4 Outcomes

The diagnostics advisory committee (section 9) considered evidence from a number of sources (section 10). Full details of all the evidence are in the committee papers.

How outcomes were assessed

4.1 The assessment consisted of:

  • A systematic review of the evidence on the diagnostic accuracy of the 4 index tests for the assessment of suspected pre‑eclampsia in the second and third trimesters of pregnancy.

  • A review of cost-effectiveness evidence on the 4 index tests for the assessment of suspected pre‑eclampsia in the second and third trimesters of pregnancy.

  • A de novo economic model designed to assess the cost effectiveness of placental growth factor (PlGF)‑based tests when used with standard clinical assessment compared with standard clinical assessment alone for the assessment of suspected pre‑eclampsia in the second and third trimesters of pregnancy.

Assessment of test accuracy

4.2 Studies were included in the systematic review if they contained information on:

  • Women with suspected pre‑eclampsia between 20 weeks and 36 weeks plus 6 days of pregnancy who had blood pressure assessment and qualitative assessment of proteinuria.

  • Triage PlGF test; Elecsys immunoassay sFlt‑1/PlGF ratio; DELFIA Xpress PlGF 1‑2‑3 test; or BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio, with standard clinical assessment, or with standard clinical assessment excluding quantitative determination of proteinuria.

  • A reference standard of clinical assessment guided by maternal hypertension, proteinuria, symptoms suggestive of pre‑eclampsia, and ultrasound fetal growth measurements.

  • Test performance outcomes, including diagnostic and prognostic test accuracy (sensitivity, specificity, incidence and related outcome measures) for pre‑eclampsia.

Overview of included studies

4.3 After searches and inclusion screening, 12 publications of 4 studies were included in the review. Two of these studies were on the Triage PlGF test and 2 studies were on the Elecsys immunoassay sFlt‑1/PlGF ratio. None of the studies included more than 1 test; so no head-to-head comparisons of the index tests were available. None of the included studies were on the Perkin Elmer DELFIA Xpress PlGF 1‑2‑3 test or the Thermo Fisher Scientific BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor ratio.

4.4 The PETRA study was a multicentre study of the Triage PlGF test. The details of this study are academic in confidence at the time of writing this diagnostics guidance.

4.5 The PELICAN study (2013) was a prospective, single cohort study of the Triage PlGF test done in 7 centres in the UK and Ireland. It included 287 women with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation, and 137 women with suspected pre‑eclampsia between 35 weeks and 36 weeks plus 6 days of gestation.

4.6 The PROGNOSIS study (2016) was a prospective, international multicentre study of the Elecsys immunoassay sFlt‑1/PlGF ratio, with 1 study centre in the UK. The study had 2 cohorts: a 'feasibility' cohort with 500 patients to derive a cut‑off value based prediction model for the sFlt‑1/PlGF ratio, and a 'validation' cohort with 550 patients to test the model. It included women with suspected pre‑eclampsia between 24 weeks and 36 weeks and 6 days of gestation.

4.7 The study by Alvarez-Fernandez et al. (2014) was a retrospective study of the Elecsys immunoassay sFlt‑1/PlGF ratio in a single cohort of patients. The study was done in Spain and included 62 women with suspected pre‑eclampsia between 20 weeks and 34 weeks of gestation.

4.8 The PELICAN study (2013) and the PROGNOSIS study (2016) defined hypertensive disorders according to the American College of Obstetrics and Gynecology practice bulletin (2002). This gives a broader definition of pre‑eclampsia than the NICE guideline on hypertension in pregnancy, because it includes superimposed pre‑eclampsia and atypical pre‑eclampsia. The definition of pre‑eclampsia used in the PETRA study is academic in confidence at the time of writing this diagnostics guidance. The study by Alvarez-Fernandez et al. (2014) used a simple definition of pre‑eclampsia, which expands on the definition in the NICE guideline on hypertension in pregnancy by including pre-existing proteinuria with superimposed pre‑eclampsia.

4.9 The PELICAN study (2013), the PROGNOSIS study (2016) and the study by Alvarez-Fernandez et al. (2014) were judged to be at low risk of bias using the Cochrane Collaboration adaptation of the QUADAS tool. However, all 3 studies were judged to be at high risk of clinical review bias because the diagnosis of pre‑eclampsia was based solely on whether index-test results were above or below the cut‑off value, whereas in clinical practice index-test results would likely be interpreted alongside clinical signs and symptoms, such as information about hypertension and proteinuria. The PETRA study is academic in confidence at the time of writing this diagnostics guidance.

Diagnostic-accuracy results for the Triage PlGF test

4.10 Diagnostic-accuracy results for the Alere Triage PlGF test are available for 3 test cut‑off values: 100 picograms/ml, the fifth centile of PlGF concentration for gestational age, and 12 picograms/ml. PlGF concentrations above 100 picograms/ml are considered normal and would be used to identify women unlikely to develop pre‑eclampsia needing delivery within 14 days. As such, a result of 100 picograms/ml or greater is used to rule‑out pre‑eclampsia.

4.11 Results from the PELICAN study (2013) show that the Triage PlGF test at cut‑off values of 100 picograms/ml and the fifth centile of PlGF concentration for gestational age gave high sensitivity with good precision for identifying women likely to develop pre‑eclampsia needing delivery within 14 days of testing, when presenting with suspected pre‑eclampsia before 35 weeks' gestation. The cut‑off value of 12 picograms/ml yielded lower sensitivity for identifying women likely to develop pre‑eclampsia needing delivery within 14 days of testing, when presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation (see table 5).

Table 5 PELICAN study results – Triage PlGF test accuracy for predicting pre‑eclampsia needing delivery within 14 days for women presenting between 20 weeks and 34 weeks plus 6 days of gestation

Test cut‑off

Sensitivity

(95% CI)

Specificity

(95% CI)

PPV

(95% CI)

NPV

(95% CI)

<100 pg/ml

0.96

(0.89 to 0.99)

0.56

(0.49 to 0.63)

0.44

(0.36 to 0.52)

0.98

(0.93 to 1.00)

≥100 pg/ml

0.96

(0.89 to 0.99)

0.56

(0.49 to 0.63)

0.43

(0.36 to 0.51)

0.98

(0.93 to 1.00)

<fifth centile

0.96

(0.89 to 0.99)

0.55

(0.48 to 0.61)

0.43

(0.36 to 0.51)

0.98

(0.93 to 1.00)

<12 pg/ml

0.63

(0.51 to 0.74)

0.90

(0.85 to 0.94)

0.70

(0.57 to 0.80)

0.87

(0.82 to 0.91)

Abbreviations: CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value; pg/ml, picograms per millilitre.

4.12 Results from the PELICAN study (2013) also show that for the cut‑off values for the fifth centile of PlGF for gestational age and 12 picograms/ml, the Triage PlGF test had poor diagnostic accuracy for predicting pre‑eclampsia needing delivery within 14 days in women presenting with suspected pre‑eclampsia between 35 weeks and 36 weeks plus 6 days of gestation (see table 6).

Table 6 PELICAN study results – Triage PlGF test accuracy for predicting pre‑eclampsia needing delivery within 14 days for women presenting between 35 weeks and 36 weeks plus 6 days of gestation

Test cut‑off

Sensitivity

(95% CI)

Specificity

(95% CI)

PPV

(95% CI)

NPV

(95% CI)

<fifth centile

0.70

(0.58 to 0.81)

0.64

(0.52 to 0.75)

0.65

(0.53 to 0.76)

0.69

(0.57 to 0.80)

<12 pg/ml

0.22

(0.13 to 0.34)

0.91

(0.82 to 0.97)

0.71

(0.48 to 0.89)

0.55

(0.46 to 0.64)

Abbreviations: CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value; pg/ml, picograms per millilitre.

4.13 The PELICAN study (2013) also reported that the Triage PlGF test at a cut‑off value of 100 picograms/ml had high sensitivity for predicting preterm pre‑eclampsia and delivery within 14 days of testing independent of the pre‑eclampsia diagnosis. A cut‑off value of 12 picograms/ml had poor sensitivity but good specificity for predicting preterm delivery independent of the pre‑eclampsia diagnosis (see table 7).

Table 7 PELICAN study results – Triage PlGF test accuracy for other outcomes for women presenting between 20 weeks and 34 weeks plus 6 days of gestation

Test cut‑off

Sensitivity

(95% CI)

Specificity

(95% CI)

PPV

(95% CI)

NPV

(95% CI)

Preterm pre‑eclampsia

<100 pg/ml

0.90

(0.83 to 0.95)

0.65

(0.58 to 0.73)

0.65

(0.57 to 0.72)

0.90

(0.83 to 0.95)

Delivery within 14 days of testing

≥100 pg/ml

0.94

(0.87 to 0.98)

0.57

(0.50 to 0.64)

0.47

(0.39 to 0.55)

0.96

(0.91 to 0.99)

Preterm delivery

<12 pg/ml

0.44

(0.36 to 0.52)

0.97

(0.93 to 0.99)

0.94

(0.86 to 0.98)

0.62

(0.55 to 0.68)

Abbreviations: CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value; pg/ml, picograms per millilitre.

Diagnostic-accuracy results for the Elecsys immunoassay sFlt‑1/PlGF ratio

4.14 The PROGNOSIS study (2016) derived a cut‑off value of 38; values below 38 were considered negative and were used to rule‑out pre‑eclampsia within 1 week; values above 38 were considered positive and used to rule‑in pre‑eclampsia within 4 weeks. Results from the combined cohort show that for women with suspected pre‑eclampsia between 24 weeks and 36 weeks and 6 days of gestation, sensitivity and specificity for ruling out pre‑eclampsia within 1 week was relatively high. Sensitivity for ruling in pre‑eclampsia within 4 weeks was lower than for ruling out pre‑eclampsia within 1 week, but specificity was relatively high (see table 8).

Table 8 PROGNOSIS study results – Elecsys immunoassay sFlt‑1/PlGF ratio accuracy for women presenting between 24 weeks and 36 weeks plus 6 days of gestation; cut‑off value 38

Sensitivity (95% CI)

Specificity (95% CI)

PPV

(95% CI)

NPV

(95% CI)

Rule‑out of pre‑eclampsia within 1 week

0.86

(0.73 to 0.94)

0.79

(0.77 to 0.82)

0.17

(0.12 to 0.22)

0.99

(0.98 to 1.00)

Rule‑in of pre‑eclampsia within 4 weeks

0.70

(0.62 to 0.78)

0.83

(0.81 to 0.86)

0.39

(0.33 to 0.45)

0.95

(0.93 to 0.96)

Abbreviations: CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value; pg/ml, picograms per millilitre.

4.15 The study by Alvarez-Fernandez et al. (2014) analysed the Elecsys immunoassay sFlt‑1/PlGF ratio at cut‑off values of 23 and 85 (see table 9). Results show that for women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks of gestation, the cut‑off value of 23 had higher sensitivity than the cut‑off value of 85 for rule‑out of pre‑eclampsia within 3 weeks (92% compared with 56%). Specificity was lower for the cut‑off value of 23 than the cut‑off value of 85 for the rule‑out of pre‑eclampsia within 3 weeks (81% compared with 97% respectively).

Table 9 Alvarez-Fernandez et al. (2014) study results – Elecsys immunoassay sFlt‑1/PlGF ratio accuracy for rule‑out of pre‑eclampsia within 3 weeks for women presenting between 20 weeks and 34 weeks of gestation

Test cut‑off

Sensitivity (95% CI)

Specificity (95% CI)

PPV

(95% CI)

NPV

(95% CI)

23

0.92

(0.73 to 0.99)

0.81

(0.64 to 0.91)

0.77

(0.57 to 0.89)

0.94

(0.78 to 0.99)

85

0.56

(0.35 to 0.75)

0.97

(0.84 to 1.00)

0.93

(0.66 to 1.00)

0.77

(0.62 to 0.87)

Abbreviations: CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value.

Review of cost-effectiveness evidence

4.16 Searches were done to identify existing studies investigating the cost effectiveness of Triage PlGF test, Elecsys immunoassay sFlt‑1/PlGF ratio, DELFIA Xpress PlGF 1‑2‑3 test and BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio for diagnosing pre‑eclampsia in the second and third trimesters of pregnancy.

4.17 Four studies described in 4 full text articles (1 unpublished) were included in the review. All studies were cost analyses; that is, they focused on potential savings and did not formally evaluate health outcomes of the mother or baby.

4.18 Two studies by Hadker et al. (2010 and 2013) used the same model, populated with identical clinical inputs, to address UK and German healthcare payer perspectives. The study population was all women assessed for pre‑eclampsia after 20 weeks' gestation. The intervention used was standard assessment plus Elecsys immunoassay sFlt‑1/PlGF ratio (cut‑off value of 85) compared with standard assessment alone. Base-case results from a UK healthcare payer perspective show that there was an overall reduction in cost of £945 per patient from £2,726 to £1,781 associated with the use of the Elecsys immunoassay sFlt‑1/PlGF ratio. From a German healthcare payer perspective, base-case results show an overall cost reduction of €637 per patient from €1,579 to €942 associated with the use of the Elecsys immunoassay sFlt‑1/PlGF ratio.

4.19 A study by Schnettler et al. (2013) included women before 34 weeks' gestation with suspected pre‑eclampsia. The intervention used was standard clinical assessment plus Elecsys immunoassay sFlt‑1/PlGF ratio (cut‑off value of 85) compared with standard clinical assessment alone. The perspective was that of a US healthcare payer. Base-case results show an overall cost reduction of $1,215 per patient, from $3,022 to $1,807, associated with the use of the Elecsys immunoassay sFlt‑1/PlGF ratio.

4.20 The fourth study was by Hunter et al. (2013) and is unpublished and commercial in confidence.

Economic evaluation

4.21 A decision tree model was developed to assess the cost effectiveness of PlGF‑based tests used with standard clinical assessment compared with standard clinical assessment alone in women with suspected pre‑eclampsia in 2 groups:

  • those presenting between 20 weeks and 33 weeks plus 6 days of gestation

  • those presenting between 34 weeks and 36 weeks plus 6 days of gestation.

Model structure

4.22 The model used a linked evidence approach in which maternal, fetal and neonatal outcomes were modelled from diagnostic test accuracy and prevalence of pre‑eclampsia data. The model had 4 components: risk stratification, management, maternal outcomes, and fetal and neonatal outcomes.

4.23 Women with suspected pre‑eclampsia were classified as being at high, intermediate or low risk of pre‑eclampsia. This was based on clinical signs, symptoms or findings with or without the addition of a PlGF‑based test. The probability of pre‑eclampsia in women with suspected pre‑eclampsia was based on the prevalence of pre‑eclampsia and the reported sensitivity and specificity of each diagnostic strategy.

4.24 Suspected pre‑eclampsia could be managedusing expectant management or immediate delivery, dependent on the risk of pre‑eclampsia (high, intermediate or low) and the number of weeks' gestation. Expectant management involves monitoring clinical signs, symptoms and findings, active management of conditions such as hypertension, and planned delivery at 37 weeks of gestation. Immediate delivery involves delivery much sooner, irrespective of gestational age because of clinical findings indicating severe risk to a pregnant woman or fetus. A low risk of pre‑eclampsia is managed on the gestational hypertension pathway (expectant management). An intermediate risk of pre‑eclampsia is managed on a modified version of the gestational hypertension pathway, which has an increased frequency of surveillance (expectant monitoring). A high risk of pre‑eclampsia presenting before 35 weeks' gestation is managed using expectant monitoring when there are no signs of increased risk for the mother or fetus. A high risk of pre‑eclampsia presenting from 35 weeks' gestation is managed by immediate delivery when there are signs of increased risk for the mother or fetus. These assumptions are in line with the NICE guideline on hypertension in pregnancy.

4.25 Maternal and fetal outcomes in the model are assumed to be related to the presence or absence of pre‑eclampsia. As a result, the outcome components of the model are preceded by an evaluation of true disease status. This is the probability of pre‑eclampsia in women in each of the risk categories (high, intermediate, low) assigned in the first stage of the model.

4.26 The maternal outcome component begins with delivery, resulting either from spontaneous labour, induced labour, or planned caesarean section. Each of these modes of delivery may be associated with a risk of conversion to assisted or instrumental vaginal delivery, or to emergency caesarean section. Each mode of delivery is associated with a risk of a severe adverse event associated with the progression of severity of pre‑eclampsia during the delivery, which can result in convulsions. These adverse events may result in admission to an intensive or high-dependency care unit and the need for anti-convulsive therapy. The model assumes that women who do not have convulsions are transferred to the ward after delivery and those who do not have any further adverse events have a normal length of stay for the given mode of delivery.

4.27 The fetal and neonatal outcome component of the model first establishes whether the labour results in a live birth or stillbirth. After a live birth, a neonate may or may not need to be admitted to a neonatal intensive care unit or high dependency unit, the probability of which is related to gestational age, presence or absence of pre‑eclampsia, principal cause of early delivery (maternal condition or fetal distress), mode of delivery, and the presence or absence of complications during delivery. The neonate may then survive or die.

Model inputs

4.28 Test accuracy data for the Triage PlGF test were taken from the PELICAN study (2013) and test accuracy data for the Elecsys immunoassay sFlt‑1/PlGF ratio were taken from the PROGNOSIS study (2016). Test accuracy data for standard clinical assessment were taken from a study included in the systematic review of economic evaluations (Schnettler et al. 2013). The data on prevalence of pre‑eclampsia were taken from the PELICAN study, which was done in the UK. Other clinical and resource use inputs were taken from a variety of published studies.

4.29 Test costs used in the base-case model were taken from economic models produced by the companies. These were slightly different from the list price test costs submitted by the companies. The list price for a single Triage PlGF test is £40. The list price for a single Elecsys immunoassay sFlt‑1/PlGF ratio is £57.23. Other costs were taken from NHS reference costs, the British national formulary and published literature.

4.30 Utility values were taken from the published literature, however, many values had to be mapped from SF‑36 to EQ‑5D (see table 10).

Table 10 Utility values used in the economic model

Parameter

Value

Source

Baseline QALYs from (vaginal) delivery to 6 months post-partum

Birth to 3 weeks post‑partum

0.0389

Jansen et al. 2007

3–12 weeks post‑partum

0.1496

Bijlenga et al. 2011

12–6 months post‑partum

0.2171

Bijlenga et al. 2011

Decrement for caesarean delivery (birth to 3 weeks post-partum)

Non-emergency caesarean section

0.0050

Jansen et al. 2007

Emergency caesarean section

0.0092

Jansen et al. 2007

Decrement for non-spontaneous delivery (induced)

3–6 months post‑partum

0.0084

Petrou et al. 2009

Abbreviation: QALYs, quality-adjusted life years.

Base-case results

4.31 Key assumptions made in the model include:

  • UK guidelines for management of suspected pre‑eclampsia, gestational hypertension and pre‑eclampsia are followed.

  • The 2 different outcomes, pre‑eclampsia needing delivery within 14 days (Triage PlGF test) and pre‑eclampsia within 4 weeks irrespective of delivery time (Elecsys immunoassay sFlt‑1/PlGF ratio) are compared as if they were the same.

  • Costs of neonatal intensive care unit stay capture the effects of neonatal morbidity for deliveries occurring between 35 and 37 weeks' gestation.

  • Tests are done in a central laboratory.

  • The unit costs associated with birth are not dependent on whether the mother has hypertension or pre‑eclampsia.

  • Utility scores for birth are assumed to last for 3 weeks.

4.32 For women with suspected pre‑eclampsia presenting before 35 weeks' gestation, in the base case, total costs varied between £6,048 for the Triage PlGF test to £8,945 for standard clinical assessment. Both the Triage PlGF test and the Elecsys immunoassay sFlt‑1/PlGF ratio were cost saving compared with standard clinical assessment. Cost reductions per patient compared with standard clinical assessment were £2,896 for the Triage PlGF test and £2,488 for the Elecsys immunoassay sFlt‑1/PlGF ratio. Total quality-adjusted life years (QALYs) for each diagnostic strategy were similar, with no more than 0.00076 QALYs separating the most clinically-effective diagnostic strategy and the least clinically-effective diagnostic strategy (see table 11). Although the PlGF‑based tests dominated standard clinical assessment, the differences in QALYs were so small that they could be considered comparable and so the calculation of incremental cost‑effectiveness ratios (ICERs) was not appropriate.

Table 11 Base-case results for women presenting with suspected pre‑eclampsia before 35 weeks' gestation

Strategy

Costs

QALYs

Total

Increment compared with standard clinical assessment

Total

Increment compared with standard clinical assessment

Triage PlGF test

£6,048

−£2,896

0.39445

−0.00076

Elecsys immunoassay sFlt‑1/PlGF ratio

£6,456

−£2,488

0.39434

−0.00066

Standard clinical assessment

£8,945

0.39368

Abbreviation: QALYs, quality-adjusted life years.

4.33 For women with suspected pre‑eclampsia presenting between 35 and 37 weeks' gestation, the cost differences were much smaller than in women with suspected pre‑eclampsia presenting before 35 weeks. In the base case, total costs varied between £3,393 for the Triage PlGF test and £3,758 for standard clinical assessment. Both strategies including PlGF‑based tests were cost saving compared with standard clinical assessment. Cost reductions per patient compared with standard assessment were £365 for the Triage PlGF test and £174 for the Elecsys immunoassay sFlt‑1/PlGF ratio. There was no difference in QALYs between any of the strategies (see table 12). Therefore, ICERs could not be calculated in this analysis.

Table 12 Base-case results for women presenting with suspected pre‑eclampsia between 35 and 37 weeks' gestation

 Strategy

Costs

QALYs

Total

Increment compared with standard clinical assessment

Total

Increment compared with standard clinical assessment

Triage PlGF test

£3,393

−£365

0.3954

0

Elecsys immunoassay sFlt‑1/PlGF ratio

£3,584

−£174

0.3954

0

Standard clinical assessment

£3,758

0.3954

Abbreviation: QALYs, quality-adjusted life years.

Scenario and sensitivity analyses

4.34 A scenario analysis looking at using PlGF‑based tests as an alternative to quantitative proteinuria testing was done. Quantitative proteinuria testing may delay the diagnostic assessment of women with suspected pre‑eclampsia, leading to some being unnecessarily admitted for overnight stays to await results of the quantitative proteinuria test. No evidence was found relating to this question, so a simple cost-based analysis was done. Results suggested that cost savings could increase slightly if PlGF‑based tests replace quantitative proteinuria.

4.35 A scenario analysis looking at the effect on costs of doing the Triage PlGF test in a near-patient setting (a midwifery day unit) rather than in a hospital laboratory was also done. The same unit cost of the test was assumed as in the base case. It was also assumed that because of the adoption of near-patient testing in the midwifery day unit, no women need to be admitted overnight while waiting for test results. For the Elecsys immunoassay sFlt‑1/PlGF ratio, it was assumed that 10% of women being assessed for suspected pre‑eclampsia need an overnight stay while waiting for test results, and for standard clinical assessment a range of 10–50% was assumed. Results suggested that cost savings could increase slightly as a result of doing near-patient testing rather than hospital laboratory testing.

4.36 A scenario analysis was done that considered using PlGF‑based tests to rule‑out (and not rule‑in) pre‑eclampsia. Results suggested that when PlGF‑based tests are used to rule‑out (and not rule‑in) pre‑eclampsia, the total costs increase compared with the base case. For women with suspected pre‑eclampsia presenting before 35 weeks' gestation, the total costs for the Triage PlGF test increased by £1,939 and the total costs for the Elecsys immunoassay sFlt‑1/PlGF ratio increased by £294. However, the Triage PlGF test and the Elecsys immunoassay sFlt‑1/PlGF ratio remain cost-effective compared with standard clinical assessment for women presenting with suspected pre‑eclampsia between 20 weeks and 34 weeks plus 6 days of gestation (see table 13).

Table 13 Rule‑out only scenario analysis results for women presenting with suspected pre‑eclampsia before 35 weeks' gestation

Strategy

Costs

QALYs

Total

Increment compared with standard clinical assessment

Total

Increment compared with standard clinical assessment

Elecsys immunoassay sFlt‑1/PlGF ratio

£6,750

−£2,195

0.39422

0.00054

Triage PlGF test

£7,987

−£958

0.39391

0.00023

Standard clinical assessment

£8,945

0.39368

Abbreviation: QALYs, quality-adjusted life years.

4.37 Deterministic sensitivity analyses were done on the following model inputs:

  • test sensitivity and specificity

  • prevalence of pre‑eclampsia in women suspected of having pre‑eclampsia

  • test cost

  • probability of admission and length of stay in neonatal intensive care

  • distribution of hypertension across women included in the model.

    PlGF‑based test strategies remained cost saving compared with standard clinical assessment in all sensitivity analyses. Changing the assumptions around neonatal intensive care had the biggest impact on costs, but strategies involving a PlGF‑based test remained cost saving compared with standard clinical assessment.

Analysis of index tests not included in the base-case model

4.38 The BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio and the DELFIA Xpress 1‑2‑3 PlGF test were included in the scope for the assessment but were not included in the base-case economic analysis because there were insufficient data. One study comparing the diagnostic accuracy of the BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio with the Elecsys immunoassay sFlt‑1/PlGF ratio was identified, but it did not meet the inclusion criteria for the systematic review of diagnostic test accuracy (Anderson et al. 2015). However, a threshold cost analysis was done for the BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio by assuming that it has equivalent diagnostic accuracy to the Elecsys immunoassay sFlt‑1/PlGF ratio.

4.39 The study by Andersen et al. (2015) compared the diagnostic accuracy of the BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio with the Elecsys immunoassay PlGF/sFlt‑1 ratio. It had 39 patients with confirmed pre‑eclampsia and 76 patients with normotensive pregnancies. The study was not done in a population of patients with suspected pre‑eclampsia, so sensitivity and specificity may be exaggerated. The BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio had a cost based on throughput, therefore the threshold analysis used regression analysis to estimate the point at which the BRAHMS sFlt‑1 Kryptor/BRAHMS PlGF plus Kryptor PE ratio would cost the same as the Elecsys immunoassay sFlt‑1/PlGF ratio, assuming equivalent sensitivity and specificity. The results of this analysis are commercial in confidence.

4.40 No relevant information on the diagnostic accuracy of the DELFIA Xpress 1‑2‑3 PlGF test was available, so no cost-analysis was done.

  • National Institute for Health and Care Excellence (NICE)