Performance of the Fetal Medicine Foundation model in the third trimester for predicting late-onset pregnancy-induced hypertension among Thai pregnant women.

OBJECTIVE

This study evaluates the performance of third-trimester screening for late-onset pregnancy-induced hypertension (PIH) using the Fetal Medicine Foundation (FMF) prediction model and compares its diagnostic accuracy with locally developed Thai predictive models.

METHODS

A prospective diagnostic accuracy study was conducted in singleton normotensive pregnancies. Mean arterial pressure (MAP), uterine artery pulsatility index (UtPI), and placental growth factor (PlGF) were measured between 30 and 34 weeks. The primary outcome was the development of PIH. The diagnostic performances of each marker, and their combinations, were evaluated using both the FMF model and Thai models specifically derived from this study.

RESULTS

Of the 213 participants meeting the inclusion criteria, 18 (8.45%) developed PIH. Based on the FMF model, MAP, UtPI, and PlGF yielded areas under curve (AUCs) of 0.827, 0.631, and 0.857, respectively. The combination of MAP and PlGF achieved the highest AUC of 0.890, with a sensitivity of 66.7% at a 10% false-positive rate, better than 55.7 and 44.4% provided by PlGF alone and MAP alone, respectively. Although UtPI alone demonstrated significant predictive value, it did not provide additional benefit when combined with other markers. Thai predictive models developed from the same dataset showed diagnostic performance comparable to that of the FMF model.

CONCLUSION

The FMF model for predicting PIH, using the combination of PlGF and MAP in the early third trimester, demonstrated better diagnostic performance than MAP alone. This finding suggests that the FMF algorithm can be applied to the Thai population, although its performance is not as robust as that reported in Western populations.