Pulmonary hypertension at admission predicts ICU mortality in elderly critically ill with severe COVID-19 pneumonia: retrospective cohort study

Abstract

Background: Point-of-care ultrasound (POCUS) is a useful diagnostic tool for non-invasive assessment of critically ill patients. Mortality of elderly patients with COVID-19 pneumonia is high and there is still scarcity of definitive predictors. Aim of our study was to assess the prediction value of combined lung and heart POCUS data on mortality of elderly critically ill patients with severe COVID-19 pneumonia. Methods: This was a retrospective observational study. Data of patients older than 70 years, with severe COVID-19 pneumonia admitted to mixed 25-bed, level 3, intensive care unit (ICU) was analyzed retrospectively. POCUS was performed at admission; our parameters of interest were pulmonary artery systolic pressure (PASP) and presence of diffuse B-line pattern (B-pattern) on lung ultrasound. Results: Between October 2020 and March 2021, 117 patients aged 70 years or more (average age 77 ± 5 years) were included. Average length of ICU stay was 10.7 ± 8.9 days. High-flow oxygenation, non-invasive ventilation and invasive mechanical ventilation were at some point used to support 36/117 (31%), 39/117 (33%) and 75/117 (64%) patients respectively. ICU mortality was 50.9%. ICU stay was shorter in survivors (8.8 ± 8.3 vs 12.6 ± 9.3 days, p = 0.02). PASP was lower in ICU survivors (32.5 ± 9.8 vs. 40.4 ± 14.3 mmHg, p = 0.024). B-pattern was more often detected in non-survivors (35/59 (59%) vs. 19/58 (33%), p = 0.005). PASP and B-pattern at admission, and also mechanical ventilation and development of VAP, were univariate predictors of mortality. PASP at admission was an independent predictor of ICU (OR 1.061, 95%CI 1.003–1.124, p = 0.039) and hospital (OR 1.073, 95%CI 1.003–1.146, p = 0.039) mortality. Conclusions: Pulmonary artery systolic pressure at admission is an independent predictor of ICU and hospital mortality of elderly patients with severe COVID-19 pneumonia.