Corporate document

2 Cost effectiveness

2 Cost effectiveness

The economic modelling exercise tested the effect of offering the Healthy Start supplements on a universal basis as outlined in the 2 scenarios above. The methods and range of data sources used to develop the economic model are described in detail in the modelling report. An ICER per quality‑adjusted life year (QALY) gained of less than £20,000 was used to indicate that an intervention was cost effective.

It should be noted that there are some significant limitations to the findings because a lack of data for some of the model inputs meant that many assumptions had to be made. Some of these assumptions could be crucial in determining cost effectiveness.

For more details see the sensitivity analyses section and the conclusion.

Universal provision (scenario 1) compared with current targeted provision

Based on the assumptions made, the model suggests that it is not cost effective to make Healthy Start supplements universally available to: pregnant women (from 10 weeks), women with a child aged under 12 months, and children aged over 6 months and under 4 years.

There are 2 reasons for this.

First, lack of data meant it was not possible to assign a quality of life measure to the effects of vitamins A, C and D supplementation. This does not mean that they do not provide benefits, but that the benefits could not be quantified as QALYs.

Second, only a small number of QALYs are gained from providing folic acid to this group to prevent neural tube defects. This is because it is too late for folic acid supplements to reduce the risk for women who are at least 10 weeks pregnant. So only the small number of women who become pregnant within a year of giving birth, while still taking the supplements gain QALYs.

Universal provision (scenario 2) compared with current targeted provision

The model suggests that it is cost effective to make Healthy Start supplements universally available to all: women planning a pregnancy, pregnant women, women with a child under 12 months, infants aged from 0–6 months and children aged from 6 months to 5 years.

But the findings have some significant limitations because a lack of data for some of the model inputs meant that many assumptions had to be made. Some of these assumptions are crucial in determining cost effectiveness. These are:

  • Women planning a pregnancy and those who are less than 10 weeks pregnant are included. This is because folic acid reduces the risk of a large number of women having a pregnancy affected by a neural tube defect. Hence there is a large gain in QALYs plus a reduction in the future costs of caring for those born with a neural tube defect. There are some cost savings from reducing the need to treat vitamin D deficiency, but these are relatively small.

  • The cost per head of including women planning a pregnancy and those who are less than 10 weeks pregnant is not considerably higher[4] than the cost per head for women already in the scheme. A mechanism would need to be identified to deliver a universal scheme to these 2 groups. In the meantime, the true cost of including these additional groups of women is unknown and may have a substantial impact on overall cost effectiveness.

Scenario 2: adding each group in turn to the current target group

The economic modelling exercise also tested the effect of offering Healthy Start supplements to the current target group on a universal basis, plus each of the following groups, in turn, on a universal basis.

Women planning a pregnancy and women less than 10 weeks pregnant

The model suggests that it is cost effective to extend provision of Healthy Start supplements to all women planning a pregnancy and to those who are less than 10 weeks pregnant.

This is because a large number of women would benefit from the reduced risk of a pregnancy affected by a neural tube defect. However, the model made some fundamental assumptions that may have a crucial impact on cost effectiveness. See the sensitivity analyses section and the conclusion.

Women less than 10 weeks pregnant

The model suggests that it is not cost effective to provide Healthy Start supplements to women who are less than 10 weeks pregnant without also extending provision to those who are planning a pregnancy.

That is because, in the model, there is a lower risk of a neural tube defect when women are given folic acid both before and during the first 10 weeks of pregnancy compared with during the first 10 weeks of pregnancy only.

In addition, fewer women will benefit because this group is smaller. Costs are also incurred from giving supplements to infants and children in the current subgroup.

So the benefits do not outweigh the costs and the best‑estimate ICER per QALY gained is above £20,000.

Infants aged 0–6 months

Based on the assumptions made, the model suggests that it is not cost effective to extend the provision of Healthy Start supplements to infants aged 0–6 months alone. This is because they do not benefit from the prevention of neural tube defects. In addition, a lack of data meant it was not possible to assign QALY gains to the effects of vitamins A, C and D supplementation.

Preventing vitamin D deficiency does lead to some cost savings and there may be some practical benefits from offering Healthy Start supplements to infants aged 0–6 months (see section 3). But based on the data available, and in particular the absence of any data on QALY gains for vitamin D, extending the offering to this group alone is not cost effective.

Children aged 4–5 years

The model suggests that extending the provision of Healthy Start supplements to children aged 4–5 years alone is not cost effective, for the same reasons as for extending it to infants aged 0–6 months.

'Spill over effect' on current target group

It was not possible to model any 'spill over' effect of a universal scenario in terms of encouraging more people in the current target group to take the supplements. This is discussed in section 3.

Estimated ICER per QALY gained of extending eligibility for universally available vitamins to the current target group plus each subgroup

Table 1 Estimated ICER for current target group plus each new subgroup

This table shows the estimated ICER per QALY gained of offering Healthy Start supplements on a universal basis to the current target group plus each of the new subgroups in turn. Bold text indicates where the offering is cost effective, that is, the estimated ICER is below £20,000 per QALY gained.

Universal offering is extended to

Incremental costs

Incremental QALYs

ICER

Current subgroups

£7,874,978

13

£620,898

Current subgroups + women planning a pregnancy and less than 10 weeks pregnant

£4,211,201

750

£5618

Current subgroups + women less than 10 weeks pregnant

£6,839,279

243

£28,185

Current subgroups + infants aged 0–6 months

£7,868,568

13

£620,392

Current subgroups + children aged 4–5 years

£8,564,095

13

£675,230

Current subgroups + all new subgroups

£4,893,907

750

£6528

Sensitivity analyses

Sensitivity analyses tested the effect of varying the model inputs. These identified 3 factors that strongly influenced cost effectiveness:

  • The extent of uptake of Healthy Start supplements, in particular by those who are not already taking other supplements that they have bought or been prescribed. Increasing uptake among these groups is key to cost effectiveness but it is not possible to predict this.

  • The proportion of pregnancies affected by a neural tube defect that result in termination. The aim of folic acid supplementation is to reduce the risk of a pregnancy being affected by a neural tube defect. It is estimated that around 80% of these pregnancies, when they do occur, are terminated. The model suggests that the more affected pregnancies that are terminated, the less cost effective the scheme becomes, because the costs for the treatment and care of those born with a neural tube defect are reduced.

  • Lack of quality of life data for vitamin D deficiency meant QALYs could not be assigned to the effects of vitamin D supplementation in the model. Because the original impetus for this work was concern about the prevalence of disease related to vitamin D deficiency, a sensitivity analysis tested the hypothetical effect of applying a QALY loss to those with symptomatic vitamin D deficiency. This showed that the higher the QALY loss, the more cost effective the scheme would be. This also reduced the impact of the factors cited above.

Conclusion

The model suggests that making Healthy Start vitamin supplements universally available is cost effective, as long as women who are planning a pregnancy and those in the first 10 weeks of pregnancy are included. This is based on the assumptions made in the model.

Although the original impetus for this work was concern about the prevalence of disease related to vitamin D deficiency, a lack of data meant it was not possible to assign QALYs to the effects of vitamin D in the model.

Consequently, folic acid has a dominant impact on the findings. This is because folic acid was the only nutrient with a measurable outcome –the prevention of neural tube defects – to which a QALY could be assigned. In addition, the number of QALYs saved by preventing pregnancies affected by a neural tube defect was high enough to outweigh the costs of including the other groups.

There are some cost savings from preventing vitamin D deficiency and although QALYs could not be assigned to vitamins A or C, increasing some people's intake of these above the lower reference nutrient intake could be an added benefit.

However, a lack of data for some of the model inputs meant that many assumptions had to be made. Some of these were fundamental and could have a crucial impact on cost effectiveness:

  • A delivery mechanism can be identified through which a universal scheme can be successfully implemented with women planning a pregnancy and those in the first 10 weeks of pregnancy. These women are not currently included in the Healthy Start scheme and it was beyond the scope of this work to identify an effective delivery mechanism for this group.

  • The cost per head of including these groups of women is not considerably higher than the cost per head for women already in the scheme[5]. The true costs are unknown and may have a substantial impact on overall cost effectiveness.


[4] Additional analyses would be needed to determine the actual cost per head beyond which it would no longer be cost effective to include these women in the scheme. This was beyond the scope of the work on which this report is based.

[5] To identify the cost per head at which it would no longer be cost effective to include these women in the scheme, would need additional analyses. This was beyond the scope of the work on which this report is based.