Using 'Degree of Extremeness' instead of 'Multiples of Median' in first trimester risk assessment for Down syndrome-an improved method or just a gimmick in face of political motivations?

摘要

INTRODUCTION: The concept of first trimester Down syndrome screening based on (NT) nuchal translucency measurement was introduced by the Fetal Medicine Foundation (FMF), London, in the late 1990s. For Germany, the mandate of NT training and auditing was assigned to the newly established German branch of the FMF in 2002. In January 2007, following the organisational split between FMF London and FMF Germany, the FMF Germany released its own, novel calculation software ("Prenatal risk calculation", PRC), for this field. It is the aim of this study to compare the modified biochemical calculation method based on Degree of Extremeness Median" (MoM). MATERIALS AND METHODS: The fbeta-hCG and PAPP-A values of 266 first trimester examinations were categorised by gestational age, nicotine consumption and body weight. For each, a scatter plot and Pearson correlation was derived. As a second step, only patients with no nicotine consumption and a body weight between 55 and 65 kg were considered. The remaining 84 cases were again classified by gestational age and a statistical analysis was performed. RESULTS: The correlation of the DoEs with the respective MoM values was found to be quite good (r = 0.76, P < 0.01). After dividing the population by gestational age, the cohort of early pregnancies showed a steeper fss-hCG correlation curve than later stages. The examination regarding nicotine consumption showed no significant differences between smokers and non-smokers. In contrast, maternal body weight was found to have a marked influence. When only body weights between 55 and 65 kg were included and smokers were excluded, the Pearson correlations clearly converged. DISCUSSION: In this study, the utilisation of DoEs instead of MoMs was not clearly superior but it also did not seem to be inappropriate. However, it is most problematic to ignore evidently meaningful factors like body weight and ethnicity. It must also be considered risky to introduce a new calculation software to the market, when it is not clear whether the test performance is in fact comparable to established software concepts, especially in view of the fact that not only a new (unproven) database, but also a new (again unproven) algorithm is used in the PRC software. As the new program has not been evaluated yet, the test performance parameters should be validated urgently.