Reliability and validity of a psycho-social problem inventory for childhood epilepsy

In response to the need for formal psychosocial assessment in childhood epilepsy, a 25-question inventory was constructed from raw items based on multidisciplinary inputs and tested for practicality, and psychometric attributes. Weighted inter-rater and test-retest reliability estimates were 61.8% and 63.9% respectively. A measure of concurrent validity was satisfactory though a modest discrepancy was observed between two raters (P = 0.019 and 0.0064). Discriminant validity was very satisfactory (average P 0.001) based on compared means of ratings from two groups of subjects with disparate prognosis. The inventory should be useful in the assessment of similar chronic childhood disabilities but interpretations of data should be in a wider clinical context including inputs of informant and school authorities as indicated.