Introduction
Many surgical procedures have been developed for the treatment of GSI but there is little consensus as to the most effective. Although meta-analyses have suggested that colposuspension is associated with the best surgical results (1, 2), there is only limited data from randomised trials on which to base clinical practice. The TVT (Gynecare, Edinburgh) is a recent development for the treatment of GSI and has potential advantages as a short stay, short convalescence procedure. Cure rates of over 80% have been consistently reported (3). Aims of study A a multi centred prospective randomised trial comparing TVT with colposuspension as a primary treatment for GSI.

Methods
The trial was conducted in 14 centres in the UK and Ireland and included urologists and gynaecologists, university and district general hospitals. Women with urodynamically proven GSI were randomised to TVT or colposuspension. None had detrusor instability (DI), major voiding problems, prolapse requiring treatment or previous surgery for incontinence or prolapse. TVT was performed under local anaesthesia and sedation.(3). Colposuspension was performed under general or regional anaesthesia. Assessment performed prior to treatment and at six months post operatively included the following: Bristol female lower urinary tract symptoms questionnaire (BFLUTS), SF-36 health survey, Euroqol, 1 week urinary diary, 1 hour pad test, filling and voiding cystometry and urethral pressure profilometry (in some centres).

Results
344 patients were recruited to the study and randomised. There were 28 withdrawals prior to surgery, 23 in the colposuspension group and 5 in the TVT group. 170 underwent TVT and 146 colposuspension, as per protocol. The two groups were comparable in terms of age, BMI, parity, menopausal status. At six months, 63 (40%) of TVT patients and 48 (38%) of the colposuspension group reported no urinary leakage whatever. However, 103 (66%) and 90 (71%) patients respectively, reported subjective cure of stress incontinence. 115 (68%) of the TVT group and 97 (66%) of the colposuspension group were objectively cured, based on a negative 1 hour pad test, together with no leakage seen on CMG. On CMG alone, the numbers cured were 142 (89%) for TVT and 114 (85%) for colposuspension. The median blood loss was 50 ml and 135ml for TVT and colposuspension, respectively. Bladder perforation occurred in 15 (9%) the TVT group and in 3 (2%) following colposuspension. Opiate analgesia usage was significantly lower following TVT (21%) than colposuspension (91%). The mean duration of hospital stay was 2.2 days for TVT and 6.5 for colposuspension. 9 patients were rehospitalised in the TVT group and 18 in the colposuspension group. There was no significant increase in the findings of uterovaginal prolapse and no evidence of an increase in symptoms of dyspareunia or prolapse in either group during the follow up period. 12 (7%) patients in the TVT group and 13 (9%) in the colposuspension group developed DI post operatively, however symptomatic urgency that was bothersome was only reported as a new symptom by 3 patients. Six (7%) patients in each group had evidence of voiding disorder. In one patient the mesh was found exposed in the vagina at follow-up; this was managed simply by trimming of the mesh and closure of the overlying vaginal mucosa.

Discussion
The definition of cure used in this study is stricter than that used in other studies in the literature, hence the cure rates for both procedures appear low. On the basis of leakage demonstrated at cystometry and subjective reporting of stress incontinence, cure rates are comparable with previous series. The cure rate at six months on this definition shows no significant difference between the two procedures. The trial was designed to compare the two procedures rather than to demonstrate equivalence. The number of patients recruited was less than initially expected and equivalence cannot therefore be assumed. Some of the initial concerns about the TVT procedure have centred on bladder perforation rate, vascular injury, tape erosion and de novo detrusor instability. Although 15 perforations are reported here, none of these led to any adverse effects. Overall blood loss was less with TVT than colposuspension, although one patient in the TVT group had a significant vascular injury. Whilst both procedures show an incidence of de novo DI, bothersome urgency as a new symptom appears uncommon. TVT shows promise as a minimal access, rapid recovery procedure for the treatment of GSI. Cure rates and adverse events appear to be comparable with colposuspension. The follow up within this protocol will continue for a minimum of two years. KW was supported by a grant from Johnson and Johnson, who also provided materials and additional support to collaborating centres.

1. BJU 1996;78(4):497-510.
2. BJOG 1994;101(5):371-374.
3. IUGJ and PFD 1996;7(2):81-85.