Use of an Unguided, Web-Based Distress Self-Management Program After Breast Cancer Diagnosis: Sub-Analysis of CaringGuidance Pilot Study

Background: Unguided, web-based psychoeducational interventions are gaining interest as a way to reach patients while reducing pressure on clinical resources. However, there has been little research on how patients with cancer use these interventions. Objective: The objective of this analysis was to evaluate how women newly diagnosed with breast cancer used the unguided web-based, psychoeducational distress self-management program CaringGuidance After Breast Cancer Diagnosis while enrolled in a pilot feasibility study. Methods: Women with stage 0 to II breast cancer diagnosed within the prior three months were recruited from clinics primarily in the Northeastern United States for participation in a 12-week pilot study of CaringGuidance plus usual care versus usual care alone. Usage prompts included sets of emails sent weekly for 12 weeks; standardized congratulatory emails after every two hours of program use, and informative emails for each cognitive-behavioral exercise. Individual user activity on the site was automatically tracked by an analytics system and recorded directly in the CaringGuidance database. Results: Complete usage data were available for 54 subjects. Ninety-eight percent of the intervention group logged into CaringGuidance independently at least once. Thirty-eight (70%) logged in during all three months, 15 (28%) were intermittent users, and one (2%) was a non-user. Users (n=53) averaged 15.6 (SD 9.85) logins. Mean logins were greatest in month 1 (7.26, SD 4.02) and declined in months 2 (4.32, SD 3.66) and month 3 (4.02, SD 3.82). Eleven (21%) used CaringGuidance with both the frequency and activity level intended at study outset, 9 (17%) exceeded intended frequency and activity (high-high users), and 10 (19%) were below expected usage on both login frequency and activity (low-low users). Low-low users and high-high users differed significantly (P<.001) in the total number of views and unique views of all program components. Change in depressive symptoms and the number of sessions (r=.351) and logins (r=.348) between study months 1 and 2 were significantly correlated (P=.018, .019). Higher baseline distress was associated with more unique views of program resources (r=.281, P=.043). Change in intrusive/avoidant thoughts from baseline to month 3, and the number of users' unique exercise views were negatively correlated (r=-.319, P=.035) so that more unique exercise views, equated with greater decline in intrusive/avoidant thoughts from baseline to month 3. Conclusions: These findings favor the hypothesis that the key ingredient is not the amount of program use, but each user's self-selected activity within the program. More research is needed on the ideal ways to maintain use, and capture and define engagement and enactment of behaviors by people with cancer accessing unguided, self-management web-based programs.