We assessed 109 reports of 60 potentially relevant trials; 31 reports of 12 trials were eligible for inclusion with a total of 1473, predominantly female, participants. In four trials not all participants with overactive bladder, in four trials had urinary incontinence. Data from eight trials with 858 participants with urinary incontinence at baseline, mostly female, are therefore included in the review. The quality of trials was variable. Few data describing long term follow up are available.
Bladder training compared to no bladder training: Data were available for 172 women from three trials comparing bladder training with no bladder training. These described only a limited number of prespecified outcomes, which varied across the three trials. Point estimates of effect favoured bladder training; however, confidence intervals were wide and no statistically significant differences were found for primary outcome variables.
Bladder training compared to other treatments: Three trials including 159 women compared bladder training with drugs: two with oxybutynin and one with imipramine plus flavoxate. In the former trials the only outcomes demonstrating a statistically significant difference were participant's perception of cure at six months (RR 1.69; 95% CI 1.21 to 2.34), quality of life (general physical measure) (WMD 9.00; 95% CI 1.64 to 16.36) and adverse events, all favouring bladder training, and number of daytime micturitions per week (WMD 2.80; 95% CI 0.91 to 4.69) favouring drug treatment. In the latter trial participant's perception of cure immediately after treatment just achieved statistical significance (RR 1.50; 95% CI 1.02 to 2.21) favouring bladder training, and this difference was maintained at approximately two months post treatment. Two comparisons of bladder training with pelvic floor muscle training plus biofeedback included 164 women: none of the differences in the primary outcomes achieved statistical significance.
Combining bladder training with another treatment compared to that other treatment alone: Two trials including 331 participants compared the combination of bladder training plus an anticholinergic drug with the drug alone. For the largest trial, data for only one prespecified outcome were available: the median number of incontinent episodes was the same for both treatment groups. One trial compared pelvic floor muscle training plus biofeedback supplemented with bladder training versus pelvic floor muscle training plus biofeedback alone and included 125 women. Of the primary outcomes, both participants' perception of improvement and quality of life, both immediately after treatment, achieved statistical significance, favouring the bladder training combined with pelvic floor muscle training and biofeedback group (perception of improvement: RR 1.18; 95% CI 1.01 to 1.39; quality of life: MD -47.20; 95% CI -87.03 to -7.37), this was not sustained at three months.
