Comparative effectiveness of eight machine learning algorithms and the FMF competing risks model for first-trimester preeclampsia prediction: a multicenter cohort study with external validation

Aim: to systematically compare the discrimination and calibration of eight machine learning (ML) algorithms with the Fetal Medicine Foundation (FMF) competing risks algorithm for first-trimester prediction of early-onset (< 34 weeks), late-onset (> 34 weeks), severe, and preterm (< 37 weeks) preeclampsia (PE) by assessing the incremental value of markers with a nested model architecture (M0 [model includes only clinical and anamnestic data] → M2 [model supplemented with biophysical markers] → M4 [model supplemented with biochemical markers]) and external validation in an independent dataset (n = 7,581).

Materials and Methods. A retrospective cohort study (TRIPOD 2b + 3) was conducted. Development: a development cohort (n = 7,581) with 10-fold stratified cross-validation. External validation: an independent cohort (n = 7,581). A total of 96 benchmark configurations were developed (8 algorithms × 3 predictor levels × 4 outcomes): M0 – 15 clinical and history-based factors; M2 – adding mean arterial pressure (MAP) and uterine artery pulsatility index (UtAPI); M4 – adding placental growth factor (PlGF) and pregnancy-associated plasma protein-A (PAPP-A). Evaluation metrics: AUC-ROC (Area Under the Receiver Operating Characteristic curve), calibration (O:E ratio (observed-to-expected ratio), calibration slope, Brier score), net reclassification improvement (NRI). integrated discrimination improvement (IDI), decision curve analysis (DCA), and SHapley Additive exPlanations (SHAP).

Results. At the M4 level, logistic regression demonstrated the highest discrimination: early-onset PE – AUC = 0.976 (an optimistic estimate with EPV (events per variable) = 3.7; detection rate 92.9 % at 10 % FPR (false positive rate); late-onset PE – 0.837; severe PE – 0.885, preterm PE – 0.908; with the most robust calibration among all algorithms (O:E = 0.99–1.00), confirmed in the external cohort for M0 and M2 levels. Nonlinear algorithms and stacking did not achieve a significant advantage. For preterm PE (the target outcome of the FMF algorithm), logistic regression and FMF posterior were comparable in discrimination (ΔAUC = +0.002; p > 0.05) level, with the former showing superior calibration in the Russian population. For late-onset PE (the dominant disease phenotype not modeled by FMF), logistic regression provided significantly better risk stratification (AUC = 0.837 vs. 0.807; p < 0.05). The FMF algorithm exhibited miscalibration: risk overestimation for its target outcomes (O:E = 0.55 for early-onset PE) and structural underestimation for non-target outcomes (O:E = 2.84 for late-onset PE). Biophysical markers were critical for early-onset PE, whereas biochemical markers were critical for late-onset PE; this pattern was consistent across algorithms. External validation confirmed robust discrimination at the M0 level (AUC = 0.790–0.823; n = 7,581) and M2 level (AUC = 0.801–0.931; n = 4,080); results at the M4 level (n = 1,010; 11–40 events) are preliminary.

Conclusion. Nonlinear ML algorithms do not outperform logistic regression, which demonstrated the most robust calibration confirmed by external validation at the M0 and M2 levels. Logistic regression with an M0 → M2 → M4 architecture is recommended for clinical decision support systems in first-trimester PE screening. The FMF algorithm requires population-specific recalibration for its target outcomes and is structurally not designed to predict late-onset PE.

1. Abalos E., Cuesta C., Grosso A.L. et al. Global and regional estimates of preeclampsia and eclampsia: a systematic review. Eur J Obstet Gynecol Reprod Biol. 2013;170(1):1–7. https://doi.org/10.1016/j.ejogrb.2013.05.005.

2. Brown M.A., Magee L.A., Kenny L.C. et al. Hypertensive disorders of pregnancy: ISSHP classification, diagnosis, and management recommendations for international practice. Hypertension. 2018;72(1):24–43. https://doi.org/10.1161/HYPERTENSIONAHA.117.10803.

3. Preeclampsia. Eclampsia. Edema, proteinuria and hypertensive disorders during pregnancy, childbirth and the postpartum period: clinical guidelines. [Klinicheskie rekomendacii – Preeklampsiya. Eklampsiya. Oteki, proteinuriya i gipertenzivnye rasstrojstva vo vremyaberemennosti, v rodah i poslerodovom periode – 2024-2025-2026 (05.09.2024)]. Moscow: Ministerstvo zdravoohraneniya Rossijskoj Federacii, 2024. 53 p. (In Russ.). Available at: http://minzdravrm.ru/wp-content/uploads/2024/09/Preeklampsiya_Eklampsiya_Oteki-ptoteinuriya_i_gipertenx_rasstroistva_vo_vremya_berem_v_rodakh_i_poslerodovom_periode.pdf. [Accessed: 20.01.2026].

4. Order of the Ministry of Health of the Russian Federation No. 1130n dated 20.10.2020 "On the approval of the Procedure for providing medical care in the field of obstetrics and gynecology". [Prikaz Ministerstva zdravookhraneniya Rossiyskoy Federatsii ot 20.10.2020 No 1130n «Ob utverzhdenii Poryadka okazaniya meditsinskoy pomoshchi po profilyu 'akusherstvo i ginekologiya"»]. Moscow: Ministerstvo zdravoohraneniya Rossijskoj Federacii, 2020. 688 p. (In Russ.). Available at: http://publication.pravo.gov.ru/Document/View/0001202011130037. [Accessed: 20.01.2026].

5. Staff A.C. The two-stage placental model of preeclampsia: an update. J Reprod Immunol. 2019;134–135:1–10. https://doi.org/10.1016/j.jri.2019.07.004.

6. Redman C.W., Staff A.C., Roberts J.M. Syncytiotrophoblast stress in preeclampsia: the convergence point for multiple pathways. Am J Obstet Gynecol. 2022;226(2S):S907–S927. https://doi.org/10.1016/j.ajog.2020.09.047.

7. Khodzhaeva Z.S., Kogan E.A., Klimenchenko N.I. et al. Clinical and pathogenetic features of early and late preeclampsia. Akusherstvo i ginekologiya. 2015;(1):12–7. (In Russ.).

8. Wright D., Akolekar R., Syngelaki A. et al. A competing risks model in early screening for preeclampsia. Fetal Diagn Ther. 2012;32(3):171–8. https://doi.org/10.1159/000338470.

9. Akolekar R., Syngelaki A., Poon L. et al. Competing risks model in early screening for preeclampsia by biophysical and biochemical markers. Fetal Diagn Ther. 2013;33(1):8–15. https://doi.org/10.1159/000341264.

10. Wright D., Syngelaki A., Akolekar R. et al. Competing risks model in screening for preeclampsia by maternal characteristics and medical history. Am J Obstet Gynecol. 2015;213(1):62.e1–62.e10. https://doi.org/10.1016/j.ajog.2015.02.018.

11. O'Gorman N., Wright D., Syngelaki A. et al. Competing risks model in screening for preeclampsia by maternal factors and biomarkers at 11–13 weeks gestation. Am J Obstet Gynecol. 2016;214(1):103.e1–103.e12. https://doi.org/10.1016/j.ajog.2015.08.034.

12. Wright D., Wright A., Nicolaides K.H. The competing risk approach for prediction of preeclampsia. Am J Obstet Gynecol. 2020;223(1):12–23.e7. https://doi.org/10.1016/j.ajog.2019.11.1247.

13. Tan M.Y., Syngelaki A., Poon L.C. et al. Screening for pre-eclampsia by maternal factors and biomarkers at 11–13 weeks' gestation. Ultrasound Obstet Gynecol. 2018;52(2):186–95. https://doi.org/10.1002/uog.19112.

14. Rolnik D.L., Wright D., Poon L.C. et al. Aspirin versus placebo in pregnancies at high risk for preterm preeclampsia. N Engl J Med. 2017;377(7):613–622. https://doi.org/10.1056/NEJMoa1704559.

15. Poon L.C., Wright D., Rolnik D.L. et al. Aspirin for Evidence-Based Preeclampsia Prevention trial: effect of aspirin in prevention of preterm preeclampsia in subgroups. Am J Obstet Gynecol. 2017;217(5):585.e1–585.e5. https://doi.org/10.1016/j.ajog.2017.07.038.

16. Chaemsaithong P., Pooh R.K., Zheng M. et al. Prospective evaluation of screening performance of first-trimester prediction models for preterm preeclampsia in an Asian population. Am J Obstet Gynecol. 2019;221(6):650.e1–650.e16. https://doi.org/10.1016/j.ajog.2019.09.041.

17. Rolnik D.L., Wright D., Poon L.C. et al. ASPRE trial: performance of screening for preterm pre-eclampsia. Ultrasound Obstet Gynecol. 2017;50(4):492–5. https://doi.org/10.1002/uog.18816.

18. Kholin A.M., Muminova K.T., Balashov I.S. et al. Prediction of preeclampsia in the first trimester of pregnancy: validation of screening algorithms in the Russian population. Akusherstvo i ginekologiya. 2017;(8):74–84. (In Russ.). https://doi.org/10.18565/aig.2017.8.74-84.

19. Kholin A.M., Ivanets T.Yu., Khodzhaeva Z.S., Gus A.I. Screening of early preeclampsia in the first trimester of pregnancy on the basis on combined estimate of maternal serum placental growth factor and uterine artery Doppler. Akusherstvo i ginekologiya. 2015;(5):42–8. (In Russ.).

20. Kholin A.M., Muminova K.T., Nagoev T.M. et al. First trimester prediction of preeclampsia based on maternal factors, biomarkers and 3D power doppler of placental bed vascularization. Akusherstvo i ginekologiya. 2018;(8):56–65. (In Russ.). https://doi.org/10.18565/aig.2018.8.56-65.

21. Lemeshevskaya T.V., Pribushenya O.V. Prediction of preeclampsia during extended first-trimester combined prenatal screening. Akusherstvo i ginekologiya. 2017;(12):52–9. (In Russ.). https://doi.org/10.18565/aig.2017.12.52-59.

22. Gabidullina R.I., Ganeeva A.V., Shigabutdinova T.N. Predictors of preeclampsia. Screening and prophylaxis in the first I of pregnancy. Ginekologiya. 2021;23(5):428–34. (In Russ.). https://doi.org/10.26442/20795696.2021.5.201213.

23. Ansbacher-Feldman Z., Grinstein E., Ganer Herman H. et al. Machine-learning-based prediction of pre-eclampsia using first-trimester maternal characteristics and biomarkers. Ultrasound Obstet Gynecol. 2022;60(6):739–45. https://doi.org/10.1002/uog.26105.

24. Gil M.M., Cuenca-Gómez D., Rolle V. et al. Validation of machine-learning model for first-trimester prediction of pre-eclampsia using cohort from PREVAL study. Ultrasound Obstet Gynecol. 2024;63(1):68–74. https://doi.org/10.1002/uog.27478.

25. Marić I., Tsur A., Aghaeepour N. et al. Early prediction of preeclampsia via machine learning. Am J Obstet Gynecol MFM. 2020;2(2):100100. https://doi.org/10.1016/j.ajogmf.2020.100100.

26. Jhee J.H., Lee S., Park Y. et al. Prediction model development of late-onset preeclampsia using machine learning-based methods. PLoS One. 2019;14(8):e0221202. https://doi.org/10.1371/journal.pone.0221202.

27. Li T., Xu W., Ouyang F. et al. Prediction model of preeclampsia using machine learning based methods: a population based cohort study in China. Front Endocrinol (Lausanne). 2024;15:1345573. https://doi.org/10.3389/fendo.2024.1345573.

28. Ranjbar A., Montazeri F., Rezaei Ghamsari S. et al. Machine learning models for predicting preeclampsia: a systematic review. BMC Pregnancy Childbirth. 2024;24(1):6. https://doi.org/10.1186/s12884-023-06220-1.

29. Christodoulou E., Ma J., Collins G.S. et al. A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models. J Clin Epidemiol. 2019;110:12–22. https://doi.org/10.1016/j.jclinepi.2019.02.004.

30. Sufriyana H., Husnayain A., Chen Y.-L. et al. Comparison of multivariable logistic regression and other machine learning algorithms for prognostic prediction studies in pregnancy care: systematic review and meta-analysis. JMIR Med Inform. 2020;8(11):e16503. https://doi.org/10.2196/16503.

31. Andreychenko A.E., Luchinin A.S., Ivshin A.A. et al. Development and validation of models to predict total and early-onset preeclampsia in the first trimester of pregnancy using machine learning algorithms. Akusherstvo i ginekologiya. 2023;(10):94–107. (In Russ.). https://doi.org/10.18565/aig.2023.101.

32. Ivshin A.A., Bagaudin T.Z., Gusev A.V. Artificial intelligence technologies in predicting preeclampsia. Obstetrics, Gynecology and Reproduction. 2021;15(5):576–85. (In Russ.). https://doi.org/10.17749/2313-7347/ob.gyn.rep.2021.229.

33. Boldina Yu.S., Ivshin A.A. Machine learning opportunities to predict obstetric haemorrhages. Obstetrics, Gynecology and Reproduction. 2024;18(3):365–381. (In Russ.). https://doi.org/10.17749/2313-7347/ob.gyn.rep.2024.491.

34. Ivshin A.A., Malyshev N.A. Preeclampsia early risk stratification based on a multiparametric machine learning model and routinely collected clinical data. Obstetrics, Gynecology and Reproduction. 2026;20(1):111–29. (In Russ.). https://doi.org/10.17749/2313-7347/ob.gyn.rep.2025.706.

35. Collins G.S., Reitsma J.B., Altman D.G., Moons K.G.M. Transparent reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD): the TRIPOD Statement. Ann Intern Med. 2015;162(1):55–63. https://doi.org/10.7326/M14-0697.

36. Collins G.S., Moons K.G.M., Dhiman P. et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ. 2024;385:e078378. https://doi.org/10.1136/bmj-2023-078378.

37. Breiman L. Random forests. Machine Learning. 2001;45(1):5–32. https://doi.org/10.1023/A:1010933404324.

38. Chen T., Guestrin C. XGBoost: a Scalable Tree Boosting System. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016. 785–94. https://doi.org/10.1145/2939672.2939785.

39. Ke G., Meng Q., Finley T. et al. LightGBM: a Highly Efficient Gradient Boosting Decision Tree. In: Advances in Neural Information Processing Systems (NeurIPS). 2017;30:3149–57. https://doi.org/10.5555/3294996.3295074.

40. Van Smeden M., de Groot J.A., Moons K.G.M. et al. No rationale for 1 variable per 10 events criterion for binary logistic regression analysis. BMC Med Res Methodol. 2016;16(1):163. https://doi.org/10.1186/s12874-016-0267-3.

41. Riley R.D., Ensor J., Snell K.I.E. et al. Calculating the sample size required for developing a clinical prediction model. BMJ. 2020;368:m441. https://doi.org/10.1136/bmj.m441.

42. Steyerberg E.W. Clinical prediction models: a practical approach to development, validation, and updating. 2nd ed. Cham: Springer, 2019. 558 p.

43. Van Calster B., Nieboer D., Vergouwe Y. et al. A calibration hierarchy for risk models was defined: from utopia to empirical data. J Clin Epidemiol. 2016;74:167–76. https://doi.org/10.1016/j.jclinepi.2015.12.005.

44. Van Calster B., McLernon D.J., Van Smeden M. et al. Calibration: the Achilles heel of predictive analytics. BMC Med. 2019;17(1):230. https://doi.org/10.1186/s12916-019-1466-7.

45. DeLong E.R., DeLong D.M., Clarke-Pearson D.L. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44(3):837–45. https://doi.org/10.2307/2531595.

46. Pencina M.J., D'Agostino R.B. Sr., D'Agostino R.B. Jr., Vasan R.S. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27(2):157–172. https://doi.org/10.1002/sim.2929.

47. Vickers A.J., Elkin E.B. Decision curve analysis: a novel method for evaluating prediction models. Med Decis Making. 2006;26(6):565–74. https://doi.org/10.1177/0272989X06295361.

48. Lundberg S.M., Lee S.-I. A unified approach to interpreting model predictions. In: Advances in Neural Information Processing Systems (NeurIPS). 2017;30:4765–74. https://doi.org/10.5555/3295222.3295230.

49. Lundberg S.M., Erion G., Chen H. et al. From local explanations to global understanding with explainable AI for trees. Nat Mach Intell. 2020;2(1):56–67. https://doi.org/10.1038/s42256-019-0138-9.

50. Grinsztajn L., Oyallon E., Varoquaux G. Why do tree-based models still outperform deep learning on typical tabular data? In: Advances in Neural Information Processing Systems (NeurIPS). 2022;35:507–20. https://doi.org/10.5555/3600270.3600307.

51. Melinte-Popescu A.S., Vasilache I.A., Socolov D. et al. Predictive performance of machine learning-based methods for the prediction of preeclampsia – a prospective study. J Clin Med. 2023;12(2):418. https://doi.org/10.3390/jcm12020418.

52. Ivshin A.A., Bagaudin T.Z., Gusev A.V. Artificial intelligence on guard of reproductive health. Akusherstvo i ginekologiya. 2021;(5):17–24. (In Russ.). https://doi.org/10.18565/aig.2021.5.17-24.

53. Wright A., Wright D., Syngelaki A. et al. Two-stage screening for preterm preeclampsia at 11–13 weeks' gestation. Am J Obstet Gynecol. 2019;220(2):197.e1–197.e11. https://doi.org/10.1016/j.ajog.2018.10.092.

54. Wright D., Tan M.Y., O'Gorman N. et al. Predictive performance of the competing risk model in screening for preeclampsia. Am J Obstet Gynecol. 2019;220(2):199.e1–199.e13. https://doi.org/10.1016/j.ajog.2018.11.1087.

55. Sidorova I.S., Nikitina N.A. Preeclampsia in the focus of a practitioner’s attention. Akusherstvo i ginekologiya. 2014;(6):4–10. (In Russ.).

56. Kan N.E., Bednyagin L.A., Dolgushina N.V. et al. Clinical and anamnestic risk factors for preeclampsia in pregnant women. [Kliniko-anamnesticheskiye faktory riska razvitiya preeklampsii u beremennykh]. Akusherstvo i ginekologiya. 2016;(6):39–44. (In Russ.). https://doi.org/10.18565/aig.2016.6.39-44.