Is Barthel Index a relevant measure for measuring prevalence of urinary incontinence in stroke patients?

AIMS To compare the prevalence of urinary incontinence (UI) measured by Barthel Index and the Danish Prostate Symptom Score (DAN-PSS-1) questionnaire in stroke patients. METHODS A cross-sectional, hospital based survey was initiated whereby 407 stroke patients, average age 67 (SD 12) years with a mean interval of 101 days since onset of last stroke at inclusion. The stroke patients self-reported the prevalence of UI by Barthel Index and the DAN-PSS-1 questionnaire. RESULTS The prevalence of UI was 10.5% measured by the Barthel Index and 49% by the DAN-PSS-1 questionnaire. Furthermore, the DAN-PSS-1 questionnaire distinguished between types of UI (38% urge UI, 28% stress UI, and 14% other UI). There were significantly fewer stroke patients who reported UI by the Barthel Index compared to the DAN-PSS questionnaire in term of urge UI (P < 0.001), stress UI (P < 0.001), and other UI (P < 0.001). CONCLUSIONS The results showed that DAN-PSS-1 questionnaire contains more specific UI information compared to the Barthel Index. Fewer stroke patients reported UI by the Barthel Index compared to the DAN-PSS-1 questionnaire and moreover the DAN-PSS-1 questionnaire enables identification of different UI-types. The Barthel Index is not sensitive to the UI prevalence; instead the authors suggest using a more specific instrument such as the DAN-PSS-1 questionnaire.