Pre-eclampsia is a major cause of mortality and morbidity in mothers and their offspring. Screening involves testing, usually in the first half of pregnancy to identify women at higher risk of pre-eclampsia; in contrast, testing later in pregnancy is more akin to diagnosis and is not considered here. There have been some advances in screening[1] but these remain to be introduced into widespread clinical practice. This largely reflects our lack of detailed understanding of the pathophysiology. Effective screening would enable women at high risk to be targeted for improved management, prevention and treatment. Therefore effective screening could lead to a major improvement in healthcare. It is reassuring that new biomarkers such as placental growth factor (PlGF) described by Myers et al. in this issue show some promise, distinguishing those at higher risk of pre-eclampsia from a low-risk population. However, it is disappointing that the sensitivity and post-test probability do not warrant introduction into clinical practice. However, PlGF may increase the predictability in mathematical models that combine a number of risk factors.[2]

The early identification of women at risk of pre-eclampsia could also enable antenatal management to be appropriately directed. At present, a huge amount of antenatal resources is targeted towards routine clinic visits including measurement of blood pressure and urinalysis in women at low or moderate risk. An early screening test that would enable those at low risk to be identified, leading to a reduction in frequency of antenatal care, could enable significant cost savings. Another major benefit of discovering an effective screening biomarker is that it may give important clues to the pathophysiological mechanism of pre-eclampsia, leading to improved management, diagnosis and treatment.

Hence, screening for pre-eclampsia seems to make sense. Yet currently, guidance in the UK on screening is mixed. The National Screening Committee does not recommend it.[3] However, recent National Institute for Health and Care Excellence (NICE) guidance on the management of hypertensive diseases during pregnancy promotes the determination of risk factors and using these to inform decisions.[4] Women with at least one high-risk factor for pre-eclampsia or at least two moderate-risk factors are advised to take aspirin 75 mg/day from 12 weeks of gestation. Although screening is not specifically mentioned, and the risk factors are not based on tests normally promoted as ‘screening tests’, risk factors such as body mass index ≥35 kg/m2 are screening tests in all but name. Both of these sources of guidance were influenced by health economic analyses so the remainder of this editorial reflects on three recent key examples and the perspectives they provide. All use the technique of economic modelling, which can be a daunting prospect. In the cases that follow, the approaches are all straightforward. Information about the accuracy of potential screening tests are simply combined with information about the effectiveness of different management strategies to give an estimate of the impact on outcome, such as the number of women with pre-eclampsia. This is illustrated diagrammatically in Figure 1.


Figure 1. Generic decision tree underlying cost-effectiveness models of screening for pre-eclampsia.

Download figure to PowerPoint

The first analysis assessed the cost-effectiveness of a screening strategy relative to no screening.[5] Shmueli et al. undertook a decision model in the Israeli healthcare system. They modelled a testing strategy costing US $112 with a false-positive rate of 10% and a false-negative rate of 23%. This achieved an 18% reduction in pre-eclampsia cases in a population where the prevalence of pre-eclampsia was 1.7%. The cost per case of pre-eclampsia averted was US $67,000, which was calculated to be equivalent to US $19,000/quality-adjusted life-years, which is generally considered cost-effective. The cost-effectiveness improved markedly when the prevalence of pre-eclampsia increased to 3%.

The second analysis by Hadker et al.[6] 2010 compared a novel diagnostic test for pre-eclampsia with standard practice in the NHS. Although the study refers to a ‘diagnostic test’, the test is a screening test using PlGF and soluble fms-like tyrosine kinase-1 (sFt-1). Standard practice also included an element of screening, but used tests with poor accuracy. The novel screening test cost £31.13, had a false-positive rate of 5% and a false-negative rate of 18%. The modelled population had a rate of pre-eclampsia of 4%. The result was that costs per patient were reduced by £945 per pregnancy, which would be considered cost-effective.

Finally, in 2008 the National Institute for Health Research Health Technology Assessment (HTA) programme published a detailed consideration of the evidence relating to screening in pre-eclampsia.[7] It involved a multi-author, multi-disciplinary, international team and at the time of its commission was one of the most ambitious secondary research projects to have been undertaken by the HTA programme. The accuracy of 27 potential screening tests was assessed and, working with the Cochrane Collaboration, the evidence on the effectiveness of 16 potential interventions was examined. Like the other economic evaluations it used modelling to assess the cost-effectiveness in a population at low risk of pre-eclampsia (2.5%). Unlike other modelling exercises it systematically considered all possible tests and management interventions in a variety of different strategies rather than a single one. In particular it considered strategies in which treatments were applied without any previous testing (‘No test/Treat all’). The results led to the conclusion that the most cost-effective approach to reducing pre-eclampsia was the provision of an effective, affordable and safe intervention applied to all mothers without previous testing.

Using the same model as that developed in the HTA programme project, Table 1 illustrates the conclusion. It considers the different ways of combining a hypothetical test with optimal accuracy and least cost and aspirin (the best established of the available treatments).

Table 1. Modelled outcomes and costs of all possible test/treat strategies per 1000 low-risk women
Strategy (ranked by outcome and then cost)Cases of PEWomen without PECosts (000s £)
  1. PE, pre-eclampsia.

  2. Hypothetical test with good accuracy and modest cost; treatment strategy based on preventive aspirin.

  3. Hypothetical screening test: sensitivity 0.82 (false-negative rate 18%); specificity 0.95 (false-positive rate 5%); cost £30.

  4. Aspirin; relative risk 0.81, 19% reduction in risk of PE; cost £2.69.

1. No test/treat all20.25980185.12
2. Test/treat all20.25980215.12
3. Test/treat if positive21.11979220.32
4. No test/no treat25975225.23
5. Test/no treat25975255.23

The key strategies to compare in Table 1 are (1) No test/Treat all and (3) Test/Treat if positive (screening). The No test/Treat all strategy is most effective, and will always be so because all women who could benefit from an intervention receive it. The Test/Treat if positive strategy can only match this if its sensitivity is 100%. The key proviso is that there are no adverse events from the intervention. Cost is also lowest in the No test/Treat all strategy. This is invariably the case where the cost of the intervention is less than the cost of the test.

The results also confirm that both (1) No test/Treat all and (3) Test/Treat if positive (screening) are cost-effective relative to (4) No test/No treat (no screening). This is consistent with the finding of a fourth modelling exercise used to underpin the recent NICE guidance (CG107 Appendix H)[4] and Shmueli et al.[5]

So does the fact that screening for pre-eclampsia is not the most effective or cost-effective option, particularly relative to a No test/Treat all strategy, mean that screening for pre-eclampsia does not make sense?

The answer is no, because of the potential of the approach. Further, if one accepts that the comparators in the models by Shmueli et al.[5] and Hadker et al.[6] are realistic, then the cost-effectiveness of screening may already have been proved enough to suggest that the guidance by the National Screening Committee needs to be re-considered. The findings on cost-effectiveness are certainly not an argument for stopping research on new tests for pre-eclampsia. As well as use in screening, such tests may also improve diagnosis later in pregnancy and advance the understanding of the pathophysiology. Also, just because a screening testing may not make sense in the context of maximising the benefit from inexpensive aspirin, it may still have a place in the targeting of future more expensive single or combined interventions.

This analysis does however highlight that greater attention needs to be given to research overcoming the natural reticence of pursuing No test/Treat all strategies by answering questions such as:

  1. Is aspirin effective in low-risk populations?
  2. Are there likely to be currently adverse effects if aspirin were universally used for the prevention of pre-eclampsia?
  3. Is a No test/Treat all strategy, particularly with aspirin, feasible and acceptable to mothers?
  4. Are there other interventions, such as calcium supplementation, that may be sufficiently effective and inexpensive to be amenable to No test/Treat all strategies?

It is undeniable that pre-eclampsia is a sufficiently important challenge to health internationally that no stone should be left unturned irrespective of whether it involves widespread application of preventive treatments known to be of value or continuing the quest for biomarkers that accurately measure the presence of pre-eclampsia either early or late in its development.


We are grateful to the mothers, researchers and funders who made the studies underpinning these comments possible.

Disclosure of interests

ST provides consultancy advice to the pharmaceutical industry.

Contribution to authorship

Both authors contributed equally.

Details of ethics approval

Not applicable.




  1. Top of page
  2. References