Page 66 - JSOM Spring 2026

Convergent Validity and Test-Retest Reliability
of the Blast Exposure Threshold Survey Among
Army Special Operations Forces

1
Eric Shumski, PhD *; Deanna Pinnow, PhD ; Christy Mote, MA ;
3
2
Steven Wilcox, BA ; Dale Welsh, PA-C ; Jamie N. Hershaw, PhD 6
4
5

ABSTRACT
Introduction: The Blast Exposure Threshold Survey (BETS) Introduction
is a self-report questionnaire assessing lifetime generalized
blast exposure value (GBEV). This study examined GBEV The impact of blast exposure on brain function in active-duty
convergent validity and GBEV and BETS item-by-item test- Servicemembers is still not fully understood due to the com-
retest reliability. Methods: In this prospective-cohort study, plex nature of the injury and the compounding effects of blast
Army Special Operations Forces (ARSOF) trainers and train- over time. Typically, blast exposure is examined through a
ees completed the BETS and a demographic intake form traumatic brain injury (TBI) lens, despite the fact that blast ex-
1
one week before Special Forces Advanced Urban Combat posure does not always result in a TBI. Even without an acute
(SFAUC) training. BETS was also administered 1 and 8 weeks injury (e.g., TBI) following blast, repetitive blast exposure ac-
post-training. Spearman’s rho correlations assessed convergent cumulated over years of military service can negatively affect
2
validity from pre-SFAUC time points. A two-way mixed ef- cognitive capabilities and neurobehavioral symptoms (e.g.,
3,4
fect, consistency, single rater intraclass correlation coefficient anger, anxiety, fatigue). These effects are also observed in
(ICC [95%CI]) and percent agreement analyzed test-retest animal models where repetitive blast exposure has been linked
3,1
5
reliability for post-SFAUC time points. Results: Forty-one to depression- and anxiety-like behaviors for up to 52 weeks.
participants (17.1% trainers; mean age: 32.8 [SD 5.7] y), and Further studies have attributed the alterations in behavior and
13 participants (7.7% trainers; mean age: 33.2 [SD 5.6] y; emotional states in both humans and animals to changes in
8
6,7
49.7 [SD 6.5] days between test-retest) were included in con- brain structure and function, white matter integrity, and
9
vergent validity and test-retest analyses, respectively. GBEV brain network connectivity.
displayed moderate convergent validity with all proxy mea-
sures of blast exposure (rho-range=0.595–0.672, P<.001 all). Tracking lifetime blast exposure is important for monitoring
GBEV displayed moderate test-retest reliability (ICC =0.576 the potential negative consequences of repeat blast exposure;
3,1
[95% CI 0.160–0.818]). Individual weapon categories were yet, accurate reporting of exposures continues to be problem-
poor-moderate (ICC -range=0.070–0.766). Items within each atic due to the variability of sensor and self-reported measures.
3,1
section displayed poor-excellent test- retest reliability (ICC - Blast gauges, or sensors, are the most reliable and valid meth-
3,1
range=–0.145-0.843). The difference and average from both ods for tracking blast exposure variables (e.g., impulse, over-
post-SFAUC timepoint were strongly correlated indicating pressure) but come with substantial drawbacks. For example,
worse recall with higher GBEV (n=13, rho=0.907, P<.001). blast-related variables differ by number of sensors used, sensor
Conclusions: GBEV has strong convergent validity but poor- placement and body location, and adherence to wearing the
10,11
12
moderate test-retest reliability. Caution should be used when sensors. Sensors are well-suited for short-term studies,
interpreting BETS results, especially in populations experienc- but not currently in a state for multi-year research projects.
ing excessive blast exposures. However, our small sample size Other options for assessing blast exposure include computa-
13
limits the interpretability of the results. tional modeling completed post-blast exposure.
Previously used questionnaires for lifetime blast exposure
Keywords: concussion; mild traumatic brain injury; generalized 4,14
blast exposure value; Blast Ordnance Occupation Exposure were single-item offering little to no granularity for the
Measure; psychometrics types of blasts, sources of blasts, duration of blasts, or other
relevant details. Recently developed questionnaires including
the Blast Exposure Threshold survey (BETS), Blast Ordnance
15
*Correspondence to eric.j.shumski.ctr@health.mil
1 Dr. Eric Shumski is affiliated with the Traumatic Brain Injury Center of Excellence, Silver Spring, MD, the General Dynamics Information Tech-
2
nology, Falls Church, VA, and the Womack Army Medical Center, Intrepid Spirit Center, Fort Bragg, NC. Deanna Pinnow is affiliated with the
Traumatic Brain Injury Center of Excellence, Silver Spring, MD, the Compass Government Solutions, Annapolis, MD, and the Evans Army Com-
3
munity Hospital, Fort Carson, CO. Christy Mote is affiliated with the Traumatic Brain Injury Center of Excellence, Silver Spring, MD, the Evans
Army Community Hospital, Fort Carson, CO, and 9Line LLC, Tampa, FL. Steven Wilcox is affiliated with the Traumatic Brain Injury Center
4
of Excellence, Silver Spring, MD, the General Dynamics Information Technology, Falls Church, VA, and the Evans Army Community Hospital,
Fort Carson, CO. Dale Welsh is affiliated with the United States Public Health Service. Dr. Jamie N. Hershaw is affiliated with the Traumatic
6
5
Brain Injury Center of Excellence, Silver Spring, MD, the General Dynamics Information Technology, Falls Church, VA, and the Joint Base San
Antonio-Brooke Army Medical Center, San Antonio TX.
64