Ward–Hilton: Tension-Free Vaginal Tape versus Colposuspension for Primary Stress Incontinence
Ward–Hilton was an open-label, multicentre randomized comparison of retropubic tension-free vaginal tape (TVT) and colposuspension in women undergoing first surgery for urodynamic stress incontinence; six-month objective cure did not differ significantly, while TVT brought faster recovery and more bladder injuries.[1]
- Design and population
- Multicentre randomized comparative surgical trial; six-month primary report. Women with urodynamic stress incontinence who had completed childbearing and were selected for first continence surgery at 14 UK and Irish centres.
- Arms
- Retropubic tension-free vaginal tape (TVT) versus colposuspension using each centre's standard technique
- Report and follow-up
- 2002 · Six months after surgery; missing outcomes counted as failures in the randomized denominator
| Endpoint | Results | Effect | Finding |
|---|---|---|---|
| Six-month negative one-hour pad test plus negative cystometric stress test; risk difference in percentage points (TVT minus colposuspension) | 115/175 (66%) vs 97/169 (57%) | difference 8.3; CI -4.7–21.3; P 0.099 | Not met |
Harms: Bladder injury, including perforation or visible trauma: 15/170 (9%) vs 3/146 (2%); Vaginal perforation: 5/170 (3%) vs 0/146; Urinary tract infection within six weeks: 38/170 (22%) vs 46/146 (32%)
Registry not recorded · Report DOI
Design
The 14 UK and Irish centres enrolled women selected for surgery after conservative treatment who had completed childbearing. Exclusions included detrusor overactivity, prolapse requiring treatment, prior incontinence or prolapse surgery, substantial voiding dysfunction, neurological disease and allergy to local anaesthetic. Gynaecologists, urogynaecologists and urologists participated; each trained in TVT at a recognised centre before enrolling patients. Allocation used central telephone randomization. Participants, surgeons and postoperative assessors knew the assignment.[1]
The six-month primary endpoint required both a negative one-hour pad test and a negative cystometric stress test. Missing outcomes were counted as failures in the randomized denominator. The sample-size plan treated a 10-percentage-point cure difference as clinically important, but recruitment stopped below the planned 197 women per arm. The follow-up reports used a negative pad test alone as their primary endpoint, so their cure percentages have a different definition.[1][2][3]
Results
At six months, combined objective cure occurred in 115/175 women assigned TVT (66%) and 97/169 assigned colposuspension (57%): an 8.3-percentage-point raw difference (reported 95% CI −4.7 to 21.3; P = 0.099). A negative pad test alone occurred in 128/175 (73%) and 109/169 (64%). Complete patient-reported continence under all circumstances was less common than objective cure: 63/175 (36%) and 48/169 (28%).[1]
In the two-year report, objective cure with missing data counted as failure was 63% of women assigned TVT and 51% assigned colposuspension (OR 1.67, 95% CI 1.09–2.58).[2]
At five years, the negative pad-test rate among women tested was 58/72 (81%) after TVT and 44/49 (90%) after colposuspension (OR 0.47, 95% CI 0.16–1.4; P = 0.21). A last-observation-carried-forward sensitivity analysis of all randomized women gave 131/175 (75%) and 117/169 (69%) (OR 1.32, 95% CI 0.82–2.12). On the symptom questionnaire, 38/98 (39%) and 36/79 (46%) reported no leakage under any circumstance.[3]
Surgery and treatment delivery
TVT was placed by the retropubic Ulmsten technique under local anaesthesia and sedation. The bladder was not routinely catheterized afterward; an injury prompted temporary catheter drainage. Colposuspension followed each centre's standard technique, with suture number, material and postoperative catheterization standardized across centres; it was usually performed under general anaesthesia. Assigned operations were performed in 170/175 TVT and 146/169 colposuspension participants, with more women withdrawing before colposuspension.[1]
Among operated women, median theatre time was 40 versus 50 minutes, postoperative stay 1 versus 5 days, and return to normal activity 3 versus 6 weeks, respectively. Patient knowledge of the assigned operation may have influenced the reported recovery interval.[1]
Adverse events
Bladder perforation or visible trauma occurred in 15/170 TVT operations (9%) versus 3/146 colposuspensions (2%). There were five vaginal perforations after TVT and none after colposuspension. One TVT passage injured an anomalous obturator artery, requiring laparotomy and transfusion. Urinary tract infection within six weeks affected 38/170 (22%) and 46/146 (32%).[1]
By five years, further stress-incontinence surgery had occurred in 4/170 TVT recipients and 5/146 colposuspension recipients (3.4%); prolapse surgery in 3/170 (1.8%) and 11/146 (7.5%). Six TVT recipients had tape-related complications: three vaginal erosions, one obstructed voiding case requiring tape division, one suprapubic extrusion, and one intravesical tape removed by cystotomy. The intravesical finding and two vaginal erosions were first detected at five years.[3]
Limitations
The primary trial recruited fewer women than planned for its prespecified cure difference, and its confidence interval included clinically important effects in either direction. More patients assigned colposuspension withdrew before surgery. The five-year extension included 11 of the original 14 centres; pad-test data were available for only 72 TVT and 49 colposuspension participants. The different endpoint definitions and substantial missing follow-up affect comparisons across reports. Colposuspension followed each centre's usual operation within the protocol's suture and catheterization rules. Ethicon supported the trial coordinator and supplied materials and support to participating centres.[1][3]
In conversation
Same setting SISTEr (2007)
SISTEr also used Burch colposuspension as a comparator, but tested an autologous rectus fascial pubovaginal sling in a different surgical population and used different success measures.
Guidelines and approvals
The EAU 2026 guideline offers open or laparoscopic colposuspension to women seeking surgical management of stress urinary incontinence after discussion of relative risks and benefits . It also offers a midurethral sling after that discussion . Its evidence discussion reports the higher long-term vault and posterior-compartment prolapse findings from the TVT–colposuspension comparison.[4]
Report history
| Published | Report | Cutoff | Headline results | Citation |
|---|---|---|---|---|
| 2002-07-13 | Six-month randomized report · primary | — | Six-month negative one-hour pad test plus negative cystometric stress test; risk difference in percentage points (TVT minus colposuspension): 115/175 (66%) vs 97/169 (57%); difference 8.3; CI -4.7–21.3; P 0.099 Not met Six-month negative one-hour pad test: 128/175 (73%) vs 109/169 (64%); — Six-month BFLUTS patient-reported complete continence: 63/175 (36%) vs 48/169 (28%); — | Ward K, Hilton P; UK and Ireland TVT Trial Group. "Prospective multicentre randomised trial of tension-free vaginal tape and colposuspension as primary treatment for stress incontinence." *BMJ.* 2002;325(7355):67. doi:10.1136/bmj.325.7355.67 DOI |
| 2004-02-01 | Two-year update · updated | — | Two-year objective cure, defined as a negative one-hour pad test (less than 1 g weight change); participants with missing data counted as failures: 63% of 175 randomized vs 51% of 169 randomized; OR 1.67; CI 1.09–2.58 | Ward KL, Hilton P; UK and Ireland TVT Trial Group. "A prospective multicenter randomized trial of tension-free vaginal tape and colposuspension for primary urodynamic stress incontinence: two-year follow-up." *Am J Obstet Gynecol.* 2004;190(2):324-331. doi:10.1016/j.ajog.2003.07.029 DOI |
| 2007-10-26 | Five-year extension · updated | — | Five-year negative one-hour pad test, less than 1 g gain, among participants with available tests: 58/72 (81%) vs 44/49 (90%); OR 0.47; CI 0.16–1.4; P 0.21 Not met Five-year sensitivity analysis using last observed pad test for missing data: 131/175 (75%) vs 117/169 (69%); OR 1.32; CI 0.82–2.12; P 0.28 Five-year BFLUTS patient-reported complete continence among questionnaire respondents: 38/98 (39%) vs 36/79 (46%); — | Ward KL, Hilton P; UK and Ireland TVT Trial Group. "Tension-free vaginal tape versus colposuspension for primary urodynamic stress incontinence: 5-year follow up." *BJOG.* 2008;115(2):226-233. doi:10.1111/j.1471-0528.2007.01548.x DOI |
References
1. Ward K, Hilton P; UK and Ireland TVT Trial Group. "Prospective multicentre randomised trial of tension-free vaginal tape and colposuspension as primary treatment for stress incontinence." BMJ. 2002;325(7355):67. doi:10.1136/bmj.325.7355.67
2. Ward KL, Hilton P; UK and Ireland TVT Trial Group. "A prospective multicenter randomized trial of tension-free vaginal tape and colposuspension for primary urodynamic stress incontinence: two-year follow-up." Am J Obstet Gynecol. 2004;190(2):324-331. doi:10.1016/j.ajog.2003.07.029
3. Ward KL, Hilton P; UK and Ireland TVT Trial Group. "Tension-free vaginal tape versus colposuspension for primary urodynamic stress incontinence: 5-year follow up." BJOG. 2008;115(2):226-233. doi:10.1111/j.1471-0528.2007.01548.x
4. European Association of Urology. EAU Guidelines on Management of Non-neurogenic Female Lower Urinary Tract Symptoms (LUTS). Limited update, March 2026. Section 4.2.4.c.2.a.