Résumé
Abstract Aims To assess the prognostic value of expiratory carbon monoxide (CO) levels in patients admitted for myocardial infarction (MI). Methods and results In this prospective study, expiratory CO levels were measured upon admission in consecutive patients hospitalized for MI across 39 centres. The primary outcome was 1-year all-cause death. Secondary outcomes included 1-year major adverse cardiac events (MACEs: cardiovascular death and recurrent MI) and in-hospital major adverse events (MAEs: death, severe ventricular arrhythmia, cardiogenic shock, and the need for mechanical ventilation). The prognostic value of expiratory CO levels was further evaluated using machine learning (ML) analysis. Among 717 patients (64 ± 13 years; 75% males; 33% active smokers; 43% ST-elevation MI), elevated expiratory CO levels (>11 ppm) were found in 79 patients (11%). Patients with elevated CO levels had a higher rate of 1-year all-cause mortality compared with those without (16.5% vs. 5.2%, P < 0.001). Elevated CO levels were independently associated with 1-year all-cause death across various adjustment models: comorbidities [odds ratio (95% confidence interval): 3.6 (1.7–7.5)], clinical parameters of in-hospital severity [4.5 (2.3–8.8)], and respiratory parameters [6.3 (3.1–12.8)]. Elevated CO levels were also independently associated with a significant increase in 1-year MACEs and in-hospital MAEs. Machine learning analysis identified CO level as one of the most important predictors of adverse events, compared with other known prognosticators. Conclusion This is the first study to demonstrate that elevated expiratory CO levels upon admission are independently associated with an increased risk of 1-year all-cause mortality, 1-year MACEs, and in-hospital MAEs in patients hospitalized for MI.