Asthma severity scores for preschoolers displayed weaknesses in reliability, validity, and responsiveness

Objective: To evaluate the measurement properties of asthma severity scores for use in preschool children. Methods: A Medline search was used to identify published asthma severity scores for use in preschool children. The measurement properties of the scores (item development, reliability, validity, responsiveness, and usability) were evaluated using a published framework. Results: Ten asthma severity scores were identified, with 19 different clinical variables used as items. Interrater agreement was assessed by five scores. Only two scores-Clinical Asthma Score (CAS) and Respiratory Distress Assessment Index (RDAI)-reported good agreement based on weighted kappa-statistics (0.64-0.90). Construct validity was reported by the CAS, Clinical Asthma Evaluation Score (CAES), the Clinical Symptom Grading System (CSGS), and the Preschool Respiratory Assessment Measure (PRAM). Correlation coefficients between asthma severity scores and clinical measures (length of stay, drug closing interval, O-2 saturation, health professional assessment, PaO2, PaCO2) ranged from 0.47 to 0.70. Responsiveness was formally demonstrated for two scales (PRAM, CAS). Conclusions: Most asthma severity scales for use in preschool children have been informally developed. Recently developed scores (CAS, PRAM) have more rigorously evaluated their measurement properties. Research is needed to directly compare the asthma severity scores developed for use in preschool children. (C) 2004 Elsevier Inc. All rights reserved.