Online health research faces a growing threat from imposter participants who provide deceptive data, with one cancer treatment survey finding 94 per cent of responses fraudulent.
Research published in the BMJ found that imposter participants fall into two categories: humans who provide deceptive responses such as lying about having the condition under investigation, and increasingly sophisticated automated computer software that mimics human behaviour and responses.
A 2025 scoping review found that 96 per cent of identified studies describing methods to detect imposter participants had been published within the past five years. The review reported that 18 of the 23 studies which looked for imposter participants in their datasets found them, with detected prevalence ranging from 3 per cent to 94 per cent.
An online survey investigating communication during ovarian cancer treatment received 576 responses within seven hours, with most submitted between midnight and 4 am. The authors judged 94 per cent of responses to be fraudulent and the remaining 6 per cent suspicious, with no participant deemed unquestionably legitimate. The survey was closed and relaunched with stricter protocols but continued to detect fraudulent responses.
The problem extends beyond survey research. In the iDEAS randomised controlled trial evaluating an alcohol reduction app, 76 per cent of online enrolments were identified as bots at screening. A further 4 per cent of participants were identified as deceptive human respondents.
The motivations of imposter participants remain unknown, although a focus on financial incentives suggests monetary benefit is a driver. However, not all studies that identified imposter participants offered financial incentives, indicating that other motives such as boredom, curiosity, or ideological intent to disrupt research may contribute.
Detection strategies include checking for implausible home addresses or submissions from multiple formulaic email addresses. Prevention strategies include identity verification procedures or CAPTCHA tests, though their efficacy in preventing or identifying imposter participants is largely untested.
The researchers stated that imposter participants are more than a nuisance and represent a systemic threat to health research. Their effect is demonstrable and their detection inconsistent, risking the undermining of health research integrity and the decisions built on it.
