There are several ethically relevant insights that emerge from this survey. The large support for challenge studies from this cross section of the public provides support for introducing them in Singapore. Medical research has a notoriously troubled history with human experimentation, including in Singapore, and implementing a policy of purposefully infecting people with a virus to test a novel vaccine without public support might have detrimental effects not only in terms of support of vaccination policies of public health importance, but of trust in science more generally (Abdul Majid 2021). After all, medical research depends on people’s trust for its funding and success (Wilson and Hunter 2010). This study suggests that CHI studies in Singapore will be able to obtain social license.

One major limitation of this survey is that despite engagement in other forms of research by many individuals, most respondents reported limited pre-survey knowledge about CHI studies. This emphasizes the need for public education and engagement during the launch of the CHI program. However, even with public education, it would be useful to understand why a small but significant minority disagreed with them. This may best be done with qualitative studies such as focus group discussions where time is provided to share detailed information about CHI studies and the nuances of participant responses can be understood.

While the majority supported the concept of CHI studies (almost two-thirds of respondents), only a minority would be prepared to enroll in a SARS-CoV-2 challenge study themselves (less than one-third) or would be supportive if one of their family members were to enroll. Even among those with previous experience as research participants, the majority—although a small one—said they would not be willing to volunteer as participants.

This discrepancy might reflect the population surveyed where most would not meet eligibility criteria for typical CHI studies due to age, medical conditions, or other study requirements. The Barker survey was conducted in October 2020 during the height of the public health emergency prior to vaccines being available, and with many COVID-19-related restrictions in place. This context may explain why more than 60% of respondents in the UK would be supportive of an immediate family member participating in a SARS-CoV-2 challenge study, compared with only 38% in Singapore. It might also contribute to the high percentage of support for CHI in Singapore in 2023 as most of the population had been either infected or vaccinated or both by then. Decision-making on participation in research studies has been observed to be different in parts of Asia compared with Western countries, with a greater involvement of family members in this process in Asia (Susilo et al. 2019). Whether this applies to potential CHI study participants in Singapore is not known, but the survey highlights the importance of a carefully managed informed consent process, including providing ample time, information, and privacy for potential participants to consider and discuss their participation with people close to them, and ensuring consent is treated as a continuous, dynamic process.

That most survey respondents would not want to participate in a challenge study is not surprising. The tension between third-person, ethical judgments (“this is the right/a good thing to do”) and first person, self-interested judgments (“I am not going to do it myself/I am not willing to pay the cost of it myself”) is both a common feature of human thinking and something that has been observed with other proposed health policies. For instance, people’s views on antimicrobial prescribing, in light of the problem of antimicrobial resistance, present the same tension (Dao et al. 2019): when asked whether doctors should prioritize patients or public interest in antibiotic prescription, respondents largely sided with the public, but when asked the same question in the case they themselves were the patient, support for the society-first approach was significantly lower. This is similar to the “Not-in-my backyard” situation where people recognize the need for a common good but are not willing to pay any price to contribute to it themselves.

CHI studies present the same problem: they have the potential to benefit public health by delivering better interventions such as vaccines in a faster way. But there will be some risk and inconvenience to individuals who enroll in these studies for the sake of societal benefit. This raises two issues. One, strictly utilitarian in nature, is about creating enough incentives for enough individuals to enroll. The second, fairness-based, is about ensuring that those who do enroll are adequately compensated and not induced to participate against their best interests such as by concealing underlying medical conditions. The first issue does not seem to raise particular concerns: as said at the beginning, one advantage of challenge studies is that they require fewer participants than traditional trials. If one-third of the population was willing to enroll, even with strict study eligibility criteria for participation, enough people are likely to be identified to be able to run many successful challenge studies. However, the second issue is more complex, as is the question of what counts as fair treatment.

In our view, fair compensation for participating in a CHI study might justifiably take into account more than time and expenses—as is normally the case for traditional trials—but also the risks regarding participant physical and mental health. This is particularly the case when participants may not be expected to directly benefit from their participation (Lynch et al. 2021). This is a complex issue, but is arguably no different from what is already the case in other contexts, for example in the case of certain jobs, where the risks associated with that job are reflected in the remuneration.

What is striking in this study is that over 60% of Singaporeans supported either a Market Model or a Payment for Risk Model of payment. Neither of these represents the status quo when research ethics committees are assessing compensation to participants in clinical research, which is based on subjective estimates of compensation for time and burden (e.g. pain). This is interesting in and of itself as it might be an example of the more general discrepancy, observed elsewhere, between the ethical views of the public and those of bioethics expertise (Pierson et al. 2024).

Some would see payments for risks as a form of coercion or undue inducement—that is, as something that would circumvent people’s rational capacity for autonomous choice. A lot of these objections depend very much on what is meant by “coercion”. On some accounts of coercion (Millum and Garnett 2019), offers of payments and incentives could count as coercive, because they put excessive psychological pressure on individuals—making “offers that they cannot reasonably refuse”. Thus, a payment for risks would add monetary value to enrolment in challenge studies, which would make it too tempting for some people. Other accounts of coercion (Wertheimer and Miller 2008), however, argue that only threats of penalties can meaningfully coerce, as they restrict the range of options open to a person. For instance, the paradigmatic case of “your money or your life” is a case of coercion because I cannot keep both my money and my life, so my range of options compared to the relevant baseline is restricted. These divergent accounts of coercion, though, share a common ethical thread: certain actions can significantly undermine autonomy and rational capacity. As such, whatever one’s view of coercion, what matters is whether paying people for risks, or indeed paying them higher sums than what is normally done in the case of traditional trials, significantly undermines their autonomy and rational capacity.

Again, this largely comes down to how autonomy and rationality are defined, which is a complex philosophical question beyond the scope of this paper, and to empirical evidence about how monetary incentives affect participants’ decision-making—for instance, evidence suggesting limited influence of monetary incentives on decisions to enroll in studies (Largent et al. 2022) would count against the claim that autonomy and rational capacity are undermined. For the purpose of this paper, we are examining whether any perceived limitation of autonomy would be considered acceptable by the society implementing that policy. The general support in our sample for market- or risk-based payment is indeed consistent with other generally socially acceptable practices around compensation, even if there is some tradeoff in terms of autonomy. Here is why.

Let’s assume, for the sake of argument, that higher payments for risk in some sense limit autonomy in deciding whether to enroll (Savulescu 2001). This would entail that any additional payments for the risks associated with more hazardous occupations—such as special military duties or firefighting—also limit people’s autonomy to take up those jobs. But we normally do not think in those terms in the case of hazardous occupations, so why then should it be a reason for preventing people from accepting larger payments for risks and other costs in challenge trials? To say that payments undermine autonomy in one case, but not the other, is inconsistent. Some have further suggested that preventing people from enrolling in paid research out of concerns for undue inducement would be a form of unjustifiable paternalism (Savulescu 2001), which is normally not adopted in the case of other choices made for the sake of money. To the extent that payments undermine autonomy, they do it across the board, whether we are talking of medical research or of the choice of hazardous occupations.

Some would appeal to “research exceptionalism” (Wilson and Hunter 2010), that is, the idea that medical research is somehow special and different from other contexts, such that ethical considerations that apply elsewhere are not straightforwardly transferrable to research. Perhaps medical research is different and ethical constraints apply that would not apply to, say, the case of choosing highly paid, hazardous occupations. Research exceptionalism—assuming it is justified—is based in part on the idea mentioned earlier that research depends on public trust and obtaining a social license, and that public trust could be undermined in a context where those who take on themselves the risks do not receive any significant benefit (Wilson and Hunter 2010). However, public support for a CHI study would significantly weaken the case for research exceptionalism in this context, as it seems plausible to suppose that trust would not be undermined by practices enjoying social license.

In this respect, our study clearly suggests that a significant majority of the Singapore population (around 80%) would not only allow participants to enroll in challenge studies with varying levels of risk, but also that risk of serious side effects should be factored in when determining the level of payment for participants, alongside consideration of pain involved, time required, number of invasive and non-invasive investigations, and inconvenience. It would be difficult on this basis to suggest that payment for risks in challenge trials would undermine public trust, since it is precisely what the public would support. Interestingly, this is consistent with the finding by Grimwade and colleagues (2020) among the UK public, suggesting that objections to CHI studies on the grounds of excessive payments and undue inducement may be out of touch with public values. Moreover, among both Singaporean and UK respondents, the Market Model and the Risk Model of payment—the former based on supply/demand of participants and the latter on the level of risk involved—enjoy wide public support (with Grimwade et al.’s study showing slightly larger support for the Payment for Risk over the Market Model).

Consistently with the ethical analysis carried out in that study, it is worth pointing out that an ethically relevant consideration is the risk of exploitation. While some are concerned about too high payments as a possible threat to autonomy—a concern we have suggested is not well founded—there are risks with paying people too little, too. Too small payments might result in exploitation of their labour and good will to contribute to valuable research or of their decision to enroll in the study because of the need for money. Thus, one increased benefit of larger payments for participation in research trials is that it could prevent exploitation (Anomaly and Savulescu 2019), a point which seems further supported by this study’s findings on Singapore’s positive public attitudes towards payments for risks, and the observed correlation between expected reimbursement and income. The counter argument to this is commonly expressed in Singapore where “local” Singaporeans are not willing to do dangerous and risky jobs such as construction or undersea ship repair so they are outsourced to migrant workers who come from low-income countries and are prepared to take risks for salaries which would not attract Singaporeans. These workers are recognized by local institutional review boards as vulnerable populations and the same arguments would likely be applied to participants in CHI studies to ensure that they are protected under standard guidelines.

This survey has several limitations. A relatively large number of respondents completed the survey and across a wide range of ages and ethnicities, however, this sample was self-selected, and hence biased towards individuals with an interest in medical research, including a substantial proportion who had experience as participants. The survey was conducted solely in English and individuals who were not sufficiently fluent in this language were excluded from participation. Further studies are required to determine if the results of the survey are representative of the general population. Respondents may have interpreted some of the questions or their response selection differently to the survey writer’s intention, and as a self-directed questionnaire, participants did not have the opportunity to confirm their understanding. The order of questions may also have affected responses, though given the detailed set of questions respondents completed, we do not believe the magnitude of this effect would have been large. At the time of the survey in February 2023, the last COVID-19 pandemic public health measures in Singapore were lifted; movement and contact limitations had been removed in Singapore less than a year previously. While fear and anxiety related to COVID-19 remain elevated, the urgency for development of new vaccines and therapeutics has waned (Broockman et al. 2021). Attitudes towards CHI studies with SARS-CoV-2 are likely to have evolved over the course of the pandemic along with the judgment of individual risks from participation vs societal benefits (Holm 2020). Whether that will have changed the publics conception of their overall acceptability is not known.