Development of the Japanese version care planning assessment tool (J-CPAT): Reliability and validity studies

Abstract

Aim: To develop a Japanese version Care Planning Assessment Tool (CPAT), a comprehensive geriatric assessment instrument for dementia care, originally developed in Australia. Methods: The J-CPAT is composed of 61 items, and 8 domains:Communication, Physical problems, Self-help skills, Confusion, Behavior, Social interaction, Psychiatric observation, and Care dependency. The development of the J-CPAT included translation into Japanese, assessment of item comprehension, back translation, production of final version, and its validity and reliability analysis. Through this process, an item about family interaction was added in the J-CPAT. The participants were 199 clients in residential and day care. Raters were professional carers with 2.5 hours training about the J-CPAT. We assessed the validity of the J-CPAT, using correlation between the scores of the J-CPAT, MMSE, NM-scale, N-ADL, and levels of care needs (Kaigodo). Ten pairs of carers were included in the inter-rater reliability analysis. Fourteen carers participated in the intrarater reliability study. Results: Cronbach's alpha values in each J-CPAT domain were 0.74-0.95. Mean difference of each domain between test and retest was 0.4-3.6%. Weighted kappa values for all items for 10 pairs of raters were over 0.6. The correlation coefficients between the domain score of 'Confusion' in the J-CPAT and MMSE was -0.90 (<0.01). Those between the domain scores of 'Physical problems', 'Self-help skills', 'Dependency on care' in the J-CPAT, and Kaigodo were sufficient (>0.62), and those between the item scores of the J-CPAT, NM-scale, and N-ADL were relevant (>0.60). Conclusions: The J-CPAT is considered to be an appropriate assessment instrument for dementia care in Japan.