5 Considerations

5 Considerations

5.1 The committee considered the potential effects of pre‑eclampsia on a pregnant woman's life and on the life of her partner. The committee heard from a patient expert that women who have had pre‑eclampsia fear that they will get pre‑eclampsia again if they have another baby. It heard further that a few women decide not to complete their family because of the experience of pre‑eclampsia in a previous pregnancy. The committee also heard that there are currently no tests that can be used to confidently rule‑out the presence of pre‑eclampsia. Therefore pregnant women with suspected pre‑eclampsia often need increased monitoring or admission to hospital, which can be inconvenient and can cause anxiety. The committee concluded that tests that could help to assess suspected pre‑eclampsia could help to reduce anxiety and unnecessary monitoring, in women with suspected pre‑eclampsia or pre‑eclampsia.

5.2 The committee considered the role of placental growth factor (PlGF)‑based tests and noted that PlGF is also a biomarker for placental disease. It heard from clinical experts on the committee that the occurrence of placental disease does not mean that pre‑eclampsia is present. The committee also heard that the requirements for the surveillance and management of placental disease are not well defined and noted that surveillance and management of this disease were beyond the scope of this guidance.

5.3 The Committee considered the definitions of pre‑eclampsia used in the studies compared with the definition of pre‑eclampsia used in the NICE guideline on hypertension in pregnancy. It noted that the definition in the NICE guideline is narrower than the definitions used in the studies. It heard from clinical experts on the committee that the decision on whether to deliver the baby in a woman with pre‑eclampsia is based on clinical symptoms suggesting risk to the mother or baby, which aligns with the expanded definitions of pre‑eclampsia used in the studies. The committee concluded that the expanded definitions of pre‑eclampsia used in the studies do not affect the generalisability of the studies to clinical practice in the NHS.

5.4 The committee further considered the generalisability of the studies included in the systematic review of diagnostic accuracy to clinical practice in the NHS. The committee noted that 3 of the 4 studies were done mainly outside the UK. It heard from clinical experts on the committee that in the NHS, women with suspected pre‑eclampsia would be referred from community care to a day unit for further assessment before a decision on whether to admit them to hospital is made. It heard further that unlike in the NHS, some countries do not have day units so suspected pre‑eclampsia would be managed differently. In countries where there are no day units, women with suspected pre‑eclampsia would be referred from community care directly to hospital for further assessment. This results in suspected pre‑eclampsia being managed for longer in the community than in the NHS, but if symptoms of suspected pre‑eclampsia become worse, admission to hospital would happen earlier. The committee concluded that this may affect the generalisability of some of the studies to clinical practice in the NHS.

5.5 The committee considered the test accuracy reported in the studies. It noted that test accuracy of the Triage PlGF test was better in women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation than in women presenting with suspected pre‑eclampsia between 35 weeks and 36 weeks plus 6 days of gestation. It heard from clinical experts on the committee that decisions on managing suspected pre‑eclampsia are more difficult in women at earlier gestations. For example, in women with pre‑eclampsia between 35 weeks and 36 weeks plus 6 days of gestation, the average time to delivery is about 4–5 days from diagnosis of pre‑eclampsia. However, in women at earlier gestations the risk to the baby from premature delivery is higher than in women at later gestation, so the average time to delivery of a woman at 28 weeks' gestation is about 13 days. The committee heard from a clinical expert that the reason for delaying delivery at early gestations is to reduce the risk of adverse neonatal outcomes. For a woman at 26 weeks' gestation, every day that the pregnancy can be extended leads to a 3% improvement in neonatal outcomes. The committee concluded that PlGF‑based tests would be more clinically useful in women presenting before 35 weeks' gestation.

5.6 The committee considered the 2 tests that were not included in the base-case economic evaluation (the DELFIA Xpress PlGF 1‑2‑3 test and the BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio). It noted that the DELFIA Xpress PlGF 1‑2‑3 test does not have validated cut‑off values, and neither test has diagnostic accuracy data for a population of women presenting with suspected pre‑eclampsia between 20 weeks and 36 weeks and 6 days of gestation. The committee concluded that the diagnostic accuracy of the DELFIA Xpress PlGF 1‑2‑3 test and the BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio could not be assumed to be equivalent to the diagnostic accuracy of the Triage PlGF test and the Elecsys immunoassay sFlt‑1/PlGF ratio.

5.7 The committee considered the use of PlGF‑based tests to rule‑out pre‑eclampsia. The committee noted that the negative predictive value for the Triage PlGF test for the rule‑out of pre‑eclampsia needing delivery within 14 days in women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation, using a cut‑off value of 100 picograms/ml, was 98% (PELICAN study 2013). It also noted that the negative predictive value for the Elecsys immunoassay sFlt‑1/PlGF ratio for the rule‑out of pre‑eclampsia within 1 week in women presenting with suspected pre‑eclampsia between 24 weeks and 36 weeks plus 6 days of gestation, using a cut‑off value of 38, was 99% (PROGNOSIS study 2016). The committee heard from clinical experts that PlGF‑based tests indicate that the placenta is functioning correctly, which gives them confidence to return a woman with a negative PlGF‑based test result (Triage PlGF test result of 100 picograms/ml or more; Elecsys immunoassay sFlt 1/PlGF ratio of less than 38) to community care. It heard further that all women who return to community care, including the small number of women with false negative PlGF‑based test results, would still be monitored, so if symptoms of pre‑eclampsia reappeared they would be picked up by community midwives and the women would be referred back to the day unit for further assessment. The committee therefore concluded that the use of PlGF‑based tests for the rule‑out of pre‑eclampsia in women presenting with suspected pre‑eclampsia between 20 weeks and 36 weeks plus 6 days of gestation was safe and would help to avoid unnecessary hospital admissions. The committee noted its conclusion that PlGF‑based tests would be more clinically useful in women presenting before 35 weeks' gestation (see section 5.5). It concluded further that PlGF‑based tests for the rule‑out of pre‑eclampsia should be used for women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation.

5.8 The committee considered the end point used to determine the accuracy data for the rule‑in of pre‑eclampsia using the Triage PlGF test. It noted that the external assessment group used the end point of pre‑eclampsia needing delivery within 14 days, and considered if an alternative end point could be preterm delivery. The committee heard from clinical experts that the end point of preterm delivery is not equivalent to the end point of pre‑eclampsia needing preterm delivery, because preterm delivery can be for various reasons, not just pre‑eclampsia. The committee concluded that the correct end point had been used in the assessment for the rule‑in of pre‑eclampsia.

5.9 The committee considered the use of PlGF‑based tests for the rule‑in of pre‑eclampsia. The committee noted that the positive predictive value for the Triage PlGF test for the rule‑in of pre‑eclampsia needing delivery within 14 days in women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation, using a cut‑off value of 12 picograms/ml, was 70% (PELICAN study 2013). It also noted that the positive predictive value for the Elecsys immunoassay sFlt‑1/PlGF ratio for the rule‑in of pre‑eclampsia within 4 weeks in women presenting with suspected pre‑eclampsia between 24 weeks and 36 weeks plus 6 days of gestation, using a cut‑off value of 38, was 39% (PROGNOSIS study 2016). It heard from clinical experts on the committee that although the PlGF‑based tests are helpful for ruling-in pre‑eclampsia because they indicate the presence of placental disease, they do not help with the decision on when to deliver the baby. The decision on when to deliver is based on clinical symptoms that indicate risk to the mother or baby, rather than the presence of pre‑eclampsia alone. The committee was concerned that in women with suspected pre‑eclampsia and a positive PlGF‑based test result (Triage PlGF test result of 12 picograms/ml or less; Elecsys immunoassay sFlt 1/PlGF ratio of greater than 38), a decision may be made to deliver the baby sooner on the basis of the PlGF‑based test result alone, rather than on clinical symptoms indicating risk to the mother or baby. The committee considered the unpublished PreOS study of the Elecsys immunoassay sFlt‑1/PlGF ratio in women with suspected pre‑eclampsia. The committee noted that the study was done in Germany and Austria, and heard from a clinical expert on the committee that although the care pathways on admission and surveillance are likely to be slightly different to those in the UK, the results are probably generalisable to clinical practice in the NHS. The results of this study are academic in confidence at the time of writing this diagnostics guidance. The committee concluded that the use of PlGF‑based tests to diagnose (rule‑in) pre‑eclampsia in women presenting with suspected pre‑eclampsia between 20 weeks and 36 weeks plus 6 days of gestation could lead to more unnecessary medical intervention resulting in a greater number of premature babies being delivered. The committee noted its conclusion that PlGF‑based tests would be more clinically useful in women presenting before 35 weeks' gestation (see section 5.5). It concluded that further research on PlGF‑based tests to diagnose (rule‑in) pre‑eclampsia should be carried out for women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation.

5.10 The committee considered how the results of PlGF‑based tests would be used in clinical practice. It was concerned that although PlGF‑based tests could be used to safely rule‑out pre‑eclampsia in women presenting with suspected pre‑eclampsia (see section 5.7), the interpretation of positive test results to rule‑in pre‑eclampsia is more difficult. The committee was concerned that if a PlGF‑based test result were available and the result were positive (Triage PlGF test result of 12 picograms/ml or less; Elecsys immunoassay sFlt 1/PlGF ratio of greater than 38) to rule‑in pre‑eclampsia, too much emphasis might be placed on this result, and not enough emphasis on clinical assessment, which could result in the unnecessary early delivery of the baby (see section 5.9). The committee heard from clinical experts that low levels of PlGF probably indicate placental disease and this is normally managed by close clinical surveillance of the woman and the baby. The committee emphasised the importance of clarity of laboratory reporting on PlGF‑based tests, which should include the relevant negative predictive values of the tests and explain to clinicians that a positive test result (Triage PlGF test result of 12 picograms/ml or less; Elecsys immunoassay sFlt 1/PlGF ratio of greater than 38) when using PlGF testing to rule‑out pre‑eclampsia, does not mean that pre‑eclampsia should be ruled-in. The committee concluded that careful laboratory reporting of PlGF‑based test results combined with targeted medical education for midwives, obstetricians and laboratory staff is important and should prevent diagnostic drift, in which the results of a test designed to make a rule‑out decision start to be used to make a rule‑in decision. The committee concluded that PlGF‑based testing could be used for the rule‑out of pre‑eclampsia in NHS clinical practice but could not be used to rule‑in pre‑eclampsia because further evidence is needed.

5.11 The committee considered the results of the economic analysis and noted that the model included short‑term outcomes. It also noted that there was very little difference in quality-adjusted life years (QALYs) for PlGF‑based test strategies compared with standard clinical assessment in women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation. Further, there was no difference in QALYs for PlGF‑based strategies compared with standard clinical assessment in women presenting with suspected pre‑eclampsia between 35 weeks and 36 weeks plus 6 days of gestation. The committee concluded that PlGF‑based testing strategies could have comparable short‑term clinical benefit to standard clinical assessment, and cost-saving benefits (see section 5.13).

5.12 The committee considered long‑term outcomes after premature birth. It heard from a clinical expert on the committee that babies born very prematurely are at higher risk of long‑term adverse outcomes, such as cerebral palsy. It heard further that these adverse outcomes are very rare, but would have a significant impact on the quality of life of both the child and the parents. Clinicians would take this into consideration when making decisions on when to deliver a fetus in a woman with pre‑eclampsia at less than 34 weeks' gestation. The committee noted that long‑term adverse outcomes for the baby were not included in the economic model and concluded that the QALYs in the economic model were therefore likely to have been underestimated. However, the committee concluded further that any underestimation was likely to be small because of the rarity of these adverse events.

5.13 The committee considered the cost savings in the economic analysis. It heard from the external assessment group that the high negative predictive value of the PlGF‑based tests to rule‑out pre‑eclampsia in women with suspected pre‑eclampsia was driving the cost savings. This happens because using the PlGF‑based tests to assess pre‑eclampsia results in fewer false positive results compared with standard clinical assessment, so fewer women are admitted to hospital unnecessarily. Immediate delivery of the fetus is also reduced, which results in fewer premature babies needing time in a neonatal intensive care unit. The committee noted that most of the cost savings came from using the PlGF‑based tests to rule‑out pre‑eclampsia, through monitoring women in a community care setting rather than admitting them to hospital unnecessarily. The committee concluded that these cost savings are plausible if the PlGF‑based test results are implemented correctly in clinical practice.

5.14 The committee considered the difference in cost savings between the different gestational age groups. It noted that for women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation, cost savings of PlGF‑based test strategies compared with standard clinical assessment ranged between £2,488 and £2,896. However, for women presenting with suspected pre‑eclampsia between 35 weeks and 36 weeks plus 6 days of gestation, cost savings of PlGF‑based test strategies compared with standard clinical assessment ranged between £174 and £365. The committee further noted its conclusion on the clinical utility of PlGF‑based tests at different gestation ages (see section 5.5). The committee concluded that the tests show most promise in those women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation because, on the basis of limited evidence, both clinical utility and cost savings appear likely to be greater in this group compared with women presenting with suspected pre‑eclampsia between 35 weeks and 36 weeks plus 6 days of gestation.

5.15 The committee considered whether repeat testing would be done in clinical practice. It noted that the economic evaluation did not include repeat testing, and that no diagnostic accuracy data were available on repeat testing. However, the economic evaluation did include sensitivity analyses doubling and tripling the cost of the test. Results showed that PlGF‑based testing remained cost saving compared with standard clinical assessment. The committee concluded that the costs of PlGF‑based tests are not prohibitive to repeat testing, but the effect on clinical outcomes is unknown. The committee heard from a clinical expert that repeat testing is often not indicated because a woman who presented with suspected pre‑eclampsia, but had a negative PlGF‑based test result (Triage PlGF test result of 100 picograms/ml or more; Elecsys immunoassay sFlt 1/PlGF ratio of less than 38), may have no symptoms of pre‑eclampsia a week later based on standard clinical assessment, and therefore would not need a PlGF‑based test. In other cases, repeat testing is indicated and would normally be done 2 weeks later unless a woman presents again with suspected pre‑eclampsia before 2 weeks after the previous test. The committee concluded that research on when repeat testing should be done, and the diagnostic accuracy of repeat testing would be useful.

5.16 The committee considered whether PlGF‑based testing could replace the need for quantitative proteinuria testing in the assessment of pre‑eclampsia. It noted that the assessment did not identify any evidence on this question and an exploratory cost analysis suggested that cost savings may only increase slightly if PlGF‑based testing replaced quantitative proteinuria compared with using both tests together. The committee heard from clinical experts that PlGF‑based testing could potentially replace quantitative proteinuria testing in the future.

5.17 The committee considered the advantages and disadvantages of near-patient testing compared with laboratory-based testing. It noted that laboratory-based testing was used in the base-case economic evaluation, and a scenario cost analysis was done on near-patient testing. The committee heard from a clinical expert that the advantages of near-patient testing include an instant result to inform patient care, which may result in a reduction in waiting times for patients and a reduction in unnecessary admittance to hospital while waiting for a result. It also heard that the disadvantages of near-patient care include: midwife time to do the test which may take them away from patient care; the risk of lack of compliance with the standard operating procedure; the risk of inadequate training in doing the test, and interpreting the results. The committee noted further that an exploratory cost analysis on near-patient testing suggested there may be a slight increase in costs savings compared with laboratory-based testing. The committee concluded that each hospital should make their own decision on whether to implement laboratory-based testing or near‑patient testing.

Research considerations

5.18 The committee considered the ongoing and planned research on PlGF‑based tests. It heard from an expert on the committee that PlGF‑based testing is being used with data collection at Liverpool Women's NHS Foundation Trust and that these data may be available soon. It heard further that a small study was due to start in January 2016: Placental growth factor to assess and diagnose hypertensive pregnant women (PARROT). This is a study of PlGF‑based testing for the assessment of suspected pre‑eclampsia, which will be done across 6 centres in England and is expected to report in 2018. The aim of the study is to evaluate the implementation of PlGF‑based tests by investigating whether PlGF measurement in women presenting with suspected pre‑eclampsia between 20 weeks and 37 weeks' gestation decreases the time to a diagnosis of pre‑eclampsia. The study will also collect data on clinical outcomes, but is not powered to show a significant difference in outcomes. The committee also noted that the INSPIRE study, a randomised controlled study, is currently running at the Oxford University Hospitals NHS Foundation Trust. This study aims to determine whether the Elecsys immunoassay sFlt‑1/PlGF ratio will reduce unnecessary hospital admissions and monitoring for patients with suspected pre‑eclampsia by ruling‑out pre‑eclampsia within 1 week. The committee also heard about the ASPRE study, which includes testing with the DELFIA Xpress PlGF 1‑2‑3 test, and noted that this study is a screening study to identify women at high risk of pre‑eclampsia, rather than to help diagnose pre‑eclampsia in women presenting with suspected pre‑eclampsia. The committee concluded that women with suspected pre‑eclampsia who are eligible to enrol in the PARROT and INSPIRE studies should be encouraged to do so. The committee concluded further that it is difficult to carry out research into adverse outcomes arising from early delivery because the outcomes are rare and therefore a study would need to be very large.

  • National Institute for Health and Care Excellence (NICE)