Short Report: Measuring the Behavioral Disturbance of Autism Spectrum Disorder by Cohen-Mansfield Agitation Inventory
Sirkku Pullinen, Tiina Luukkaala, Tommi Salokivi, Nina Bjelogrlic
Introduction: Cohen-Mansfield Agitation Inventory (CMAI) has originally been developed for the
evaluation of agitation in Alzheimer’s disease. Reliable tools for measuring behavioral symptoms in
autism spectrum disorder (ASD) are missing.
Methods: The reproducibility of the CMAI was evaluated in 26 patients. The CMAI-questionnaire
was filled for each participant by three different nurses. The measurement was repeated by the primary
caregiver for 15 patients. Agreement between nurses was measured using Cohen’s weighed kappa,
intraclass correlation coefficient (ICC) was used to evaluate intra-rater reliability and Cronbach’s alphas
to quantify internal consistency.
Results: The inter-rater reliability was found to be substantial for all items ICC 0.64 (95%CI 0.43-0.80).
Cronbach’s alpha was good for the whole CMAI score (α=0.87 with 95% CI 0.82-0.91). No statistically
significant differences were seen in the CMAI-scores between repeated measurements.
Conclusions: The CMAI may be a reasonably reliable tool to follow-up changes in behavioral symptoms
of ASD.
