Advertisement

33: Assessment of risk for fetal loss by maternal characteristics with first and second trimester maternal serum markers

      Objective

      To develop and evaluate a method of estimating patient-specific risk for fetal loss by combining maternal characteristics with 1st and 2nd trimester serum markers.

      Study design

      Data was obtained on 35,603 women from the FASTER trial. Separate likelihood ratios (LRs) were estimated directly for significant maternal characteristics and, using a multivariate Gaussian model of the frequency distributions, for significant serum markers. Patient-specific risk was calculated by multiplying the incidence of fetal loss, as an odds, by the LRs for each maternal characteristic and for different serum marker combinations. Risks were aggregated to estimate the detection rate (DR) for a fixed 1%, 5% and 10% false-positive rate (FPR).

      Results

      193 women had fetal loss <20 weeks (early) and 189 >20 weeks (late). The significant characteristics were age, BMI, race, parity, threatened abortion and previous abortions; with a low DR when risk assessment was based on all of them. The significant markers for early loss were AFP, uE3 and PAPP-A; risk assessment based on all of them yielded 40-54% detection, and only increased slightly when characteristics and markers were combined (Table). Only inhibin was a significant marker for late loss and the DR was poor at 8%, 20% and 32% at FPRs of 1%, 5% and 10%, respectively, in combination with maternal characteristics.
      Tabled 1
      1% FPR5% FPR10% FPR
      characteristics4%16%24%
      marker
      AFP32%44%50%
      uE327%33%36%
      PAPP-A8%17%24%
      All40%49%54%
      characteristics + marker
      AFP30%46%55%
      uE328%36%43%
      PAPP-A4%21%54%
      All40%50%55%

      Conclusion

      Patient-specific risk assessment for early fetal loss using serum markers, with or without maternal characteristics, has a reasonably high detection.