This is a cohort study. From 1st January to 31st December of 2019, all successive patients that were hospitalized with AMI at Timisoara Institute of Cardiovascular Disease were evaluated for enrolling in the study. In the absence of contraindications, percutaneous coronary intervention (PCI) was performed in the first 12 hours from the symptoms onset in all STEMI and in the high-risk NSTEMI patients. As high-risk NSTEMI patients were stated those with at least one of the following criteria: a GRACE-score $>$140, evolutive changes of the ST-segment/T-wave, or a relative rise or fall in cardiac enzymes [4]. High-risk NSTEMI patients in whom PCI could not be performed in the first 12 hours of the symptoms onset were not enrolled in the study.

The initial evaluation was based on the analysis of the parameters resulting from clinical presentation, cardiac troponin, and resting 12-lead ECG [4, 5]. The diagnosis of AMI with persistent ST-segment elevation (STEMI) was stated in the presence of at minimum 2 of the subsequent 3 parameters: (1) typical angina lasting over 20 minutes; (2) ST-segment elevation $\ge$1 mV, lasting over 0.08 sec after the J point, seen in at least 2 contiguos leads; (3) transitory rise in myocardial necrosis markers exceeding twofold the normal laboratory limit [5].

AMI without persistent ST-segment elevation (NSTEMI) was diagnosed in the presence of an appropriate clinical situation (typical angina or angina equivalent) with ST‑segment depression or prominent T‑wave inversion, without ST‑segment elevation and/or positive biomarkers of necrosis (e.g., troponin I $\ge$1 $\mu$g/L in our laboratory) [6].

The PCI and the adjunctive pharmacological medication was performed according to the ESC guidelines. The patients were administered standard loading doses of 300–600 mg clopidogrel, 300–500 mg aspirin and 5000 IU unfractionated heparin before the PCI. Glycoprotein IIb/IIIa inhibitors were given when the operator considered it necessary. If a coronary stent was implanted, clopidogrel was prescribed for 12 months, associated with aspirin.

The inclusion criteria were the confirmed diagnosis of AMI hospitalized within the first 12 hours of the symptoms onset and the absence of exclusion criteria.

Exclusion criteria were: PCI‑ or CABG- related AMI [7], and the association of diseases worsening the long‑term prognosis such as severe primary cardiomyopathy, severe valvular diseases congenital heart diseases, kidney dysfunction, liver cirrhosis, a malignant tumor, and severe infection.

Baseline data were taken from hospital records and comprised gender, age, Killip functional class on admission, medical history, 12 leads resting electrocardiogram, laboratory data, echocardiographic data, and the results of the coronary angiography.

Medical history integrated data regarding obesity, smoking, old myocardial infarction, hypertension, peripheral vascular disease, chronic obstructive pulmonary disease, diabetes, chronic kidney disease, stroke. The laboratory data determined at admission were creatine kinase-MB isoenzyme and cardiac troponin levels, blood cell count, serum hemoglobin, serum glucose, liver enzymes, serum creatinine, serum electrolytes, and lipogram.

Echocardiographic evaluation was done within the first 24 hours of hospitalization, using a VIVID S5 echocardiograph. Mono- and two dimensional echocardiography, as well as pulsed and continuous Doppler imaging, were performed in all patients. Left ventricular ejection fraction (LVEF) was calculated using the Simpson method, evaluating the end-diastolic volume and left ventricular end-systolic volume [8].

Medical treatment reports were accomplished at discharge and 1 year follow up.

The cause of death was determined from hospital records, or by asking the patient‘s physician for those who died at home.

All causes of readmissions were noted during the 1-year follow-up period. The causes of readmissions were determined by utilizing the hospital records.

The primary endpoint was in-hospital mortality, defined as the death of any cause during the hospitalization for AMI.

Cardiac deaths were considered those due to AMI, heart failure, acute pulmonary edema, cardiogenic shock, ventricular fibrillation, or cardiac rupture.

Noncardiac deaths were considered those having an extra-cardiac cause, such as stroke, acute renal failure, or sepsis.

The secondary endpoints were the 1-year death and readmission rates. 1-year mortality included both cardiac and non-cardiac deaths. 1-year readmissions had as possible causes recurrent myocardial infarction (RMI), stroke, stent thrombosis, and bleeding. RMI was defined according to the Academic Research Consortium (ARC) criteria [9]. Bleeding complications were defined according to the BARC and the TIMI bleeding classification [10, 11]. Stroke was stated as an irreversible neurological impairment, as confirmed by the neurologist, and based on sustaining information, such as brain images.

Data were collected and analyzed by means of the MedCalc Statistical Software version 19.1.7 for Windows. Ostend, Belgium. Data are given as mean $\pm$ SD when normally distributed, and as frequencies and percentages for categorical variables. For comparisons between two groups of continuous data, t‑tests were used, while for comparisons of categorical data, Chi‑square tests were done. To evaluate the contribution of each variable in the studied outcomes, univariate and multivariate logistic regression was used. Variables considered for potential prognostic impact included age, sex, current smoking, obesity, diabetes mellitus, systemic hypertension, chronic obstructive pulmonary disease, chronic kidney disease, peripheral artery disease, history of coronary heart disease and of congestive heart failure, left ventricular ejection fraction (LVEF) at admission, Killip class, laboratory data (serum creatinine, brain natriuretic peptide, peak creatine kinase‑MB and troponin‑I levels) Multivariate logistic regression analysis with backward stepwise method was used for all initial parameters that were associated with the endpoints in univariate analysis. ROC curves were utilized to determine the optimal sensitivity and specificity of the analyzed parameters. The discrimination ability of the analyzed parameters was estimated by the C-statistic, which is equivalent to the area under the ROC curve. A model with C-statistic $>$0.65 is considered to have a meaningful discriminatory ability. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using Cox proportional hazards models. The threshold for statistical significance was established at a p-value of 0.05. All P-values were results of two-tailed tests.

Of the 297 patients hospitalized with AMI, 277 were enrolled in the study. The mean age was 67.38 $\pm$ 13.4 years (32–95 years). 173 (63%) were males. LVEF at admission was 44 $\pm$ 13 %. According to the age, the AMI patients were separated into 2 groups: group I ($\ge$80 years, n = 63), and group II (80 years, n = 214). The baseline characteristics, the medical history and the cardiovascular risk factors of the patients are shown in Table 1. Compared with group II, the group patients were more often women, non-smokers, and presented more frequent systemic hypertension, chronic kidney disease, stroke, congestive heart failure. They had more often a LVEF 40% at admission and a higher Killip functional class.

Table 2 describes the findings of the coronary angiography performed within the first 12 hours from the onset of symptoms and the therapeutical interventionsf. No angiography could be done in 3 (4%) of group I patients and in 1 (0.5%) patients of group II (P = 0.03), because of severe kindney failure. Group I patients presented a significantly higher prevalence of 3-vessel coronary disease (P = 0.01), and a significantly lower prevalence of myocardial revascularization by PCI (48% vs. 78%, P 0.0001). The rate of coronary artery bypass graft (CABG) was 2% in group I and 3% in group I (P = 0.67). Regarding the concomitant medication, diuretics and oral anticoagulants were more often administered in group I patients (P 0.001).

The total all-cause mortality rate was 12.6% (n = 35). The number of deaths was 18 (28%) in the octogenarians, besides 17 (8%) in the non-octogenarians, P 0.0001. The relative risk for total mortality was 3.5 (95% CI: 1.94 to 6.46, P 0.0001) in the octogenarians with AMI.

During the hospitalization for AMI, 26 patients (9.3%) died, 12 being from group I (19%), and 14 from group II (6.5%), P 0.003. Both the cardiac causes of death and the noncardiac causes of death were more often present in group I (P 0.05), as shown in Table 3. The relative risk for in-hospital mortality in the octogenarians was 3.2 for all-cause deaths (95% CI: 2.08 to 5.18, P 0.0001), and 2.9 for cardiac deaths (95% CI: 1.11 to 7.83, P = 0.029). Fig. 1 displays the Kaplan-Meier curve for in-hospital survival.

Fig. 1.

Kaplan-Meyer curves for in-hospital mortality in AMI patients.
Abbreviations: AMI, acute myocardial infarction.

In univariate analysis, the variables associated with in-hospital death were age $\ge$80 years (P 0.0001), male sex (P 0.0001), diabetes (P 0.0001), previous myocardial infarction (P 0.001), a Killip class $\ge$3 (P 0.0001), an LVEF 40% at admission (P 0.0001), and a three-vessel CAD (P 0.01).

The multivariate logistic regression selected three parameters as independent predictors for in-hospital mortality. These parameters were: an age $\ge$80 years [OR = 7.03, 95% CI: 3.05 to 16.20, P 0.0001], a LVEF 40% at admission (OR = 2.1, 95% CI: 0.99 to 4.59, P 0.0001), and a Killip class 3 or 4 (OR = 2.41, 95% CI: 1.06 to 5.51, P 0.0001). When comparing the receiver operating characteristic curves for these variables, as shown in Fig. 2, we found that an LVEF 40% at admission had the best discriminative power, with an AUC of 0.687, followed by the age $\ge$80 years (AUC = 0.672) and the Killip class $\ge$3 (AUC = 0.665). The differences between the AUCs of the ROC curves were not significant.

Fig. 2.

Comparison between ROC curves of the independent predictors for in-hospital death in AMI patients.
Abbreviations: AMI, acute myocardial infarction; AUC, area under the curve; LVEF, left ventricular ejection fraction; ROC, receiver operating characteristic.

251 AMI patients were discharged alive (90%) and followed-up for 1 year. Group I patients received at discharge more often oral anticoagulants (P = 0.01) and diuretics (P = 0.04). During the follow-up interval further 9 patients died, thus the 1-year mortality was 3.6%. 1-year mortality was notably higher in group I patients (12% vs. 1.5%, P 0.001). The relative risk for death during the 1-year follow-up in the elderly group was 7.8 (95% CI: 2.03 to 30.29, p = 0.002). Fig. 3 displays the Kaplan-Meier curve for the 1-year survival in the two AMI patient groups. The most frequent causes of death were recurrent myocardial infarction as well as congestive heart failure (Table 3).

Fig. 3.

Kaplan-Meyer curves for 1-year mortality in AMI patients.
Abbreviations: AMI, acute myocardial infarction.

The 1-year mortality risk was associated in univariate analysis with triple-vessel disease (P = 0.002), Killip class $\ge$3 (P = 0.028), LVEF 40% at admission (P = 0.0009) and age $\ge$80 years (P = 0.04).

The multivariate logistic regression selected two parameters as independent predictors for the 1-year mortality: the triple-vessel CAD (OR = 3.3, 95% CI: 1.47 to 7.75, P = 0.004) and an LVEF 40% at admission (OR = 5.0, 95% CI: 1.85 to 13.91, P = 0.001). The comparison of the ROC curves for these parameters (Fig. 4) showed no significant differences regarding the AUCs (0.662 for LVEF 40% ROC curve, 0.643 for the triple-vessel disease ROC curve, P = 0.75).

Fig. 4.

Comparison between ROC curves of the independent prognosticators for 1-year mortality in AMI patients.
Abbreviations: AMI, acute myocardial infarction; AUC, area under the curve; LVEF, left ventricular ejection fraction; ROC, receiver operating characteristic.

During the 1-year follow-up phase, 22 patients (8.7%) were rehospitalized. The relative risk for 1-year readmissions was 2.7 in the octogenarians when compared with those aged 80 years (95% CI: 1.22 to 5.99, P = 0.01). The readmission rate was higher in the octogenarians (P 0.01), and congestive heart failure (CHF) was a significantly more frequent cause of readmission in these patients (P = 0.02), as shown in Table 4. Fig. 5 displays the Kaplan-Meier curves for the 1-year readmissions in the two AMI patient groups.

Fig. 5.

Kaplan-Meier curves for 1-year readmission in AMI patients.
Abbreviations: AMI, acute myocardial infarction.

The readmission rate was notably associated in univariate analysis with the history of CHF (P 0.0001), permanent atrial fibrillation (P 0.001), and an age $\ge$80 years (P 0.001). All these variables were identified by the multivariate logistic regression as independent predictors for 1-year readmission. The comparison between the ROC curves showed no significant differences between their discriminative capacities (Fig. 6).

Fig. 6.

Comparison between ROC curves of the independent prognosticators for 1-year readmission in AMI patients.
Abbreviations: AMI, acute myocardial infarction; AUC, area under the curve; ROC, receiver operating characteristic.