Does Patient Preference for Mode of Intervention Delivery Impact Intervention Efficacy and Attrition?

Purpose: To explore whether there was a difference in objectively measured physical activity and study participation between people who received their preferred study group allocation (matched) and those who did not receive their preferred study group (mismatched). Design: Secondary data from the NewCOACH randomized controlled trial. Setting: Insufficiently active patients in the primary care settings in Sydney and Newcastle, Australia. Participants: One hundred seventy-two adults aged 20 to 81 years. Intervention: Participants indicated their intervention preference at baseline for (1) five face-to-face visits with an exercise specialist, (2) one face-to-face visit and 4 telephone follow-ups with an exercise specialist, (3) written material, or (4) slight-to-no preference. Participants were then allocated to an intervention group and categorized as either “matched” or “mismatched” based on their indications. Participants who reported a slight-to-no preference was categorized as “matched.” Measures: Daily step count as measured by pedometers and study participation. Analysis: Mean differences between groups in daily step count at 3 and 12 months (multiple linear regression models) and study participation at baseline, 3 months, and 12 months (χ2 tests). Results: Preference for an intervention group prior to randomization did not significantly (all P’s > .05 using 95% confidence interval) impact step counts (differences of <600 steps/day between groups) or study participation. Conclusion: Future research should continue to address whether the strength of preferences influence study outcome and participation and whether the study preferences change over time.