Maternal serum screening for trisomy 18: assessing different statistical models to optimize detection rates

Area of Science:

• Prenatal screening
• Maternal serum marker analysis
• Chromosomal abnormality detection

Background:

• Trisomy 18 screening is crucial for prenatal diagnosis.
• Maternal serum markers like AFP and hCG are used in screening protocols.
• Evaluating different analytical models can optimize screening accuracy.

Purpose of the Study:

• To evaluate three alternative models for trisomy 18 screening.
• To compare the performance of bivariate and univariate analyses using AFP and hCG.
• To assess the impact of using population-specific versus published distribution parameters.

Main Methods:

• Utilized data from 46 affected and 48,150 unaffected pregnancies.
• Calculated AFP and hCG multiples of the median (MoMs) and their product.
• Compared detection and false positive rates for different models and parameter sets.

Main Results:

• A fixed cut-off for AFP and hCG yielded a 28.3% detection rate at 0.5% FPR.
• The univariate model showed a higher detection rate (32.6%) at 0.5% FPR using published parameters.
• Locally derived parameters improved the bivariate model's detection rate at higher FPRs but not lower ones.

Conclusions:

• Trisomy 18 screening using AFP and hCG is a valuable addition to Down syndrome screening.
• The univariate analysis model appears slightly more effective for trisomy 18 screening.
• The choice of distribution parameters can influence screening performance, particularly for the bivariate model.