A new approach to calculating the risk of chromosomal abnormalities with first-trimester screening data.

PURPOSE

First-trimester screening at 11 - 14 weeks has been proven to be very useful in the early detection of chromosomal defects. The aim of this project was to develop a CE-certified new risk calculation program (PRC = Prenatal Risk Calculation) using a nationwide database.

MATERIALS AND METHODS

The database underlying the new risk calculation procedure was established in Germany from 2003 through 2006. Overall, the database includes measurements from 70,030 pregnant women having given birth to healthy children. Following consideration of all pregnancies associated with a chromosomally abnormal outcome, the sample size was 451. The algorithm used for calculating the risk of a chromosomally abnormal outcome comprises the following variables: maternal age, crown-rump length (CRL) (restricted to a range from 45 - 84 mm or, equivalently, 11 + 1 - 14 + 0 weeks of gestation), nuchal translucency (NT), as well as the maternal serum parameters PAPP-A (pregnancy associated plasma protein A) and free beta-hCG (free human chorionic gonadotropin). In a preliminary cross-validation study, we applied both the new algorithm and the FMF UK program to an independent sample containing n = 40,568 pregnancies with negative outcome, n = 187 cases of trisomy 21, n = 34 trisomies 18 and n = 13 trisomies 13.

RESULTS

Using the primary sample of 70,030 pregnancies with a negative outcome, reference bands were constructed for the sonographic parameter fetal nuchal translucency and the biochemical parameters PAPP-A and free beta-HCG. Instead of MoM values we used "degree of extremeness" (DoE) values. This statistical parameter has been proven to give more precise results than the MoM measure because it assesses the deviation of the actual measurement value from the centre of the reference band expressed as a multiple of the width of the respective band section. The result of the risk calculation is visualized by means of a traffic light graph which allows the patient to comprehend her individual risk at first glance. The red color indicates a high risk, green a low risk, and yellow represents a moderate risk. In our preliminary cross-validation study the detection rate obtained for the German algorithm was 86.6 % for trisomy 21, 94.1 % for trisomy 18 and 92.4 for trisomy 13. The corresponding detection rates obtained with the same data by the FMF UK program were 86.1 %, 82.3 % and 69.2 % throughout. The false-positive rate was 5.0 % throughout.

CONCLUSION

The new risk calculation procedure of the FMF Germany (PRC) has been made available as a CE-certified computer program. In screening for trisomy 21 it yields results comparable to those of the program used by the FMF UK. Regarding the diagnosis of trisomy 13 and 18, even higher detection rates are currently achieved with the German algorithm. Program, data base and license key are available free of charge to registered members of the FMF Germany.