ABC | Volume 114, Nº6, June 2020

Special Article Barros and Silva et al. Brazilian Registry of Acute Coronary Syndromes Arq Bras Cardiol. 2020; 114(6):995-1003 Table 3 – Relationship between the revascularization procedure and clinical outcomes in the 3 types of acute coronary syndrome Events in revascularized patients compared to non-revascularized patients Unstable Angina HR [95% CI] AMI without ST elevation HR [95% CI] AMI with ST elevation HR [95% CI] Severe Bleeding 2.03 [0.75 ; 5.44] 1.15 [0.55 ; 2.41] 1.28 [0.37 ; 4.50] Cardiorespiratory Arrest 0.27 [0.09 ; 0.79] 0.54 [0.34 ; 0.87] 0.54 [0.36 ; 0.83] Myocardial Reinfarction 1.69 [1.03 ; 2.76] 1.28 [0.85 ; 1.90] 0.87 [0.53 ; 1.43] Cerebrovascular Accident (CVA) 1.18 [0.26 ; 5.28] 0.80 [0.30 ; 2.13] 1.02 [0.34 ; 3.11] Death 0.33 [0.17 ; 0.65] 0.53 [0.37 ; 0.76] 0.45 [0.33 ; 0.63] Cardiovascular death 0.45 [0.20 ; 1.06] 0.43 [0.28 ; 0.66] 0.43 [0.31 ; 0.62] Composite endpoint 0.97 [0.66 ; 1.42] 0.75 [0.57 ; 0.98] 0.64 [0.48 ; 0.85] Composite endpoint: Death, Myocardial reinfarction and CVA. HR: Hazard Ratio. Table 4 – Multivariate analysis of factors associated with the occurrence of composite events (CVA, reinfarction or death). Multivariate Variables HR [95% CI ] p-value Age Age (5-year increase) 1.16 [1.11;1.20] <0.001 Sex Female 1.10 [0.91;1.33] 0.328 Healthcare (Supplemental Insurance/Private) Supplemental Insurance/Private 0.57 [0.47;0.69] <0.001 Dyslipidemia Yes 0.98 [0.81;1.19] 0.826 AMI Yes 1.29 [1.03;1.63] 0.030 Angina Yes 0.95 [0.78;1.16] 0.613 Hypertension Yes 1.08 [0.85;1.36] 0.534 CVA Yes 1.38 [1.06;1.80] 0.017 Renal Failure Yes 2.08 [1.59;2.71] <0.001 Diabetes Yes 1.48 [1.23;1.78] <0.001 CHF Yes 1.10 [0.83;1.45] 0.502 Percutaneous coronary intervention Yes 1.00 [0.80;1.27] 0.961 CABG YES 0.94 [0.72;1.25] 0.684 ASA use Yes 1.18 [0.96;1.47] 0.120 Smoking Never ref ref Ex-smoker 1.22 [0.99;1.50] 0.055 Current smoker 1.27 [1.00;1.62] 0.047 Complete therapy Yes 0.72 [0.61;0.86] <0.001 Final Diagnosis Unstable Angina ref ref AMI with ST elevation 1.76 [1.39;2.23] <0.001 AMI with ST elevation 2.04 [1.59;2.62] <0.001 *Variables with p values < 0.15 in the univariate analysis were included in the multivariable model. ** The variables with p values > 0.15 in the univariate analysis were: Transfer from another service, Family History of Coronary Disease,Abdominal Obesity, Sedentary Lifestyle and PeripheralArterial Disease. for the external validity of the effects identified in controlled clinical trials. The explanation for the difference in the outcomes identified between the patients from public or private-sector, could be owed to the difference in healthcare quality. However, since the multivariate model identified that private healthcare is associated with better outcomes independently of the quality of therapy, a possible explanation would be the patients’ social/ educational level itself. These data were not collected for direct inclusion in the multivariate model of this analysis. However, in previous studies, they were identified as factors associated with clinical outcomes in this population. 15,20 Study limitations One limitation of this study regards the patients’ profile, since this is a voluntary registry, whose participant services showed clinical research capacity. Therefore, the results may not be applicable to populations that do not fit these characteristics (for instance, hospitals with more limited structure). In any case, even in centers with potential for high-quality care, relevant gaps were identified when applying scientific evidence. Another limitation is related to assessment of adherence to evidence- based therapies, because this analysis was based on medical adherence in terms of the prescription of evidence-based therapies. We did not collect data on the eligibility, the actual administration of the therapies prescribed and the reasons for prescription discontinuity. Thus, considering that the adherence on the part of the patients was not assessed in this registry, the gap on the use of evidence-based therapies could be even bigger than that found in the ACCEPT registry, which evaluated the medical prescription. Finally, the clinical outcome assessment presents limitations regarding the absence of events adjudication and missing data of the 12-month follow-up of 410 patients. Nevertheless, the assessment of clinical outcomes in pragmatic observational studies is usually performed by notification of the investigator physician, without the use of a specific committee for adjudication, which would represent a scenario closer to the identification of events in real clinical practice. As for the follow-up, taking into account that the follow-up losses occurred at different moments, the analyses were performed using the Cox model. Consequently, the patients were censored in the last registered contact, in order to minimize the differences in follow-up length. Conclusion In the largest prospective study ever published on patients with ACS in Brazil, we identified a mean rate of major 1001