Does a computerized neuropsychological test predict prolonged recovery in concussed children presenting to the ED?

Abstract

ABSTRACT Computerized neuropsychological tests (CNTs) are widely used internationally in concussion management. Their prognostic value for predicting recovery post-concussion is poorly understood, particularly in pediatric populations. The aim of the present study was to examine whether cognitive functioning (measured by CogSport) has prognostic value for predicting rapid versus slow recovery. This is a prospective longitudinal observational cohort study conducted at a state-wide tertiary pediatric hospital. Data were collected at 1–4, 14, and 90 days post-injury. Eligible children were aged ≥5 and <18 years presenting to the Emergency Department having sustained a concussion within 48 h. Concussion was defined according to the Zurich/Berlin Consensus Statement on Concussion in Sport. Dependent variables were reaction times and error rates on the CogSport Brief Battery. In total, 220 cases were analyzed; 98 in a rapid recovery group (asymptomatic at 14 days post-injury, mean age 11.5 [3.2], 73.5% male) and 122 in a slow recovery group (symptomatic at 14 days post-injury, mean age 12.0 [3.1], 69.7% male). Longitudinal GEE analyses modeled the trajectories of both mean log10-transformed reaction time and error rates between groups over time (1–4, 14 and 90 days). Both group main and interaction (time by group) terms for all models were non-significant (p > .05). Cognitive functioning, measured by CogSport and assessed within 1–4 days of concussion, does not predict prolonged recovery in a pediatric sample. Further, there were no significant group differences at any time point. Considering the widespread use and promotion of CNTs, it is important that clinicians understand the significant limitations of the CogSport battery.