Development, reliability and validation of a neurogenic bowel dysfunction score in pediatric patients with spina bifida.

Abstract

To develop a reliable and valid questionnaire to monitor neurogenic bowel symptoms in children. Thirty-four children aged 6-18 with neurogenic bowel and their caregivers. Eighteen control patients. An expert panel generated a domain of observables and formative/reflective content. Response options were scaled following Likert-type items. Key informant interviews revised the measures. A final questionnaire was given to patients twice to calculate intra-rater reliability using Cohen's Kappa Coefficient (k) and paired t-test. Blinded interviews were conducted after physical examination and health assessment and questionnaires completed by a nurse to determine construct validity and inter-rater reliability using k and Spearman's rank-order correlation. Control patients completed the questionnaire once, their results were used to determine discriminate validity and a receiver operating characteristic (ROC) curve. Intra-rater reliability showed 85% of the questionnaires having k >0.6. Paired t-test results of t(33) = 1.997, P = 0.054, d = 0.53, confirmed there was not a significant difference between the scores of the two completed questionnaires. Inter-rater reliability showed 97% of the questionnaires having k >0.6 between the nurse and the patient/caregiver responses. Scores had a strong positive correlation at rs (32) = 0.943, P < 0.0005. Mean score with neurogenic bowel was 15.18(STD ± 5.77) and control group 4.68(STD ± 2.98). ROC analysis showed an area under the curve of 0.9. A score of 8.5 correlated with presence of neurogenic bowel with sensitivity of 94% and specificity of 87%. The questionnaire shows positive reliability and validity when used for pediatric neurogenic bowel patients. The questionnaire differentiates between normal and neurogenic patients. Larger studies are necessary to conduct further validation.