Accuracy in reporting incontinence in adults with spina bifida: a pilot study

Abstract

Background: The accuracy of self-reported urinary incontinence (UI) and fecal incontinence (FI) among adults with spina bifida (SB) is unknown. We aimed to quantify the accuracy of self-reported recall incontinence in the last 4 weeks using prospective diary data. Methods: Adults with SB were enrolled via patient advocacy groups in a larger 30-day smartphone-based ecological momentary assessment study of daily well-being and incontinence. We examined agreement between yes/no questions on exit questionnaires ("In the last 4 weeks, did you leak any urine and get your underwear, pads, pull-ups or disposable underwear wet?") and 30-day diaries. Collected data included a non-validated 4-item UI Negativity scale (UIN: 0-100, 0 = no impact). Inter-rater reliability was assessed with Cohen's kappa (>0.60 = substantial). FI was analyzed similarly. Sub-groups were too small for statistical analysis. Results: Median age of 88 adults was 35 years old (70% female, 53% shunted, 71% community ambulators). Among 81 adults reporting UI in the last month, 79 (98%) had diary-documented UI (UI agreement), 2 (2%) did not (Summary Table). Among 7 adults reporting no UI in the last month, 5 (71%) did not record UI in their diaries (agreement in no UI), 2 (29%) did. Both adults reporting no UI on exit questionnaires, despite contrary diary data, experienced single UI episodes (negativity: 0 and 6). In contrast, the UI agreement group recorded UI on median 18 days (median negativity: 22). Sensitivity of the recall UI question was 0.98 (specificity 0.71, kappa 0.69). Among 66 adults reporting FI in the last month, 65 (98%) had diary-documented FI (FI agreement), 1 (2%) did not. Among 22 adults reporting no FI in the last month, 17 (77%) did not record FI in their diaries (agreement in no FI), 5 (23%) did. Five adults reporting no FI on exit questionnaires, despite contrary diary data, experienced a median single episode (median negativity: 8). In contrast, the FI agreement group recorded FI on median 5 days (median negativity: 31). Sensitivity of the recall FI question was 0.93 (specificity 0.94, kappa 0.81). Results were unchanged when analyzing the first and last 28 days of data. Comment: Sensitivity/specificity of single incontinence questions approach those reported for women with UI but without SB. Diaries may best serve SB adults with bothersome incontinence. Conclusions: Self-reported incontinence accurately and reliably captures UI and FI among adults with SB. It may minimize less bothersome incontinence, supporting its use in screening for clinical practice and research.