Personalized Trials for Stress Management Against Standard of Care

Stress will be assessed via 3 daily text message prompts sent at random times throughout the day asking individuals to rate their stress on a scale of 0 (low) to 10 (high). Levels of EMA stress will be aggregated within the baseline assessment period and during the follow-up period to generate an overall mean and standard deviation for each period. The change in EMA stress between baseline and follow-up will be calculated by subtracting these aggregated means. The difference in change in EMA stress between baseline and follow-up periods will be compared between the personalized arms (Arm 1 and 2) and the standard of care arm (Arm 3) using a two-sample t-test and Generalized Linear Mixed Model analyses.

Weekly perceived stress will be assessed with the 10-item Perceived Stress Scale (PSS-10), modified to be delivered to assess the prior week rather than the prior month. The minimum total score possible is 0 and the maximum total score possible is 40. Higher values represent higher level of stress. Levels of weekly perceived stress will be aggregated within the baseline assessment period and during the follow-up period to generate an overall mean and standard deviation for each period. The change in weekly perceived stress between baseline and follow-up will be examined using Generalized Linear Mixed Model analyses.

At the end of the intervention in the personalized trial arms (Arm 1 & Arm 2), personalized trial data will be used to identify which intervention (mindfulness, yoga, or walking) was most effective at reducing stress. This recommendation will be presented in a personalized report sent after completion of the intervention. The number of participants in the personalized trial arms who select this recommended intervention during follow-up relative to the total number of participants in Arms 1 & 2 will be presented as a count of participants/percentage, with a higher count of participants/percentage indicating greater levels of agreement.

The SUS is a validated 10-item questionnaire that asks users to score each item on a Likert scale from Strongly Disagree (rating of 1) to Strongly Agree (rating of 5). The SUS has a minimum score of 0 and a maximum score of 40. The SUS will be presented to the participant as addressing the ease of use, complexity, consistency of the Personalized Trials system as a whole, from recruitment to receipt of the report.

Odd items are scored by subtracting one from the user response. Even-numbered items are scored by subtracting the user responses from 5. This scales all values from 0 to 4 (with four being the most positive response). Converted responses are summed and multiplied by 2.5. This converts the range of possible values to a composite measure from 0 to 100, reported below. Participant SUS scores will be averaged together and an overall mean will be reported with standard deviation. Higher scored values correlate to a more usable system, and therefore a better outcome.

Participants will rate their satisfaction with the Trial overall and with individual elements of the trial in a satisfaction survey. Satisfaction items are not part of an already existing scale but were developed to assess participant satisfaction with specific elements of the current study and the trial overall. Participants will rate their satisfaction on a scale of 1 to 5, with higher numbers indicating greater levels of satisfaction. Means and standard deviations will be reported for each element of satisfaction.

Stress will be assessed via 3 daily text message prompts sent at random times throughout the day asking individuals to rate their stress on a scale of 0 (low) to 10 (high). Levels of EMA stress will be aggregated within the baseline assessment period and in each of the treatment periods (yoga, meditation, guided walking), and then averaged across each treatment period, to generate an overall mean and standard deviation for each period. Changes in EMA stress from baseline to each treatment period will be evaluated using Generalized Linear Mixed Model analyses.

Fatigue will be assessed via 3 daily text message prompts sent at random times throughout the day asking individuals to rate their stress on a scale of 0 (low) to 10 (high). Levels of EMA fatigue will be aggregated within the baseline assessment period and in each of the treatment periods (yoga, meditation, guided walking), and then averaged across each treatment period, to generate an overall mean and standard deviation for each period. Changes in EMA fatigue from baseline to each treatment period will be evaluated using Generalized Linear Mixed Model analyses.

Pain will be assessed via 3 daily text message prompts sent at random times throughout the day asking individuals to rate their stress on a scale of 0 (low) to 10 (high). Levels of EMA pain will be aggregated within the baseline assessment period and in each of the treatment periods (yoga, meditation, guided walking), and then averaged across each treatment period, to generate an overall mean and standard deviation for each period. Changes in EMA pain from baseline to each treatment period will be evaluated using Generalized Linear Mixed Model analyses.

Mood will be assessed via 3 daily text message prompts sent at random times throughout the day asking individuals to rate their stress on a scale of 0 (low) to 10 (high). Levels of EMA mood will be aggregated within the baseline assessment period and in each of the treatment periods (yoga, meditation, guided walking), and then averaged across each treatment period, to generate an overall mean and standard deviation for each period. Changes in EMA mood from baseline to each treatment period will be evaluated using Generalized Linear Mixed Model analyses.

Concentration will be assessed via 3 daily text message prompts sent at random times throughout the day asking individuals to rate their stress on a scale of 0 (low) to 10 (high). Levels of EMA Concentration will be aggregated within the baseline assessment period and in each of the treatment periods (yoga, meditation, guided walking), and then averaged across each treatment period, to generate an overall mean and standard deviation for each period. Changes in EMA Concentration from baseline to each treatment period will be evaluated using Generalized Linear Mixed Model analyses.

Confidence will be assessed via 3 daily text message prompts sent at random times throughout the day asking individuals to rate their stress on a scale of 0 (low) to 10 (high). Levels of EMA Confidence will be aggregated within the baseline assessment period and in each of the treatment periods (yoga, meditation, guided walking), and then averaged across each treatment period, to generate an overall mean and standard deviation for each period. Changes in EMA Confidence from baseline to each treatment period will be evaluated using Generalized Linear Mixed Model analyses.

Daily step counts will be assessed by a Fitbit device. Daily step counts will be aggregated within the baseline assessment period and in each of the treatment periods (yoga, meditation, guided walking) to generate an overall mean and standard deviation for each period. Mean daily steps will be compared to baseline using three paired-samples t-tests (yoga vs baseline, meditation vs baseline, guided walking vs baseline). Daily steps values over the course of the study will also be evaluated using Generalized Linear Mixed Model analyses.

Nightly sleep duration will be assessed by a Fitbit device. Sleep duration will be aggregated within the baseline assessment period and in each of the treatment periods (yoga, meditation, guided walking) to generate an overall mean and standard deviation for each period. Mean sleep duration will be compared to baseline using three paired-samples t-tests (yoga vs baseline, meditation vs baseline, guided walking vs baseline). Sleep duration values over the course of the study will also be evaluated using Generalized Linear Mixed Model analyses.

For each participant of the proportion of surveys measures completed (both daily and weekly surveys) will be calculated. Completion rates across all participants will be reported with means and standard deviations.

For each participant of the proportion of EMA measures completed will be calculated. Completion rates across all participants will be reported with means and standard deviations.

For each participant, the proportion of days where the Fitbit device was worn will be calculated. Completion rates across all participants will be reported with means and standard deviations.

For each participant, the average days of participant-observed sleep data, defined as recorded sleep and wake cycles, will be calculated across all participants with means and standard deviations.

For each participant, the intervention adherence rate, defined as unique video views of the appropriate recorded intervention during the assigned interventions, will be calculated and reported across all applicable individuals with means and standard deviations.

For each participant, the intervention adherence rate, defined as unique video views of the appropriate recorded intervention during the assigned interventions, will be calculated and reported across all applicable individuals with means and standard deviations.

For each participant, the follow-up period adherence rate, defined as unique video views of the appropriate recorded intervention during the assigned follow-up period, will be calculated and reported across all individuals with means and standard deviations.

After completion of the trial, a random sample of 10% of participants from each arm will be asked to participate in 60-minute qualitative interviews to discuss their experiences. Descriptive content from these recorded and transcribed interviews will be analyzed and reported.