Validation of a bowel dysfunction instrument for adolescents with spina bifida

Abstract

Existing survey instruments for bowel dysfunction in the pediatric population are either parent-reported or focus on non-neurogenic bowel dysfunction. The purpose of this study was to develop and validate an adolescent-reported survey to assess the severity of bowel dysfunction in spina bifida patients and examine its impact on quality of life (QOL). We performed a cross-sectional study of patients in our Myelodysplasia Program, aged 11-17 years, with a history of constipation and/or fecal incontinence (FI) from November 2010 to June 2013. Control patients, aged 11-17 years, were recruited from the stone clinic. Exclusion criteria were lack of English fluency, insufficient reading skills, or an incontinent fecal diversion. A 29-item version of the Adolescent Fecal Incontinence and Constipation Symptom Index (A-FICSI) was developed with five domains (Figure). Test re-test reliability and correlation with the total global health-related QOL score from the Parkin survey were measured using the Pearson correlation coefficient. A factor analysis model with four-fold correlations was tested. Of the 65 study-eligible individuals approached, 25 (11 boys) completed the A-FICSI (median age 12.6 years, IQR 11.8-14.7 years) and 17 completed another survey on urinary incontinence (64.6% response rate). Twenty-one control patients with nephrolithiasis completed the A-FICSI. Nine of the 25 completed a second administration of the survey. The mean correlation between repeated administrations of the survey was r = 0.43. There was a significant negative correlation between severity of constipation (r = -0.299, p < 0.05) and severity of FI (r = -0.316, p < 0.05) with the total Parkin QOL score. The Comparative Fit Index (CFI) was 0.864. Most items loaded significantly on their respective factors. Between-factor correlations were all significant (>0.30) in the predicted direction. Unstandardized residuals were 8.7% (95% CI 6.4-10.9%). Item reduction was performed on the 29-item instrument based on results of the factor analysis. The finalized instrument contained 21 items. This is the first adolescent-reported bowel dysfunction instrument to undergo formal psychometric assessment in the spina bifida population. The instrument demonstrated adequate reliability and the five-factor structure fit the data well. This study highlights the negative impact of bowel dysfunction on the QOL of spina bifida patients. It is limited by the moderate sample size that is a common issue in relatively rare diseases. The A-FICSI possesses desirable psychometric properties for the measurement of bowel dysfunction in the spina bifida population.