Characterizing the impact of snowfall on patient attendance at an urban emergency department in Toronto, Canada.

OBJECTIVES: We sought to determine whether addition of a snowfall variable improves emergency department (ED) patient volume forecasting. Our secondary objective was to characterize the magnitude of effect of snowfall on ED volume. METHODS: We used daily historical patient volume data and local snowfall records from April 1st, 2011 to March 31st, 2018 (2542 days) to fit a series of four generalized linear models: a baseline model which included calendar variables and three different snowfall models with an indicator variable for either any snowfall (>0 cm), moderate snowfall (≥1 cm), or large snowfall (≥5 cm). To evaluate model fit, we examined the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). Incident rate ratios were calculated to determine the effect of snowfall in each model. RESULTS: All three snowfall models demonstrated improved model fit compared to the model without snowfall. The best fitting model included a binary variable for snowfall (<1 cm vs. ≥1 cm). This model showed a statistically significant decrease in daily ED volume of 2.65% (95% CI: 1.23%-4.00%) on snowfall days. DISCUSSION: The addition of a snowfall variable results in improved model performance in short-term ED volume forecasting. Snowfall is associated with a modest, but statistically significant reduction in ED volume.