Maternal Weight Correction for α-Fetoprotein: Mathematical Truncations Revisited

Abstract
The objective of this study was to determine if maternal weight correction as implemented 20 years ago for alpha-fetoprotein (AFP) testing is still appropriate, and what should be the weight cutoff for correction. AFP measurements from 44,951 biochemical screening cases from a single laboratory were divided by maternal weight, and separated into high >/=2.5 multiples of the median (MOM), and low /= 2.5 MOM). Using a truncated range to 200 lbs. (for women greater than 200), produces more uniform results than using an upper cutoff of 250 for truncation (2.7% vs 5.0% >/= 2.5 MOM). The same relationship is seen for "lows." Without correction, higher weight lowers AFP, correction to 250 lbs. overcompensates, and 200 lbs. gives the most even distribution, and, therefore, optimizes the distribution of highs.