4 Evidence and interpretation
4.1.1 The Assessment Group identified eight trials from the previous appraisal 'Guidance on the use of routine antenatal anti-D prophylaxis for RhD-negative women' (NICE technology appraisal guidance 41) that compared a group receiving RAADP using one of the currently licensed regimens with a control group, and that were suitable for inclusion in the current review. Four trials used a dose of 500 IU at both 28 and 34 weeks gestation, one trial used a dose of 1500 IU at both 28 and 34 weeks gestation, and three trials used a single dose of 1500 IU at 28 weeks gestation. Five new publications were identified by the Assessment Group: one of these presented follow-up data from a trial included in the original review, relating to women during subsequent pregnancies; a further three related to two trials included in the original review but did not present new data; and one was a new randomised controlled trial (RCT) comparing intravenous with intramuscular Rhophylac. This new RCT was included in the current review, along with the eight studies from the previous appraisal.
4.1.2 Only one study was an RCT (the new study that compared intramuscular and intravenous Rhophylac), but this was not powered to detect differences between the two treatment arms. One study was a quasi-RCT with year of birth used to allocate participants to treatment groups. One of the other studies was a community intervention trial (controlled before-and-after study), one was a retrospective before-and-after trial, and five were non-randomised studies with historical or geographical controls. Studies that use historical controls may overestimate the effectiveness of RAADP because changes in obstetric care may have led to a decrease in sensitisation. Alternatively, the use of historical controls may underestimate the effectiveness of RAADP because newer assays for maternal anti-D antibody are more sensitive. The use of geographical controls may be confounded by variations in obstetric practice.
4.1.3 Five studies recruited only primigravidae (women with a first pregnancy) and four recruited primigravidae and non-sensitised multigravidae (women with a second or subsequent pregnancy). Most studies reported the rate of sensitisation according to the presence of maternal anti-D antibody at the time of delivery and 6 months after delivery. However, the true rate of sensitisation is higher because of the phenomenon of silent sensitisation. Silent sensitisation means that women have no detectable anti-D antibody, but have developed the template for antibody production and will mount an augmented immune response after FMH in any future pregnancy with an RhD-positive baby. Four studies done in primigravidae reported sensitisation rates in subsequent pregnancies, and one of these provided data on the incidence of sensitisation in subsequent pregnancies in which RAADP was not given. Only two studies took as their primary endpoint the number of women who were RhD negative, who had a baby that was RhD positive and who were found to be sensitised during a subsequent pregnancy with an RhD-positive fetus.
4.1.4 The results across the control groups (women who did not receive RAADP but may have received anti-D immunoglobulin for other indications) were broadly similar; the proportion sensitised ranged from 1.2% to 1.8% (0.8–1.6% in primigravidae and 1.4–2.2% in multigravidae). Over time there was a reduction in the number of women in the control group found to be sensitised during a subsequent pregnancy with an RhD-positive fetus. This could have been a result of improved obstetric practice as well as an improved uptake of anti-D immunoglobulin for potentially sensitising events. In all studies, the rate of sensitisation was lower in the intervention (RAADP) arm. Because the new RCT identified was not sufficiently similar to the previous studies to allow its inclusion in a meta-analysis, the results of the meta-analysis from the original review were considered valid.
4.1.5 This meta-analysis divided the trials into three groups: group 1 comprised four studies that used a dose of 500 IU at both 28 and 34 weeks gestation in primigravidae; group 2 comprised three studies that used a dose of 1500 IU at 28 weeks gestation and included primigravidae and multigravidae; and group 3 comprised two community-based trials in the UK that used a dose of 500 IU at both 28 and 34 weeks gestation in primigravidae. Group 3 was considered the most representative for the cost-effectiveness analysis. This group consisted of UK-based community trials that took as their primary endpoint the number of women who were RhD negative, who had a baby that was RhD positive and who were found to be sensitised in a subsequent pregnancy with an RhD-positive fetus. These trials used an intention-to-treat analysis of all women within a geographical area and demonstrated the likely reduction in sensitisation rate achievable in practice. The sensitisation rate from the meta-analysis of these two trials was 0.95% in the control group (95% confidence interval [CI] 0.18 to 1.71%) and 0.35% in the treatment group (95% CI 0.29 to 0.40%). This represents an absolute reduction in risk of sensitisation in women who are RhD negative and at risk (that is, carrying an RhD-positive baby) of 0.6%, with an odds ratio of sensitisation for RAADP of 0.37 (95% CI 0.21 to 0.65).
4.1.6 There were no trials comparing the two-dose regimen (doses given at 28 and 34 weeks) with a single dose given between 28 and 30 weeks, and no evidence of a difference in efficacy between these regimens.
4.2.1 A systematic review of the cost effectiveness of RAADP identified 11 papers relating to nine studies. Five studies used UK costs, but only two evaluations were applicable to the NHS. One study calculated incremental costs per case of HDN and fetal loss prevented. The results suggested that for most anti-D regimens the use of RAADP in primigravidae would be cost saving in terms of prevention of sensitisation and fetal loss. When RAADP for all women who are RhD negative was compared with RAADP for primigravidae who are RhD negative, the additional cost per incident of sensitisation prevented ranged from £2900 to £8300 depending on the regimen used. The cost per HDN-associated fetal loss avoided was between £42,000 and £120,000. Another study suggested that a programme of RAADP would be cost saving if HDN was eradicated. Similar cost savings were predicted in a study of RAADP in England and Wales. The independent economic evaluation for the previous appraisal (NICE technology appraisal guidance 41) calculated that the incremental cost-effectiveness ratio (ICER) for RAADP was £11,000–13,000 per quality-adjusted life year (QALY) gained for primigravidae compared with no prophylaxis. For multigravidae compared with primigravidae, the ICER was £46,000–52,000 per QALY gained. The evaluation also suggested that adding a utility gain for avoiding fetal loss and interventions in the next pregnancy could reduce the ICER for multigravidae.
4.2.2 No economic models were submitted from the manufacturers.
4.2.3 The Assessment Group modelled a cohort of RhD-negative primigravidae and multigravidae. It assumed the UK birth rate to be 12.1 per 1000 women and that 16% of the population is RhD negative. Each regimen for RAADP was compared with no RAADP. It was assumed that women in their second and subsequent pregnancies had received RAADP in their first pregnancy. The base-case sensitisation rate was assumed to be 0.95% and the odds ratio for each of the regimens of RAADP was assumed to be 0.37. It was assumed that in their first pregnancy 61% of women who are RhD negative will have an RhD-positive fetus and are therefore at risk. This figure was calculated based on the fact that 84% of men are RhD positive, of whom 55% are heterozygous and have a 50% chance of fathering a baby who is RhD positive. Of the 61% of RhD-negative women who are at risk, 0.35% will be sensitised during their first pregnancy. A certain proportion of these (85%) are then expected to go on to have a second baby. Approximately 70% of these babies will be RhD positive because a mother who has had one RhD-positive child is more likely to have another. These babies are at risk of developing HDN. A further 0.35% of women who are not sensitised during their first pregnancy will be sensitised during their second. A smaller proportion (40%) of these either go on to have a third pregnancy where the baby may be at risk of HDN as a result of sensitisation in the first or second pregnancy, or are at risk of sensitisation during their third pregnancy. Similarly, 35% of women who have had three pregnancies go on to have a fourth pregnancy.
4.2.4 The Assessment Group assumed that the probability of fetal loss in pregnancies of sensitised women is around 4%, and that 6% of babies with HDN will have minor developmental problems. Within the model, a child with minor developmental problems had a health utility score of 0.85 and was assumed to incur a cost of £100 per year until 16 years of age. The Assessment Group assumed that 3% of babies with HDN would have major developmental problems. For these children, a health utility score of 0.42 and a cost of £458 per year, over a life expectancy of 60 years, were assumed. The costs of the preparations of anti-D immunoglobulin were taken from the 'British national formulary' (edition 53). Each anti-D injection was assumed to incur an administration cost of £5. The cost of managing a pregnancy in a sensitised mother was estimated to be £2885.
4.2.5 The model assumed that the first RhD-positive child born to an RhD-negative mother is unaffected, and that the risk of sensitisation and the effectiveness of RAADP remain the same in successive pregnancies. In the base case, the model assumes that the loss of a fetus or neonate because of HDN is associated with a 10-QALY loss, in keeping with the previous appraisal (NICE technology appraisal guidance 41).
4.2.6 In the base-case analysis for primigravidae who are RhD negative, comparison of RAADP with no prophylaxis resulted in ICERs of £14,802 (Rhophylac), £19,438 (D-Gam), £25,372 (Partobulin SDF) and £113,827 (WinRho SDF) per QALY gained. For all women who are RhD negative (multigravidae and primigravidae) compared with RhD-negative primigravidae, the ICERs for RAADP were £34,336 (Rhophylac), £45,172 (D-Gam), £59,043 (Partobulin SDF) and £265,807 (WinRho SDF) per QALY gained.
4.2.7 One-way sensitivity analyses suggested that the model results were sensitive to the base-case sensitisation rate and the odds ratio for the sensitisation rate associated with RAADP. The number of QALYs lost because of fetal loss, the rate of fetal loss owing to HDN and the rate of major disability owing to HDN also had significant impacts on the ICER.
4.2.8 The Assessment Group conducted additional analyses that combined primigravidae and multigravidae into one group. Treating the combined group with RAADP was compared with giving no RAADP. This comparison resulted in ICERs of £21,156 for Rhophylac, £27,810 for D-Gam, £36,326 for Partobulin SDF and £163,268 for WinRho SDF per QALY gained.
4.3.1 The Appraisal Committee reviewed the data available on the clinical and cost effectiveness of RAADP for women who are RhD negative, having considered evidence on the nature of the condition and the value placed on the benefits of RAADP by pregnant women who are RhD negative, those who represent them, and clinical specialists. It was also mindful of the need to take account of the effective use of NHS resources.
4.3.2 The Committee considered the evidence on the clinical effectiveness of RAADP. The Committee acknowledged that the clinical trials showed that the use of RAADP reduced the rate of sensitisation in primigravidae and multigravidae who are RhD negative. The Committee heard testimony from the clinical specialists that RAADP is considered to be an effective intervention, although it is not possible to determine what proportion of the total benefit of the use of anti-D immunoglobulin treatment (including use as prophylaxis after potentially sensitising events) is derived from RAADP. The Committee was aware that treatment with RAADP carries the usual risks associated with blood products. However, it heard from the clinical specialists and patient experts that the benefits of using RAADP are much greater than the risks, and that the use of anti-D immunoglobulin, including routine prophylaxis, provides reassurance for pregnant women who are RhD negative. The Committee recognised that anti-D immunoglobulin has no clinical benefit for women who have been previously sensitised. The Committee concluded that RAADP is clinically effective in reducing sensitisation and therefore in reducing the risk of HDN and its consequences.
4.3.3 The Committee considered the costs included in the economic model. The Committee noted that the costs of managing a severe disability were derived from a study of young adults with hemiplegic cerebral palsy and were limited to NHS costs for inpatient, outpatient and emergency care as well as primary care and healthcare in the community. The Committee heard from the clinical specialists that children with a severe disability resulting from HDN were likely to require more NHS resources and a greater range of services than those provided to the young adults in the study. The Committee also heard from the clinical specialists that the cost of managing a pregnancy in a sensitised woman was likely to have been underestimated in the economic model. The Committee concluded that the actual costs to the NHS and to personal social services both of managing a severe disability and of managing a sensitised pregnancy were likely to be much greater than those included in the economic model.
4.3.4 The Committee considered the value that the economic model attached to the loss of a fetus or neonate owing to HDN. The Committee recognised that it was difficult to ascribe a precise utility value to this aspect of the use of the technology, and noted that during the original appraisal (NICE technology appraisal guidance 41) the loss of a fetus late in pregnancy or at birth was assumed to be equivalent to a loss of at least 10 QALYs. The Committee concluded that, in the absence of other data, a disutility for fetal loss of 10 QALYs should also be applied in this appraisal. The Committee noted that parents also experience disutility as a result of the intensive intervention necessary in subsequent pregnancies if sensitisation has occurred, as well as in caring for a child with disability owing to HDN.
4.3.5 The Committee considered the evidence that treatment with RAADP is most cost effective in the first pregnancy and becomes less cost effective with each subsequent pregnancy. The Committee was aware of the need to consider subgroups with differential cost effectiveness and that multigravidae could be further subdivided according to parity and therefore the cost-effectiveness of RAADP. The Committee considered that separate consideration of each pregnancy would cause practical difficulties in refusing women an effective intervention that they had received in an earlier pregnancy. The Committee therefore concluded that it would be more appropriate to consider the cost effectiveness of offering RAADP to all pregnant women who are RhD negative.
4.3.6 The Committee considered the results of the cost-effectiveness analysis. The cost-effectiveness analysis for three of the products resulted in ICERs of between £21,000 and £36,000 per QALY gained for giving RAADP to all women who are RhD negative, irrespective of the number of previous pregnancies, compared with not using RAADP. The Committee acknowledged that the costs associated with the management of a pregnancy in a sensitised woman and with caring for a child with severe disability had been underestimated in the model, and that the disutility of caring for a child with disability was not included in the model. The use of more realistic values for these parameters in the model would decrease the ICERs. The Committee concluded that RAADP is therefore a cost-effective use of NHS resources. The Committee noted that the product WinRho SDF, although licensed, is not marketed for RAADP; in addition, the ICERs for this product would be unacceptably high and it could not therefore be considered as cost effective.
4.3.7 The Committee acknowledged that it is also standard practice to give anti-D immunoglobulin within 72 hours to all women who are RhD negative who give birth to RhD-positive babies, and after potentially sensitising events to all women who are RhD negative. RAADP is given in addition to the anti-D treatment given in these situations. Moreover, use of RAADP is not affected by the administration of anti-D immunoglobulin for other indications earlier in the pregnancy.
4.3.8 The Committee considered the use of single-dose and two-dose regimens. The Committee was aware that there was no evidence of a difference in effectiveness between the regimens. The Committee acknowledged that the differences in cost effectiveness were solely a result of the differences in price of the products, and that the two-dose regimen was associated with higher administration costs than used in the economic model because the second dose may require an extra clinic visit. The Committee noted that use of a single-dose regimen may improve compliance by avoiding logistical failures associated with a second dose, but this would have no effect if the reason for non-compliance is a woman's refusal of treatment. The clinical specialists informed the Committee that, depending on local practice, it may not be possible to administer either the single-dose or the two-dose regimen at routinely scheduled antenatal visits. This may necessitate setting up additional clinics specifically to administer anti-D immunoglobulin, which would incur additional costs. The Committee also heard theoretical concerns that a single-dose regimen may not provide protection towards the end of pregnancy. In addition, the Committee was aware that consideration should be given to limiting a woman's exposure to different batches of anti-D immunoglobulin. Finally, the Committee heard that there are occasionally supply problems with individual products, and that the option of having a range of suppliers was important to ensure the continued availability of anti-D treatment. In summary, the Committee decided that it could not make a firm recommendation for either the single-dose or the two-dose regimen. The Committee also concluded that, although it was not possible to recommend a particular product, individual purchasers should use the product with the lowest cost available locally, taking into account the acquisition cost as well as the costs associated with administration.
4.3.9 The Committee was aware that there maybe circumstances in which a woman cannot receive treatment with anti-D immunoglobulin because of strongly held beliefs that make it impossible for her to accept treatment with blood products. The Committee recognised that passive immunisation is not possible for such women, and that no alternative treatment options exist. The Committee also acknowledged that in certain circumstances it may be unnecessary for a woman who is RhD negative to receive RAADP; for example, if she is planning to have no more children or is in a stable relationship with a man known to be RhD negative. The Committee concluded that a woman eligible for RAADP should be given the opportunity to discuss the benefits and risks so that she can make an informed choice about the use of the treatment.