Abstracts

Rationale: The impact of Seizure Monitoring Units (SMU) on health resource utilization has been reported in individual centres, but data for geographic regions or populations are sparse. Using data linkage in a large health region, we investigated the temporal trend in utilization of health care services before and after admission to a SMU in patients with epilepsy (PWE).Methods: The Calgary SMU provides specialized epilepsy care for a population of ~1.3 million, and systematically collects clinical data on all admissions. Alberta Health Services (AHS), the single health care provider, collects administrative health data on hospital-based inpatient and outpatient care, including emergency services. We linked SMU clinical data with administrative datasets to assess use of hospital-based services pre and post admission to SMU in PWE. Health care utilization data abstracted for SMU patients included hospitalizations, emergency visits, hospital-based ambulatory clinics and day surgeries. Databases used included: the AHS Discharge Abstract Database (hospitalizations) and the AHS hospital-based ambulatory care dataset (emergency visits, hospital-based ambulatory clinics and day surgeries). Health care services were categorized as either emergency or non-emergency. Mean and standard deviation of health care services per patient were calculated per visit type and by year using SMU admission year as the index date to determine if health care utilization changed after admission to the SMU.Results: The 4-year SMU database contained linkable health services utilization records on 408 patients. Before SMU admission, the 5-year annual average (SD) of non-emergency and emergency health services encounters was 8.3(9.9) and 3.5(5.0), respectively; the corresponding figures for the index year were 11.4(10.9) and 2.3(2.5) respectively, and the average for the 4-year period following admission fell to 6.9(8.8) and 1.9(1.6), respectively. Tukey adjusted pair-wise comparisons indicate a significant peak in non-emergency utilization on the index year and significant differences pre and post SMU admission, for emergency and non-emergency encounters (all p < 0.001).Conclusions: In preliminary analyses, SMU admissions were associated with a reduction of emergency and non-emergency hospital-based health service utilization, suggesting that investigation and care in the SMU results in improved health and more efficient health resource use in these patients.