Performance of the Asthma Clinical Score and the Pediatric Respiratory Assessment Measure in obese children presenting to the emergency department for asthma exacerbations

Our primary objective was to evaluate the interrater reliability (IRR), responsiveness and predictive validity of two asthma severity scores in obese and normal weight (NW) children. Our secondary objective was to compare asthma interventions in these groups. This was a planned secondary analysis of a prospective observational study of children aged 2-to- < 18 years treated for asthma exacerbations in the emergency department. Obesity was defined as >95^th% sex-specific weight-for-age. Clinicians completed paired assessments of the Asthma Clinical Score (ACS) and Pediatric Respiratory Assessment Measure (PRAM). IRR was analyzed using Cohen’s kappa coefficient. Reliable change index was used to determine the responsiveness. Predictive validity of the pretreatment ACS and PRAM was analyzed using Receiver Operating Characteristics and Akaike Information Criterion. The likelihood of receiving an intervention was described using odds ratios (OR). 399 children were enrolled, 111 classified as obese. In obese children, the ACS showed moderate IRR, while the PRAM showed weak IRR (κ_w 0.73 and 0.57 respectively). Both scores better detected responsiveness in NW versus obese children (ACS 34% vs 23%, PRAM 28% vs 3.7% respectively) but were better predictors of hospitalization in obese children (ACS 0.85, vs 0.74 and PRAM 0.79 vs 0.72 respectively). Obese children had higher odds of receiving intravenous methylprednisolone, continuous albuterol, and non-invasive ventilation. Performance of the ACS and PRAM varied between obese and NW children. Future studies should aim to fully elucidate the performance and utility of asthma severity scores in obese children.