Fresh-frozen plasma (FFP) is sometimes administered to nonbleeding preterm neonates who are judged to be at risk for bleeding because they have abnormal coagulation tests. The benefits/risks of this practice are not well defined. One limitation to progress is lack of reference intervals for the common coagulation tests, thus limiting precision about whether coagulation tests are indeed abnormal.

Study Design and Methods:

In a sequential observational study, fetal blood was drawn at preterm birth (≤34 weeks) from the umbilical vein near the placenta. Fibrinogen, prothrombin time, activated partial thromboplastin time, D-dimer, platelet (PLT) count, and mean PLT volume were measured. Reference intervals were constructed using 5th and 95th percentile values. Associations were then sought between abnormal coagulation values at birth and bleeding problems identified during the first week.


Coagulation tests were drawn at 175 preterm deliveries and the results were organized into reference intervals by gestational age. No abnormal coagulation value, either alone or in combination, predicted hemorrhage (intraventricular, gastrointestinal, or pulmonary) during the first week. However, fibrinogen exceeding the 95th percentile was associated with evidence of in utero infection/inflammation (correlations with elevated C-reactive protein, p < 0.01, and elevated immature to total neutrophil ratio, p < 0.001).


Abnormal coagulation values at preterm birth do not predict bleeding during the first week. This suggests to us that bleeding in the days after preterm birth is not generally the result of in utero coagulopathy. These findings bring into question the value of coagulation screening of nonbleeding preterm infants and prophylactic FFP administration to those with abnormal values.