User-Centered Refinement of a Digital Tool for Tuberculosis Treatment Support: Iterative Mixed Methods Study

Background Despite the potential of digital adherence technologies to support patient-centered monitoring for tuberculosis (TB), there is limited research on incorporating indirect and direct adherence monitoring or assessing patients' experiences with these technologies. The TB Treatment Support Tools (TB-TST) includes a comprehensive mobile app for patients and health care providers and a direct adherence metabolite test to report and monitor adherence. Objective This paper describes the iterative refinement process of the TB-TST intervention. Methods To refine the TB-TST intervention, we used an iterative approach involving multiple embedded mixed methods studies guided by the Information Systems Research framework and Design Thinking Process. Embedded studies included a randomized controlled pilot study, interviews, usability testing, and surveys with patients and experts to inform ongoing refinements. The project consisted of interface evaluation, high-level system design, and iterative redesign. Results The TB-TST intervention was refined through 3 iterative phases. In Phase 1, based on feedback from pilot study participants and 4 experts in TB, improvements included an in-app discussion board, submission confirmations, and enhanced account recovery. Cultural adaptation was based on Hofstede's dimensions. Phase 2 involved 4 Directed Research Groups and 19 stakeholders to redesign user flows, simplify reporting, and transition the app to a progressive web app, improving device compatibility. Phase 3 included usability testing cycles with 48 participants (26 patients and 22 health care professionals), yielding high satisfaction scores: patient app Mobile Health App Usability Questionnaire, mean 5.96 (SD 0.46); provider mobile dashboard IT Usability Evaluation Scale scores ranged from 5.83 to 6.23 out of 7, and optimization of interface and dashboard. Refinements included larger icons, streamlined onboarding, symptom summary enhancements, and a new cohort-level adherence graph. These modifications improved navigation, usability, and remote monitoring for patients with TB and providers in preparation for a multisite clinical trial. Conclusions Combining multiple methods guided by the Information Systems Research framework and elements of the Design Thinking Process can help researchers and developers leverage the strengths of mixed methods iterative designs to create highly personalized and effective digital health interventions.