Individual risk assessment of adverse pregnancy outcome by multivariate regression analysis may serve as basis for drug intervention studies: retrospective analysis of 426 high-risk patients including ethical aspects

Abstract

Objective: To identify patients at very high risk for adverse pregnancy outcome (APO) at the 20- to 23-week scan and to assess the effectiveness of Aspirin (ASS) and low molecular weight heparin (LMWH) starting after this examination.

Patients and methods: By applying an algorithm based on multivariate logistic regression analysis using the parameters maternal age, parity, body mass index (BMI), mean pulsatility index of both uterine arteries (meanPI), presence of uni- or bilateral notch, and depth of notch (mean notch index (meanNI), we retrospectively calculated the individual risk for APO of 21,302 singleton pregnancies. We isolated a subgroup of 426 patients with the highest calculated probability for APO (cpAPO > 27.8 %). 147 had been treated with ASS; 73 with LMWH, 15 patients with a combination of ASS and LMWH, and 191 patients had not received anticoagulants.

Results: Administration of ASS starting after 20 gestational weeks in comparison to non-treated patients significantly reduced the frequency of intrauterine/neonatal death (IUD/NND), preeclampsia <33 weeks (PE < 33), and preterm delivery <33 weeks (PD < 33), while the frequency of IUGR showed a tendency to be elevated (P = 0.061). The subgroup of high-risk patients treated with LMWH was characterised by a higher a priori risk for APO and showed no significant reduction of any form of APO but an increased frequency of PE.

Conclusion: Individual assessment of risk for APO by applying a simple algorithm based on biometrical/biographical as well as sonographic parameters may serve as basis for drug intervention studies. The administration of ASS in high-risk patients starting after 20 gestational weeks reduced the frequency of most of the severe forms of adverse pregnancy outcome in high-risk patients. A complication-reducing effect of LMWH starting after 20 weeks of gestation in patients could not be proven. From an ethical point of view, it may not be justified any more to preclude high-risk patients from administration of ASS or to perform studies of ASS against placebo.

Fig. 1

Relation of observed and calculated complications depending on mode of anticoagulation visualized by LOESS—regression analysis of 1,254 complications in 21,258 patients (excluding 44 patients with LMWH plus ASS, 14 of them with complications)