Fig. 2

Imbalances in confounders of mortality by influenza infection status before and after propensity score matching. Based on the matched sample, there was no evidence of any difference in hospital mortality across groups (OR = 1.01, 95%CI 0.90–1.13, p = 0.85). We developed a propensity score (PS) logistic model to have flu then matched the individuals on the basis of their PS using a 1:1 matching algorithm without replacement within a caliper of 0.15 standard deviation of the logit of the propensity score. To handle missing values in confounders, multiple imputation with chained equation was used for the PS model, where propensity score for each patient was averaged across 30 completed datasets while propensity score matching used these averaged scores to estimate the treatment effect. Only 1 patient with influenza could not be matched. Imbalances in confounders were reduced after matching