429 - Association of Anemia, Red Blood cell transfusions (RBCT) and Retinopathy of Prematurity (ROP)

Poster Number: 429
Publication Number: 429.137

Rakesh Rao, Washington University in St. Louis School of Medicine, Saint Louis, MO, United States; Tiffany J. Wu, Washington University in St. Louis School of Medicine, Thornton, CO, United States; Hongjie gu, Washington University in St. Louis School of Medicine, St. Louis, MO, United States; Margaret Reynolds, Washington University in St. Louis School of Medicine, Saint Louis, MO, United States

Background: ROP is associated with both anemia (low hemoglobin) and RBCTs in preterm infants. It is unclear whether a higher hemoglobin (Hgb) threshold ( >8 g/dl) or transfusions are associated with the development of severe ROP (sROP).

Objective: To determine the association of severe anemia (Hgb≤8 gm/dl or Hct≤23%), RBCT and sROP (laser photocoagulation or bevacizumab treatment).

Design/Methods: Single center cohort of patients screened for ROP over a 6 year epoch. Infants with sROP were matched with gestational age (GA) and gender matched controls. Clinical and demographic factors, weekly Hgb, hematocrit (Hct) and RBCT data were collected. Wilcoxson rank sum, Chi square and logistic regression tests were used for analyses. Cox proportional-hazards regression models to estimate cause-specific hazard and magnitude of effect of each risk factor, and random effects Cox proportional-hazards regression models were performed for covariates with repeated measurements. A P-value < 0.05 was considered significant.

Results: 311 patients with ROP were included, with 132 cases of sROP and 179 controls. Infants with sROP (Table 1) were of lower birth weight (p < 0.001) but not GA, intubated more often in the delivery room, develop bronchopulmonary dysplasia (BPD), receive steroids for BPD, undergo patent ductus arteriosus ligation (PDAL) and receive nitric oxide (iNO). Infants with sROP had more RBCT, started iron and reached full enteral feeds later. Infants with sROP had lower Hgb, lower Hct, received more RBCTs and a higher proportion had severe anemia (Table 2). In univariate regression analyses (OR, 95% CI), the total number of RBCT (1.06, 1.03-1.09, p< 0.001), day of starting iron supplementation (1.007, 1.003-1.012, p=0.002), steroids for BPD (2.12, 1.48-3.07, p< 0.001), iNO (1.78, 1.11-2.72, p=0.02) and PDAL (1.76, 1.21-2.53, p=0.0036) increased risks of sROP, whereas higher birth weight (per 100 gram increase) was protective (0.86, 0.77-0.96, p= 0.005). In random effects regression model, a higher hemoglobin (0.79, 0.65-0.97, p=0.02) but not Hct (0.95, 0.87-1.03, p=0.20) was protective, while number of RBCT (1.87, 1.15-3.03, p=0.01) increased risk of sROP. In multivariate random effects model, steroids for BPD increased the risks (OR, 95% CI = 3.5, 1.48-8.26, p=0.004) whereas a higher Hgb was protective (OR, 95% CI = 0.79, 0.64-0.97, p=0.02) against sROP.Conclusion(s): Low Hgb, not RBCT or Hct, increases risks of sROP. The use of steroids for BPD in the presence of low Hgb increases the risks of sROP over three fold whereas every 1g/dl increase in Hgb (above 8g/dl) decreased the risks of sROP by 21%.
Table 1: Clinical and Demographic characteristics
Table 2: Hemoglobin, Hematocrit (Hct), RBCT and Severe Anemia