RESUMO
OBJECTIVE: The objective is to determine whether unsupervised machine learning identifies traumatic brain injury (TBI) phenotypes with unique clinical profiles. METHODS: Pilot self-reported survey data of over 10,000 adults were collected from the Centers for Disease Control and Prevention (CDC)'s National Concussion Surveillance System (NCSS). Respondents who self-reported a head injury in the past 12 months (n = 1,364) were retained and queried for injury, outcome, and clinical characteristics. An unsupervised machine learning algorithm, partitioning around medoids (PAM), that employed Gower's dissimilarity matrix, was used to conduct a cluster analysis. RESULTS: PAM grouped respondents into five TBI clusters (phenotypes A-E). Phenotype C represented more clinically severe TBIs with a higher prevalence of symptoms and association with worse outcomes. When compared to individuals in Phenotype A, a group with few TBI-related symptoms, individuals in Phenotype C were more likely to undergo medical evaluation (odds ratio [OR] = 9.8, 95% confidence interval[CI] = 5.8-16.6), have symptoms that were not currently resolved or resolved in 8+ days (OR = 10.6, 95%CI = 6.2-18.1), and more likely to report at least moderate impact on social (OR = 54.7, 95%CI = 22.4-133.4) and work (OR = 25.4, 95%CI = 11.2-57.2) functioning. CONCLUSION: Machine learning can be used to classify patients into unique TBI phenotypes. Further research might examine the utility of such classifications in supporting clinical diagnosis and patient recovery for this complex health condition.
Assuntos
Concussão Encefálica , Lesões Encefálicas Traumáticas , Aprendizado de Máquina , Fenótipo , Humanos , Masculino , Feminino , Projetos Piloto , Adulto , Lesões Encefálicas Traumáticas/diagnóstico , Lesões Encefálicas Traumáticas/epidemiologia , Pessoa de Meia-Idade , Concussão Encefálica/diagnóstico , Concussão Encefálica/epidemiologia , Estados Unidos/epidemiologia , Adulto Jovem , Adolescente , Autorrelato , Aprendizado de Máquina não Supervisionado , Idoso , Centers for Disease Control and Prevention, U.S. , Análise por ConglomeradosRESUMO
OBJECTIVES: To quantify norms and changes in eye-tracking proficiency, and determine vestibular symptom correlations in varsity college athletes following acute mild traumatic brain injury (mTBI). We hypothesized that mTBI impacts central coordination between the vestibular and oculomotor systems with resultant changes in eye-tracking proficiency that are correlated with vestibular symptom provocation. DESIGN: Retrospective cohort study. SETTING: Sports medicine care at a single institution. PATIENTS: One hundred and nineteen college athletes diagnosed with mTBI by a physician between 2013 and 2019. INTERVENTIONS: N/A. MAIN OUTCOME MEASURES: Standard deviation of tangential error, standard deviation of radian error, mean phase error, and horizontal gain from virtual reality-based, circular eye-tracking goggles used at baseline and within 72 hours post-mTBI. Headache, dizziness, nausea, and fogginess provocation after the Vestibular Ocular Motor Screening (VOMS) smooth pursuits subtest compared with pretest baseline, assessed within 72 hours post-mTBI. RESULTS: One hundred and nineteen college athletes (N = 56 women and 63 men) aged 18 to 24 years sustained a total of 177 mTBI. Forty-four percent of athletes displayed abnormal eye-tracking on at least 1 eye-tracking measure following acute mTBI compared with their baseline. From the VOMS, horizontal gain showed medium-sized to large-sized positive correlations with headache ( r = 0.34) and dizziness ( r = 0.54), respectively. Mean phase error showed a medium-sized negative correlation with nausea ( r = -0.32) on the VOMS. CONCLUSIONS: Eye-tracking proficiency was impaired and correlated with vestibular symptom provocation following acute mTBI in college athletes. Future research should examine eye-tracking proficiency testing in other acute care settings to support mTBI diagnosis.
RESUMO
OBJECTIVE: To report the symptom burden of anxiety and mood-related indicators following mTBI in collegiate student-athletes. STUDY DESIGN: Retrospective cohort study of varsity collegiate athletes. SETTING: University sports medicine at a tertiary care center. PATIENTS: Division I college varsity athletes diagnosed with mTBI at a single institution between 2016 and 2019. INDEPENDENT VARIABLES: Pre- and post-injury. MAIN OUTCOME MEASURES: Comparisons between baseline testing and post-mTBI symptom scale assessments were made to determine changes in scores at the individual and group levels. The primary outcome was the prevalence of post-mTBI symptoms from within 72 h of injury through return to play. Associations with sport, sex, age, and return-to-play time were included. RESULTS: Compared to baseline, mood and anxiety symptom scores were significantly higher acutely following mTBI (2.1 ± 3.3 vs. 14.3 ± 12.2; p < 0.001). A family history of migraine was significantly associated with higher mood and anxiety symptom scores (20.0 ± 14.9 with history vs. 13.3 ± 11.3 without history; p = 0.042). Mood and anxiety symptom scores were highly correlated with non-mood and anxiety symptom scores for all athletes, including the subgroup with prolonged symptoms (r = 0.769; p < 0.001). CONCLUSIONS: Symptoms of anxiety or mood disruption are common during the acute period post-injury in varsity college athletes. Risk factors for higher symptom reports immediately following mTBI and for prolonged symptoms (>10 days) included female sex, those with a family history of migraine, and those with an overall higher symptom burden post-injury.