PURPOSE OF THE STUDY.
To develop and internally validate an effective clinical prediction rule regarding need for hospitalization in pediatric patients presenting to the emergency department with acute asthma exacerbations.
Children (n = 928) aged 5 to 17 years (mean age 8.8 years, 61% boys, 59% African American) presenting to a tertiary children’s hospital emergency department with an acute asthma exacerbation between April 2008 and February 2013 were included using a prospective convenience design.
Fifteen predictor variables including demographics, asthma symptom history and control, pulmonary examination findings, and measures of lung function and inflammation were evaluated for the development of an asthma prediction rule. Penalized maximum likelihood estimation logistic regression models were used to evaluate the association between the need for hospitalization and hospitalization decision by the clinical team treating the patient. A bootstrapping algorithm was implemented to assess for internal validity.
Of the 15 predictor variables evaluated, oxygen saturation on room air (odds ratio 2.4, 95% confidence interval 1.9–3.1) and inspiratory-to-expiratory ratio (odds ratio 1.9, 95% confidence interval 1.1–3.1) correlated most with need for hospitalization. Oxygen saturation on room air, intercostal retractions, inspiratory-to-expiratory ratio, and wheezing were variables most associated with the hospitalization decision of the clinical team.
The asthma prediction rule is an internally validated model for predicting need for hospitalization in children with acute asthma exacerbations presenting to the emergency department. Further studies are needed to assess the external validity and impact on patient care and clinical outcomes of this tool.
This study provides a useful clinical prediction tool that may expedite hospitalization decision-making in the emergency department of children presenting with acute asthma exacerbations. A limitation of the study is that it was done in a specialized setting (academic, urban); external validation in other settings is needed before the impact on patient care and resource utilization can be determined. The take-home point for the clinician is that the key predictor variables of the model (oxygen saturation and expiratory phase prolongation) are fortunately readily perceptible at the bedside.
- Copyright © 2015 by the American Academy of Pediatrics