Safety and efficacy of restrictive red blood cell transfusion strategy in extremely low birth weight infants: a before-after comparative study.

BACKGROUND: Extremely low birth weight (ELBW) infants frequently require red blood cell (RBC) transfusions, yet optimal hemoglobin thresholds remain debated. Recent evidence supports restrictive transfusion strategies, but real-world implementation data in Chinese neonatal intensive care units (NICUs) are limited. This study aimed to evaluate the safety and efficacy of implementing a restrictive RBC transfusion protocol in ELBW infants. METHODS: This multicenter before-after comparative study enrolled ELBW infants (birth weight < 1,000 g) admitted to two tertiary NICUs between January 2021 and December 2025 Infants admitted during the liberal transfusion period (January 2021-December 2022 n = 82) were compared with those admitted after implementation of a restrictive transfusion protocol (January 2023-December 2025, n = 94). Liberal thresholds were hematocrit < 35% with respiratory support and < 30% without respiratory support; restrictive thresholds were hematocrit < 30% with respiratory support and < 25% without respiratory support. Primary outcomes included in-hospital mortality and composite neonatal morbidity. Secondary outcomes included the number of RBC transfusions, donor exposure, bronchopulmonary dysplasia (BPD), retinopathy of prematurity (ROP), necrotizing enterocolitis (NEC), and intraventricular hemorrhage (IVH). RESULTS: Baseline characteristics were comparable between groups regarding gestational age (26.8 ± 1.6 vs. 26.9 ± 1.5 weeks, P = 0.674), birth weight (811 ± 124 vs. 804 ± 130 g, P = 0.713), and initial hematocrit (48.2 ± 5.1 vs. 47.8 ± 4.9%, P = 0.601). The restrictive group received significantly fewer RBC transfusions per infant (3.2 ± 2.1 vs. 5.8 ± 2.7, P < 0.001) and had reduced donor exposure (1.5 ± 1.2 vs. 2.9 ± 1.4 donors, P < 0.001). In-hospital mortality did not differ significantly between groups (9.6% vs. 11.0%, P = 0.955). No significant differences were observed in the incidence of BPD (38.3% vs. 42.7%, P = 0.662), severe ROP requiring treatment (8.5% vs. 9.8%, P = 0.981), NEC ≥ stage II (6.4% vs. 7.3%, P = 1.000), or severe IVH (grade III-IV) (10.6% vs. 12.2%, P = 0.931). Length of hospital stay was similar between groups (78.4 ± 24.3 vs. 81.2 ± 26.6 days, P = 0.465). CONCLUSION: Implementation of a restrictive RBC transfusion strategy in ELBW infants significantly reduced transfusion requirements and donor exposure without increasing mortality or major neonatal morbidities. These findings support the adoption of restrictive transfusion thresholds in Chinese NICUs caring for ELBW infants.