Sequential and contingent prenatal screening for Down syndrome.

OBJECTIVE: To compare the Integrated test in three policies for prenatal Down syndrome screening: Integrated screening for all women, sequential screening (first-trimester tests allowing early completion of screening for high-risk pregnancies), and Contingent screening (early completion of screening for high- and low-risk pregnancies). DESIGN AND METHODS: Estimation of detection rates (DRs) and false-positive rates (FPRs) using Monte Carlo simulation and cost effectiveness for each method. SETTING AND POPULATION: Down syndrome affected and unaffected pregnancies studied in the Serum Urine and Ultrasound Screening Study (SURUSS). RESULTS AND MAIN OUTCOMES: Integrated screening has the best screening performance. The performance of the other two policies approached that of Integrated screening as the first-trimester test FPR decreased. If the first-trimester FPR is set to 0.5% (risk >or= 1 in 30) with an overall DR of 90%, sequential and contingent screening yield overall FPRs of 2.25% and 2.42%, respectively, and 66% of the affected pregnancies are detected by the first-trimester test. The Integrated test on all women yields an FPR of 2.15%. With sequential screening, 99.5% of women would proceed to an Integrated test, or 30% with contingent screening if those with first-trimester test risks of <or=1 in 2000 are classified screen-negative and receive no further testing. About 20% of affected pregnancies identified in the first trimester using sequential or contingent screening would have unnecessary terminations (they would miscarry before the early second trimester). Contingent screening is the most cost-effective if there is no alphafetoprotein screening for neural tube defects, otherwise Integrated screening is more cost-effective.
CONCLUSIONS: Integrated screening for all women is the simplest, most effective, and the safest policy. Contingent screening is the most complex with the lowest screening performance. Making an earlier diagnosis with sequential and contingent screening has adverse consequences that are sufficient to discourage their use.