719.45 Pain, joint, pelvic region 729.6 Foreign body in soft tissue 936 Foreign body in intestine or colon 938 Foreign body in alimentary tract 939.0 Foreign body in bladder or urethra 939.2 Foreign body vulva or vagina 939.9 Foreign body in genitourinary tract 959.9 Foreign body 996.30 Mechanical complication of genitourinary device implant and graft 996.65 Infection and inflammatory reaction due to genitourinary device, implant, or graft 996.76 Mesh erosion V58.32 Removal of suture |
Medical records of patients who met the eligibility criteria were reviewed and demographic data, medical history, information about the index surgery, nature of the mesh complication(s), and management of the mesh complication(s) were collected. The investigators then classified the patient’s mesh complication(s) according to the IUGA/ICS, TVM, Dindo et al, and Accordion classification systems. The TVM classification system is specific to vaginal mesh complications. It includes 3 categories, which include the following complications: infection, vaginal exposure or erosion, and periprosthetic retraction. Because some of these complications may be seen together, the system allows multiple classifications. Dindo et al and Accordion systems are based on the therapeutic consequences of surgical complications irrespective of the type of surgery or surgical specialty. Accordion severity grading system consists of 2 components: expanded and contracted. Contracted version is designed for application to the smaller studies, while expanded version contains additional categories and is meant for use in larger studies, particularly if they involve complex procedures. To ensure proper adherence to the various classification systems’ rules and directions, both reviewers were required to review the publication materials for each classification system. Reviewers were allowed to refer to the written instructions for each complications system when classifying complications. The charts were reviewed independently and the reviewers were blinded to each other’s assessment.
The ICS/IUGA classification system contains 3 domains : organ involved/severity, time, and site. The first domain contains 26 categories; for our analysis, categories 1A-3D were grouped as vaginal complications; 4A-4C as urologic complications; 5A-5D as bowel complications; 6A-6D as skin and/or musculoskeletal complications; and 7A-7C as patient compromise ( Figure ).
Data are presented as percent agreement and interrater reliability for each group, as well as for time and site domains of the ICS/IUGA system was assessed using a weighted κ with 95% confidence intervals (CIs). Weighted κ statistic is a more robust measure of interrater agreement than a simple percent agreement calculation since it takes into account the agreement occurring by chance. The κ is always ≤1. A value of 1 implies perfect agreement and a value of 0 indicates agreement that is no greater than chance.
Results
A total of 1236 patients were identified as potential subjects using the Current Procedural Terminology and International Classification of Diseases, Ninth Revision codes listed in Table 1 . We identified 133 eligible patients with 195 mesh-related complications. The mean age at the time of presentation was 55.3 years (SD ± 11.9). The majority of patients were Caucasian (96%). The median time from the index surgery to presentation with symptoms was 8 months (6 days-3.5 years). The median follow-up time after treatment for the mesh complication was also 6 months (0 days-3.8 years).
Seven (5.3%) patients had mesh-related complications after a sacrocolpopexy; in 54 (40.6%), after a sling procedure; in 65 (48.9%), in relation to the placement of TVM; and in 7 (5.3%), from sling and TVM when placed during the same procedure. Of subjects, 69% had their mesh procedure performed at an outside hospital and were referred to our center for management of their complication. We identified 77 patients with mesh erosions, 75 with lower urinary tract complications, 19 with bowel complications, and 4 with vesicovaginal fistulas. Pelvic pain and/or dyspareunia were listed for 55.6% of women with mesh-related complications. Of patients, 26% presented with vaginal bleeding or bothersome discharge.
When applying the ICS/IUGA classification, there was disagreement between the reviewers in the number of complications that should be classified in 29% of subjects. Three (2.2%) complications could not be classified into one of the organ/severity categories and for 42 (38%) complications the location could not be classified by ≥1 reviewers because of lack of clarity of the system.
ICS/IUGA categories 1A-3D were grouped as vaginal complications. The percent agreement between the 2 observers for this category was 44.09%, with κ of 0.32 (95% CI, 0.21–0.45) ( Table 2 ). For urologic complications (categories 4A-4C) the percent agreement was 96.1%, with κ of 0.78 (95% CI, 0.55–1.01). For bowel complications (categories 5A–5C), there was 100% agreement. No musculoskeletal injuries as specified by the ICS/IUGA criteria were identified by either reviewer and therefore the agreement could not be calculated for this category. Classification of the organ/severity domain by each reviewer is demonstrated in Table 3 . The percent agreement on the classification of time was 47.6%, with κ of 0.28 (95% CI, 0.17–0.39). The reviewers agreed on the classification of the complication site 29.7% of the time, with κ of 0.15 (95% CI, 0.07–0.23) ( Table 2 ).
| Category | Agreement (%) | κ (95% CI) |
|---|---|---|
| Organ/severity | ||
| Vaginal (1A-3D) | 44.1 | 0.32 (0.21–0.44) |
| Urologic (4A-4C) | 96.1 | 0.78 (0.55–1.01) |
| Bowel (5A-5D) a | 100 | NA |
| Skin and/or MSK (6A-6D) a | NA | NA |
| Patient compromise (7A-7C) a | NA | NA |
| Time | 47.6 | 0.28 (0.17–0.39) |
| Site | 29.7 | 0.15 (0.07–0.23) |
| Reviewer 1 | 1A | 1B | 1C | 1D | 2A | 2B | 2C | 2D | 3A | 3B | 3C | 3D | 4A | 4B | 4C | 5A | 5B | 5C | Insufficient information a |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Reviewer 2 | |||||||||||||||||||
| 1A | 1 | 1 | 1 | ||||||||||||||||
| 1B | 1 | 17 | 1 | 1 | 1 | ||||||||||||||
| 1C | |||||||||||||||||||
| 1D | |||||||||||||||||||
| 2A | 1 | 1 | |||||||||||||||||
| 2B | 2 | 5 | 1 | 4 | 7 | 1 | 6 | ||||||||||||
| 2C | |||||||||||||||||||
| 2D | |||||||||||||||||||
| 3A | 1 | 3 | 1 | 1 | |||||||||||||||
| 3B | 3 | 1 | 5 | 5 | 12 | 2 | 1 | 2 | |||||||||||
| 3C | 1 | 1 | |||||||||||||||||
| 3D | |||||||||||||||||||
| 4A | |||||||||||||||||||
| 4B | 46 | ||||||||||||||||||
| 4C | 2 | ||||||||||||||||||
| 5A | |||||||||||||||||||
| 5B | 2 | ||||||||||||||||||
| 5C | |||||||||||||||||||
| Insufficient information | 3 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
