BackgroundThere is limited information about the mortality and neurodevelopmental outcomes of very preterm infants (= 1 year corrected age were extracted as the main outcome measure. Data were pooled separately for adjusted and unadjusted ORs using the random-effects model. The evidence level was assessed using the GRADE (Grading of Recommendations, Assessments, Development and Evaluations) framework. ResultsEighteen cohort studies (13 606 infants) were included. Meta-analysis of unadjusted ORs showed that SIP was significantly associated with increased odds of mortality, cerebral palsy, composite outcome of death or disability, visual impairment and hearing impairment. However, pooling of adjusted ORs (aOR) found significant associations only for mortality (aOR (95% CI) 2.27 (2.07 to 2.49); I-2: 0%; four studies (n=10 695)), severe disability (aOR (95% CI) 2.06 (1.38 to 3.08); I-2: 0%; two studies (n=321)) and composite outcome of 'death or disability' (aOR (95% CI) 2.18 (1.55 to 3.06); I-2: 0%; two studies (n=321)). The level of evidence was 'low' or 'very low'. LimitationsLack of information on aORs from many studies. ConclusionsSIP in very preterm infants is associated with higher odds of mortality, severe disability, and death or disability.