Study objective: To estimate the appropriateness of decision-making by emergency physicians regarding coronary thrombolysis.
Methods: We conducted a retrospective chart review of patients admitted over a period of 13 months from a tertiary care center emergency department with a diagnosis of an acute ischemic coronary syndrome. Both thrombolysed and nonthrombolysed patients were eligible for inclusion. The decisions of emergency physicians to use or not use thrombolytics were compared with standard Canadian guidelines, based on the blinded assessments of two reviewers. Appropriateness was estimated with the use of adjusted kappa statistics, and a hierarchical statistical model was developed to estimate the distribution of appropriate decision-making rates for individual emergency physicians.
Results: The overall adjusted kappa for appropriateness was .85 (95% confidence interval [CI], .76 to .94). The appropriateness rate for thrombolysed patients was 80.6% (95% CI, 62.5 to 92.5), and for nonthrombolysed patients it was 97.2% (95% CI, 91.9 to 99.4). The distribution of individual emergency physician appropriateness rates had an estimated mean of 91.3% and a 95% CI of 81.3% to 97.7%. Complication rates were not significantly different from previously published rates.
Conclusion: This study demonstrates excellent agreement between emergency physicians' decisions regarding thrombolysis and standard Canadian guidelines, based on an adjusted kappa statistic. The distribution of individual emergency physician appropriateness rates and the appropriateness rate for nonthrombolysed patients are estimated for the first time.