Objective: Describe short-term mortality and hospitalization outcomes in patients undergoing troponin testing in Western Australia tertiary public hospital ED.
Methods: Retrospective cohort study of adults presenting to urban tertiary ED between July 2000 and December 2004, and undergoing troponin inhibitory or tropomyosin binding testing. Patient data relating to ED attendance, troponins, hospitalization and death were linked using probabilistic data linkage. Troponin results were dichotomized (normal/abnormal) according to hospitals' cut-offs across various time periods. Presentations undergoing ED troponin testing, provided that it was not within 28 days of a previous discharge involving ED troponin testing, met study inclusion criteria. Logistical regression analyses quantified risk of subsequent hospitalization and death.
Results: A total of 119 646 troponin tests were performed during 83 635 ED presentations involving 54 214 patients. For the 73 965 ED presentations meeting the study criteria, 6779 (9.2%) recorded an abnormal peak troponin level and 168 (2.5%) of these presentations were discharged directly from ED. After adjusting for covariates, those discharged directly from ED with abnormal troponin results were significantly more likely to die within 2 days (OR 60.4, 95% CI 31.9-114.2), 7 days (OR 31.2, 95% CI 17.9-54.3) and 28 days (OR 19.4, 95% CI 12.5-30.1) of discharge, compared with those with normal troponin results. Adjusted odds of subsequent hospital admission within 28 days following discharge from ED were 3.3 (95% CI 2.4-4.5) times greater for presentations with abnormal versus normal troponin results.
Conclusion: An abnormal ED troponin result appears to be associated with increased risk of subsequent hospitalization and death within 28 days of discharge.