Registry design

RENASICA III was established by the Mexican Cardiology Society, with unrestricted support provided by Sanofi. The full protocol, including definitions, has been published.⁵ In brief, RENASICA III was a prospective, multicenter registry involving adult men and women hospitalized with ACS in Mexico. In-hospital data were collected from patients in tertiary and community hospitals across both the public and private healthcare systems. The principal investigators were selected from community and tertiary hospitals located throughout Mexico.

The study population comprised a consecutive, prospective cohort of patients with a high clinical suspicion of ACS who were admitted to hospital. The initial diagnosis was established by the physician-in-charge on the basis of the patient's clinical and electrocardiographic characteristics. Four groups of patients were identified according to the electrocardiographic findings: normal or non-specific electrocardiogram (ECG), ST depression, transient ST elevation,⁶ and persistent ST elevation. Patients with a final diagnosis of ACS and objective evidence of ischemic heart disease (identified using invasive or non-invasive tests) were included in the registry and those with type II infarction were excluded. ACS nomenclature (ST-elevation myocardial infarction [STEMI] and non-ST-elevation myocardial infarction [NSTEMI] or unstable angina [UA]) was used at admission and, for epidemiologic reasons, UA or myocardial infarction at discharge.

The registry structure, data collection, and data analysis were based on current quality recommendations, including 11 criteria proposed by Gitt et al.⁷ The registry was conducted in accordance with national and international regulations. The protocol was approved by the institutional ethics committees in all participating centers and all patients provided informed consent.

The data were sent to the registry-coordinating center via a web site (http://www.renasica.mx/). Electronic case report forms were reviewed by the coordinating center to determine data quality. All participating sites had regular access to a data-entry clerk. Data-quality monitoring was conducted at 20% of the centers, chosen at random. Technical support was provided through periodic software updates. Investigators at all sites underwent training in the use of the electronic database and in the study definitions, and had real-time access to technical advice from the coordinating center through a 01-800 number. Follow-up by telephone and during site visits was performed through the coordinating center. In specific cases follow-up was done by the physician-in-charge. An executive committee set out the publication policies to allow an expedited process for publication or presentation at national and international meetings.

Sample size determination

A sample size with statistical significance was obtained using data from RENASCA³ and RENASICA II.² Fifty percent of the patients in RENASCA³ did not receive adequate treatment and only 8% underwent percutaneous coronary intervention (PCI); consequently, 30% of patients developed left ventricular dysfunction. In RENASICA II, 15% of the patients had PCI and 17% developed left ventricular failure as an early complication. This difference between the two studies allowed us to calculate the smaller sample size necessary to evaluate this outcome (i.e. left ventricular dysfunction). A sample of 182 patients in each group (STEMI and NSTEMI/UA) would give a statistical power of 80%, with an alpha value of 0.05, delta value of 13%, 1:1 ratio of exposed: unexposed, and a confidence interval (CI) of 95%. Allowing for an estimated 20% loss to follow-up, the final number of patients chosen per group was 220. A lower delta value for mortality of 2% required 3975 patients to be treated (or not) with PCI; thus 8000 subjects had to be enrolled in the registry.

Statistical analysis

Data are expressed as percentages, mean ± standard deviation (SD), and odds ratios (ORs) with 95% CIs. Differences between continuous variables with a normal distribution were examined using Student's t test. The Wilcoxon rank-sum test was used for non-normal continuous variables. The chi-square test, Fisher's exact test, or Yates' correction was used to analyze categorical variables, as appropriate. A two-tailed test with a p-value ≤0.05 was considered statistically significant. Logistic regression analysis was used to identify independent predictors of hospital mortality on the basis of those with a p-value of ≤0.05 on univariate regression analysis, with adjustment for potential confounders to avoid confusion, the relationship between historical variables for atherosclerosis and cardiovascular events was examined through logistic regression. With the Cox proportional risk multivariable model, the relationship between each of these variables was analyzed. Kaplan-Meier survival curves were used to determine the risk of mortality, and differences between the curves were determined using the log-rank statistic. The Cox proportional risk model was used for adjusted survival analysis.

Study population

Between November 2012 and November 2013, 8296 consecutive ACS patients with proven ischemic heart disease were enrolled by 123 investigators at 29 tertiary and 44 community hospitals (63% government health system centers, 29% private practices, and 8% other health systems; hospitals were located across 91% of the Mexican territory). Of these patients, 4038 were diagnosed with NSTEMI/UA and 4258 with STEMI. All tertiary hospitals had the capability to perform percutaneous coronary intervention (PCI) and coronary artery bypass graft (CABG) surgery.

The mean ± SD age of the population was 62.0 ± 12 years, 73.8% were male, and the mean body mass index was 27.8 ± 4.0 kg/m² (Table 1). The prevalence of major risk factors for ischemic heart disease was high in both ACS groups. Patients with STEMI were younger than those with NSTEMI/UA and a higher percentage smoked tobacco. Patients with NSTEMI/UA were more likely to have angina, dyslipidemia, heart failure, hypertension, previous myocardial infarction, or to have undergone a previous PCI or CABG (all p < 0.0001).

Presenting symptoms, clinical characteristics, and biomarkers

On admission 80.5% of patients had ischemic chest pain lasting >20 min and clinical stability (Table 1). Killip class II-IV⁵ (Appendix B) was more frequent in patients with NSTEMI/UA than in those with STEMI (p < 0.0001). Mitral and septal rupture murmur and cardiac rub were infrequent in NSTEMI/UA (1.5%, 0.1%, 0.1%, respectively) and STEMI (1.1%, 0.2%, 0.1%).

Patients with STEMI more likely have had anterior, inferior, and lateral ST-elevation, whereas anterior, inferior, and lateral ST depressions were more frequent in patients with NSTEMI/UA (Table 1). The prevalence of atrial fibrillation was low in both groups, although it was numerically more frequent in patients with NSTEMI/UA.

The plasma concentrations of various biomarkers, including hemoglobin, hematocrit, leucocytes, glucose, creatine phosphokinase, CK-MB, troponin I and NT-proBNP, were higher in patients with STEMI compared to those with NSTEMI/UA, whereas concentrations of fibrinogen and creatinine were similar (Table 2). BNP was obtained in 28% and NT-proBNP in 22.9% of the overall population.

Management during hospitalization and at discharge

During hospitalization patients with STEMI were more likely than those with NSTEMI/UA to receive aspirin, clopidogrel or prasugrel, a glycoprotein IIb/IIIa inhibitor, low-molecular weight heparin, and statins. Patients with NSTEMI/UA were more likely to receive angiotensin receptor blockers, calcium antagonists, and oral nitrates (Table 3). Non-vitamin K oral anticoagulants were used in <1% of patients in both groups.

At discharge from hospital, aspirin, nitrates, calcium antagonists, and angiotensin receptor blockers were more frequently used in patients with NSTEMI/UA, whereas patients with STEMI were more likely to receive clopidogrel and angiotensin converting enzyme inhibitors (Table 3). The use of beta-blockers and statins was similar in both groups. In patients discharged alive, the use of ACE inhibitors, aspirin, beta-blockers, clopidogrel, and statins was higher at discharge than during hospitalization (all p < 0.0001) (Table 4).

The overall mean ± SD length of hospitalization was 7.1 ± 7.8 days (6.8 ± 7.2 days in NSTEMI/UA vs. 7.4 ± 8.5 days in STEMI).

Reperfusion approaches

The median (25th-75th percentiles) ischemia times in patients with STEMI were 45.0 (20.0-90.0) min for door-to-needle and 100.0 (50.0-270.0) min for door to balloon (Table 5). In the STEMI group, 37.6% of patients received thrombolytic therapy and 15.0% underwent primary PCI (Table 5). Rescue PCI was performed in 8.6% of patients with STEMI. Pharmacological reperfusion was conducted primarily in the emergency department (70.3%) and less frequently in the prehospital setting (17.9%) or coronary care unit (10.0%). Tenecteplase was the most frequently used fibrinolytic agent, followed by streptokinase and alteplase.

PCI was performed in 39.6% of patients with NSTEMI/UA, with an early strategy used in 10.8% and an urgent strategy in 3.0% (Table 6). Drug-eluting stents were used more frequently than bare-metal stents in these patients. Coronary artery bypass grafting was more frequently performed in patients with NSTEMI/UA (5.9% vs. 2.1% in STEMI).

Clinical outcomes

The overall hospital death rate was 6.4%, and was lower in the NSTEMI/UA population (3.9% vs. 8.7%, p < 0.0001) (Table 7). The rates of heart failure, cardiogenic shock, myocardial infarction or reinfarction, ventricular arrhythmias, pericarditis, and major bleeding were higher in the STEMI population. Angina or recurrent ischemia was more frequent in patients with NSTEMI/UA (p < 0.0001). The incidences of renal failure, sepsis, and cerebrovascular events were similar.

Multivariable analysis

Independent predictors of hospital mortality are shown in Table 8. The strongest predictors of hospital death, based on risk factors and cardiovascular history, were STEMI diagnosis, left ventricular dysfunction, and age >65 years (Fig. 1). The strongest predictor of death in terms of clinical data and complications was any expression of left ventricular dysfunction (Fig. 2). Ventricular and supraventricular arrhythmias, myocardial infarction or reinfarction, acute pulmonary edema, and major bleeding also had close associations with death.

Figure 1 The strongest predictors of hospital death, based on risk factors and cardiovascular history, were STEMI diagnosis, left ventricular dysfunction, and age >65 years. 

Figure 2 The strongest predictor of death in terms of clinical data and complications was any expression of left ventricular dysfunction. 

Lessons learned from RENASICA I

RENASICA I included 4353 ACS patients with or without ST elevation.¹ This registry highlighted practices and therapeutic approaches in the transition from the 20th to the 21st century. Non-ST-elevation ACS was a common cause of hospital admission and cardiac biomarkers were not available. In non-ST-elevation ACS, age >65 years, ST depression, necrosis, and extensive coronary disease detected on angiography were risk markers of mortality. Ischemic chest pain, dyspnea, and diaphoresis were closely associated with a diagnosis of STEMI. Less than half of the patients received reperfusion therapy despite the availability of appropriate facilities in 90% of hospitals. Standard antiplatelet and antithrombin treatments were used in 70% of patients; low-molecular-weight heparin and glycoprotein IIb/IIIa receptor blockers were used fewer patients. The RENASICA I registry provided valuable information for Mexican health authorities to better apply health resources for the management of ACS in the 21st century.¹

Lessons learned from RENASICA II

The second RENASICA registry included 8098 patients with final diagnosis of ACS. This registry had a higher ratio of STEMI to NSTEMI/UA (1.3:1) compared with RENASICA I (35% vs. 65%), attributed to the availability of mechanical reperfusion facilities in participating hospitals, as well as a high prevalence of diabetes in patients with STEMI. The average length of hospital stay was 8.1 days, independent of the final diagnosis. The incidence of diabetes (42%) was slightly lower than that in RENASICA I (50%). This metabolic disorder was a strong predictor of hospital mortality in patients with NSTEMI/UA. Tests to assess patients' glucometabolic state were not done, so the true prevalence of diabetes, impaired glucose tolerance, and insulin resistance are likely to be higher than indicated, and the rate of obesity was not determined.² The clinical presentation of ACS was similar to that observed in RENASICA I and use of troponin was low. Optimal pharmacological treatment (e.g. glycoprotein IIb/IIIa inhibitors and statins) was lower than expected in both groups of ACS patients. The rate of pharmacological thrombolysis decreased from 50% in RENASICA I to 37% in RENASICA II, with streptokinase being the most commonly used drug. The rate of PCI was 15%. Consequently, a considerable proportion of patients with STEMI received no form of reperfusion therapy. Left ventricular dysfunction (Appendix B) was the most important hospital complication and the strongest predictor of death, which occurred with a frequency of 10% in STEMI and 4% in NSTEMI/UA.²

RENASICA III

The profile of patients in RENASICA III in terms of age and sex is similar to that in RENASICA II,² suggesting that the high incidence of ACS persists in this high-productivity social group and in male patients. Of note, the rate of tobacco smoking decreased from 60% in RENASICA I¹ and 64% in RENASICA II² to 53.3% in RENASICA III, signifying that the health policy toward reducing rates of tobacco consumption has had some effect. Conversely, the prevalence of traditional risk factors such as hypertension, diabetes, and hypercholesterolemia increased, and the rate of diabetes remained unchanged, at nearly 50%. The proportion of men and the prevalence of hypertension in RENASICA III (73.8% and 62.1%, respectively) are similar to those reported in the GRACE registry (62-72% and 50-65%, respectively),⁸ the National Registry of Myocardial infarction (NRMI),^{9 10 11} and the Euro Heart Survey (64-72% and 52-64%, respectively).¹² However, the prevalence of diabetes is higher than that reported in GRACE (21-27%),⁸ ACCESS (33%),⁴ INTERHEART (7.2%),¹³ and KERALA (37.6%).¹⁴ The Indian CREATE registry found that diabetes prevalence was related to socioeconomic status, being lowest amongst the poor (18.7%) and highest among the most affluent (40.9%).¹⁵ It is apparent that strategies for primary prevention of diabetes and ischemic heart disease are mandatory in this Mexican population.

Clinical presentation and duration of hospitalization

In terms of clinical presentation, ACS with or without ST elevation, high-clinical suspicion of ACS, and risk stratification were conducted through a clinical process that included historic risk factors, chest pain with ischemic profile, and electrocardiographic findings. Considering the low proportion of ischemic equivalents, clinical suspicion does not seem to be a problem in the emergency departments of community and tertiary hospitals. In view of the low rates of use of cardiac biomarkers, a significant number of high-risk patients may not have been identified early. In both groups of ACS patients, the high proportion with left ventricular dysfunction could be attributed to several mechanisms, including extensive and severe coronary artery disease, ischemia time, metabolic instability, and acute inflammatory state. Biomarkers of left ventricular dysfunction were also underused (BNP in 28% and NT-proBNP in 22.9% of the overall population), so it is possible that a significant proportion of "clinically stable" patients with left ventricular dysfunction without clinical expression were omitted from the study. The continuing underuse of biomarkers^{1 2} remains as an important problem in Mexico, and highlights the need to identify and establish an appropriate risk model tailored toward middle-income countries. Also encourage health authorities to provide access to these biomarkers.

The present data show that the length of hospitalization for an ACS has decreased slightly in Mexico (9.1 days in RENASICA II² and 7.1 ± 7.8 days in RENASICA III), but exceeds that reported in the NRMI registries (4.3 days).^{9 10 11}

Pharmacological therapies

The optimal treatment for ACS continues to evolve, on the basis of robust evidence emerging from randomized trials and recommendations from cardiac societies.¹⁶ Although well-conducted studies have been undertaken, their results are open to interpretation; furthermore, they are conducted within the confines of a selected population and may not therefore be applicable to the broad spectrum of patients treated in clinical practice. Furthermore, because of resource limitations new drugs may not be available in all countries and settings. Consequently, good clinical indicators that reflect our "true" practices need to be identified. In RENASICA III, use of in-hospital and discharge treatments showed improvements when compared with the findings from RENASICA II.² Most of the patients received dual antiplatelet therapy comprising aspirin plus a second-generation thienopyridine. In spite of the advantages of the new antiplatelet drugs in ACS,^{17 18} prasugrel and ticagrelor were infrequently used, most likely due to the low rate of PCI, lack of local availability, high costs, and absence of strong evidence in medically treated patients with non-ST elevation ACS.

Anticoagulation

In the acute phase, anticoagulation remains the cornerstone of treatment for ST-elevation and non-ST-elevation ACS. In our study, a large proportion of patients received heparin, most commonly enoxaparin, which is easy to administer and has a better anticoagulation profile (improved coronary patency, greater prevention of reocclusion, low incidence of bleeding complications and heparin-induced thrombocytopenia compared with unfractionated heparin). However, in spite of the advantages of heparin therapy, a sizable proportion of patients (17.3%) did not receive any anticoagulation therapy. Reasons for the failure to administer heparin include risk of bleeding complications or perception of a limited benefit in patients outside the pharmacological or mechanical reperfusion window.

Reperfusion therapy

PCI remains a key therapy for patients with symptomatic coronary artery disease, primarily those with acute myocardial infarction. PCI has garnered tremendous attention in the past 15 years, with issues relating to the risks and benefits of drug-eluting stents and the use of adjunctive antithrombotic therapies.¹⁹ The present data from Mexico, spanning nearly two decades,^{1 2 3} show that pharmacological reperfusion remains the most frequently used form of reperfusion in STEMI patients. Primary PCI in STEMI and early cardiac catheterization (within 48-72 h) in NSTEMI/UA were infrequent. When compared with the earlier RENASICA registries,^{1 2} the most notable finding from RENASICA III was the transition in STEMI patients from first-generation to third-generation thrombolytic therapy and from unfractionated heparin to low-molecular weight heparin.

In terms of ischemia time, delays from first medical contact to electrocardiogram or to starting a specific therapy were within acceptable ranges, whereas delays to reperfusion exceeded the current recommendations.¹⁶ While quality-improvement initiatives that target both physicians and patients should help to reduce these delays, a national program is far from being a reality; therefore it is important to reinforce the need for regional programs in order to improve the care of ACS patients. One such program is currently in place at the Cardiology National Institute "Ignacio Chavez" in Mexico City. In this program, staff in the emergency departments of community hospitals have been trained to administer pharmacological thrombolysis. Electrocardiograms that are difficult to interpret are sent for evaluation, and recommendations on the best treatment are provided. This program intent evaluating the efficacy and safety of a pharmaco-invasive strategy.

Hospital outcomes

The hospital mortality rate in RENASICA III was 6.4% (8.7% in STEMI and 3.9% in NSTEMI/UA), which shows a reassuring decrease from RENASICA II (10% in STEMI, 4% in NSTEMI/UA).² However, it is higher than the 30-day mortality reported in ACCESS (3.6% all ACS; 5.0% in STEMI, 2.4% in NSTE-ACS)⁴ and hospital mortality in GRACE (4.6% in STEMI, 2.2% in NSTE-ACS).²⁰ The GRACE registry also demonstrated a statistically significant decline in hospital death over the course of the study (from 8.4% to 4.6%, p < 0.001, for STEMI; and from 2.9% to 2.2%, p = 0.02, for NSTE-ACS), which corresponded with increasing use of medications and primary PCI and decreasing use of pharmacological reperfusion.²⁰ The rate of bleeding complications in RENASICA III was low and can be attributed to the low rates of use of PCI, glycoprotein IIb/IIIa inhibitors, and new antiplatelets and oral anticoagulants.

The most important hospital complication in our study was the expression of left ventricular dysfunction (Appendix B). In the multivariable analysis, as observed previously,^{1 2} cardiogenic shock was the strongest predictor of death, as well as myocardial infarction or reinfarction. All of these major adverse cardiovascular events are closely related to prolonged ischemic times and low rate of PCI. Although bleeding complications and ventricular arrhythmias were scarce, both were related with hospital mortality.

Limitations

RENASICA III was not a population-based epidemiological study; therefore, some bias may have been introduced with respect to the selection of participating centers. Other limitations include the restricted number of visits to monitor data accuracy. Patients who died early, before admission to hospital or while in the emergency department, were excluded. The inclusion of community hospitals may explain the low rate of PCI.

Protection of human subjects and animals in research

The authors declare that the procedures followed were in accordance with the regulations of the responsible Clinical Research Ethics Committee and in accordance with those of the World Medical Association and the Helsinki Declaration.

Confidentiality of data

The authors declare that they have followed the protocols of their work centre on the publication of patient data and that all the patients included in the study have received sufficient information and have given their informed consent in writing to participate in that study.

Right to privacy and informed consent

The authors have obtained the informed consent of the patients and/or subjects mentioned in the article. The author for correspondence is in possession of this document.