Background: Hematocrit falls after birth in preterm infants due to physiological factors and blood letting. Low plasma levels of erythropoietin (EPO) in preterm infants provide a rationale for the use of EPO to prevent or treat anemia.
Primary objective: To assess the effectiveness and safety of late initiation of EPO (initiated at 8 days after birth or later) in reducing the use of red blood cell transfusions in preterm and/or low birth weight infants.
Secondary objectives: Subgroup analyses of low (< 500 IU/kg/week) and high (> 500 IU/kg/week) doses of EPO and within these subgroups analyses of the use of low (< 5 mg/kg/day) and high (> 5 mg/kg/day) doses of supplemental iron, in reducing the use of red blood cell transfusions in these infants.
Search strategy: MEDLINE, EMBASE, CINAHL, abstracts from scientific meetings published in Pediatric Research and reference lists of identified trials and reviews were searched in November 2005/April 2006 and the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 2, 2006). No language restrictions were applied.
Selection criteria: Randomised or quasi-randomized controlled trials of late initiation of EPO treatment (started at eight days of age or later) vs. placebo or no intervention in preterm (< 37 weeks) and/or low birth weight (< 2500 g) neonates. For inclusion the studies needed to provide information on at least one outcome of interest.
Data collection and analysis: Data were abstracted by the two authors on pre-tested data collection forms. Data were entered by one review author (AO) and checked for accuracy by the other (SA). Data were analysed using RevMan 4.2.8. The statistical methods included relative risk (RR), risk difference (RD), number needed to treat to benefit (NNTB), number needed to treat to harm (NNTH) for dichotomous outcomes and weighted mean difference (WMD) for continuous outcomes reported with their 95% confidence intervals (CI). A fixed effects model was used for meta-analyses. Heterogeneity tests including the I squared (I(2)) statistic were performed to assess the appropriateness of pooling the data.
Main results: Twenty-eight studies enrolling 1302 preterm infants in 21 countries were included. The quality of the trials varied. Most trials were of small sample size. Only one study clearly stated that infants were excluded if they had received red blood cell transfusion prior to study entry (Samanci 1996). A total of 19 studies including 912 infants reported on the primary outcome of "Use of one or more red cell transfusions". The meta-analysis showed a significant effect [typical RR; 0.66 (95% CI; 0.59, 0.74); typical RD -0.21 (95% CI; -0.26, -0.16); typical NNTB of 5 (95% CI 4, 6)]. There was statistically significant heterogeneity [for RR (p < 0.00001), I(2 )= 74.0% and for RD (p = 0.0006), I(2 )=58.9%]. Similar results were obtained in secondary analyses based on different combinations of high/low doses of EPO and iron supplementation. There was a significant reduction in the total volume (ml/kg) of blood transfused per infant (four studies enrolling 177 infants) [typical WMD = -7 ml (95% CI -12, -3)] and in the number of transfusions per infant (nine studies enrolling 567 infants); [typical WMD -0.78 (-0.97, -0.59)]. The effect size was less in a post hoc analyses of high quality studies compared to studies in which the quality was uncertain and in studies that used strict guidelines for red blood cell transfusions vs. studies that did not. There were no significant differences in mortality, retinopathy of prematurity, sepsis, intraventricular haemorrhage, periventricular leukomalacia, necrotizing enterocolitis, bronchopulmonary dysplasia, SIDS, neutropenia, hypertension, or length of hospital stay. Long-term neurodevelopmental outcomes were not reported.
Authors' conclusions: Late administration of EPO reduces the use of one or more red blood cell transfusions, the number of red blood cell transfusions per infant and the total volume of red blood cell transfused per infant. The clinical importance of the results for the latter two outcomes is marginal (< 1 transfusion per infant and 7 ml/kg of transfused red blood cells). Any donor exposure is likely not avoided as most studies included infants who had received red cell transfusions prior to trial entry. Late EPO does not significantly reduce or increase any of many important neonatal adverse outcomes including mortality and retinopathy of prematurity. Further research of the use of late EPO treatment to prevent donor exposure is not indicated. Research efforts should focus on limiting donor exposure during the first few days of life in sick neonates, when red blood cell requirements are most likely to be required and cannot be prevented by late EPO treatment.