Development of a clinical prediction rule for iron deficiency anemia in pregnancy.

Abstract

The purpose of this study was to elucidate if anemia in pregnancy caused by iron deficiency could be predicted by simple measures obtained on a complete blood count (CBC) using ferritin as the gold standard. One hundred forty-one pregnant patients with anemia were recruited during their first prenatal visit. CBC and ferritin level were obtained and candidate-predictors identified. Receiver operator characteristic (ROC) curves, stratified analysis, and logistic regression analysis were used to develop a prediction rule with maximum specificity. A hemoglobin level < or = 9.7 and a red cell distribution width (RDW) > or = 15 (in gestational age <20 weeks) predicted iron deficiency with high specificity. Our final model correctly classified 79.43% of the patients (area under the ROC curve of 0.88 [95%CI 0.82-0.93], specificity of 95.74%). Iron deficiency anemia can be predicted in pregnancy using lower cost tests, which could be an incredibly useful tool in areas with limited resources and a high prevalence of the disease. This study was limited to a mostly African American population in an urban setting, and the results may not be generalized to other populations.