John Mathai1, Peter Anderson, Angela Bourne. 1. Royal Children's Hospital Mental Health Service, Murdoch Children's research Institute, University of Melbourne, Flemington Road, Parkville, Victoria 3052, Australia. john.mathai@rch.org.au
Abstract
OBJECTIVE: To examine the level of agreement between clinical diagnoses by a community child and adolescent mental health service (CAMHS) and diagnoses generated by the Strengths and Difficulties Questionnaire (SDQ). METHOD: A computerized algorithm developed by Goodman et al. was used at a community CAMHS in Australia to predict child psychiatric diagnoses on the basis of the symptom and impact scores derived from the SDQ completed by parents (n = 130), teachers (n = 101) and young people (n = 38). These diagnoses were compared with the diagnoses made by clinicians in a multidisciplinary community outpatient team and an independent clinician that examined the case notes and was blind to the SDQ scores. RESULTS: The level of agreement between SDQ generated diagnoses and clinical team diagnoses was moderate to high, ranging from 0.39 to 0.56. Correlations between the SDQ and an independent clinician ranged from 0.26 to 0.43. All were statistically significant. CONCLUSION: The SDQ is a useful instrument to aid clinicians in diagnosis and could be used as part of the initial assessment process.
OBJECTIVE: To examine the level of agreement between clinical diagnoses by a community child and adolescent mental health service (CAMHS) and diagnoses generated by the Strengths and Difficulties Questionnaire (SDQ). METHOD: A computerized algorithm developed by Goodman et al. was used at a community CAMHS in Australia to predict childpsychiatric diagnoses on the basis of the symptom and impact scores derived from the SDQ completed by parents (n = 130), teachers (n = 101) and young people (n = 38). These diagnoses were compared with the diagnoses made by clinicians in a multidisciplinary community outpatient team and an independent clinician that examined the case notes and was blind to the SDQ scores. RESULTS: The level of agreement between SDQ generated diagnoses and clinical team diagnoses was moderate to high, ranging from 0.39 to 0.56. Correlations between the SDQ and an independent clinician ranged from 0.26 to 0.43. All were statistically significant. CONCLUSION: The SDQ is a useful instrument to aid clinicians in diagnosis and could be used as part of the initial assessment process.