Can lung ultrasound score accurately predict surfactant replacement? A systematic review and meta-analysis of diagnostic test studies

BackgroundClinical and radiographic criteria are traditionally used to determine the need for surfactant therapy in preterm infants. Lung ultrasound is a bedside test that offers a rapid, radiation-free, alternative to this approach. ObjectiveTo conduct a systematic review and meta-analysis to determine the accuracy of a lung ultrasound score (LUS) in identifying infants who would receive at least one surfactant dose. Secondary aims were to evaluate the predictive accuracy for >= 2 doses and the accuracy of a different image classification system based on three lung ultrasound profiles. MethodsPubMed, SCOPUS, Biomed Central, and the Cochrane library between January 2011 and December 2021 were searched. Full articles enrolling preterm neonates who underwent lung ultrasound to predict surfactant administration were assessed and analyzed following Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) and QUADAS-2 guidelines. ResultsSeven prospective studies recruiting 697 infants met the inclusion criteria. Risk of bias was generally low. Oxygen requirement, clinical and radiographic signs of respiratory distress syndrome were used as reference standards for surfactant replacement. The summary receiver operator characteristic (sROC) curve for LUS predicting first surfactant dose showed an area under the curve (AUC) = 0.88 (95% confidence interval [CI]: 0.82-0.91); optimal specificity and sensitivity (Youden index) were 0.83 and 0.81 respectively. Pooled estimates of sensitivity, specificity, diagnostic odds ratio, negative predictive value, and positive predictive value for LUS predicting the first surfactant dose were 0.89 (0.82-0.95), 0.86 (0.78-0.95), 3.78 (3.05-4.50), 0.92 (0.87-0.97), 0.79 (0.65-0.92). The sROC curve for the accuracy of Type 1 lung profile in predicting first surfactant dose showed an AUC of 0.88; optimal specificity and sensitivity were both 0.86. Two studies addressing the predictive accuracy of LUS for >= 2 surfactant doses had high heterogeneity and were unsuitable to combine in a meta-analysis. DiscussionDespite current significant variation in LUS thresholds, lung ultrasound is highly predictive of the need for early surfactant replacement. This evidence was derived from studies with homogeneous patient characteristics and low risk of bias.