October 4, 2024 Brainstorm Overview
# Micro-Randomized Trial (MRT) for Upper Limb Stroke Recovery Intervention
Collaboration with Prabhat Pathak, James Arnold, and Pat Puma of the [Biodesign Lab](https://biodesign.seas.harvard.edu/lab-members); [Lynn Blaney](https://ltnr.mgh.harvard.edu/team/lynn-blaney-msot-otr-l/) is the consulting occupational therapist on our team

## Goals for the Brainstorm
Our primary goals for this brainstorm are to evaluate and improve our proposed micro-randomized trial (MRT). Our proposed MRT will investigate mobile health (mHealth) intervention options that encourage affected arm usage during everyday life for [chronic stroke survivors](https://www.neurolutions.com/after-stroke/a-guide-to-chronic-stroke-recovery-and-symptoms/). Our secondary goal is to improve the way we communicate the work for future presentations, papers, and proposals.

## Questions to Think About
- Do the proposed study details align with the research questions?
- How can we improve the proposed study?
- Is there any background information you feel would have better motivated the work?
- Are there any study details you feel would have been more important to discuss?

## Overview
### Background
[Every 40 seconds, someone suffers a stroke in the United States](https://www.cdc.gov/stroke/data-research/facts-stats/index.html). Upper limb rehabilitation following stroke requires lifelong, consistent effort. Chronic stroke survivors must invest effort into the use of their affected arm for activities of daily living (ADLs) (e.g., opening a door). This investment is critical to the long-term recovery of their affected arm, though affected arm usage is less convenient in the short-term than unaffected arm usage.

A key challenge for upper limb stroke recovery is that it is difficult to maintain motivation and remember to use one’s affected arm for ADLs during everyday life. Even for those that have access to occupational therapist support, visits to the therapist are typically once or twice a week at the maximum. Thus, the majority of survivors' lives are spent without clinical support.

MHealth interventions are a promising avenue to fill this gap in daily life support. [Work in the related physical activity space](https://doi.org/10.1093/abm/kay067) has shown that mHealth interventions can be effective in promoting physical activity for sedentary adults. In an analogous way, push notifications and other mHealth interventions could be used to promote affected arm usage for chronic stroke survivors.

Recent mHealth research for stroke survivors has shown that quantifying arm usage via wrist-worn accelerometers and providing feedback accordingly can be [encouraging for the recovery process](https://doi.org/10.1080/09593985.2018.1493759). The primary limitation of this previous approach is the dependence on an occupational therapist for feedback. Other related work includes an investigation of activity-dependent prompting of arm usage via wrist-worn accelerometers. This work differed in its goal of [evaluating an intervention in a pilot randomized controlled trial](https://doi.org/10.1177/0269215519834720). No studies to date have evaluated the effects of candidate mHealth intervention options to optimize a future just-in-time adaptive intervention.

### Research Questions
1. How do the proposed mHealth intervention options impact affected arm use in individuals post-stroke?
2. How much more or less effective are the proposed intervention options in increasing affected arm use when tailored to a stroke survivor’s data?
### High-Level Details of Proposed MRT
- 30 chronic stroke survivors who are mild to high moderate functioning
- 1-week baseline observation with no interventions followed by 5-week MRT
- In-lab sessions at the beginning, end, ~~and after baseline week~~ for functional assessments; Zoom call to finalize MRT decision times and the like after baseline week
- Three intervention components (expected 3 total interventions/day):
- ~~Goal setting vs. no goal setting each morning with 0.5 probability~~
- ~~Arm Usage feedback vs. no feedback each evening with 0.5 probability~~
- Reminders vs. no reminder before breakfast, lunch, dinner, and an additional user-specified time for exercise (total 4 times/day) with 0.5 probability
- Apple Watch with LTE as wrist-worn wearable and intervention delivery system
- Two proximal outcomes:
- ~~Affected arm usage for next 24 hours for goal setting and feedback interventions~~
- Affected arm usage for next 1 hour for reminder interventions
- Distal outcomes include the Reach Scale, FMA, stroke impact scales, etc.
- Tailoring variables being considered include goals at baseline, previous goal setting response, prior week's affected arm usage, and prior day's affected arm usage (variables listed in order of descending priority)
- Logistically, this study will proceed in phases:
- Phase 0: Asim, Prabhat, James, and Pat experiment for 6 weeks
- Phase 1: 5 stroke survivors experiment for 2 weeks
- Phase 2: 10 stroke survivors experiment for 4 weeks
- Phase 3: Full study as outlined above