Dr Julia Bailey e-Health Unit University College London * image from lac-group.com Research using digital media
Overview Examples of digital research methods (for sexual health) –Online surveys –Online randomised controlled trials –Apps for data collection –Real-time behaviour monitoring Pros and cons How useful are these methods for health research? (your views)
What are digital research methods? Electronic data, saved locally or via Internet or mobile phone
Online cross-sectional survey – Sexual health of ethnic minority men who have sex with men in Britain Recruitment –Adverts on websites including Gaydar (dating site for gay men), community venues, via informal networks and sexual health clinics Enrolment, consent and questions all online –Online questionnaire (1,200 ethnic minority MSM, 13,700 white British MSM) –Online qualitative interview using (n=67) Risk for HIV, stigma and discrimination Elford, J.Elford, J. McKeown, E. Doerner, R. Nelson, S.Low, N. Anderson J. (2010). Sexual health of ethnic minority MSM in Britain (MESH project): design and methods. BMC Public HealthMcKeown, E.
Online cross-sectional surveys Advantages –Large numbers, quickly –Can reach hard-to-reach groups (MSM, sex workers) –Can reach unusual populations (e.g. rare diseases) –Automated eligibility checks, consent –Skip patterns – only relevant questions –Sensitive questions –Automated data collection –Compulsory answers - no missing data Elford
Online cross-sectional surveys Concerns –Generalisability of non-randomly sample –Digital divide –Attention spans online (do user testing, make questions relevant, use skip patterns) –Potential meaningless answers (do consistency checks, allow n/a options, add qualitative comment boxes) –Fraud (e.g. repeat registration for an incentive) (ask for several identifiers (e.g. phone, , postal address) –Ensure data protection and data security (encryption, secure servers…) Elford
Cross-sectional surveys - Computer–assisted interviews National Survey of Sexual Attitudes and Lifestyles (NATSAL) –Recruitment Men and women aged 16–74 years, across Britain (n=15,000) Random selection by postcode; household; and of one person in each household Visit by researcher in person –Data collected via computer By researcher (computer-assisted personal interview, CAPI) Self-completed (computer-assisted self-interview, CASI) –Urine and saliva samples Erens et al. Erens et al. Methodology of the third British National Survey of Sexual Attitudes and Lifestyles (Natsal-3). Sex Transm Infect Mar;90(2):84-9. Sex Transm Infect.
Online cross-sectional surveys - computer-assisted Advantages –Face-to-face initial rapport and clarification –Allows sensitive questions – data accuracy is better than face-to-face –Privacy (from researcher and others present at home) –Technical help possible (Other pros and cons as before)
Online randomised controlled trial ( Sexunzipped) Recruitment – Facebook advert (2006 young people) £10 incentive offered (voucher by post) Online eligibility (16-20, living in the UK) Online consent and registration Online data collection (baseline) Automated randomisation Automated follow up at 3 months (by ) Additional follow up by post (Chlamydia samples) Bailey JV et al. The Sexunzipped trial: optimizing the design of online randomized controlled trials. J Med Internet Res. 2013;15(12):e278
What are the advantages of online RCTs? Large numbers, quickly Young people were already online (Facebook) Automated eligibility checks, consent Automated randomisation (blinding of researcher and participant) Sensitive questions are possible Automated data collection Automated reminders
What are the challenges of online RCTs? –Sample representativeness –Possible fraud (for incentives) (collect several participant identifiers, date of birth at two time points, incentives in post) –Possible nonsense data (do consistency checks) –Drop-out at follow-up (incentives, multiple methods of follow up, multiple prompts) –‘Gold standard’ (biological) measurements are hard
Data collection via mobile phone app Smoking cessation app – Smoke Free 28 advice on stop-smoking medication inspirational stories and videos a distraction game advice on reducing exposure to smoking cues Target quit date App use was monitored (logins) Self-reported smoking status Kaur et al. A Mobile App to Aid Smoking Cessation: Preliminary Evaluation of SmokeFree28. J Med Internet Res 2015;17(1):e17
Data collection via mobile phone app Advantages and concerns Data can be logged on-the-go Notifications can enhance response Privacy may be a concern (phone sharing is very common)
Real-time data collection Intervention usage (e.g. web pages visited, app use, targets set) Movement and physiological monitoring with sensors Geolocator e.g. visits to McDonald’s
Real-time data collection Advantages –Data collection can be continuous and automatic Concerns –Mechanisms for adequate informed consent –Measurable conditions are receiving more attention than less measurable (e.g. loads of exercise trackers) –Data needs interpreting! E.g. is high use of an app a good thing, or bad?
Qualitative research online Examples –Comment boxes within (quant) surveys –Interview questions via –Instant chat interviews or focus groups –Analysis of online data such as blogs, bulletin boards etc. –Technology-mediated interviews (e.g. Skype video calls) –User feedback and discussion on social networking sites (e.g. Facebook)
Qualitative research online Advantages –Cost –Convenience –Geographic reach Disadvantages –Typing is time-consuming –Literacy –Loss of non-verbal cues –Difficult to cover issues in depth
Questions What is your experience of digital research methods? Would digital research methods be useful for you or other researchers?
Summary - advantages of digital research methods –Potentially lower cost –Standardised procedures –Sampling some hard-to-reach populations –Convenient for participants Privacy, mobile access –Convenient for researchers Automated recruitment, data collection, follow-up
Summary - challenges of digital research methods –Sampling – representativeness; IT access and literacy –Possible fraud (for incentives) –Data protection and data security –Attrition at follow-up –‘Gold standard’ measurements may be hard (e.g. biological samples)
Recommendations –User views and user testing is essential –Gain participants’ trust – e.g. NHS or university branding –Clear information, in bite sized bits –Combine approaches (e.g. face-to-face recruitment, remote support for engagement)
Recommendations –Technical testing is essential Customisable trial and survey software which meets design, access and security standards is needed –Work with specialists Data protection and security Technical specialist on team Designers (graphic design, software design) –Assess risks Informed consent from participants Guidance for ethical committees on how to assess online research (needed)
* image from lac-group.com Thank you