Fig. 3

Funnel plot of the risk of late-onset sepsis (LOS) in preterm infants exposed to prolonged versus short antibiotic exposure. Egger’s regression analysis showed no evidence of publication bias (Egger’s test, p = 0.27)