Can the presumed risk of autosomal trisomy be decreased in fetuses of older women following a normal sonogram?

Previous studies showed that a sonographic scoring system can be used to identify women under age 35 who are at an increased risk for a fetal autosomal trisomy. We propose to use the same scoring system to select women over age 35 who are at a decreased risk of a fetal autosomal trisomy because the ultrasonogram is normal. We reviewed the sonographic scores of 97 trisomic and 694 normal control fetuses from two previously published consecutive series designed to identify women at increased risk for an affected fetus. Using the same scoring index in this study, we calculated the sensitivity and specificity of a score = 0 zero for identifying fetuses with autosomal trisomy. We then applied Bayes' theorem to determine the probability of an autosomal trisomy at various maternal ages, given a sonographic score of 0. Of the 97 fetuses with an autosomal trisomy, 83 had a score of > or = 1, for a sensitivity of 86% (95% CI 77-92%). Of the 694 control fetuses with normal karyotype, 606 had a score of 0, for a specificity of 87% (95% CI 83-89%). A 42 year old woman's probability of an autosomal trisomy if the sonographic score is 0 becomes equivalent to the age-specific probability for a 35 year old woman. Using the lower limit of the CIs for sensitivity and specificity, we calculated that the probability of having a fetus with an autosomal trisomy falls from 18.8 in 1000 to 5.3 in 1000 for a 40 year old woman with a sonographic score of 0.(ABSTRACT TRUNCATED AT 250 WORDS)