Implementation of a screening program for diagnosing open neural tube defects: selection, evaluation, and utilization of alpha-fetoprotein methodology.

We evaluated and compared three different commercial kit immunoassays for alpha-fetoprotein (AFP) before we implemented our neural tube defect screening program. Each kit can be used with either serum or amniotic fluid. Analytical recovery ranges for AFP reference sera within each kit's standard curve limits (in kilo-int. units/L) were 97-108% (7.5-180) for the Kallestad kit, 77-101% (21.8-436) for Amersham, and 92-100% (0-177) for Hybritech. CVs, within each manufacturer's standard-curve limits, for combined intra-assay (amniotic fluid pools) and inter-assay (kit serum controls) averaged 3.6-7.3% (Kallestad), 2.4-9.3% (Amersham (y) kit results showed a correlation of r = 0.97, y = 1.05x + 5.5 kilo-int. units per liter of maternal serum (n = 66; range, 2.0-98.5). Gestational age did not influence these assay correlations. The Kallestad AFP assay demonstrated a maternal serum positivity rate of 2.9% at greater than or equal to 2.5 (n = 655) and 8.9% at less than 0.5 (n = 423) multiples of the median. All kits performed well analytically.