Materials and Methods
Study setting
In Denmark, the health care system is tax financed and provides health care free of charge for all residents. The initial medical contact is with the general practitioner, who may refer the patient to a public or private hospital.
Data source
Because the Danish unique civil registration number (CPR number) enables linkage between all nationwide registers, the entire Danish population can be considered a cohort.
The Danish National Patient Registry was established in 1977 and provides information on diagnoses, minor procedures, and operations of inpatients, outpatient, and emergency room visits in Danish hospitals. It is mandatory by Danish law for all Danish hospital departments and private hospitals to report data to the Danish National Patient Registry. The registry is used for administration, quality of care, and research. Studies of procedure codes registered with the Danish National Patient Registry have shown a high validity.
The Danish Civil Registration System was established in 1968 and provides information on gender, date of birth, and continuously updated data on vital status. As a result, patients can be tracked over time with accurate accounting for censoring because of emigration or death.
Subjects
The cohort comprised all female patients recorded in the Danish National Patient Registry as having undergone surgical treatment for UI from 1998 through 2007 and with no surgery for UI 2 years prior to enrollment in the study. This period of 2 years of no exposure was inserted to ensure that no women with recent surgery for UI were included ( Figure 1 ). If 2 or more procedures were registered on the same first-time date of surgery, the observations were excluded from the cohort.
Definition and outcome
The primary objective was to evaluate the cumulative incidence of reoperation after different surgical procedures for UI within 5 years. For all operations we used the Nordic Medico-Statistical Committee procedure codes, which were divided into 6 groups ( Appendix A ): (1) retropubic midurethral tape (code KLEG10); (2) transobturator tape (code KLEG10A); (3) urethral injection therapy (codes KKDV20 and KKDV22) with polyacrylamide gel or polyacryl hydrogel; (4) pubovaginal slings (code KKDG30), which is an a.m. McGuire procedure performed with autologous fascia (rectus fascia or fascia lata); (5) Burch colposuspension (code KDG00); and (6) miscellaneous operations that separately were less frequently used procedures for UI (codes KDG01, KDG10, KDG31, KDG40, KDG50, KDG96, KDG97, KLEG00, KLEG20, and KLEG96). A reoperation was defined as any subsequent surgical treatment for UI.
Statistical analysis
Women with or without reoperation were compared by the χ 2 test and the χ 2 trend test (categorical variables) or the Student’s t test (age). The start date was set at baseline surgery and an outcome was a reoperation within the following 5 years. A Kaplan-Meier curve was used for measuring the time to reoperation for the 6 groups of surgical treatment. Information on vital status was obtained using the CPR number, and data were censored before time if the women disappeared, emigrated, or died within the 5 year follow-up period.
A Cox proportional regression hazard model assessed the hazard ratio (HR) with 95% confidence intervals (CIs) for each type of procedure (retropubic midurethral tape as the reference group), adjusting for age, department volume, and calendar effect.
To adjust for the department volume, we calculated the annual number of UI procedures for each department and computed tertiles from these. Because we included all surgical procedures for UI since the implementation of the synthetic MUS, we took the initial learning curve for retropubic midurethral tape into account by adjusting for each year (1998–2007) as well as for both periods (1998–2002 and 2003–2007) as a measure of calendar effect. Only results for adjustment for each year are presented because no differences were observed between the 2 modes of adjustment.
Data analysis was performed using Stata version 13.0 (StataCorp, College Station, TX).
Approval
This study was approved by the Danish Data Protection Agency (number 2013-41-2210). Because the study did not include patient contact, it was not necessary to obtain approval from the Health Research Ethics Committee.
Results
A total of 8671 women (mean age 56.1 years, ±12.6) underwent surgical treatment for UI from 1998 through 2007. Among these, 888 women (10%) were reoperated within a 5 year period ( Table 1 ). The lowest rate of reoperation was observed among women who had pubovaginal slings (6%), retropubic midurethral tape (6%), and Burch colposuspension (6%) followed by transobturator tape (9%) and miscellaneous operations (12%), whereas the highest observed risk was for urethral injection therapy (44%).
| Complete cohort (n = 8671) | No reoperation (n = 7783) | Reoperation (n = 888) | P value | |
|---|---|---|---|---|
| Age, y, mean ± SD | 56.1 (12.6) | 55.9 (12.4) | 58.4 (14.4) | < .001 a |
| Procedures performed, volume | ||||
| High, % | 5931 (101) b | 5249 (89) | 682 (12) | < .001 c |
| Medium, % | 2095 (100) | 1927 (92) | 168 (8) | |
| Low, % | 645 (100) | 607 (94) | 38 (6) |
b Because of rounding, percentages exceed 100
c P value was based on a χ 2 test statistic of a 3 × 3 frequency table of department volume for surgical procedures for urinary incontinence at baseline.
At baseline, women subsequently undergoing reoperation were significantly older than women not being reoperated (mean age, 58.4 vs 55.9 years) ( P < .001). However, the difference was present only for women operated with transobturator tape (58.3 vs 53.8 years, P < .004) and urethral injection therapy (64.4 vs 61.7 years, P < .009).
The number of women who underwent reoperation was significantly increasing from low-volume (6%) over medium-volume (8%) to high-volume departments (12%) ( P for trend < .001). We stratified this by the 6 groups of surgical treatments, and only women with urethral injection therapy had a higher frequency of reoperation increasing with department volume, whereas no differences were observed for the other treatment modalities.
In the first period (1998–2002), the proportion of retropubic midurethral tapes (35%) was almost equal to Burch colposuspensions (28%), but in the second period (2003–2007), the synthetic MUS had replaced Burch colposuspension in the surgical treatment of UI (81% vs 2%, Table 2 ). At low-volume departments, transobturator tape (29%) was more frequently used compared with high-volume departments (13%), whereas urethral injection therapy was more rarely used.
| Department Volume, % | Period 1998–2002 | Period 2003–2007 | ||||||
|---|---|---|---|---|---|---|---|---|
| Low (n = 221), medium (n = 519), and high (n = 1820) Total (n = 2560) | Low (n = 424), medium (n = 1576), and high (n = 4111) Total (n = 6111) | |||||||
| Retropubic midurethral tape, % | 56 (25) | 102 (20) | 731 (40) | 889 (35) | 289 (68) | 1178 (75) | 2674 (65) | 4141 (68) |
| Transobturator tap, % a | — | — | — | — | 122 (29) | 149 (10) | 519 (13) | 790 (13) |
| Urethral injection therap, % | 4 (2) | 3 (1) | 129 (7) | 136 (5) | 4 (1) | 132 (8) | 607 (15) | 743 (12) |
| Colposuspension, % | 116 (53) | 240 (46) | 372 (20) | 728 (28) | 1 (0) | 46 (3) | 56 (1) | 103 (2) |
| Pubovaginal slings, % | 1 (1) | 55 (11) | 178 (10) | 234 (9) | 0 (0) | 57 (4) | 110 (3) | 167 (3) |
| Miscellaneous, % | 44 (20) | 119 (23) | 410 (23) | 573 (22) | 8 (2) | 14 (1) | 145 (4) | 167 (3) |
a Blank cells: transobturator tape was first implemented in 2003 in Denmark, and therefore, no data appear in these cells (1998–2003).
The risk of reoperation was determined by the Kaplan-Meier curves for the 6 groups of treatments ( Figure 2 ). The majority of reoperations occurred within the first 2 years of the primary operation and then leveled off during the remaining 3 years. The median time to reoperation was 1 year for sling surgery (retropubic midurethral tape, transobturator tape, and pubovaginal slings), 2 years for Burch colposuspension, and 6 months for urethral injection therapy.
Among women with urethral injection therapy, 30% had repeat UI surgery within the first year and 14% had repeat UI surgery from 2 to 5 years.
Because 4 women emigrated before their primary operation, 8667 women were included in the Cox proportional hazard model. A total of 368 women (4%) were censored before time because of death (n = 345), emigration (n = 22), or disappearance (n = 1).
After adjusting for age, department volume, and calendar effect, the risk of repeat surgery was almost 12-fold higher for urethral injection therapy, and for transobturator tape, the risk of reoperation was significantly higher in comparison with retropubic midurethral tape (HR, 2.1; 95% CI, 1.5-2.9) ( Table 3 ). There was virtually no difference between the crude and adjusted HRs.
| Hazard ratio (95% CI), reoperation | Adjusted hazard ratio (95% CI), reoperation a | |
|---|---|---|
| Procedures | ||
| Retropubic midurethral tape | (Reference) | (Reference) |
| Transobturator tape | 1.7 (1.3–2.3) | 2.1 (1.5–2.9) |
| Urethral injection therapy | 10.7 (8.9–12.8) | 11.5 (9.3–14.3) |
| Colposuspension | 1.3 (0.9–1.8) | 1.4 (1.0–2.0) |
| Pubovaginal slings | 1.1 (0.8–1.9) | 1.2 (0.7–1.9) |
| Miscellaneous | 2.2 (1.6–2.8) | 2.1 (1.5–2.8) |
a Adjusted for age, department volume, and calendar effect (1998–2002 and 2003–2007).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
