Mechanical ventilation in patients with cardiogenic shock complicating acute myocardial infarction

Abstract

Funding Acknowledgements Type of funding sources: None. Introduction Invasive mechanical ventilation (IMV) provides up to a 30% reduction in cardiac output requirements and is frequently used in patients with cardiogenic shock following acute myocardial infarction (AMICS). However, practice of IMV in the setting of AMICS is sparsely described. Purpose The aim was to evaluate the use of IMV in a contemporary cohort of patients with AMICS. Methods Between 2010 and 2017, all adult AMICS patients admitted to two tertiary heart centres, providing AMICS care for two thirds of the Danish population, were individually identified through patient records. Temporal changes in application of IMV were registered as well as patient characteristics. Real-time electronic ICU data were available for a subset of mechanically ventilated ICU patients (n=566), and were retrieved for the first 24 hours of IMV, following ICU admission, and described in relation to 30-day survival and the presence of out-of-hospital cardiac arrest (OHCA). Results A total of 1716 AMICS patients were retrospectively identified, of which 1274 (74%) received IMV during ICU admission (IMV-ICU). The proportion of IMV increased from 70% in 2010 to 78% in 2017. IMV-ICU patients were younger (67 vs 76 years), more frequently male (79% vs 61%), and more likely to have OHCA (54% vs 3%) and higher lactate at diagnosis of cardiogenic shock (5.8 vs 4.1 mmol/L) compared with non-intubated patients (p for all<0.001). Among IMV-ICU patients, 69% were intubated in the prehospital setting, of which three quarters presented with OHCA. Median PaO2 and PCO2 were both within normal ranges among 30-day survivors and non-survivors. However, non-survivors required 10-25% higher median fraction of inspired oxygen (p<0.001; FiO2), positive-end-expiratory pressure (p=0.002; PEEP), and minute ventilation (p<0.001; MV). Differences in IMV settings were mainly driven by non-survivors without OHCA (see table). Conclusion In a contemporary cohort of patients with AMICS, use of IMV increased during the observation period from 2010 to 2017. Observations did not reveal any association between 30-day mortality and IMV parameters in OHCA-patients, whereas FiO2, PEEP, and MV were significantly elevated in 30-day non-survivors without OHCA.