Anguera JA, Jordan JT, Castaneda D, Gazzaley A, & Areán PA. (2016). Conducting a fully mobile and randomised clinical control trial for depression: Access, engagement and expense. BMJ Innovations, 2(1), 14-21. PMCID: PMC4789688.
This study assessed participant access and engagement, and cost of conducting a fully remote randomized control trial (RCT) on depressed adults using three different mobile apps for treating depression. Researchers recruited 2923 participants using traditional methods (i.e. written ads and Craigslist posts), social network posts, search engine based ads, and unanticipated means of recruitment (i.e. participant search, referrals). The majority of participants were recruited using traditional means and unanticipated means; social networking and search engine ads each accounted for less than one percent of recruitments. Researchers used an online research portal to screen participants for depression ( PHQ-9). Eligible participants owned a smart phone and an iPad (since one of the apps only operated on an iPad). Of participants recruited, 1098 passed screening and were randomized into one of three groups: a cognitive intervention video game, a problem-solving therapy app, and an information-only app targeting management of depressed mood (served as control). Over 12 weeks, participants completed daily assessments of mood and enjoyment. A separate app recorded passive analytics of participants’ phone usage and mobility. At four, 8-, and 12-weeks, participants completed self-reports of mood, communication, activity and cognitive control. Researchers do not discuss the results of the intervention study but rather focus on levels of participant access and engagement and cost. While researchers were able to recruit a large sample of ethnically diverse participants that included representation of difficult to access groups, attrition 12 weeks was high (59%). Participants who received a bonus for completing all the assessments stayed in the study longer. Low adherence rates and lack of discussion of the success of the interventions make it is difficult to assess whether results truly reflect feasibility of a fully remote RCT.