Structure of the Health of the Nation Outcome Scales for Children and Adolescents: An ordinal factor analysis of clinician ratings of a sample of young people referred to community mental health services

Abstract

The Health of the Nation Outcome Scales for Children and Adolescents (HoNOSCA) have been implemented as performance and research measures for youth mental health services in the UK and elsewhere. However, the factor structure of this instrument has not yet been reported. Clinician ratings from a representative sample of 1335 young people seen by community-based Child and Adolescent Mental Health Services were randomised into exploratory and confirmatory datasets. Parallel analysis followed by an exploratory factor analysis was performed on the former ratings. The potential structures elicited were tested using confirmatory factor analyses (CFA). The resulting structures were evaluated in the confirmatory dataset. On CFA a five factor model with one crossloading item provided the best fit to the data. However, the internal reliability consistency of some of the postulated subscales was poor. We conclude that the HoNOSCA is a multidimensional measure with a three factor structure possibly providing the optimum compromise between fit and internal reliability. Thus, it may be most appropriate to summarise clinical outcomes using two symptom subscores, relating to behavioural and emotional problems, rather than using the total scores, as is current practice. Such subscales may be more sensitive to change than total summed scores.