Comparison of methods used to quantify general practice-type patients in the emergency department: A tertiary paediatric perspective.

Objective: The aim of this study was to compare methods for quantifying general practice-type (GP-type) patients in a paediatric ED to determine the robustness of current identification methods. Methods: Data was sourced from the ED Information System for 5 years (2010-2014) from the tertiary paediatric hospital in Perth, Western Australia, Australia. A further detailed analysis of 2014 presentations established an independent sample for valid statistical comparison of the three methods used to identify GP-type patients: Australian Institute of Health and Welfare, ACEM and the Diagnosis. Results: A total of 348 020 patients presented in 2010-2014, with the percentage of GP-type patients identified varying from 38% to 68% by the three methods. Fewer GP-type presentations occurred over each consecutive year and had significantly lower median medical consultation times and ED length of stay when compared with the total presentation sample. GP-type presentations were evenly spread across the days and times of the week, with ACEM demonstrating only fewer presentations during office hours (P = 0.02). In 2014 there was only overlap in 36.6% cases for all three methods -6.9% exclusively flagged by ACEM, 22.7% by Australian Institute of Health and Welfare and 0.2% by the Diagnosis method. Conclusions: We demonstrated that current models to quantify perceived GP-type presentations were inconsistent in paediatric patients. All methods flagged a high proportion of GP-type presentations, and although the reasons for these presentations are multifactorial and not easily explained by access to GP services alone, they do represent a significant workload in the paediatric ED.