Hemoglobinopathy screening in pregnancy: Comparison of two protocols

This study was designed to determine the ability of a hemoglobinopathy scre ening protocol involving sickle solubility testing and red blood cell (RBC) indices to identify at-risk pregnancies. Retrospective chart review of all patients registering for prenatal care at the New York Hospital/Cornell Me dical Center prenatal clinic in 1996 was the study design. All patients had RBC indices as well as hemoglobin electrophoresis. RBC indices of those wi th normal and abnormal electrophoreses were compared. Comparison of protoco ls involving universal hemoglobin electrophoresis and selective use of hemo globin electrophoresis were compared. Student's t-test was used for statist ical analysis. There were 36 carriers of hemoglobinopathy traits in 631 pat ients screened (5.7%). Four (three with hemoglobin C trait and one with hem oglobin D trait) had normal RBC indices and presumably would have had negat ive sickle solubility testing. The sensitivity, specificity, positive predi ctive value (PPV), and negative predictive value (NPV) of a protocol with s elective use of hemoglobin electrophoresis would have been 88.9, 79.4, 20.8 , and 99.2%, respectively. Cost analysis reveals a difference of $11,384 or $18/patient less in a protocol with selective use of hemoglobin electropho resis. Although a protocol involving solubility testing with RBC indices wi ll not identify every carrier of a hemoglobinopathy trait, it may be approp riate in some populations.