Restrictive platelet thresholds reduce transfusions without increasing death or bleeding risk in thrombocytopenic neonates

Rationale Neonatal thrombocytopenia is frequently treated with platelet transfusion. Thresholds prompting platelet transfusions vary greatly between providers, institutions, and countries. Thresholds for stable non‐bleeding infants have ranged from 25,000 to 150,000 per microliter. Objectives To assess the benefits and harms of more restrictive platelet transfusion thresholds compared to more liberal thresholds in thrombocytopenic neonates. Search methods We searched CENTRAL, MEDLINE, Embase, Epistemonikos, and trial registries until June 2025. We checked the reference lists of included studies and systematic reviews where subject matter related to the intervention or population examined in this review. Eligibility criteria We included randomized controlled trials (RCTs), quasi‐RCTs, and cluster‐RCTs that compared platelet transfusion thresholds in thrombocytopenic neonates with or without evidence of major bleeding. We excluded infants with inherited coagulopathies or active extracorporeal membrane oxygenation support. Outcomes Our outcomes were death prior to hospital discharge, major bleeding (onset after study entry) or worsening of existing major bleeding, any hemorrhage (all grades of hemorrhage/bleeding), neurodevelopmental impairment: cerebral palsy and global developmental delay, and any platelet transfusion. Risk of bias We used the Cochrane RoB 1 tool to assess risk of bias in the included studies. Synthesis methods We calculated risk ratio (RR) and risk difference (RD) with 95% confidence interval (CI) for each outcome. We analyzed and interpreted individual trials separately, and summarized the results narratively, as each comparison was informed by only a single trial. We used GRADE to assess the certainty of evidence for the prespecified outcomes. Included studies We included three studies (enrolling a total of 856 participants) comparing different platelet transfusion thresholds in thrombocytopenic neonates. Two of the studies were conducted in high‐income settings and one in a lower‐middle‐income country. All three studies varied with regard to their platelet transfusion thresholds, and were included in separate comparisons based on platelet transfusion thresholds. Liberal or high threshold in the: 1. mild thrombocytopenia range: transfusion for platelet count ≤ 100,000 per microliter compared to transfusion for mild thrombocytopenia (platelet count > 100,000 to 150,000 per microliter); 2. moderate thrombocytopenia range: transfusion for platelet count ≤ 50,000 per microliter compared to transfusion for moderate thrombocytopenia (> 50,000 to ≤ 100,000 per microliter); and 3. severe thrombocytopenia range: transfusion for platelet count < 30,000 per microliter (very severe thrombocytopenia) compared to transfusion for severe thrombocytopenia (30,000 to ≤ 50,000 per microliter). Synthesis of results Liberal or high threshold in the mild thrombocytopenia range The evidence suggests that transfusion for platelet count ≤ 100,000 per microliter may result in little to no difference in the risk of death prior to hospital discharge compared to transfusion for platelet count > 100,000 to 150,000 per microliter (RR 0.72, 95% CI 0.36 to 1.46; 152 participants; 1 study; very low‐certainty evidence); however, the evidence is very uncertain. Using a more restrictive platelet transfusion threshold may result in a large reduction in platelet transfusions when compared to using a more liberal threshold (RR 0.36, 95% CI 0.26 to 0.51; 152 participants, 1 study; low‐certainty evidence). Major bleeding, any hemorrhage, cerebral palsy, and global developmental delay were not reported in the studies. Liberal or high threshold in the moderate thrombocytopenia range The evidence suggests that transfusion for moderate thrombocytopenia may result in little to no difference in the risk of death prior to hospital discharge compared to transfusion for severe thrombocytopenia (RR 1.13, 95% CI 0.53 to 2.37; 44 participants; 1 study; very low‐certainty evidence), while there may be little to no difference in risk of any hemorrhage (RR 1.00, 95% CI 0.52 to 1.91; 44 participants, 1 study; very low‐certainty evidence); however, the evidence for both outcomes is very uncertain. Major bleeding, cerebral palsy, global developmental delay, and any platelet transfusion were not reported in the studies. Liberal or high threshold in the severe thrombocytopenia range The evidence suggests that using a more restrictive transfusion threshold may result in little to no reduction of death prior to discharge compared to transfusion for severe thrombocytopenia (RR 0.89, 95% CI 0.63 to 1.25; 1 study, 656 participants; low certainty‐evidence). The evidence suggests that transfusion for severe thrombocytopenia may result in little to no difference in major bleeding or worsening of existing major bleeding (RR 0.77, 95% CI 0.51 to 1.17; 1 study, 658 participants; 1 study, low‐certainty evidence), any hemorrhage (RR 1.02, 95% CI 0.92 to 1.13; 652 participants; 1 study, low‐certainty evidence), cerebral palsy (RR 0.78, 95% CI 0.46 to 1.33; 432 participants; 1 study, low‐certainty evidence), or global developmental delay (RR 0.67, 95% CI 0.45 to 1.01; 432 participants; 1 study, low‐certainty evidence). Using a more restrictive transfusion threshold probably results in a large reduction in platelet transfusions when compared to transfusion for severe thrombocytopenia (RR 0.59, 95% CI 0.53 to 0.66; 659 participants; 1 study, moderate‐certainty evidence). All three studies were considered to be at high risk of performance bias due to their unblinded designs. Two of the studies were small, and all studies lacked precision for the critical outcome of death before hospital discharge. Authors' conclusions The available evidence is of moderate to very low certainty and comes from only three studies. We found no evidence of a benefit of higher transfusion thresholds for any outcomes of interest, including death or bleeding, when giving a platelet transfusion to babies with higher platelet count levels (when there is less risk of bleeding). Of note, prior individual studies using different methods of analysis have raised concerns regarding the use of higher thresholds. This suggests that the use of lower platelet transfusion thresholds likely reduces platelet transfusions without a concomitant increased risk of death or major bleeding, although there is some uncertainty for critical outcomes. Funding This Cochrane review is funded in part by Vermont Oxford Network. Registration Protocol (2024) DOI: 10.1002/14651858.CD015341 PICOs PICOs Population Intervention Comparison Outcome