UCSF

Background

Internet-based surveys are increasingly used for health research because they offer several advantages including greater geographic reach, increased participant anonymity, and reduced financial/time burden. However, there is also a need to address inherent challenges, such as the likelihood of fraudulent responses and greater difficulty in determining eligibility.

Methods

We conducted an online nationwide survey of 18-29 year olds living with HIV in the United States, to assess willingness to participate in HIV cure research. To ensure that respondents met age and HIV serostatus inclusion criteria, we instituted screening procedures to identify ineligible respondents using tools that were built into the survey platform (eg, reCAPTCHA, geolocation) and required documentation of age and serostatus before providing access to the incentivized study survey.

Results

Of 1308 eligibility surveys, 569 were incomplete or ineligible because of reported age or serostatus. Of the remaining 739 potentially eligible respondents, we determined that 413 were from fraudulent, bot, or ineligible respondents. We sent individual study survey links to 326 participants (25% of all eligibility survey respondents) whose eligibility was reviewed and confirmed by our study team.

Conclusion

Our multicomponent strategy was effective for identifying ineligible and fraudulent responses to our eligibility survey, allowing us to send the study survey link only to those whose eligibility we were able to confirm. Our findings suggest that proactive fraud prevention can be built into the screening phase of the study to prevent wasted resources related to data cleaning and unretrievable study incentives and ultimately improve the quality of data.