Noninferiority Randomized Clinical Trial Comparing Three Delivery Formats of a Rehabilitation Intervention to Reduce Fatigue Among Individuals With Multiple Sclerosis

Abstract

Objective: To compare 3 delivery formats of a rehabilitation intervention in reducing fatigue: (1) synchronous group sessions via audio-only teleconference, (2) asynchronous group sessions delivered online, and (3) synchronous 1-to-1 sessions delivered in-person or videoconference. The primary hypothesis was that group-based teleconference and internet formats would be noninferior to the 1-to-1 format.

Design: Decentralized, randomized, noninferiority trial with assessments at baseline, 2, 3, and 6 months.

Setting: Community-based adults in Midwestern United States.

Participants: A total of 590 adults with multiple sclerosis.

Interventions: An occupational therapist-led intervention aimed at supporting self-management of fatigue.

Main outcome measures: Generalized estimating equations tested the primary hypothesis using a 10-point noninferiority margin on the Fatigue Impact Scale (FIS). Moderation and mediation analyses explored the effects of demographic, health, and psychosocial factors.

Results: Generalized estimating equation analysis showed that the teleconference format had higher (worse) estimated FIS scores compared with the 1-to-1 format, with values of 5.96 (1.80-10.12) at 2 months, 8.73 (4.52-12.94) at 3 months, and 4.89 (0.67-9.12) at 6 months. The internet format also had higher FIS scores, with estimates of 9.94 (5.68-14.20) at 2 months, 9.09 (4.80-13.38) at 3 months, and 6.12 (1.81-10.44) at 6 months. Moderation analysis indicated that participants from racial and ethnic minority groups, and individuals experiencing more anxiety, sleep disturbance, or comorbidities derived better improvements in FIS from the 1-to-1 format. College graduates (4-year degree) typically did worse in the internet format compared to the 1-to-1 format. The number of logins significantly mediated fatigue improvements in the internet format, whereas outcome expectations significantly mediated the teleconference format.

Conclusions: The upper bound confidence interval for teleconference and internet formats exceeded the noninferiority margin at most timepoints, indicating the hypothesis was neither supported nor rejected. Although teleconference and internet formats may offer broader accessibility, potential effectiveness trade-offs should be considered, especially for individuals with complex health profiles needing intense, personalized care.

Keywords: Behavior; Nervous system diseases; Occupational therapy; Rehabilitation; Self-management.