Minimisation of blood sampling losses in preterm neonates: a systematic review and meta-analysis

Objective To evaluate the effect of minimising blood sampling losses on red blood cell (RBC) transfusion-related outcomes in preterm infants <37 weeks' gestation. Study design We searched PubMed, Embase, Web of Science and Google Scholar from inception to October 2024 for studies that evaluated sampling stewardship practices (SSP) in preterm infants during initial hospitalisation. Two authors independently screened articles that evaluated one or more sampling approaches to minimise blood loss or non-invasive methods to avoid sampling losses. Meta-analysis was conducted using a random effects model. Results Eighteen studies (4 randomised controlled trials (RCTs) and 14 non-randomised studies) were included. Five studies used umbilical cord blood sampling, four used protocol-based sampling and two used retransfusion of sampled blood back to the infant as an SSP. Sampling care bundles were used in seven studies. Meta-analysis showed that SSP reduced early RBC transfusion rates (RCTs: Relative risk(RR) =0.50, 95% CI 0.36, 0.68; non-RCTs: RR=0.78, 95% CI 0.69, 0.90), the average number of transfusions per infant (RCTs: mean difference=-0.4 transfusions, 95% CI -0.68, -0.05; non-RCTs: standardised mean difference=-0.40, 95% CI -0.55, -0.25) and the rates of multiple transfusions (non-RCTs: RR=0.51, 95% CI 0.42, 0.62). There were no significant effects on mortality and other morbidities. Certainty of evidence was high for transfusion-related outcomes and moderate for other outcomes. Conclusion SSPs are associated with a significant reduction in RBC transfusion rates among very and extremely preterm infants. Large RCTs are required to assess the effects of SSP on other important outcomes.