Data Analysis
Descriptive statistics were conducted for the sample overall and by hospital, with mean and standard deviation (SD) being reported for age and frequencies and percentages being reported for all other patient characteristics, which were categorical. Hospitals were compared via an ANOVA test for age and chi-square tests for all other variables when the sample size was sufficient. The primary outcome of D2B time, as well as secondary outcomes of time of ED arrival to first ECG, time from last ECG to activation, and time from activation to transport to the CCL was examined with medians and interquartile ranges (IQR) and compared across hospitals via Kruskal-Wallis tests. Percent of patients who met the guidelines of first ECG in < 10 minutes and of D2B time in < 90 minutes was calculated overall and by hospital and compared via chi-square tests. Also, a binominal test compared the overall percentage of patients with D2B times ≤90 minutes to the American Heart Association minimal adherence score of 75%.3
Predictors for D2B times were identified via multivariate linear regression. Each characteristic was analyzed in an unadjusted model predicting D2B time then included in one or both of two adjusted models. Both models contain age, gender, mode of arrival (walk-in versus not), scenario of arrival, day of the week, patient confounder (yes/no), and non-patient confounder (yes/no). One model then contained indicators for the individual hospitals while the other contained the number of steps from the ED to the CCL. All analysis was conducted in IBM SPSS at a significance level of 0.05.30