To evaluate the clinical utility of "rebaseline" testing (ie, new baseline) before the season after an athlete's concussion recovery by (1) determining differences between baseline and rebaseline scores, and (2) comparing differences to clinical reliable change indices (RCIs). Retrospective cohort. Clinical research center. Thirty-four Division I collegiate athletes (16 females, 18 males; baseline age = 18.38 ± 0.78 years). Neurocognitive, balance, and symptom testing was conducted at 3 times: (1) original baseline, (2) final postinjury before return-to-play, and (3) rebaseline completed before the next competitive season. Physicians diagnosed concussions, and all concussed athletes' returned-to-play during the same season. Dependent variables included all neurocognitive domains, balance composite, and total symptom scores. Mean differences between sessions were compared with 80% RCIs to clinically interpret statistical findings. Statistically significant improvements in neurocognitive performance were observed between baseline and rebaseline sessions: psychomotor speed (F1,33 = 13.79, P = 0.001), cognitive flexibility (F1,33 = 7.99, P = 0.008), processing speed (F1,32 = 15.93, P < 0.001), executive functioning (F1,33 = 9.61, P = 0.004), and reasoning (F1,32 = 7.63, P = 0.009). We observed significant improvement in balance (F1,29 = 22.26, P < 0.001), but no difference in total symptom score (P = 0.719). Participants performed statistically better at last postinjury compared with rebaseline in visual memory (F1,29 = 5.64, P = 0.024) and complex attention (F1,31 = 11.38, P = 0.002), but worse in processing speed (F1,30 = 8.92, P = 0.006). Participants reported more symptoms (F1.28 = 10.83, P = 0.003) at rebaseline than last postinjury. No observed mean differences exceeded RCIs. All between-session improvements were within RCIs and therefore within expected test-retest variation. Our results suggest limited clinical utility for rebaseline assessments in the collegiate setting. Despite statistically significant differences between test sessions, no differences exceeded RCIs, suggesting the differences were of little clinical utility in the interpretation of neurocognitive, balance, and symptom scoring.
Read full abstract