Heart Rate Fluctuation and Mortality in Critically Ill Myocardial Infarction Patients: A Retrospective Cohort Study

Abstract

Background: Whether heart rate (HR) fluctuation after admission has an impact on the outcomes of critically ill myocardial infarction (MI) patients in intensive care unit remains unknown. Methods: A total of 2,031 MI patients were enrolled from the Medical Information Mart for Intensive Care (MIMIC-III) database. HR fluctuation was calculated as the maximum HR minus the minimum HR in the initial 24 h after admission. Participants were divided into 3 groups, namely, low HR fluctuation [<30 beats per minute (bpm)], medium HR fluctuation (30-49 bpm), and high HR fluctuation (≥ 50 bpm). The main outcomes were 30-day and 1-year mortality. Cox regression and restricted cubic spline model were used. Results: Each 10-bpm increase in HR fluctuation was associated with a higher risk of 30-day mortality and 1-year mortality, with adjusted hazard ratios of 1.122 (95% CI, 1.083-1.162) and 1.107 (95% CI, 1.074-1.140), respectively. Compared with the low HR fluctuation group, the high HR fluctuation group suffered a significantly higher risk of mortality after adjustment, with hazard ratios of 2.156 (95% CI, 1.483-3.134) for 30-day mortality and 1.796 (95% CI, 1.354-2.381) for 1-year mortality. A typical J-type curve was observed in restricted cubic splines for the association between HR fluctuation and 30-day or 1-year mortality of MI patients, with the lowest risk on the HR fluctuation of 30 bpm. Sensitivity analyses emphasized the robustness of our results. Conclusions: This retrospective cohort study revealed an independent positive association between HR fluctuation and 30-day and 1-year mortality in critically ill MI patients, which warrants further investigation.

Keywords: heart rate fluctuation; intensive care unit; mortality; myocardial infarction; risk factor.

Figure 1

Association between HR fluctuation and outcomes of MI patients. Crude hazard ratio and 95% CI for HR fluctuation in 30-day mortality (A) and 1-year mortality (B). Adjusted hazard ratio and 95% CI for HR fluctuation in 30-day mortality (C) and 1-year mortality (D). The analyses used a model with restricted cubic splines. The reference (hazard ratio = 1, horizontal dotted line) was an HR fluctuation of 30 bpm (vertical dotted line). Adjusted variables included age, male sex, weight, SOFA score, admission HR, sedative use, ventilation use, vasopressor use, CHF, AF, hypertension, and liver disease, namely, model 5 described above. HR, heart rate; SOFA, sequential organ failure assessment; CHF, congestive heart failure; AF, atrial fibrillation; CI, confidence interval.

Figure 2

Adjusted survival rate by 3 HR fluctuation groups. Survive curves presented adjusted survival rates of 30-day mortality (A) and 1-year mortality (B) by 3 HR fluctuation groups. Adjusted variables included age, male sex, weight, SOFA score, admission HR, sedative use, ventilation use, vasopressor use, CHF, AF, hypertension, and liver disease, namely, model 5 described above. The hazard ratio and 95% CI were calculated compared with those of the low HR fluctuation group. HR, heart rate; SOFA, sequential organ failure assessment; CHF, congestive heart failure; AF, atrial fibrillation; CI, confidence interval.

Figure 3

Hazard ratio and 95% CI of HR fluctuation (continuous) for mortality in subgroups. Hazard ratio and 95% CI were calculated along as 10 bpm HR fluctuation increase by Cox hazard regression model using backward process. Model adjusted for age, male, weight, SOFA score, admission HR, sedative, ventilation, vasopressor, CHF, AF, hypertension, liver disease, namely model 5 described above. HR, heart rate; SOFA, sequential organ failure assessment; CHF, congestive heart failure; AF, atrial fibrillation; CI, confidence interval.