In general, the analysis of the existing evidence on the economic evaluation of health promotion and primary prevention actions addressing older population groups showed huge differences in the methods applied as well as the overall quality of the studies [
4]. The economic evaluations performed are very heterogeneous in relation to cost categories included and the presentation of outcomes. Therefore, the comparability of results is on the whole quite limited. Although the societal perspective is recommended for the economic evaluation of health promotion interventions, only 13 studies include this approach. Still, some of these studies do not differ from studies that claim to assume a patient/provider perspective with regard to costs and benefits included. These findings corroborate results discussed by Davis et al. [
34] who argue that comparability of fall prevention studies is limited and seek to establish guidelines for conducting and reporting respective economic evaluations.
Heterogeneity is particularly obvious when it comes to the consideration of cost categories and outcomes that need special attention for health promotion interventions aiming at older people. This clearly indicates the need for a standard practice for conducting economic evaluations for health promotion interventions for older people. Theoretical debates on specific requirements of economic evaluations for this target group are rarely reflected on in economic evaluation practice so far.
Research on the economic valuation of informal care has developed significantly over the last 10–15 years, but as yet this is not sufficiently reflected in applied health economic evaluations studies. Given the heterogeneity of methods used to include informal care, more research on the impact of different methods and an increasing consensus on how to value informal care, i.e. by the development of guidelines and/or reference cases, is necessary to attain a better comparability of results [
11,
12,
69]. Given that it is often a defined target of health promotion interventions for older people to sustain self-sufficiency and avoid dependency on long-term care, costs for informal care should be included as part of the direct non-medical costs that may be avoided by an intervention. This comprises costs that are incurred to facilitate informal care and the valuation of the time spent by the uncompensated care-giver. Against the background of the missing consensus concerning the methods to assess informal care, and only a limited knowledge so far on the influences of the different methods on the cost–effectiveness-ratios, it is advisable to assess the impact of the inclusion by sensitivity analyses. If costs of informal care are not included in the study a sufficient explanation should be provided [
12]. Effects on the health or well-being of the care-giver should be considered on the effect side of the economic evaluation.
If productivity costs are included (see below) informal care has to be considered as a value of ‘work-time’ lost or gained due to an intervention. This refers to indirect ‘programme costs’, i.e. value of time spent on the intervention itself as well as paid or unpaid ‘production loss’ or gain.
Productivity costs
The inclusion of productivity costs in health economic evaluations is controversial. As mentioned above, their inclusion is dependent on the normative framework of a study, and national health economic guidelines differ widely in their requirements and recommendations with regard to productivity costs in general and unpaid work in particular [
70,
71]. A study by Krol and Brower on the general role of productivity costs related to unpaid work in current health economic evaluations confirms a general lack of awareness concerning these costs. This is reflected by a lack of clear guidance on how to measure and value lost unpaid work in major health economic textbooks and national health economic guidelines [
70]. Concerning studies on elderly patient populations they conclude that it seems to be common reasoning “that productivity costs are irrelevant since patients are too old to be in paid profession” (ibid. 128).
Thus there are three reasons why the representation of productivity costs is unsatisfactory in health economic evaluations for older people. Firstly, the inclusion of productivity cost is contentious per se and not always required by national guidelines. Secondly, there is no consensus and a lack of guidance on how to measure and value productivity appropriately—in particular with regard to unpaid work. This is especially challenging if productivity costs are not only considered as part of the implementation costs of an intervention, but also comprise the long-term effects of an intervention. Thirdly, there is a disregard of productivity costs, because it is assumed that they are not relevant for this specific target group, which neglects the societal value of seniors’ unpaid work.
A discussion of the implications of different options of including productivity cost and especially unpaid work as part of productivity costs is beyond the scope of this paper. For future research it may be a starting point to report productivity costs separately from direct costs in order to allow a comparison of outcomes between studies. It may not be necessary—or due to lacking data feasible—to include productivity costs in every study; nevertheless, the non-inclusion of productivity costs should be justified very carefully. The omission of a differentiated analysis of productivity costs in the economic evaluation of interventions for older age groups will be problematic if study results are compared to those of interventions aiming at younger age groups that include productivity costs.
The study in the review that considers costs for long-term care and health care in gained years in one version of its evaluations [
62] shows that this has a significant effect on the net benefit of the intervention. These aspects and their possible discriminatory effects on interventions aiming at older people will need further attention in future research, as the inclusion or exclusion of unrelated future medical and long-term care costs may have notable distributional consequences [
19], especially for interventions that increase the longevity of older people.
‘Beyond-health’ benefits and specific preferences of older people
The assessment of the studies under review shows that beyond-health benefits are taken into account only to a very limited extent; and for the most part respective problems are not even mentioned. Limitations of the QALY to cover health benefits of older people or specific preferences of older people that may limit the comparability of the study results were not mentioned at all.
As it is methodologically challenging to include diverse effects in an economic evaluation compared to the assessment of a simple proxy indicator such as ‘number of falls prevented’, and social or beyond-health benefits are rarely included in health economic evaluations in general, it would be desirable that authors at least mention these problems, which is done in only very few studies.
One main reason for the omission of beyond-health effects is the methodical demand for economic studies to condense effects to a quantifiable outcome indicator. As yet, there is only scant methodological guidance on how to include social benefits in economic evaluations beyond cost–benefit analyses. Methodological guidelines were originally developed for clinical interventions, where social benefits are not usually taken into account. Most of the studies included in this review concern fall prevention—interventions with a clearly defined primary objective. As these are relatively similar to clinical interventions—compared to complex multi-layered health promotion or public health interventions—the necessity to include further-reaching effects may have appeared negligible. This might be true if the cost–effectiveness of fall prevention interventions is proven without considering social benefits, but the comparability of these results nevertheless remains limited. Thus, even when comparing results of different fall-prevention interventions it is beneficial to take into account in the cost–effectiveness assessment as to which interventions entail additional social benefits.
The QALY that comprises multidimensional health benefits is an approved outcome indicator in many national health economic guidelines. Its inclusion as an outcome indicator reflects the endeavour to achieve comparability with other interventions. Methodological debates on their limited capability to adequately capture quality-of-life gains for older people have only in recent years been reflected in attempts to develop more adequate outcome indicators. It is essential that concepts are developed that include beyond-health benefits and specific preferences of older people in health economic studies. One promising approach is the use of instruments such as ICECAP-O or ASCOT, which have been developed in recent years to measure the well-being of older people on the basis of their preferences (e.g. [
24]). These instruments are recommended for use, for example, in economic evaluations of social care by the British National Institute for Health and Care Excellence [
72]. Some authors recommend the use of cost consequence analysis in economic assessments in the field of health promotion, public health or other complex interventions [
1,
73,
74]. Benefits such as, for example, general well-being, maintenance of independence, social inclusion, and safety are very relevant potential outcomes of health promotion interventions targeting older people; their ignorance might lead to an underestimation of their positive effects in economic evaluations.
Walter et al. [
75] published a feasibility study in December 2017 on a home-based health promotion intervention for older people with mild frailty, that included an economic evaluation. Their conclusion is that a full randomized controlled trial including an economic evaluation of this intervention is feasible. The study included ICECAP-O and general well-being measures, considered informal care as part of the cost assessment, and while not explicitly referring to productivity costs, the time of the participants spent on caring responsibilities was at least reported. The study thus supports our claim for the inclusion of the categories discussed in this article in health economic evaluations and gives reason for hope for future studies with such a perspective.