UCSF

Aims

To develop a clinically relevant, easy to use, and validated instrument for assessing severity and bother related to urinary incontinence.

Methods

Survey items were piloted and refined following psychometric principles in five separate patient cohorts. Patient and expert endorsement of items, factor analyses, Spearman rank correlations and response distributions were employed for item selection. Formal reliability and validity evaluation were conducted for the final questionnaire items.

Results

Expert physicians and patient focus groups confirmed face and content validity for the measure. A 10-item measure called the Michigan Incontinence Symptom Index (M-ISI) was developed with two domains: a Total M-ISI Domain consisting of subdomains for stress urinary incontinence, urgency urinary incontinence, and pad use, and a Bother Domain. High construct validity was demonstrated with a Cronbach's alpha for the Total M-ISI Domain (items 1-8) of 0.90 and for the Bother Domain (items 9-10) of 0.82. Cronbach's alpha for the subdomains were all > 0.85. Construct validity, convergent and divergent validity, internal discriminant validity, and predictive validity were all robust. The minimally important difference for the measure was determined to be 4 points (out of 32) for the Total M-ISI Severity Domain, and 1-2 points (out of 8-12) for the individual subdomains.

Conclusions

The M-ISI is a parsimonious measure that has established reliability and validity on several levels and complements current clinical evaluative methods for patients with urinary incontinence.