To examine whether patients with mild traumatic brain injury (mTBI) receiving text messaging–based education and behavioral support had fewer and less severe postconcussive symptoms than those not receiving text-message support. Our secondary objective was to determine the feasibility of using text messaging to assess daily symptoms and provide support to patients with mTBI.
Randomized controlled trial with 14-day follow-up.
Convenience sample of 43 adult emergency department patients with mTBI.
Fourteen days of timed SMS (short-message service) symptom assessments (9 AM: headaches; 1 PM: difficulty concentrating; 5 PM: irritability or anxiety) with self-care support messages.
SMS symptom reports, Rivermead Postconcussion Symptoms Questionnaire.
Compared with the control group, intervention participants trended to lower odds of reporting headaches (odds ratio [OR] = 0.38; 95% confidence interval [CI]: 0.07-1.99), concentration difficulty (OR = 0.32; 95% CI: 0.04-2.24), and irritability or anxiety (OR = 0.33; 95% CI: 0.05-2.35). There were also trends of lower mean scores for headaches (0.99 vs 1.19; P = .5), difficulty concentrating (0.88 vs 1.23; P = .2), and irritability/anxiety (1.00 vs 1.62; P = .06). There were high response rate to SMS symptom assessments and high satisfaction with the intervention.
Those receiving the text messaging–based education and support had fewer and less severe postconcussive symptoms than the controls but none of the differences reached statistical significance. Further evaluation of more robust mobile interventions and larger sample of participants are still needed.
Departments of Emergency Medicine (Drs Suffoletto, Calabria, Kingsley, Kristan, and Callaway) Physical Medicine and Rehabilitation (Drs Wagner and Arenth), University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania.
Corresponding Author: Brian Suffoletto, MD, MS, University of Pittsburgh, Department of Emergency Medicine, Iroquois Bldg, Ste 400A, 3600 Forbes Ave, Pittsburgh, PA 15261 (email@example.com).
The authors thank Jack Doman in the Office of Academic Computing at the Western Psychiatric Institute at the University of Pittsburgh for all computer programming support.
The authors declare no conflicts of interest.