BT22 Online dermatological surveys: are your participants who you think they are?
Natalie King Stokes,Sarah McCusker,Aoife Daly,Manrup Hunjan,Ashima Lowe,Carolyn Charman,Susannah George,Lea Solman,Esther Burden-Teh
DOI: https://doi.org/10.1093/bjd/ljae090.419
IF: 11.113
2024-06-28
British Journal of Dermatology
Abstract:Abstract Surveys are a valuable research method in dermatology. They have a multitude of applications, particularly in understanding the impacts of skin disease, treatments and services on patients. Increasingly researchers are utilizing online surveys and using social media to advertise. We describe our experience conducting an online survey to explore the experiences of parents and adolescents having teleconsultations for eczema in the UK. The survey was distributed through the Jisc online survey platform and advertised via social media platforms including Facebook, Instagram and Twitter in November 2022. As with many patient surveys, there was a prize draw incentive to win a £50 shopping voucher. Within 24 h, 138 responses were received, a higher response rate than expected. On review of the data, only six responses appeared genuine, with the rest fraudulent or suspicious for being fraudulent (95.7%). Red flags for fraudulent responses included the following: 94 responses had supplied an email address containing a name that did not match the participant’s name, while 30 responses gave an email address consisting of a random combination of letters and numbers. Clusters of survey responses appeared to have been started and completed at the same time, with duplicate answers in free-text boxes. In total, 105 responses contained a duplicate free-text response. Fraudulent responses were inconsistent and illogical; for example, 41 respondents selected that they only had experience of face-to-face appointments, despite having previously selected that they did have experience of a teleconsultation in the survey eligibility screening question. Several aspects of the overall results did not fit with expected demographics; for example, a high proportion of patients had had private appointments (42.8%), a high proportion of children were on systemic treatment (29%) and only 20.3% had been prescribed emollients. Immediate action was taken to close the survey. On review of the literature, fraudulent participation in online research is an increasing problem, which threatens the integrity of data, in addition to wasting time and funding. We discuss measures that researchers can take to prevent fraudulent responses, such as use of CAPTCHA to stop bots and not offering monetary incentives. Researchers should include measures to identify and deal with fraudulent responses in the study protocol, for example including an instruction question (such as select answer ‘b’), free-text questions and repeated questions that can be checked for consistency and logic. We hope that our experience will raise awareness among researchers conducting online surveys in dermatology, to be able to prevent and identify fraudulent data, to protect data validity.
dermatology