Home > Clinical Guides > Hypertension in Pregnancy: The... > Cost effectiveness of aspirin compared...

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Women's and Children's Health (UK). Hypertension in Pregnancy: The Management of Hypertensive Disorders During Pregnancy. London: RCOG Press; 2010 Aug. (NICE Clinical Guidelines, No. 107.)

Appendix HCost effectiveness of aspirin compared with no aspirin in preventing pre-eclampsia in women at risk of developing pre-eclampsia

Introduction

Pre-eclampsia is associated with high maternal and neonatal morbidity and mortality. Worldwide, pre-eclampsia and eclampsia are estimated to be responsible for approximately 14% of maternal deaths per year (50 000–75 000).238 Pre-eclampsia is estimated to account for one-fifth of antenatal admissions, two-thirds of referrals to day-care assessment units and one-quarter of obstetric admissions to intensive care units in the UK.239 Interventions that aim to reduce the risk of pre-eclampsia may be cost effective or even cost saving if the intervention leads to lower overall health service costs by reducing the need for continuing assessment and admission and thereby freeing up scarce NHS resources to be used to improve health in other ways.

An economic model was developed to consider the use of aspirin in the prevention of pre-eclampsia. The question was partially addressed by Meads et al. (2008).39 However, the GDG considered that the reported test accuracy and effectiveness data in that study were not sufficiently robust to be used in this model since the data were obtained from heterogeneous populations. Instead, the economic model developed for this guideline used data from the PARIS study,42 which showed that aspirin was clinically effective in preventing pre-eclampsia.

Objectives

To determine the cost effectiveness of aspirin versus usual management in the prevention of pre-eclampsia and its complications in women at risk of developing pre-eclampsia.

Model structure and assumptions

A probabilistic model was developed in Microsoft Excel™. The analytic structure is illustrated by the schematic in Figure H.1. In the model, all women have pregnancy-related hypertension at week 12 of their pregnancy and are at risk of developing pre-eclampsia. For simplicity, Figure H.1 shows only a sub-tree of the whole model representing those women who develop pre-eclampsia. The pathway is identical for women who do not develop pre-eclampsia. The model includes the following maternal outcomes: delivery before 34 weeks, delivery of babies who are small for gestational age (SGA), and death. Outcomes for the neonatal infant are: delivered healthy with no admission, delivered healthy and admitted, and delivered healthy but die before discharge.

Figure H.1. Model structure for the cost effectiveness of aspirin in preventing pre-eclampsia (pre-eclampsia subtree shown).

Figure H.1

Model structure for the cost effectiveness of aspirin in preventing pre-eclampsia (pre-eclampsia subtree shown).

Model event rates

The incidence of maternal outcomes were taken from the placebo arm of the PARIS study.42 Neonatal admissions were taken from Habli et al. (2007).145 The baseline data with no treatment and the treatment effectiveness data (both taken from the PARIS study42), are shown in Tables H.1 and H.2, respectively. The outcomes of interest were pre-eclampsia, perinatal and maternal deaths, SGA babies, birth before 34 weeks, hospitalisation, maternal and neonatal quality of life, and healthcare costs. The side effects of aspirin were not explicitly considered as the GDG felt that the aspirin dose recommended for use in this population is sufficiently small (75 mg) and treatment duration sufficiently short not to have any significant side effects such as internal bleeding.

Table H.1. Baseline event rates with no treatment: all women with pregnancy-related hypertension who are at risk of developing pre-eclampsia, and by gestational age.

Table H.1

Baseline event rates with no treatment: all women with pregnancy-related hypertension who are at risk of developing pre-eclampsia, and by gestational age.

Table H.2. Treatment effects of aspirin in all women with pregnancy-related hypertension who are at risk of developing pre-eclampsia.

Table H.2

Treatment effects of aspirin in all women with pregnancy-related hypertension who are at risk of developing pre-eclampsia.

Cost inputs

In accordance with NICE methods for clinical guidance,38 a public sector, NHS and Personal Social Services (PSS) perspective was adopted.

A systematic review of the economic literature to search for costs was undertaken as part of the guideline development process. All costs are presented in GB pounds, at 2008–09 prices. Drug costs were taken from the British National Formulary198 and the cost of other outcomes were taken from NHS reference costs.240 The model’s cost inputs are shown in Table H.3. It was assumed that women who did not develop pre-eclampsia gave birth in an obstetric unit and no assumptions were made about the mode of delivery since the GDG consensus was that aspirin had no impact on this. For simplification, it was assumed that each woman had an uncomplicated vaginal delivery.

Table H.3. Health service costs incurred by women with pre-eclampsia, 2008–09.

Table H.3

Health service costs incurred by women with pre-eclampsia, 2008–09.

Valuing outcomes

The Harvard Cost-Effectiveness Registry was searched for quality of life values associated with normotensive pregnant women. One study was identified that evaluated the cost effectiveness of contraception methods in women of average health and fertility, ranging from 15 to 50 years of age compared with non-use of contraception.243 The authors found that short-term loss of quality of life due to pregnancy was 0.0375.

For this guideline, no quality of life data associated with pre-eclampsia could be identified and therefore it was assumed that those who developed pre-eclampsia had the same quality of life as normotensive pregnant women, based on GDG opinion. It was assumed that all children discharged alive would live a normal healthy life up to 80 years and have 27.7 discounted quality-adjusted life years (QALYs). Thus the total QALYs lost was the sum of maternal and neonatal QALY loss. The model’s QALY value are shown in Table H.4.

Table H.4. Quality of life loss assigned to pregnant women and neonatal death (QALYs).

Table H.4

Quality of life loss assigned to pregnant women and neonatal death (QALYs).

Probabilistic sensitivity analysis

Probabilistic sensitivity analysis (PSA) was undertaken to explore to what extent the results were affected by the uncertainty surrounding the model input parameters. In PSA, each model parameter is assigned a distribution reflecting the expected sampling variation. Costs and effects are determined after simultaneously selecting random values from each distribution. The process is repeated many times in a Monte Carlo simulation to give an indication of the extent to which model input parameter uncertainty affects the incremental cost-effectiveness ratio (ICER), that is, change the relative order of cost effectiveness between alternatives. Distributions were not applied to cost parameters as there was generally little uncertainty associated with this data, but treatment costs of pre-eclampsia were varied in one-way sensitivity analysis.

One-way sensitivity analysis

In addition to the probabilistic sensitivity analysis, one-way sensitivity analyses were undertaken to assess the impact of changing input parameter values on the base-case results. This was restricted to parameters where there was uncertainty that the GDG felt could possibly alter the results. Using ranges suggested by the GDG and incorporating the 95% confidence interval (CI), the treatment effect was varied on various outcomes, the short-term utility loss and the cost of treating pre-eclampsia.

Results

Table H.5 shows the results of the deterministic (static) economic model for a cohort of 100 pregnant women. A cohort of 100 was chosen for illustrative purposes representing a typical GP practice.

Table H.5. Outcomes in both treatment strategies per 100 pregnant women at risk of developing pre-eclampsia.

Table H.5

Outcomes in both treatment strategies per 100 pregnant women at risk of developing pre-eclampsia.

There were more adverse outcomes in women who did not take aspirin compared with those who did. There were more cases of pre-eclampsia, more babies were delivered before 34 weeks, more babies were SGA and there were more neonatal admissions, all of which require additional NHS resources. The costs of these adverse events offset the initial costs of giving aspirin to all pregnant women at risk of developing pre-eclampsia.

The total costs per woman were £270,663 for those who received aspirin compared with £278,515 for those not taking aspirin (Table H.6). Aspirin generated less QALY loss compared with no aspirin (13.66 versus 14.18) and was the cheaper strategy overall, resulting in savings of £7,852 per pregnancy and 0.52 additional QALYs per pregnancy. In this scenario, cost effectiveness was unequivocal and aspirin is said to dominate no aspirin in women at risk of developing pre-eclampsia (that is, giving aspirin is cheaper and results in more health benefits). The results demonstrate that, using these baseline data for cost and effectiveness, the policy of giving all pregnant women at risk of developing pre-eclampsia aspirin is cost saving when compared with no aspirin.

Table H.6. The cost effectiveness of aspirin versus no aspirin for a pregnant women at risk of developing pre-eclampsia.

Table H.6

The cost effectiveness of aspirin versus no aspirin for a pregnant women at risk of developing pre-eclampsia.

Probabilistic analysis

The results of 1000 iterations of the model are illustrated on the cost-effectiveness/decision plane in Figure H.2. Each point represents the ICER of aspirin compared with no aspirin derived from one iteration of the model and shows that, in 99.8% of the iterations, aspirin was cost saving and resulted in more QALYs, as shown by the close bunching of points in the south-east quadrant. All points lie below the black line that represents the willingness to pay threshold, in this case £20,000/QALY.

Figure H.2. Cost-effectiveness plane comparing aspirin use in pregnant women at risk of developing pre-eclampsia with no aspirin.

Figure H.2

Cost-effectiveness plane comparing aspirin use in pregnant women at risk of developing pre-eclampsia with no aspirin.

One-way sensitivity analysis

Varying the treatment effect on pre-eclampsia

In the base-case analysis, aspirin was found to reduce the incidence of pre-eclampsia by about 10%. The 95% CI ranged between 84% (lower) to 97% (upper). The 95% CI was used in sensitivity analysis and the results did not change (that is, aspirin was always the preferred strategy). There were more savings and high QALY gain when treatment effects were higher. When the lower values in the CI were put in to the model, the savings increased to around £12,643 per pregnancy and QALY gain to about 0.59 per pregnancy compared with savings of only £2,263 and QALY gain of about 0.44 when a 3% reduction in the incidence of pre-eclampsia was assumed. The effect of treatment effect size on cost savings is shown in Figure H.3.

Figure H.3. Sensitivity analysis showing cost savings of aspirin compared with no aspirin in women at risk of developing pre-eclampsia, varying treatment effect on the incidence of pre-eclampsia across the 95% CI (0.84–0.97).

Figure H.3

Sensitivity analysis showing cost savings of aspirin compared with no aspirin in women at risk of developing pre-eclampsia, varying treatment effect on the incidence of pre-eclampsia across the 95% CI (0.84–0.97).

Varying the aspirin treatment effect on the incidence of neonatal death, maternal death, SGA and birth before 34 weeks (these outcomes were varied one at a time)

Aspirin remained cost saving when treatment effects on neonatal outcomes were varied across the 95% CI. When the lower end of the 95% CI was used (suggesting a bigger treatment effect) the aspirin strategy generated more savings than when the upper end of the CI was used. In all scenarios, the strategy was cost saving. A worst-case scenario was also considered where all parameters were set at their upper limit of the 95% CI at once and the model remained cost saving, although the savings fell from £79 to £19 per person.

Varying the short-term utility loss from pre-eclampsia

In the base case we assumed that short-term utility loss due to pre-eclampsia was the same as that of normotensive pregnant women, which was 3.75%. The GDG suggested a range of 1–15% and this was tested in sensitivity analysis. The results, illustrated in Figure H.4, demonstrate the relationship between short-term utility loss due to pre-eclampsia and overall QALY loss when aspirin is not taken. Aspirin remained dominant even at low short-term utility loss.

Figure H.4. Sensitivity analysis showing QALY loss for women not taking aspirin compared with those taking aspirin in women at risk of developing pre-eclampsia, varying short-term utility loss from pre-eclampsia over a range suggested by the GDG.

Figure H.4

Sensitivity analysis showing QALY loss for women not taking aspirin compared with those taking aspirin in women at risk of developing pre-eclampsia, varying short-term utility loss from pre-eclampsia over a range suggested by the GDG.

Varying the cost of treating pre-eclampsia

The cost of treating pre-eclampsia was varied between £500 and £10,000. The cost of pre-eclampsia did not affect model results across this wide range. There were fewer cases of pre-eclampsia in the aspirin strategy than the no aspirin strategy, meaning that the reduced cost of treating pre-eclampsia more than offset the increased cost of aspirin treatment.

Discussion

The model demonstrated that, in a wide range of scenarios, the aspirin strategy was cost saving compared with a no aspirin strategy for women at risk of developing pre-eclampsia. This is essentially because aspirin is a very low-cost intervention that works effectively. The savings were driven by cost savings due to a lower risk of adverse events requiring hospitalisation in the aspirin group. The model suggested that the aspirin strategy would result in fewer cases of pre-eclampsia, fewer neonatal admissions; fewer women delivering before 34 weeks and fewer SGA babies. Probabilistic sensitivity analysis suggested that there is a 99.8% probability that giving aspirin is cost saving.

The effectiveness data were taken from a high-quality individual-patient meta-analysis. The analysis demonstrated that, on average, aspirin will reduce the incidence of adverse morbidity by about 10%. No published economic evaluations of aspirin in women at risk of pre-eclampsia were identified. However, it is acknowledged that aspirin has been widely evaluated in the cardiovascular field, where it has also been shown to be cost saving.

Quality of life weightings derived from normotensive pregnant women were used. A conservative assumption was also made about the quality of life of women who develop pre-eclampsia, which was assumed to be the same as that seen normotensive women. The GDG felt it was difficult to measure quality of life of children and thus neonatal morbidity was not considered explicitly in this model. A conservate approach was takenn by excluding quality of life of children as this would have strengthened the cost effectiveness of aspirin conclusion. Sensitivity analysis showed that aspirin still generated more QALYs whether the utility loss from pre-eclampsia or pregnancy was low or high.

Conclusion

This model shows that aspirin strategy is cost saving compared with no aspirin in women who are at risk of developing pre-eclampsia across a wide range of assumptions.

Copyright © 2011, Royal College of Obstetricians and Gynaecologists.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK [www.cla.co.uk]. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

Cover of Hypertension in Pregnancy
Hypertension in Pregnancy: The Management of Hypertensive Disorders During Pregnancy.
NICE Clinical Guidelines, No. 107.
National Collaborating Centre for Women's and Children's Health (UK).
London: RCOG Press; 2010 Aug.

NICE (National Institute for Health and Care Excellence)

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...