Objective: Acute respiratory distress syndrome (ARDS) is considered the main cause of COVID-19 associated morbidity and mortality. Early and reliable risk stratification is of crucial clinical importance to identify patients developing a severe course of the disease. Deceleration capacity (DC) of heart rate as a marker of cardiac autonomic function predicts outcome in post-infarction and heart failure patients. We hypothesized that reduced modulation of heart rate can be suitable to detect patients with COVID-19 at risk for development of ARDS.

Methods: We prospectively enrolled 60 consecutive COVID-19 positive patients presenting at the University Hospital of Tuebingen between March and April 2020. Arterial blood gas analysis and 24h-Holter ECG recordings were performed and analyzed at admission. The primary end point was defined as development of ARDS with regards to the Berlin classification.

Results: 61.7% (37 of 60 patients) developed an ARDS. In patients with ARDS DC was significantly reduced when compared to patients with milder course of infection (3.2 ms vs. 6.6 ms, p < 0.001). DC achieved a good discrimination performance (AUC = 0.76) for predicting ARDS in COVID-19 patients. In multivariate analysis, DC was an independent predictor of ARDS (HR, 0.7; 95% CI, 0.6–1.0; P = 0.04) and superior to well-known predictive markers like troponin I, LDH or hypertension.

Conclusion: Our data suggest a promising role of DC for identification of high risk COVID-19 patients.