Additional file
2 presents the details of the studies reported in the analyzed publications. Among the 15 studies selected for the review, eight studies [
23‐
25,
27,
29,
32,
34,
35] specifically focus on older people (sample includes only 50 years old or over). There is a great diversity in the socio-demographic characteristics of the participants. There are studies targeted at veterans [
26,
27,
37], individuals with specific health problems [
27,
33,
37], sedentary adults [
24,
25,
32] or those from low-income rural households [
29]. In most studies, the sample is dominated by women. This is especially the case of studies for which the recruitment was based on self-selection, i.e. the voluntary responses to the media advertisement [
23‐
25]. The samples in the studies on veterans [
26,
27,
37] are however, largely dominated by men.
Quantitative studies
The review of quantitative studies indicates a great heterogeneity in the design of the studies. We identify six randomized controlled trials [
23‐
28], among them 5 from the USA, with a study period from 1 month [
25,
26] to 6 months [
23,
24,
27,
28] (including post-intervention fallow-up in some studies) and sample size from 45 participants [
24] to 1549 participants [
26]. In addition to the randomized controlled trials, there are three non-interventional studies aiming at evaluating the effects of governmental or insurer programs, i.e. Oportunidades program in Mexico [
29], Vitality wellness program in South Africa [
30], and the preventive bonus program of German Sickness Fund [
31]. We also identified two studies [
32,
33], which present the quantitative results from stated preference studies on the willingness of older people to participate in the prevention programs with financial incentives.
The behavior most often targeted by the financial incentives is physical activity. The effectiveness of financial incentives in motivating adults to walk more is evaluated in three randomized controlled trials [
23‐
25] (mostly women participating), and also in two stated preference studies [
32,
33]. Yet, this incentivized behavior is evaluated differently in the studies, e.g. as number of steps per day or week, minutes of continuous walking per week, days of walking certain number of minutes.
Another frequently analyzed health-related behavior in quantitative studies is screening [
26,
28,
30]. One study focuses on the effectiveness of explicit financial incentives to increase immunization rates, evaluating the effects of Oportunidades governmental program in Mexico [
29]. In two studies [
31,
33], incentives are used to motivate more than one health prevention activity. For example in the German Sickness Fund program [
31], physical activity, screening, immunization and check-ups are taken into account when granting financial incentives.
The explicit financial incentives in the quantitative studies reviewed include only positive incentives (rewards). None of the studies presents the research on penalties (deposits). Most of the evaluated incentives are guaranteed rewards [
25,
27‐
33] as oppose to non-guaranteed rewards, such as lottery or raffle [
23,
24]. One study looks at both, guaranteed and non-guaranteed rewards [
26]. Further, the incentives include largely cash rewards, with only few studies analyzing non-cash rewards, such as shopping voucher [
28], discounts for goods [
30] or in-kind benefits such as bag, watch etc. [
31]. The value of the incentives also differs significantly across the studies with the higher values for non-guaranteed rewards, e.g. participation in $500 raffle for screening completion [
26]. Furthermore, financial incentives are combined with other measures to change health-related behavior, such as peer network [
23], motivational meetings [
24] or information brochure [
28].
The findings from the quantitative studies identified in the review, do not give a clear answer to the question on the effectiveness of explicit financial incentives (rewards) in changing consumer health-related behavior. For example, from the three randomized controlled trials on the effects of financial rewards on physical activity, two studies [
24,
25] indicate that the incentives are effective in increasing walking among older adults, while one study [
23] shows no effect of monetary incentive on meeting walking goals. However, these studies present a great heterogeneity, in terms of duration of the intervention, magnitude and type of the reward, as well as population targeted and sample size. For example, in the study that shows no effects [
23], relatively active older adults (
n = 92) during a 16-week intervention could win weekly monetary reward of max. $200 if they increase their baseline number of daily steps by 50 % in 5 of the past 7 days. On the other hand, two studies which indicate the effectiveness of rewards, specifically focus on inactive older adults, they present a shorter intervention, i.e. 12 weeks [
24] and 4 weeks [
25], and they are based on a smaller sample size (45 and 51 participants respectively). In the former, participants enter into a lottery with a change of winning up to $100 for each day in a week, during which the recommended number of steps was met [
24]. In the latter, the reward is guaranteed i.e. participants receive weekly variable payment (up to $25) depending on the average daily number of aerobic minutes during a week [
25].
Similarly, there is no conclusive evidence on the effects of explicit financial incentives on screening participation studied in the randomized controlled trials. For example, the results of one 6-month randomized controlled trial [
28] indicate that receiving $25 shopping voucher for attending cardiovascular risk assessment with GP, does not significantly increase screening attendance, while in another 30-day randomised controlled trial [
26], the effectiveness of lottery (1 in 10 changes of $50) in increasing the rates of faecal occult blood test completion, was found. Also, we observe mixed results in the studies on the stated willingness to participate in health promotion activities when receiving incentives [
32,
33].
On the other hand, non-interventional studies on selected effects of governmental and insurer programs using financial incentives show rather promising results. Namely, in Mexico, a higher immunisation rate was observed among those older individuals who received cash transfer conditional on adherence to various activities, including attendance at a monthly health seminar and compliance to scheduled preventive health check-ups [
29]; in South Africa, financial incentives (discount on selected goods) increase the likelihood of colorectal cancer screening (but not prostate and osteoporosis screening) among insured older persons [
30], while in Germany older adults in bonus payment program when they receive rewards for participation in various prevention activities, generated significantly lower health care expenditure, leading to cost savings of sickness fund [
31]. The German study is the only among the identified studies which presents the economic analysis of the program. All other quantitative studies focus on the effectiveness of financial incentives without analysing cost-effectiveness or cost-benefits of the programs.
As mentioned in the methods section, we assessed the study designs in a qualitative manner by reviewing the study limitations reported in the publications as well as based on our assessment of the study design. Overall, common shortcomings of the randomized controlled trials are a small study sample [
23‐
25] and its non-representativeness of the general populations (e.g. studies include largely highly educated participants with high health status [
23,
25], women [
23‐
25], male veterans [
26,
27] or patients of a given health care facility [
26‐
28]). The researchers also acknowledge a short duration of these studies (no longer than 6 moths). The main limitation of identified non-interventional studies [
29‐
31] is the possibility of unmeasured confounders and difficulties to investigate the causal effects of incentive on the outcomes measured. Two stated preference studies on the willingness of older people to participate in the prevention programs with financial incentives [
32,
33] suffer mainly from their hypothetical nature, i.e. respondents’ hypothetical statements might not be reflected in real-life situations.
Qualitative studies
Our review also included four qualitative studies on the attitudes of older individuals towards financial incentives. In three of these studies [
34‐
36], data are collected through focus group discussions and in one study this is done through semi-structured interviews [
37]. In two studies, focus group discussions are narrowed down to a specific group of older adults and their health-related behaviour, i.e. physical activity among cardiac rehabilitation patients [
34] and adherence to colorectal cancer screening among the individuals eligible for screening [
35]. In both studies, the views on positive (rewards) incentives are studied. The third study based on focus group discussions [
36], explores the opinions of the general older adults population on both, positive (rewards) and negative (penalties) incentives to modify health behavior. Semi-structured interviews are, on the other hand, conducted among the participants of a randomized controlled trial (veterans) to collect data on their attitudes towards financial reward used in this trial [
37].
All studies reveal the lack of trust among older adults about explicit financial incentives. The main concerns identified in the studies are: immorality and unfairness towards those who take care of their own health and might have to finance rewards for those who engage in risky health behavior [
34‐
37], perception of incentives as bribery [
34‐
36], questionable effectiveness and waste of scarce resources [
34,
36], risk of abusing the scheme [
36], harm to the physician-patient relationship or undermining individual autonomy and intrinsic motivation [
35,
37].
Some forms of incentives seem to be more accepted than others. For example, respondents showed a preference for positive rewards rather than negative penalties or deposits [
36], for in-kind (shopping or gym vouchers) rather than cash incentives [
34,
36], for guaranteed reward rather than lottery [
36], privately sponsored rather than government funded incentives [
34]. In the opinion of older people, providing more tailored and meaningful incentives [
34,
36] can prove better results. The respondents also acknowledge the importance of the size of incentives which in their opinion, should be sufficient for the incentive to be effective [
34,
36,
37]. The acceptability of financial incentives is greater if they prove to be effective [
36]. In one study [
36], education and peer support were mentioned as being more appropriate strategy than financial incentives, to change people's behavior using public resources.
The limitations of the identified qualitative studies include: restriction of the results to a specific population group - only veterans [
37] or individuals with middle and low socio-economic status [
35], small number of focus groups [
34], limited openness of the respondents to discuss this issue [
36] and typical for focus group discussions - the possibility of moderator bias. Moreover, it should be acknowledged that the studies were conducted in various countries with a specific health system environment, including Canada [
34], Israel [
35], the UK [
36] and the USA [
37], which may have affected the opinion of the respondents.