Review Article

Obstetric Medicine
2021, Vol. 14(4) 220–224
Prediction of pre-eclampsia ! The Author(s) 2021
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1753495X20984015
journals.sagepub.com/home/obm
Konstantinos Giannakou

Abstract
Pre-eclampsia is a leading cause of neonatal and maternal mortality and morbidity that complicates approximately 2–8% of all pregnancies worldwide.
The precise cause of pre-eclampsia is not completely understood, with several environmental, genetic, and maternal factors involved in its pathogenesis
and pathophysiology. An accurate predictor of pre-eclampsia will facilitate early recognition, close surveillance according to the individual risk and early
intervention, and reduce the negative consequences of the disorder. Current evidence shows that no single test predicts pre-eclampsia with sufficient
accuracy to be clinically useful. A combination of markers into multiparametric models may provide a more useful and feasible predictive tool for pre-
eclampsia screening in the routine care setting than a test of either component alone. This review presents a summary of the current advances on
prediction of pre-eclampsia, highlighting their performance and applicability. Key priorities when conducting research on predicting pre-eclampsia are
also analyzed.

Keywords
Biological markers, pre-eclampsia, prediction, screening, pregnancy complications
Date Received: 3 July 2020; Revised 18 November 2020; accepted: 6 December 2020

Introduction would be 100% sensitive and 100% specific, henceforth would be

Pre-eclampsia (PE), a complex, multisystem, pregnancy-associated positive for all those with the disease and negative for all those
hypertensive disorder, typically developing after the 20th week of who did not.18 In clinical terms, with a high sensitivity test, most
gestation, that complicates 2–8% of pregnancies, is a leading cause patients with PE will correctly be identified as having the condition.
of neonatal and maternal mortality and morbidity.1–4 PE is consid- While with a high specificity test, most healthy women (without PE)
ered to be the most prevalent hypertensive disorder of pregnancy, will correctly be identified as not having the condition. The test
associated with an increased risk of both short- and long-term com- should also be simple, rapid, non-invasive, inexpensive as well as
plications for both the mother and the neonate, contributing to more valid, reliable, and reproducible with a high positive and a low neg-
than 60,000 maternal deaths per year.3,5–7 There is growing evidence ative likelihood ratio.19 The cost-effective test should ideally identify
highlighting that in addition to the immediate risk, PE is also con- women with an increased risk early in pregnancy, who could be
sidered a cardiovascular risk factor for both mother and child in later offered potential treatment to prevent the disorder and thereby min-
life.8–10 Despite an intensive research effort to elucidate the origin of imize its negative impact. However, the optimal timing for PE screen-
PE, the exact underlying pathophysiological mechanisms still remain ing is controversial. While the conventional approach to PE diagnosis
complex and unclear with several environmental, genetic and mater- is based on the incidence of clinical symptoms, usually discovered
nal factors contributing.11 An accurate prediction of PE and early during routine obstetric visits in the second or third trimester of preg-
identification of women at increased risk is a clinical priority to min- nancy, an alternative screening method has been proposed in 2011 by
imize complications by both careful monitoring and early treatment Dr. Nicolaides, which involves a comprehensive screening in the first
trimester for stratification for all major obstetric complications, and
(i.e. low-dose aspirin started prior to 16 weeks of gestation).12–16
then contingent screening based on the risk reassessment at each
Several studies have evaluated the predictive ability of different
visit.20,21 By following this proposed inverted pyramid of pregnancy
tests to predict PE using individual or a combination of clinical char-
care, low-risk pregnancies would attend a standard care program
acteristics, biomarkers, and ultrasound markers. Despite this, no con-
with fewer visits, while a more accurate monitoring of high-risk
sensus on the optimal screening strategy has been reached, mostly
groups and possible prophylactic pharmacological interventions
because of the lack of adequate performance.17 The aim of this review
(e.g. low-dose aspirin) would possibly lead to a reduction of
is to provide an overview of the current knowledge regarding PE
prediction.

Department of Health Sciences, School of Sciences, European University

Screening for pre-eclampsia Cyprus, Nicosia, Cyprus

Considering pre-eclampsia’s prevalence and clinical importance, an Corresponding author:

effective screening test would be most useful in its ability to detect Konstantinos Giannakou, Department of Health Sciences, School of
women who require close monitoring. To assess a screening tool’s Sciences, European University Cyprus, Nicosia, Cyprus, 6 Diogenes Str.
predictive ability, sensitivity, specificity, positive predictive value, and Engomi, 2404, P.O. Box 22006, Nicosia 1516, Cyprus.
negative predictive value should be evaluated. A perfect screening test Email: k.giannakou@euc.ac.cy
Giannakou 221

complicated pregnancies, but also could lead to fewer long-term com- a sample of 210 intermediate-risk women at 11–14 weeks as an 88%
plications for both mother and child.20–23 detection rate for a 10% false positive for the detection of early onset
PE was found.35 These findings indicate that this pulse wave analysis
may offer a more accurate evaluation of central vascular pressure
Patient identification: Risk factors than conventional blood pressure. Despite several years of research
in the field, a single test, accurate enough to predict PE sufficiently
Due to lack of knowledge of possible underlying pathophysiological
well, has not yet been found.17,19,36,37 In actual practice, the use of
mechanisms responsible for PE, there are not yet any reliable and
clinical and lifestyle factors in the prediction of PE has shown an
validated predictors to identify most women who will develop the
overall predictive potential of 30% to 37% for early-onset PE and
disease. The diagnosis of PE is usually made by detection of signs
29% for late-onset PE and had a 5% of false-positive results.33,38
including hypertension and proteinuria during antenatal monitoring
that could indicate development of the disease. However, this method
is not appropriate for early prediction or identification of high-risk
women that are likely to develop the disease.24,25 The uncertainty of
Biochemical, biophysical, or ultrasound
clinical diagnosis, with a fraction of pregnant women developing markers
subjective symptoms and signs of PE (e.g. headache, abdominal
Amongst other alternatives of potential predictors is the use of bio-
pain, visual disturbances) and only 20% of these are diagnosed, indi- chemical markers. Extensive research has identified a range of poten-
cates that there is a clear need for improved testing methods.26 Even
tial biophysical and biochemical predictors of PE.39–42 Several of
though the precise etiology of PE remains complex and unclear, dis- these markers are measurable in maternal blood and have therefore
tinguishing between women of moderate- and high-risk is possible. It
been evaluated as biomarkers for PE prediction. These include serum
was proposed by the National Institute for Health and Clinical
and plasma markers of placental endocrine function, maternal endo-
Excellence (NICE) that there is a classification for moderate risk thelial dysfunction, renal dysfunction, general metabolic status, oxi-
and high risk factors, which could be used to identify the women
dative stress, and hemolysis and inflammatory markers.43 Screening
most likely to benefit from aspirin prophylaxis.27 Women viewed markers that have recently undergone investigation for PE include
as high risk include those with chronic hypertension, history of any
factors related to angiogenesis, coagulation, lipids, placental hor-
hypertensive disorder or PE in previous pregnancies, diabetes (type 1
mones, cell adhesion, fetal DNA, inflammation, and growth factors.
or type 2), chronic kidney disease and autoimmune disorders, includ- Despite rigorous research efforts to identify potential biochemical
ing systemic lupus erythematosus or antiphospholipid antibody syn-
markers, no factor established a sufficient degree of accuracy for
drome.22,28,29 Factors considered to be of moderate risk are the prediction of PE.17 A recent review of different biochemical
primiparity or pregnancy interval greater than 10 years, extreme age
markers for PE before the 25th week of gestation did not reveal a
(below 20 years old or above 40), BMI of 35 kg/m2 or more, a family
single test with a sensitivity and specificity over 90%.36 Likewise,
history of PE, polycystic ovarian syndrome, and multiple pregnan- another study that reviewed 27 different tests for PE prediction
cies.22,30 According to this classification, aspirin is advised, if two
found only a few reached specificities above 90%. These were BMI
moderate-risk or a single high-risk factor are present.15 The of 34 kg/m2 or higher, a-fetoprotein, and bilateral uterine artery
American College of Obstetricians and Gynecologists (ACOG) con-
Doppler notching.44 As far as genetic markers are concerned, there
siders the same risk factors, with the exception of a value higher than
were numerous suggestions of gene mutations playing a role in the
30 kg/m2 for the BMI at the beginning of the pregnancy and makes development of PE, although no single polymorphism has shown any
no distinction between their severity but rather considers them all as
predictive value.17 Amongst the genes being investigated are methyl-
‘high-risk’ factors.31 The main advantage of using such risk factors is enetetrahydrofolate reductase and endothelial nitric oxide synthase,
to allow the identification of those women that should be offered
while PAI-1 4 G/5G (recessive model) polymorphism showed evi-
subspecialty referral very early in their pregnancy. However, despite
dence of contribution to PE.11
the fact that organizations such as ACOG and NICE endorse eval-
uation of clinical and demographic factors as the best and only rec-
ommended screening approach for PE, yet, they are neither sensitive Combination of markers
or specific enough to be used alone and therefore, they cannot be
used reliably for PE prediction. Likewise, this screening approach is The absence of a single robust, sensitive marker is not surprising
likely to identify many women as requiring additional monitoring, since PE is characterized by a complex etiology with a range of het-
and increase the resources required with potentially limited diagnostic erogeneous clinical and laboratory findings. Hence, it is unlikely that
utility.25,32 a single marker could predict the mixed presentations and potential
causes of the disorder.40,41 It is established that combinations of
markers that reflect different aspects of pathogenesis are needed to
Maternal hemodynamic and vascular improve the possibility for predicting PE with a high degree of accu-
racy.19 Potential components of such combinations could be anam-
markers nestic risk factors, angiogenic, inflammatory and other biochemical
No early and reliable first trimester marker is available for early factors, uterine artery Doppler, and MAP. In the last decades, many
prediction of pre-eclampsia development. Blood pressure remains studies have combined one or more biochemical markers with uterine
an important clinical predictor used as part of routine antenatal Doppler to assess prediction rates for early and late PE separately.
assessment. However, in an evaluation of a heterogenous population, Yet, until now, there is no general acceptance of these combinations
mean arterial pressure (MAP) has shown prediction rates of 58% for in clinical practice.40,41 A previous large study that combined mater-
early PE and 44% for late PE, and a 5% of false positives.33,34 An nal characteristics, including MAP, uterine artery pulsatility index
additional non-invasive screening tool that is used for the evaluation and several biochemical markers (PAPP-A, PlGF, PP13, sEndoglin,
of the uteroplacental circulation is the study of uterine arteries using Inhibin-A, Activin-A, Pentraxin 3, and P-Selectin) has shown 95%
Doppler velocimetry. The use of it on a large scale is restricted by the specificity for early-onset 91% sensitivity, intermediate onset 79%
cost, availability of the ultrasound device and the expertise of the sensitivity, and late onset PE 61% sensitivity.45 Similarly, a study
professional performing the examination.33 Pulse wave analysis has that examined a combination of different maternal characteristics
been found to be a promising noninvasive technique when applied in and biochemical markers in the first trimester (MAP, uterine artery
222 Obstetric Medicine 14(4)

pulsatility index, PAPP-A, and PlGF) has revealed predictive values Acknowledgements
with 95% specificity and 93% sensitivity for early-onset and 36% for Not applicable.
late-onset PE.46 Another example of such combination is the fetal
hemoglobin (HbF)/hemoglobin ratio and a1-microglobulin, that has
Declaration of conflicting interests
showed 90% sensitivity and 77% specificity for prediction of PE in
early pregnancy.47 Findings of systematic reviews and meta-analyses The author(s) declared no potential conflicts of interest with respect to the
that evaluated the predictive capabilities of combinations of biochem- research, authorship, and/or publication of this article.
ical and ultrasonographic markers demonstrated that such combina-
tions predicted PE in a better way than a single predictor would. This Funding
insight might improve the prediction of PE, especially in high-risk
This research received no specific grant from any funding agency in the public,
populations.37,48–50 However, even though these multiparametric
models have shown promising results, when applied to populations, commercial, or not-for-profit sectors.
other than the population from which they were derived, they have
shown a poorer performance.51 This finding is also supported by Ethical approval
several studies which found that only a minority of these models Not applicable.
were externally validated or calibrated.52–57
Informed consent
Priorities for future research Not applicable.

There is a need to broaden our understanding of PE and particularly

Guarantor
its prediction and prevention. Optimising the early identification of
women at high risk will increase the opportunities for earlier inter- KG is the guarantor of the present work.
vention which may alter the prognosis and reduce the chance of
adverse pregnancy outcomes. Thus, establishing a reliable screening Contributorship
program will enable the improvement of clinical guidelines, better
KG is the sole author.
surveillance, and efficient prophylactic measures as well as support
for high risk women and offspring.
The challenge of PE prediction requires additional research stud- ORCID iD
ies to define the best and most accurate models and methods of pre- Konstantinos Giannakou https://orcid.org/0000-0002-2185-561X
diction that could be used in clinical practice. More emphasis should
be targeted at predicting PE in the first trimester so that prophylactic
interventions such as low-dose aspirin can be initiated. Also, when References
evaluating the prediction of a given test or combination of tests for 1. Duley L. The global impact of pre-eclampsia and eclampsia.
PE, a clear distinction must be made between early-onset and late- Semin Perinatol 2009; 33: 130–137.
onset phenotypes. Thus, standardization of definitions, methods, and 2. Duley L. Pre-eclampsia, eclampsia, and hypertension. BMJ Clin
statistical analyses will be useful for multi-national comparison and Evid 2011; 2011: 1402.
meta-analysis of results as well as development and evaluation of 3. Armaly Z, Jadaon JE, Jabbour A, et al. Preeclampsia: novel
prediction models, which ultimately promote clinical health prac- mechanisms and potential therapeutic approaches. Front
tice.58 Furthermore, well-designed large prospective studies are Physiol 2018; 9: 973–908.
needed to assess the accuracy of risk prediction models according 4. Stevens W, Shih T, Incerti D, et al. Short-term costs of pre-
to disease onset and population risk in comparison to the current eclampsia to the United States health care system. Am J Obstet
single risk factor screening. Further research is necessary to deter- Gynecol 2017; 217: 237–248.e216.
mine whether predictive models can be further improved with the 5. Khan KS, Wojdyla D, Say L, et al. WHO analysis of causes of
addition of novel biomarkers implicated in the pathophysiology of maternal death: a systematic review. Lancet 2006; 367:
PE as well as identify additional combinations of markers that may 1066–1074.
predict the occurrence of PE. Finally, multiparametric models need to 6. Witlin AG, Saade GR, Mattar F, et al. Predictors of neonatal
be validated externally to assure validity and reliability of their pre- outcome in women with severe pre-eclampsia or eclampsia
dictive performance before they can be used in clinical practice. between 24 and 33 weeks’ gestation. Am J Obstet Gynecol
2000; 182: 607–611.
7. Wu P, Haththotuwa R, Kwok CS, et al. Preeclampsia and future
Conclusions cardiovascular health: a systematic review and meta-analysis.
Clinical history remains important in the assessment of pregnant Circul Cardiovasc Qual Outcomes 2017; 10: 2–24.
women, to identify features that put them at higher risk of developing 8. Brown MC, Best KE, Pearce MS, et al. Cardiovascular disease
pre-eclampsia. Considering the rising prevalence, as well as the great risk in women with pre-eclampsia: systematic review and meta-
social and economic impact of PE, the early identification would analysis. Eur J Epidemiol 2013; 28: 1–19.
mean plans for enhanced monitoring could be instituted in a sub- 9. Ghossein-Doha C, van Neer J, Wissink B, et al. Pre-eclampsia:
group of women. Overall, no reliable single predictor for PE exists an important risk factor for asymptomatic heart failure.
and the clinical tools are restricted to subjective symptoms with poor Ultrasound Obstet Gynecol 2017; 49: 143–149.
specificity and sensitivity. A combination of maternal characteristics, 10. Sehgal A, Skilton MR and Crispi F. Human fetal growth restric-
biophysical, biochemical, and ultrasound markers may provide a tion: a cardiovascular journey through to adolescence. J Dev
more useful and feasible predictive tool for PE screening in the rou- Orig Health Dis 2016; 7: 626–635.
tine care setting than a test of either component alone. However, 11. Giannakou K, Evangelou E and Papatheodorou SI. Genetic and
further validation and calibration of promising multivariable predic- non-genetic risk factors for pre-eclampsia: umbrella review of
tion models in required before can be clinically implemented systematic reviews and meta-analyses of observational studies.
universally. Ultrasound Obstet Gynecol 2018; 51: 720–730.
Giannakou 223

12. Bujold E, Roberge S, Lacasse Y, et al. Prevention of pre-eclamp- 32. O’Brien WF. Predicting pre-eclampsia. Obstet Gynecol 1990; 75:
sia and intrauterine growth restriction with aspirin started in 445–452. 1990/03/01.
early pregnancy: a meta-analysis. Obstet Gynecol 2010; 116: 33. Poon LC, Kametas NA, Chelemen T, et al. Maternal risk factors
402–414. for hypertensive disorders in pregnancy: a multivariate approach.
13. Roberge S, Bujold E and Nicolaides KH. Aspirin for the preven- J Hum Hypertens 2010; 24: 104–110.
tion of preterm and term pre-eclampsia: systematic review and 34. Poon LC, Kametas NA, Valencia C, et al. Hypertensive disor-
metaanalysis. Am J Obstet Gynecol 2018; 218: 287–293.e281. ders in pregnancy: screening by systolic diastolic and mean arte-
14. Roberge S, Nicolaides KH, Demers S, et al. Prevention of peri- rial pressure at 11-13 weeks. Hypertens Pregnan 2011; 30: 93–107.
natal death and adverse perinatal outcome using low-dose aspi- 35. Khalil AA, Cooper DJ and Harrington KF. Pulse wave analysis:
rin: a meta-analysis. Ultrasound Obstet Gynecol 2013; 41: a preliminary study of a novel technique for the prediction of
491–499. pre-eclampsia. BJOG 2009; 116: 268–276; discussion 276–277.
15. Rolnik DL, Wright D, Poon LC, et al. Aspirin versus placebo in 36. Cnossen JS, ter Riet G, Mol BW, et al. Are tests for predicting
pregnancies at high risk for preterm pre-eclampsia. N Engl J Med pre-eclampsia good enough to make screening viable? A review
2017; 377: 613–622. of reviews and critical appraisal. Acta Obstet Gynecol Scand
16. Wright D and Nicolaides KH. Aspirin delays the development of 2009; 88: 758–765.
pre-eclampsia. Am J Obstet Gynecol 2019; 220: e580–e586. 37. Kuc S, Wortelboer EJ, van Rijn BB, et al. Evaluation of 7 serum
17. Townsend R, Khalil A, Premakumar Y, et al. Prediction of pre- biomarkers and uterine artery Doppler ultrasound for first-
eclampsia: review of reviews. Ultrasound Obstet Gynecol 2019; 54: trimester prediction of pre-eclampsia: a systematic review.
16–27. Obstet Gynecol Surv 2011; 66: 225–239.
18. Trevethan R. Sensitivity, specificity, and predictive values: foun- 38. Leslie K, Thilaganathan B and Papageorghiou A. Early predic-
dations, pliabilities, and pitfalls in research and practice. Front tion and prevention of pre-eclampsia. Best Pract Res Clin Obstet
Pub Health 2017; 5: 307. Gynaecol 2011; 25: 343–354.
19. Conde-Agudelo A, Villar J and Lindheimer M. World Health 39. Akolekar R, Syngelaki A, Poon L, et al. Competing risks model
Organization systematic review of screening tests for pre-eclamp- in early screening for pre-eclampsia by biophysical and biochem-
sia. Obstet Gynecol 2004; 104: 1367–1391. ical markers. Fetal Diagn Ther 2013; 33: 8–15.
20. Nicolaides KH. Turning the pyramid of prenatal care. Fetal 40. Anderson UD, Gram M, Akerstrom B, et al. First trimester pre-
Diagn Ther 2011; 29: 183–196. diction of pre-eclampsia. Curr Hypertens Rep 2015; 17: 584–588.
21. Nicolaides KH. A model for a new pyramid of prenatal care 41. Anderson UD, Olsson MG, Kristensen KH, et al. Review: bio-
based on the 11 to 13 weeks’ assessment. Prenat Diagn 2011; chemical markers to predict pre-eclampsia. Placenta 2012;
31: 3–6. 33(Suppl): S42–47.
22. Duckitt K and Harrington D. Risk factors for pre-eclampsia 42. Wright D, Akolekar R, Syngelaki A, et al. A competing risks
at antenatal booking: systematic review of controlled studies. model in early screening for pre-eclampsia. Fetal Diagn Ther
BMJ 2005; 330: 565–503. 2012; 32: 171–178.
23. McDonald SD, Malinowski A, Zhou Q, et al. Cardiovascular 43. Cetin I, Huppertz B, Burton G, et al. Pregenesys pre-eclampsia
sequelae of pre-eclampsia/eclampsia: a systematic review and markers consensus meeting: what do we require from markers,
meta-analyses. Am Heart J 2008; 156: 918–930. risk assessment and model systems to tailor preventive strategies?
24. North RA, McCowan LM, Dekker GA, et al. Clinical risk pre- Placenta 2011; 32(Suppl): S4–16.
diction for pre-eclampsia in nulliparous women: development of 44. Meads CA, Cnossen JS, Meher S, et al. Methods of prediction
model in international prospective cohort. BMJ 2011; 342: and prevention of pre-eclampsia: systematic reviews of accuracy
d1875–04. and effectiveness literature with economic modelling. Health
25. Wallenburg HC. Prevention of pre-eclampsia: status and per- Technol Assess 2008; 12: iii–iiv, 1–270.
spectives 2000. Eur J Obstet Gynecol Reproduct Biol 2001; 94: 45. Akolekar R, Syngelaki A, Sarquis R, et al. Prediction of early,
13–22. intermediate and late pre-eclampsia from maternal factors, bio-
26. Chaiworapongsa T, Romero R, Korzeniewski SJ, et al. Maternal physical and biochemical markers at 11-13 weeks. Prenat Diagn
plasma concentrations of angiogenic/antiangiogenic factors in 2011; 31: 66–74.
the third trimester of pregnancy to identify the patient at risk 46. Poon LC, Kametas NA, Maiz N, et al. First-trimester prediction
for stillbirth at or near term and severe late pre-eclampsia. Am J of hypertensive disorders in pregnancy. Hypertension 2009; 53:
Obstet Gynecol 2013; 208: e281–287.e215. 812–818.
27. Visintin C, Mugglestone MA, Almerie MQ, et al. Management 47. Anderson UD, Olsson MG, Rutardottir S, et al. Fetal hemoglo-
of hypertensive disorders during pregnancy: summary of NICE bin and alpha1-microglobulin as first- and early second-trimester
guidance. BMJ 2010; 341: c2207–2208. predictive biomarkers for pre-eclampsia. Am J Obstet Gynecol
28. Bramham K, Parnell B, Nelson-Piercy C, et al. Chronic hyper- 2011; 204: 520.e1–5.
tension and pregnancy outcomes: systematic review and meta- 48. Giguere Y, Charland M, Bujold E, et al. Combining biochemical
analysis. BMJ 2014; 348: g2301–2304. and ultrasonographic markers in predicting pre-eclampsia: a sys-
29. Tranquilli AL, Dekker G, Magee L, et al. The classification, tematic review. Clin Chem 2010; 56: 361–375.
diagnosis and management of the hypertensive disorders of preg- 49. Hui D, Okun N, Murphy K, et al. Combinations of maternal
nancy: a revised statement from the ISSHP. Pregnan Hypertens serum markers to predict pre-eclampsia, small for gestational
2014; 4: 97–104. age, and stillbirth: a systematic review. J Obstet Gynaecol Can
30. Mol BWJ, Roberts CT, Thangaratinam S, et al. Pre-eclampsia. 2012; 34: 142–153.
Lancet 2016; 387: 999–1011. 50. Wu P, van den Berg C, Alfirevic Z, et al. Early pregnancy bio-
31. O’Gorman N, Wright D, Poon LC, et al. Multicenter screening markers in pre-eclampsia: a systematic review and meta-analysis.
for pre-eclampsia by maternal factors and biomarkers at 11-13 Int J Mol Sci 2015; 16: 23035–23056.
weeks’ gestation: comparison with NICE guidelines and ACOG 51. Cuckle HS. Screening for pre-eclampsia–lessons from aneuploidy
recommendations. Ultrasound Obstet Gynecol 2017; 49: 756–760. screening. Placenta 2011; 32(Suppl): S42–48.
224 Obstetric Medicine 14(4)

52. Kleinrouweler CE, Cheong-See FM, Collins GS, et al. Prognostic 56. Tan MY, Wright D, Syngelaki A, et al. Comparison of diagnos-
models in obstetrics: available, but far from applicable. Am J tic accuracy of early screening for pre-eclampsia by NICE guide-
Obstet Gynecol 2016; 214: 79–90.e36. lines and a method combining maternal factors and
53. Allen RE, Zamora J, Arroyo-Manzano D, et al. External valida- biomarkers: results of SPREE. Ultrasound Obstet Gynecol 2018;
tion of preexisting first trimester pre-eclampsia prediction 51: 743–750.
models. Eur J Obstet Gynecol Reprod Biol 2017; 217: 119–125. 57. De Kat AC, Hirst J, Woodward M, et al. Prediction models for
54. Mosimann B, Amylidi-Mohr S, Holand K, et al. Importance of pre-eclampsia: a systematic review. Pregnan Hypertens 2019; 16:
timing first-trimester placental growth factor and use of serial 48–66.
first-trimester placental growth factor measurements in screening 58. Giannakou K. Perinatal epidemiology: issues, challenges, and
for pre-eclampsia. Fetal Diagn Ther 2017; 42: 111–116. potential solutions. Obstet Med 2020; 䊏: 䊏.
55. Park FJ, Leung CH, Poon LC, et al. Clinical evaluation of a first
trimester algorithm predicting the risk of hypertensive disease of
pregnancy. Aust N Z J Obstet Gynaecol 2013; 53: 532–539.