Utilization of a Text-Messaging Robot to Assess Intraday Variation in Concussion Symptom Severity Scores.
Anthony CA, and Peterson AR. Clin J Sport Med. 2014. [Epub ahead of Print].
Take Home Message: Athletes who have sustained a concussion exhibit a wide range of symptom severity scores when surveyed through text-messaging at various times during the day.
Following a concussion, athletes are often assessed with the Post Concussion Symptom Score (PCSS), a symptom severity checklist. While the PCSS has been shown to be valid and reliable, there has been no evaluation of the effect that survey frequency has on results and whether the scores vary overtime within a day. Therefore, Anthony and Peterson completed a prospective cohort study of 14 athletes (14 to 22 years of age) who sustained a concussion to determine how much PCSS scores vary within a day. All athletes were concussed while playing sports, symptomatic with a Concussion Symptom Severity Score of 10 or more, referred by a concussion clinic, English speaking, and owners of a cellular phone for text messaging. No demographic information was collected due to the inability to securely transmit patient information. An automated text-messaging system (robot) was created to automatically deliver PCSS questions to each athlete. Athletes chose 5 times during the day, which had to be spread out during waking hours, to have the questions sent. The robot would then send 1 question from the PCSS with a corresponding Likert scale. Once the athlete responded, another question was sent until all 22 PCSS questions were completed. If the athlete did not respond to the first text message, or did not complete all 22 questions, the robot would restart the questioning at the next scheduled time. Incomplete surveys were not included in the final analysis. Any athlete who did not complete more than 2 surveys on 3 or more days were excluded leaving 11 athletes for analysis. The robot sent text messages every day for 30 days or until the athlete completed the PCSS on 7 consecutive days. Overall, 804 survey were completed and used for analysis. The authors found that the PCSS score could often vary within one day by 23.1 points (a common range of PCSS scores in this study was 0 to 80).
Ultimately, the data presented in the current study suggests that using a text-messaging robot to assess athletes with a concussion has poor repeatability and cannot be considered reliable. While not reliable, this method of surveying athletes may be of interest to clinicians, especially in a growing technologically-dependent society. While not the purpose of the study, perhaps integrating concussion symptom surveys into a more technological format may be of interest to clinicians. Secondly, the study did not attempt to address why there may be such variability in PCSS answers, which may also be of interest to clinicians. These findings could be related to a limitation of the PCSS via text message or it could demonstrate that athletes with a concussion experience a wide range of symptom severity depending on the time of day, location, current activity, etc. Future research may benefit from a similar study which includes questions regarding the athlete’s mood, activity at the time of survey, and/or medications. This could help gain a better understanding of why such variability existed. While we need to gain a better understanding of the utility of PCSS via text messaging and why these scores fluctuate during the day clinicians should continue to administer the PCSS to patients to get an objective measurement of concussion symptoms and severity.
Questions for Discussion: Do you see a lot of variability in PCSS answers from your athletes within one day? What steps might you take to limit variability in the current format of the PCSS you administer?
Written by: Kyle Harris
Reviewed by: Jeffrey Driban