BACKGROUND: Management of a patent ductus arteriosus (PDA) after pharmacological therapy failure in preterm neonates is controversial and shows marked practice variation. To evaluate which factors motivate the decision to ligate a PDA in clinical practice we examined several clinical and echocardiographic variables.
METHODS: We conducted a retrospective single center cohort study. We included infants born at less than 37 weeks of gestation, admitted to our neonatal intensive care between 01.01.2008 and 31.12.2015 with a PDA detected on echocardiography after two or three courses of medical therapy. Logistic regression analyses were used to predict surgical ligation for twelve clinical and nine echocardiographic variables separately. We used the multiple imputation technique for missing values.
RESULTS: A total of 89 neonates were included of which forty (45%), underwent surgical ligation of their PDA. In our final multivariate regression model, invasive respiratory support (OR 3.6, 95% CI 1.29–10.03), left atrial/aortic root ratio (OR 5.48, 95% CI 1.66–18.11) and presence of ductal steal (OR 3.82, 95% CI 1.47–9.91) were significant predictors for surgical ligation. The prediction model using clinical and echocardiographic variables explained 9% and 24% of the variability to ligate respectively, indicating significant residual variation due to unmeasured factors.
CONCLUSIONS: Our results indicate that invasive respiratory support, increased left atrial/aortic root ratio and the presence of ductal steal were important predictors for surgical ligation in our center. However, this explained only a small proportion of the variability, which emphasizes the need for evidence-based guidelines in the management of preterm neonates after failed pharmacological therapy for a PDA.