Asthma hospital admission and readmission spikes, advancing accurate classification to advance understanding of causes

Abstract

Abstract Background: Days of high asthma admissions (HAADs) or readmissions (HARDs) create substantial acute burdens on hospital systems. An important component of asthma care is understanding potential causes of these admission spikes with potential of mitigation of risk. Crucial to this research is accurately distinguishing these events from background seasonal changes and time trends. To date, classification methods have been based on ad hoc and un-tested definitions which may hamper understanding of causes due to HAADs and HARDs misclassification. The aim of this article is to introduce an easily applied robust statistical approach, demonstrated to have high classification accuracy in other settings. It is referred to as the Seasonal Hybrid Extreme Studentized Deviate (S-H-ESD) method.Methods: We demonstrate S-H-ESD on a time series between 1996 to 2009 of all daily paediatric asthma hospital admissions from the Victorian Admitted Episodes Data set (VAED).Results: The S-H-ESD method clearly identified HAADs and HARDs without applying ad-hoc classification definitions, while appropriately accounting for seasonality and time trend. Furthermore, this was able to be done with statistical testing to provide evidence in support of their identification. Conclusion: The S-H-ESD is useful and statistically appropriate for the accurate classification of HAADs and HARDS. This method removes the need for ad-hoc approaches and presents as a means of systemizing the accurate classification and detection of high asthma admission days. This will strengthen synthesis and efficacy of research in order to understand causes of high asthma hospital admission and readmission.