Variability of the Behavioral Assessment Screening Tool for Mobile Health (BASTmHealth) for ecological momentary assessment in chronic traumatic brain injury

Research Objectives: To describe the variability of behavioral symptoms in individuals with chronic traumatic brain injury using subscale scores from the Behavioral Assessment Screening Tool for Mobile Health (BASTmHealth).

Design: Prospective, repeated measures design over 14 days

Setting: Community

Participants: Forty-four individuals with chronic traumatic brain injury (TBI)

Interventions: We randomized individuals to receive one of three ecological momentary assessment sampling frequencies: a) once daily, random schedule; b) once daily, time of choice; c) every other day at a time of choice. Participants received a link via text to report neurobehavioral symptoms in REDCap.

Main Outcome Measures: We collected responses from the BASTmHealth, which includes 10 items and five subscales (Impulsivity, Negative Affect, Fatigue, Substance Use, Executive Dysfunction).

Results: An examination of Root Mean Square of Successive Differences (RMSSD) showed that the fatigue scale had most variability from one day to the next, while the substance use had the least day-to-day variability. RMSSDs were higher in the group who reported symptoms every other day. The intraclass correlation coefficients (ICC) revealed that the substance use scale had the least within-person variability over the measurement period. The average ICC was greatest for the once daily, random schedule group. We found that variation in executive dysfunction was significantly correlated with total lifetime TBIs with loss of consciousness (r=.33, p=.03), variation in fatigue was significantly correlated with total lifetime TBIs (r=- .31, p=.04), and variation in substance use was significantly correlated with TBI before the age of 15 (r=.40, p=.01).

Conclusions: We were able to use EMA to capture symptom variability within the chronic TBI population. Based on our results, every other day sampling frequency provided the most information about behavioral symptom variability, while the once daily sampling using a random schedule provided the most information about within-person variation. Identifying the best sampling frequency to capture within-person symptom variability could improve symptom assessment and monitoring in both clinical research trials and practice.

Author(s) Disclosures: All authors have no disclosures.