In-hospital Outcomes and Early Hemodynamic Management According to Echocardiography Use in Hypotensive Preterm Infants: A National Propensity-Matched Cohort Study

International audienceBackground Hypotension is a common condition during the first postnatal days of very preterm infants and has been associated with an increased risk of adverse outcomes but its management remains controversial. There is a consensus to promote the use of neonatologist-performed echocardiography (NPE) in hypotensive very preterm infants, although no clinical trial ever assessed this practice. Methods We conducted a retrospective analysis of prospectively collected data from the French national EPIPAGE-2 cohort to evaluate the association of NPE with survival, severe morbidity, and therapeutic management in very preterm infants with early hypotension. Reasons for administering antihypotensive treatments were also analyzed. We included infants born before 30 weeks of gestation with hypotension within 72 h of birth. Infants managed with (NPE group) or without (no-NPE group) NPE use were compared after matching on gestational age and a propensity score, reflecting each patient's probability of having an NPE based on his/her baseline covariates. This matching procedure intended to control for the indication bias of NPE. Results Among 966 eligible infants, 809 were included (NPE group, n = 320; no-NPE group, n = 489), and 229 from each group could be matched. The NPE group did not differ significantly from the no-NPE group for survival (OR 1.01, 95% CI 0.64 to 1.60; p = 0.95) or survival without severe morbidity at discharge (OR 0.92, 95% CI 0.63 to 1.34; p = 0.66), but received more antihypotensive treatments [144/229 (62.9%) vs. 99/229 (43.0%), p < 0.001]. Isolated hypotension was the main reason for treatment in both groups. Among treated infants, volume expansion was administered at equal rates to the NPE and no-NPE groups [118/144 (82.1%) vs. 79/99 (80.1%), p = 0.67], but the NPE group received inotropic drugs more often [77/144 (53.7%) vs. 37/99 (37.8%), p = 0.023]. Conclusion NPE use in hypotensive preterm infants was not associated with in-hospital outcomes and had little influence on the nature of and reasons for antihypotensive treatments. These results suggest the need to optimize NPE use