When research participation pays, some people lie, Penn study suggests
Offering compensation can be an important tactic to attract potential participants for enrollment in research studies, but it might come at a cost. A new study conducted by researchers in the Perelman School of Medicine at the University of Pennsylvania found that up to 23 percent of respondents lied about their eligibility to participate in a survey when offered payment, even small amounts.
Anecdotal evidence and common sense suggest that offering money may encourage participants to lie about their eligibility or other aspects of study participation in order to secure payment, the authors said. But few studies have investigated whether and to what extent people will deceive, leaving a major gap in the literature. The new findings, published in JAMA Network Open, suggest the practice may be pervasive.
A total of 2,275 respondents participated in a nationally representative, randomized survey on flu vaccination status. One study group lacked motivation to lie about whether they had recently had a flu shot because their eligibility didn't depend on it. Their reported rates of flu vaccination were therefore used to determine the true rate of vaccination in the study population. Other groups were offered $5, $10, or $20 for participation and were told they were eligible only if they had (or in some groups, had not) received a recent flu shot. If no one was lying, all study groups would have reported about the same rates of flu vaccination.
"Instead, we found evidence of significant deception by participants who were not eligible, but claimed they were in order to be able to join the study," said first author Holly Fernandez Lynch, JD, MBE, an assistant professor of Medical Ethics and Health Policy. "This type of behavior not only undermines a study's integrity and its results, but in a study with eligibility criteria that are intended to protect participants, it also has the potential to put participants at risk."
In the control group, 52.2 percent of respondents reported a recent flu shot. But when told that eligibility depended on recent flu vaccination and financial compensation of $5, $10, or $20 was offered, reports of vaccination jumped to 63.1 percent, 62.8 percent, and 62.1 percent, respectively. When told that eligibility depended on having not received a recent flu vaccination, reports of vaccination dropped to 46.5 percent at $5, 41.8 percent at $10, and 46.7 percent at $20. Because the only differences between these groups were their eligibility criteria and payment amounts, the differences in their reported vaccination rates can be attributed to deception. The researchers calculated that between 10.5 and 22.8 percent of participants engaged in deception about their eligibility to participate.
The authors say one of the study's most interesting findings is that more money wasn't associated with higher rates of deception. "This suggests that keeping payments low will not necessarily prevent deception," said study co-author Steven Joffe, MD, MPH, chief of the division of Medical Ethics, "It also suggests that higher payments may encourage recruitment without posing a greater risk to the study's integrity."
The study's senior author Emily A. Largent, PhD, JD, RN, an assistant professor of Medical Ethics and Health Policy, emphasized the importance of finding ways to minimize deception by study participants. "Rather than relying on self-reporting by participants, investigators should use objective metrics whenever possible," Largent said. "The best response to our findings is not necessarily to reduce or eliminate payment offers for participation. Payment can help boost legitimate enrollment; in addition, investigators may owe payment as compensation for the time, effort, and burden that participants assume by joining the study."
Although this study was conducted in the context of survey research, its findings support the need for further studies in clinical trials, where payment amounts and other benefits, like access to investigational drugs, as well as study risks, are often more substantial.