Patterns and impact of arterial CO2 management in patients with Acute Respiratory Distress Syndrome: Insights from the LUNG SAFE study

Background: Considerable variability exists regarding CO2 management in early ARDS, with the impact of arterial CO2 tension on management and outcomes poorly understood. Research question: To determine the prevalence and impact of hypo- and hypercapnia on the management and outcomes of patients with early ARDS enrolled in the Large observational study to UNderstand the Global impact of Severe Acute respiratory FailurE (LUNG SAFE) study, an international multicenter observational study. Study design and methods: Our primary objective was to examine the prevalence of Day 1 and sustained (day 1 and 2) of hypocapnia (PaCO2 < 35mmHg), normocapnia (PaCO2 35-45mmHg) and hypercapnia (PaCO2 > 45mmHg) in patients with ARDS. Secondary objectives included elucidating the effect of CO2 tension on ventilatory management, and examining the relationship with ARDS outcome RESULTS: Of 2,813 patients analyzed, 551 (19.6%, 95%CI: 18.1-21.1) were hypocapnic, 1,018 (36.2%, 95%CI: 34.4-38.0) were normocapnic, while 1,214 (43.2%, 95%CI: 41.3-45.0) were hypercapnic, on day 1. Sustained hypocapnia was seen in 252 (9.3%, 95%CI: 8.2-10.4), sustained normocapnia in 544 (19.3%, 95%CI: 17.9-20.8) and sustained hypercapnia in 654 (24.1%, 95%CI: 22.5-25.7) patients. Hypocapnia was more frequent and severe in patients receiving non-invasive ventilation but was also observed in patients on controlled mechanical ventilation. Sustained hypocapnia was more frequent in middle income countries, while sustained hypercapnia was more frequent in Europe. ARDS severity profile was highest in sustained hypercapnia, and these patients received more protective ventilation. There was no independent association between arterial CO2 and outcome. In propensity matched analyses, hospital mortality was 36% in both sustained normocapnic and hypercapnic patients (P=1.0). ICU mortality was higher in patients with mild to moderate ARDS receiving sustained hypocapnia (38.1%) compared to normocapnia (27.1%). Interpretation: There was no evidence for benefit or harm with hypercapnia. Of concern, ICU mortality was higher with sustained hypocapnia in mild-moderate ARDS.