Dynamic Model Predicts Individualized Benefits and Risks of Platelet Transfusion in Preterm Infants with Thrombocytopenia: JAMA
Netherlands: Researchers have found in a new modeling study that in severely thrombocytopenic preterm infants, there is a significant individual variation in the predicted benefits and harms of prophylactic platelet transfusion at platelet counts below 50 × 10⁹/L, depending on clinical characteristics. The dynamic prediction model showed strong validation performance and may aid personalized transfusion decision-making pending further evaluation.
The study, published in JAMA by Hilde van der Staaij and colleagues from the Department of Clinical Epidemiology, Leiden University Medical Center, the Netherlands, aimed to determine when and in which preterm infants prophylactic platelet transfusions reduce the risk of major bleeding or death. Despite the frequent use of platelet transfusions in infants with severe thrombocytopenia, clinical guidance on their individualized benefit has been limited.
The researchers developed a dynamic prediction model using data from 1,042 preterm infants (<34 weeks’ gestation) treated across 14 neonatal intensive care units in the Netherlands, Sweden, and Germany between 2017 and 2021. The model compared two transfusion strategies: receiving a platelet transfusion within six hours (prophylaxis) versus no transfusion for three days (no prophylaxis). Predictions were updated every two hours during the first week following the onset of severe thrombocytopenia.
Key predictors included gestational and postnatal age, small-for-gestational-age status, necrotizing enterocolitis, sepsis, mechanical ventilation, vasoactive support, platelet counts, and prior transfusions.
Key Findings:
- The model was externally validated using 637 infants from a separate Dutch cohort (2010–2014).
- Across both development and validation cohorts, the median gestational age was 28 weeks, and the median birth weight was 900 g.
- Major bleeding or death occurred in 23% of infants in the development cohort and 21% in the validation cohort.
- The model showed strong discriminatory performance, with a time-dependent AUC of 0.69 for the prophylaxis strategy and 0.85 for no prophylaxis.
- Calibration analyses confirmed that the model provided reliable predictions across different clinical scenarios.
The study highlighted substantial variation in predicted outcomes depending on an individual infant’s clinical status at the time of transfusion. While some infants were likely to benefit from prophylactic platelet transfusion, others faced minimal benefit and potential harm, emphasizing the importance of individualized decision-making rather than uniform transfusion thresholds.
The authors note that the model could serve as a decision-support tool, helping clinicians balance the benefits and risks of platelet transfusion for each infant. However, they caution that further research is needed to evaluate the model’s impact in clinical practice. Limitations include the reliance on observational data, potential residual confounding, and restriction to preterm infants in high-resource neonatal units, which may affect generalizability.
"The study provides evidence that personalized platelet transfusion strategies based on an infant’s dynamic clinical characteristics could optimize outcomes in severely thrombocytopenic preterm infants. By accounting for individual risk factors, the dynamic prediction model has the potential to guide safer, more effective transfusion decisions and reduce unnecessary exposure to transfusions," the authors concluded.
Reference:
van der Staaij H, Prosepe I, Caram-Deelder C, et al. Individualized Prediction of Platelet Transfusion Outcomes in Preterm Infants With Severe Thrombocytopenia. JAMA. 2025;334(14):1267–1277. doi:10.1001/jama.2025.14194
Next Story