Natural history of postpartum hematocrit recovery in an urban, safety-net population

Abstract

Postpartum anemia is common after delivery, and postpartum blood transfusion is the leading indicator of severe maternal morbidity in the United States. Although hematologic changes during the antepartum period are well understood, little is known about postpartum hematologic changes. Therefore, we investigated the hematologic changes in the postpartum period in a large, contemporary cohort. This study aimed to characterize hematologic recovery in the postpartum period and evaluate the effect of demographics, medical conditions, and pregnancy characteristics on the recovery. In a contemporary cohort of deliveries in 2019 at a single institution, the hematocrit of postpartum women before hospital discharge was compared with the hematocrit of women at the postpartum follow-up. Our population was composed of a predominantly Hispanic population at an urban, safety-net hospital. All women received a complete blood count on postpartum day 1 and a spun hematocrit at their postpartum follow-up visit in our hospital system. All women were scheduled for a routine postpartum visit 2 to 3 weeks after delivery. All deliveries of a live-born infant with available postpartum hematocrit before hospital discharge and at postpartum follow-up were included. Demographics and pregnancy characteristics, along with medical conditions, were evaluated. To evaluate an uncomplicated cohort, those with multiple gestation, preeclampsia with severe features, chronic hypertension, and diabetes mellitus were excluded in the secondary analysis. Statistical analysis included chi-square, paired Student t test, Student t test for independent groups, and analysis of variance. Of 12,456 deliveries, 9003 met the inclusion criteria. The average number of days from discharge to follow-up was 21.73±10.39 days. The average hematocrit levels were 30.77±3.61 before discharge and 38.70±3.61 at postpartum follow-up. From discharge to postpartum follow-up, the hematocrit levels increased by an average of 7.93±3.24. In the cohort without complications, the average increase in hematocrit levels was 8.19±3.09. The rise in hematocrit levels was significantly lower for those with chronic hypertension (6.9±3.6), diabetes mellitus (7.3±3.3), and preeclampsia with severe features (6.9±3.7). The severity of anemia influenced the rise in hematocrit levels in the postpartum period. In women with postpartum anemia (hematocrit<30%), the rise in hematocrit levels was 9.49±2.97 in the uncomplicated cohort. Postpartum hemorrhage did not influence the rise in postpartum hematocrit levels; women receiving blood transfusion had a greater rise in hematocrit levels (9.01±3.29). Our study establishes the natural course of hematologic recovery in the postpartum period, and we found that women with asymptomatic postpartum anemia may have a hematocrit level of 37% to 39% at their postpartum follow-up approximately 3 weeks after hospital discharge. Women with preexisting and obstetrical complications experienced less hematologic recovery and adapted more slowly to postpartum physiological changes.