ABC | Volume 112, Nº3, March 2019

Original Article Castro et al Troponin I after acute coronary syndrome Arq Bras Cardiol. 2019; 112(3):230-237 study of ACS patients enrolled between February 2009 and December 2013 at a community hospital in Sao Paulo, Brazil. All patients with suspected acute coronary syndrome in the emergency department were screened for participation in the study. ERICO participants must fulfill diagnostic criteria for ST elevation myocardial infarction (STEMI), non-ST elevation myocardial infarction (NSTEMI), or unstable angina (UA); the criteria used to define acute coronary syndromes were: 14 1) Myocardial infarction (MI): presence of symptoms consistent with cardiac ischemia within 24 hours of hospital presentation, and troponin I levels above the 99th percentile with a test-specific coefficient of variation < 10%. 1a) STEMI: presence of criteria for MI plus one of the following: persistent ST segment elevation ≥ 1 mm in two contiguous electrocardiographic leads, or the presence of a new or presumably new left bundle branch block. 1b) NSTEMI: presence of criteria for MI, but not STEMI. 2) UA: symptoms consistent with cardiac ischemia 24 hours prior to hospital admission, absence of MI criteria, and at least one of the following: history of coronary heart disease; positive coronary disease stratification test (invasive or noninvasive); transient ST segment changes ≥0.5mm in two contiguous leads, new T-wave inversion ≥ 1 mm, and/or pseudonormalization of previously inverted T-waves; troponin I ≥0.4 ng/ml; or diagnostic concordance of two independent physicians. During the in-hospital phase, all subjects were treated at the discretion of the hospital staff with standard procedures, without influence from the study. The study protocol was approved by the Institutional Review Board addressing research in human participants. All participants provided written informed consent for the study. Participants were interviewed during admission to the hospital and provided data regarding sociodemographic factors, medical history, and main cardiovascular risk factors (hypertension, diabetes, obesity, dyslipidemia, smoking, physical inactivity, cocaine use, menopause, and familial and personal history of coronary heart disease). Three physicians were responsible for reviewing patient information and for validating ACS cases. According to the study protocol, a blood sample was drawn for laboratory tests (troponin I, MB-creatine kinase, serum glucose, total cholesterol, HDL and LDL-cholesterol, triglycerides and total blood cell count). At approximately 30 days after the event, participants were invited to undergo a new on-site evaluation by a physician to update data on cardiovascular risk stratification, current medication use, and additional clinical data. New blood samples were also collected. At six months after the index event and annually thereafter, all participants were contacted by phone to update information about their vital status, cardiovascular history, medications and symptoms. All participants enrolled in the ERICO study who had blood samples collected 25 to 90 days after an ACS episode were included in this analysis. The lower limit of this interval was chosen to avoid confounding by the expected elevated circulating cTn levels in the first few days after an ACS episode; the upper limit of 90 days allows comparison with previous studies, 13,16 although there is currently no widely accepted definition of subacute phase after ACS in the literature. High‑sensitivity cardiac troponin I was measured in all patients at presentation and in the subacute phase after the event. The assay used to measure hs-cTnI was the Advia Centaur TnI-Ultra Assay (Siemens Medical Solutions Diagnostics, Tarrytown, NY, USA), with the 99th percentile reference value of 0.04 mcg/L in healthy subjects and coefficient of variance lower than 10% at this range. Subjects were classified into three subgroups according to hs‑cTnI tertiles in the subacute phase after the index ACS event. In this study, our endpoints were all-cause mortality and cardiovascular mortality. We searched official death records on a regular basis for information about all participants if (1) we received information that they had died or (2) we could not contact them at the time. Municipal and state health offices searched their files to obtain death certificates and returned the results of this search to the ERICO research team. Two medical doctors reviewed these data and classified the cause of death for deceased participants according to the information from the death certificates. Participants were defined to have died from a cardiovascular cause (cardiovascular mortality) if we identified a cause of death classified in Chapter IX of the 10th version of the International Classification of Diseases (ICD-10), entitled “Diseases of the circulatory system”, or if we identified a cause of death classified with the ICD-10 code R57.0 “Cardiogenic shock”. Statistical analysis The statistical analyses were performed with R for Mac version 3.5.0. Categorical variables are presented as proportions and compared using the chi-square test. To test the assumption of normality in the distribution of continuous variables, we used the Shapiro-Wilk test. Continuous variables with normal distribution are presented as means (standard deviations) and compared using one-way ANOVA. Continuous variables with non-normal distribution are presented as medians (interquartile intervals) and compared using the Kruskal-Wallis test. Cumulative survival probabilities across the tertiles are presented as Kaplan-Meier curves and compared using the log-rank test. We built Cox proportional hazard models for all-cause mortality and cardiovascular mortality, and presented them as crude, age-sex adjusted, and two multivariate models. Model 1 was adjusted for age, sex, ACS subtype, traditional cardiovascular risk factors (hypertension, diabetes, dyslipidemia, smoking history, and body mass index), and medication use at the first follow-up visit (aspirin, clopidogrel, beta blockers, angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers and statins). Model 2 was additionally adjusted for estimated glomerular filtration rate (GFR) < 60 ml/min/1.73 m 2 and left ventricular ejection fraction < 0.40, two variables associated with worse prognosis in previous studies. 13,17 All tests were two-sided, and p value < 0.05 was considered significant. Results From February 2009 to December 2013, 1085 patients were enrolled in the ERICO study. Blood samples were collected at the emergency room and between 25 and 90 days after the initial event from 525 of these participants, who were included 231