Biomechanics of ARDS in COVID-19: A prospective observation study

Document Type


Publication Date


Publication Title



PURPOSE: Early clinical observations of novel coronavirus-associated Acute Respiratory Distress Syndrome (C-ARDS) reported a higher than expected compliance discordant to the degree of hypoxemia. Our goal was to evaluate survival of patients with CARDS according to different lung mechanical parameters. METHODS: We conducted a prospective, observational study of C-ARDS patients admitted to the ICU at a tertiary care hospital in Michigan (March 1st to May 1st, 2020). Subjects were enrolled at the time of intubation and followed until death or discharge. The average lung compliance (AvC) measured in ml/cm H2O and the ratio of arterial oxygen partial pressure to fractional inspired oxygen (P/F ratio) were calculated daily over the first 7 days post-intubation. Median differences in AvC and P/F ratio between alive and deceased participants were calculated using Wilcoxon Rank Sum. Pearson correlation was used to describe the relationship between AvC and P/F ratio. Multivariate logistic regression was performed to explain co-variate effect after adjustment. Survival function was performed using the Kaplan-Meier estimator for different levels of AvC. RESULTS: 135 patients were included; mortality was 41.4% (n¼56). Baseline demographics were similar between the survivor and deceased groups, except for hypertension and coronary artery disease (CAD) trending towards significance in the latter. Survivors had longer length of stay (34.7 days vs. 17.6 days; p ¼ 0.001). There was no significant correlation between AvC and P/F ratio (Pearson’s r 0.09; p¼0.31). Surviving and deceased patients had different median P/F values (185 [IQR 242 to 135] vs 150 [IQR 196 to 109]; p¼0.01), but did not differ significantly in terms of median AvC (29 [IQR 35 to 26] vs 31 [IQR 37 to 24]; p 0.69). There was no difference in survival curves between AvC levels of 35, 40 and 45. A multivariate logistic regression analysis was done to adjust for gender, BMI, hypertension, coronary artery disease and ventilator length of stay. Only age [OR ¼1.05, 95% CI: 1.02, 1.08, p 0.003] and PF value [OR ¼ 0.98, 95% CI: 0.98, 0.99, p 0.001] were significant risk factors for mortality. CONCLUSIONS: Our results suggest that low P/F ratio and advanced age are the most significant independent predictors of mortality in C-ARDS. The severity of hypoxemia did not correlate significantly with lung elasticity, and lung compliance was not predictive of mortality. Our findings are consistent with other reports on lung mechanics in C-ARDS (Schenck et al). CLINICAL IMPLICATIONS: As with typical ARDS, degree of hypoxemia at the onset of respiratory failure may have prognostic implications in patients with C-ARDS. Pulmonary compliance by itself does not predict risk of mortality.




4, Supplement

First Page


Last Page