Resiliency in U.S. Air Force Personnel

A cluster randomized, pragmatic-explorative two-arm multi-site study is proposed to examine the effectiveness of the Stress Management and Resilience Training (SMART) in increasing the levels of resiliency in active component personnel serving in the U.S. Air Force. The SMART training will be provided through either video-teleconference (VTC), in-person or computer-based training (CBT).The VTC/in-person trainings will be provided to a maximum of 10 individuals per session. Participants in the CBT group will be provided a code to access the training website. A cluster randomization of sites will take place first to place two sites into the self-selection arm, a pragmatic trial where study participants will choose intervention arm and place the other two sites into a randomization-arm in which participants will be randomized into either the VTC or CBT arms. SMART is focused on improving the practices of gratitude, mindful presence, kindness, and developing a resilient mindset.

Four study sites have been identified for the study (Joint Base Andrews, Joint Base San Antonio-Lackland, Nellis Air Force Base, and Wright-Patterson Air Force Base). Simple randomization will be used to assign 2 sites to the randomization arm (Arm 1) and 2 sites to the self-selection arm (Arm 2). Arm randomization will be accomplished in Year1, Month 1. Initial recruitment and randomization will be completed during Month 2 of Year 1 through Month 5 of year 2 of the study.

Arm 1: Randomization: At two study site locations, the investigators will use a ratio of 2:1 to randomly assign participants to the VTC/in-person or CBT training modalities. Therefore, the investigators anticipate 84 participants (42 per site) will be assigned to the CBT modality and 168 (84 per site) will be assigned to the VTC/in-person modality. Randomization will be stratified with respect to military grade and sex to reduce potential confounding due to baseline differences. There may be important differences based upon military grade between the 5 categories of military grade or two categories of participants may select. Under stratification, 1 in every 3 participants will be assigned to complete CBT and the other two participants will be assigned to VTC/in-person training to maintain a balanced distribution between groups.

Arm 2: Self-selection: At two study site locations, the investigators aim to recruit 250 participants in the self-selection arm. Participants will be able to self-select their preferred method of SMART completion (VTC/in-person or CBT training modality).

In order to assure consistent delivery of the intervention, the study PI and the Site AIs will meet to discuss course delivery, make notations, and jointly agree to incorporate recommendations for improving training delivery. At the end of the course, participants will be offered a copy of a Resilience Journal. This book will serve as a review of the resiliency practices and allow participants to track their use of these practices for improving gratitude, mindful presence, kindness, and developing a resilient mindset.

The Site PIs and PI will work with identified local points of contact to develop a schedule to assure the maximum number of participants can receive SMART during a variety of times. After informed consent is obtained and prior to the delivery of the intervention, a survey will be administered to collect demographic information and baseline measurements of resilience, stress, anxiety, and quality of life. Participants will be provided a link to access the initial survey via the research data capture survey system (REDCap; DHHS/NIH/NCRR #8UL1TR000041) to provide demographic data, e-mail contact information, and responses to survey instruments.

REDCap is a secure and encrypted, web-based platform licensed to and managed by the University of New Mexico Health Sciences Center Clinical and Translational Science Center (UNM HSC CTSC). REDCap includes research tools for project management, survey administration, encrypted database storage and retrieval, and reporting. REDCap will be utilized to administer follow-up surveys at week 12, 24, and 36 (if applicable) after participants complete SMART. The investigators propose to collect the 36-week follow-up survey responses in order to conduct an analysis to assess the longevity of the effect of SMART on resilience and stress.

Participants' e-mail contact information will be used be utilized to send follow-up surveys through REDCap. Each survey will take approximately 10 to 15 minutes for participants to complete. If the participant does not respond to the initial REDCap invitation, a maximum of three e-mail reminders (one reminder per week after the initial e-mail is sent) will be sent to complete the survey. If the participant still has not responded, no further attempt to contact the participant will be made until the next scheduled interval (e.g. at 24 and 36 weeks). Because participants' demographic information will be associated with an individual's contact information in REDCap™, demographic information will not be requested as part of the surveys at weeks 12, 24, or 36. All data will be de-identified when it is exported from REDCap for the data analysis.

IBM SPSS Statistics (version 28 or later) and R (survey package, version 4.1.2) will be used for the statistical analysis. Initial analysis will include descriptive statistics, including means or medians, frequencies and percentages, as appropriate, to characterize demographic status, military grade, duty location, military occupation, and previous deployment status. Cronbach's α will be calculated for each multi-item scale.

The objectives of our analysis include testing and estimating the effectiveness of the intervention by comparing pre-post intervention changes (improvements) in the outcome measurements of interest at each follow-up. Analyses will be reported as point estimates with 95% confidence intervals and appropriate estimates of effect size. In this analysis, both the VTC and CBT groups will be analyzed separately and scores will also be pooled together to test for overall intervention effects as well as difference between the modalities.

Post-intervention improvements in resilience, stress, anxiety, and quality of life will be assessed by separately analyzing changes from baseline to each follow-up time (e.g. week 12, week 24, and week 36). A joint analysis of the longitudinal trend over the study period will be conducted as well. The investigators will also consider a joint analysis of resilience, stress, anxiety, or QOL, as well as, analysis for subscales of any specific domain of interest. To better understand factors that can impact the intervention effects, our regression models on the post intervention improvements over time incorporate changing time-slopes and between-subject random effects. The investigators will control for demographic characteristics, AFSC, and previous deployment. The investigators will also use random effects to take into consideration potential clustering effects among the participants from the same organizational unit.

With the two-arm randomization scheme, the results of the "self-selection" group provide a window into "potential" outcomes in real-world settings.

Comparison of the outcomes from the "self-selection" arm with the results from the "randomized assignment" (per-protocol) will enable us to conduct causal inference and use the real difference between VTC and CBT to inform scale-up implementation of SMART.