Risk analysis of hemolysis and establishment of prenatal prediction model in newborns due to maternal-fetal ABO blood type incompatibility

This study aims to investigate the impact of blood type, antibody titer, and pregnancy history on the incidence and severity of ABO hemolytic disease of the newborn (HDN). Data from 214 ABO HDN cases (2018–2021) were analyzed using elution tests for diagnosis and transcutaneous bilirubin for severity. The positivity rate was significantly higher in the OA group (86.73%) than in the OB group (53.47%, P < 0.01), although differences in bilirubin levels between the two groups were limited. An antibody titer of ≥ 64 was associated with a marked increase in HDN incidence; however no consistent correlation with bilirubin levels was observed. Pregnancy history showed no significant correlation. Logistic regression identified blood type and antibody titer as significant predictors of HDN incidence (odds ration OR = 6.09, P < 0.01; OR = 1.001, P < 0.05), whereas pregnancy history was not statistically significant (OR = 0.633, P > 0.05). The prediction model, expressed as logit(P) = lnP/(1−P) = 0.023 + 1.806 × blood type (1) + 0.001 × titer − 0.457 × pregnancy history (1), achieved an area under the curve (AUC) of 0.776, indicating moderate predictive accuracy. The study suggests that newborn blood type and maternal antibody titer are key factors for predicting ABO HDN incidence, though not its severity, while pregnancy history has limited relevance. This predictive model provides a moderately accurate tool for risk assessment.