Validation of a clinical algorithm to diagnose stress urinary incontinence for large studies.

PURPOSE: We assessed the accuracy with which a clinical diagnostic algorithm for stress urinary incontinence (SUI) based on symptoms and signs without including urodynamics predicted the observation of urodynamic SUI and the condition of SUI. Such an algorithm would have applicability to ensure that the majority of women entering large SUI clinical trials would have urodynamic SUI and/or the condition of SUI without performing urodynamic testing in all subjects.

MATERIALS AND

METHODS

A total of 1455 women with urinary incontinence at least 3 months in duration were enrolled in 3 randomized clinical trials (1 phase 2 and 2 phase 3 trials) of duloxetine vs placebo for the treatment of SUI in Europe and North America. Subjects were identified as having SUI based on a clinical algorithm that required a predominant symptom of SUI with a weekly incontinence episode frequency of 4 or greater (phase 2 study), or 7 or greater (phase 3 studies), absent predominant urge symptoms, normal diurnal and nocturnal frequency, a bladder capacity of 400 ml or greater, and a positive cough stress (sign of SUI) and stress pad test. Multichannel urodynamic studies were performed in a subset of 184 women at 23 study sites. Of these urodynamic tracings 173 (94%) were of adequate quality to make an assessment of the filling phase and assign a diagnosis of urodynamic SUI, detrusor overactivity or a normal filling phase. Two-sided 95% exact CIs for the proportions were calculated.

RESULTS

The clinical algorithm had a positive predictive value of 90.2% for urodynamic SUI with or without detrusor overactivity and 76.9% for urodynamic SUI only (pure urodynamic SUI). The positive predictive value for the condition of pure SUI was 85.0%, while for the condition of SUI in pure and mixed forms the positive predictive value was 98.3%. Patient age, previous continence surgery or the severity of incontinence did not influence algorithm accuracy.

CONCLUSIONS

The algorithm is suitably feasible and sufficiently predictive to be used in large clinical trials designed to evaluate conservative treatment for women with SUI. It ensures that the overwhelming majority of the study population would have urodynamic SUI and the condition of SUI.