Criterion validity, test-retest reliability and sensitivity to change of the St George Urinary Incontinence Score

To assess the criterion validity, test-retest reliability and the sensitivity to change after treatment of the St George Urinary Incontinence Score (SGUIS).

Women presenting with urinary incontinence completed two SGUIS tests one week apart. A frequency-volume chart (FVC) was completed before the first attendance and the women had a 1-h pad test before treatment. Patients were treated conservatively under the care of a urogynaecologist or nurse continence advisor, or surgically with an open or laparoscopic colposuspension. After treatment, the SGUIS, FVC and 1-h pad test were repeated.

The SGUIS correlated moderately well with the number of leaks per week (Spearman's r = 0.610, 95% confidence interval 0.516-0.689, P < 0.001) but less well with the 1-h pad test loss (r = 0.257, 0.124-0.380, P = 0.002). The test-retest reliability was acceptable, as the mean (SD) difference between the first and second SGUIS was 0.337 (2.675), with limits of agreement of -5.012 to -5.686. The change in the SGUIS after treatment correlated well with the improvement in the number of leaks per week (r = 0.742, 0.662-0.805, 156 samples, P < 0.001) but not as well with the change in 1-h pad test loss (r = 0.531, 0.405-0.636, 151, P < 0.001), although the trend was similar.

The criterion validity of the SGUIS appeared adequate for both the number of leaks per week on the FVC and leakage on the 1-h pad test. The statistical reproducibility of the test was adequate and appeared more responsive to change after treatment than the other measures. The score is suitable for rapid self-administration by patients with a range of incontinence types, unlike most other currently available test instruments.