Abstract

Purpose: Pulmonary vascular dysfunction has been associated with adverse prognosis in acute respiratory distress syndrome (ARDS). The diastolic pulmonary gradient (DPG) has been variably associated with adverse outcomes in pulmonary hypertension (PH) due to left heart disease, however the prevalence of elevated DPG, its determinants, and prognostic value in ARDS is unknown. We hypothesized that elevated DPG would be associated with death. Methods: We performed an analysis of the ARDS Network Fluid and Catheter Treatment Trial of pulmonary artery (PA) catheter use in ARDS, including 392 subjects with a complete hemodynamic data set on trial day 0. DPG was calculated as diastolic PA pressure minus pulmonary artery wedge pressure (PAWP). Cox proportional hazard modeling was performed with primary outcome of death. Results: Mean DPG was 8 ± 6 mmHg on study day 0. Two hundred twenty-four subjects (57.1%) had a DPG ≥ 7mmHg, and 183 of these (82%) had concomitant PH (mean PA pressure ≥ 25mmHg). The group with elevated DPG had higher PA pressure, pulmonary vascular resistance and airway pressures with lower PAWP (Table). PaO2 was lower in the high DPG group and there was a trend toward worsening acidosis (p= 0.06) and increasing heart rate (p= 0.09). Age, body-surface area, heart rate, PaO2, and pulmonary vascular resistance were predictors of DPG in a multivariate linear regression model. DPG ≥ 7 mmHg on day 0 did not predict mortality (HR 1.14 (0.78-1.67); p = 0.5), nor did DPG predict mortality when considered as a continuous variable or at various cutoffs ranging from 3 to 11 mmHg. For 274 subjects with available data on day 0 and day 3, mean DPG did not change significantly over the period (8 ± 5 mmHg on d0 v. 8 ± 5 mmHg on d3, p = 0.55). A DPG greater than 7 mmHg on day 3 did not predict mortality (HR 1.49 [.93-2.4] p = 0.1). Conclusion: Factors other than pulmonary vascular disease influence DPG in ARDS. In ARDS, an elevated DPG is common but does not predict mortality.