Fig. 2

Sensitivity analyses stratifying 30-day-mortality in subgroups for patient-specific characteristics using logistic regression producing univariate odds ratios. The 30-day-mortality was similar between patients with and without paracetamol intake regardless of treatment limitations, the use of NIV, age strata and the time from symptom onset until admission. Patients categorized as vulnerable according to CFS (OR 0.36) and without endotracheal intubation and vasopressor use (OR 0.62, respectively) were more likely to take paracetamol