LO64: Variation in Alberta emergency department patient populations

Abstract

Introduction: Increasing pressures on the health care system, particularly in emergency departments (EDs), make it critical to understand changing ED case-mix, patient demographics and care needs, and resource utilization. Our objective is to assess Alberta (AB) ED volumes, utilization and case mix, stratified by ED type. This knowledge will help identify opportunities for system change and quality improvement. Methods: Data from Alberta Health Services administrative databases, including the National Ambulatory Care Reporting System, ED Admission/Discharge/Transfer data, and Comprehensive Ambulatory Care Classification System codes, were linked for all ED visits from 2010-17. Data were stratified by seven facility categories: tertiary referral (TR), regional referral (RR), community<5,000 inpatient discharges (CL), community>600 inpatient discharges (CM), community <600 inpatient discharges (CS), community ambulatory care (CA), and free-standing EDs (FS). Results: We analyzed 11,327,258 adult patient visits: 13% at TR, 34 % at RR, 24% at CL, 16% at CM, 9% at CS, 1% at CA, and 3% at FS sites. Acuity was highest at TR and RR hospitals, with 76%, 63%, 25%, 26%, 22%, 12% and 55% of patients falling into CTAS levels 1-3, for TR, RR, CL, CM, CS, CA, and FS respectively. Admission rates were highest at TR and RR hospitals, (23%, 13%, 5%, 5%, 4%, 0% and 0%), as were left without being seen rates, (5%, 4%, 1%, 2%, 1%, 0% and 5%). The most common ICD-10 diagnoses were chest pain/abdominal pain in TR and RR centres, and IV (antibiotic) therapy in all levels of community and FS EDs. Conclusion: Acuity and case-mix are highly variable across ED categories. Acuity, admission rates and LWBS rates are highest in TR and RR centres. Administrative data can reveal opportunities for health system re-engineering, e.g. potentially avoidable IV antibiotic visits. Further investigation will clarify the type of ED care provided, variability in resource utilization by case-mix, and allocation, and will help identify the optimal metrics to describe ED case-mix.