Objective: Non-compliance in clinical studies is a significant issue, but causes remain unclear.Utilizing the Elaboration Likelihood Model of persuasion, this study assessed the psychophysical peripheral cue 'Interactive Voice Response System (IVRS) call frequency' on compliance.Methods: 71 participants were randomized to once daily (OD), twice daily (BID) or three times daily (TID) call schedules over two weeks. Participants completed 30-item cognitive function tests at each call.Compliance was defined as proportion of expected calls within a narrow window (±30 min around scheduled time), and within a relaxed window (-30 min to +4 hour). Data were analyzed by ANOVA and pairwise comparisons adjusted by the Bonferroni correction.Results: There was a relationship between call frequency and compliance. Bonferroni adjusted pairwise comparisons showed significantly higher compliance (p=0.03) for the BID (51.0%) than TID (30.3%) for the narrow window; for the extended window, compliance was higher (p=0.04) with OD (59.5%), than TID (38.4%).Conclusion: The IVRS psychophysical peripheral cue call frequency supported the ELM as a route to persuasion. The results also support OD strategy for optimal compliance. Models suggest specific indicators to enhance compliance with medication dosing and electronic patient diaries to improve health outcomes and data integrity respectively.Response to Reviewers: Sincerest thanks to the reviewers for providing feedback. Each comment has been addressed either in the manuscript or is shown below. 2. 'Participants (N=71), were randomized using an IVRS in the ratio 1:1:1 to a OD, BID or TID call schedule'.How did authors divide them according to 1:1:1? Explanatory text has been added to page 8. Please note that randomization was not stratified.3. '44 (62%) completed the study (Fig.1). Participants who prematurely discontinued were not followed up, so reasons for withdrawal are unknown'. Just 44 of 71 participants completed study, the quality of it should be susceptible.The authors recognize that the completion rate is lower than perhaps might be considered ideal. However it should be noted that 78% of participants randomized to once-daily calls completed the study compared to 43% of participants randomized to three-times daily calls. This difference in completion rates supports our findings for the primary endpoint, given that the once-daily call schedule was associated with the highest rate of compliance (with the relaxed window). Additional text has been added to the discussion (page 16) to raise this point.4. The discussion should be concise; the conclusion should be cautious.The discussion and conclusion have been shortened, to ensure that only the key points from the study are discussed, and to remove some of the aspects of the conclusions that generalised the findings of the study in a clinical study context. See pages 15-20.Reviewer 2
Define "electronic patient diaries "A definition has been added to page 5.
Define "Interactive Voice Response systems "A definition has ...