Construct validity of the incontinence severity index

Aims: To assess the construct validity of the incontinence severity index (ISI) by testing its correlation with two health-related quality of life measures, the short forms of the Urogenital Distress Inventory (UDI-6) and the Incontinence Impact Questionnaire (IIQ-7), in women with urodynamic stress incontinence. Materials and Methods: A cohort of 170 women with the urodynamic stress incontinence who underwent corrective surgery completed the ISI, IIQ-7, and the UDI-6 both pre- and post-treatment. We correlated the pre- and post-treatment responses between the ISI, the IIQ-7, the UDI-6 and their subscales. We also assessed the sensitivity of the ISI to change by correlating the percent change in score between the three instruments. The results were analyzed using a non-parametric test of correlation, the Spearman's rho. Results: The ISI scores were generally not well correlated with the pre-treatment IIQ-7 and UDI-6 scores (r < 0.40). The post-treatment scores and percent change from pre- to post-treatment of the ISI, however, were highly correlated with that of the IIQ-7 and UDI-6 (r's > 0.70, P < 0.001). The ISI was most highly correlated with the UDI-6 stress symptoms subscale and most poorly correlated with the UDI-6 obstructive/discomfort subscale. Conclusions: This study provides valuable insight into the construct validity of the ISI. Evidence of its convergent validity is found in the high correlation with the stress symptoms subscales of the UDI-6, while the poorer correlation with the obstructive subscale provides evidence of its divergent validity. These data also suggest that the ISI is highly sensitive to change seen with treatment.