{"resourceId":"usf-src-clearance-ml-2026","versions":[{"version":"external-8516fff3baabb96a1e65db42d5400ef7424e4aeca45f5586a0c3cce7cfb35da8","resource":{"id":"usf-src-clearance-ml-2026","title":"USF study: longitudinal ML predicts sport-related concussion clearance with XGBoost accuracy 0.84","organization":"University of South Florida / Digital Health (SAGE)","sector":"College athletics / sports medicine return-to-play risk modeling","geography":"United States (USF Concussion Center clinic cohort)","publishedAt":"May 20, 2026","publicationDate":"2026-05-20","eventDate":null,"sourceName":"Digital Health (PMC full text)","sourceLabel":"Academic research — measured SRC clearance prediction models","sourceUrl":"https://pmc.ncbi.nlm.nih.gov/articles/PMC13191159/","evidenceClass":"academic-research","outcomeClass":"mixed","topics":["knowledge-work","data-security","operating-model","infrastructure"],"finding":"Using 217 sports-related concussion patients (mean age 26.94) from the USF Concussion Center (2021–2025), Tran et al. train six ML classifiers to label normal (<30 days) versus prolonged (≥30 days) medical clearance. Adding Visit-2 longitudinal features improved accuracy in 66% of models; XGBoost achieved the highest accuracy (0.84, +5% over Visit 1). Specificity remained low (0.00–0.34) with 81.1% prolonged cases. VOR Vertical Headache and its change score were consistent prolonged-recovery predictors; treatment presence predicted normal recovery. Authors require external validation before clinical deployment.","sledRelevance":"Newly relevant athlete-support / sports-medicine evidence for collegiate athletic trainers considering ML on VOMS and related assessments already used in RTP—complements archived Claros post-concussion MSK risk modeling with a clearance-time outcome. Transferability is limited because the cohort is a specialty clinic sports-related sample, not an NCAA CARE multi-site labeled set.","evidence":"PMC full text of Digital Health article (doi:10.1177/20552076261450858): abstract metrics, methods (LOOCV; Visit 1 vs Visit 2 feature sets; class imbalance), Table performance figures, feature-frequency discussion, limitations, and explicit external-validation requirement; IRB STUDY003514; public code link.","architectureImplications":"Interpretation: score VOMS/BESS/ImPACT-like exports in a clinician-controlled service with longitudinal difference features and treatment flags; do not expose raw cognitive scores to coaches without policy.","governanceImplications":"Interpretation: keep model output under sports-medicine ownership for clearance support only; prohibit scholarship/roster use; require local revalidation before any prospective alert.","securityPrivacyImplications":"Interpretation: concussion and mental-health (GAD-7/PHQ-9) features are highly sensitive; minimize secondary AI-training rights and apply role-based access even inside athletics.","caveats":"Clinic cohort not exclusively NCAA-labeled; mean age ~27; low specificity; no external validation; binary 30-day threshold; self-report bias; gender imbalance.","streamIds":["college-athletics"],"roles":{"sales":"Open with return-to-play uncertainty after concussion, not generic injury AI. Athletic trainers already collect VOMS and related measures; the credible ask is whether longitudinal change scores can flag athletes likely to need prolonged care pathways without promising faster clearance. Discover who owns concussion EMR data, whether coaches see symptom detail, and what false-positive burden ATs will accept given the paper’s low specificity. Bound value to triage and rehab targeting—not win totals or automated medical clearance.","engineering":"Fit is clinical decision-support on existing assessment exports with engineered difference features between visits—not a new wearable fleet. Prefer batch scoring with documented feature schemas, class-imbalance reporting, and site-specific revalidation. Prerequisites: clean two-visit concussion records, IRB/clinical governance, and holdout seasons. Constraints: source study’s LOOCV-only validation and clinic case-mix. Proof of value is local balanced accuracy/specificity plus clinician override logs—not the 0.84 headline alone.","delivery":"Delivery is sports-medicine led: map VOMS subscore fields, validate on local historical SRC cases, set alert thresholds with ATs and team physicians, train staff that scores inform rather than replace clearance, and monitor drift each season. Dependencies include AT documentation quality and IT access to assessment stores. Ownership sits with sports medicine under AD sponsorship. Acceptance criteria: documented local validation, written coach-access rules, and zero unauthorized scholarship use of scores in the pilot window."},"roleDetails":{"sales":{"customerProblem":"Sports-medicine staff struggle to anticipate which concussed athletes will need prolonged clearance timelines using single-visit impressions alone.","stakeholders":["Head athletic trainer","Team physician","Sports performance director","Sport coaches","Compliance/privacy officer"],"discoveryQuestions":["Which VOMS/BESS/ImPACT fields are already mandated and stored longitudinally?","Who may view cognitive or symptom scores today?","What false-positive rate is tolerable given AT workload?","Would recovery-risk tiers ever influence scholarship or roster decisions?"],"valueHypothesis":"If local revalidation improves balanced accuracy beyond chart review alone, ATs could target vestibular/rehab pathways earlier using assessments they already collect—without claiming faster RTP.","potentialEngagement":"Six-to-eight-week retrospective validation on the school’s last two seasons of SRC cases, plus a draft clinical-use policy.","unsupportedClaims":["Guaranteed shorter clearance times","That 0.84 accuracy generalizes without local validation","NCAA multi-site representativeness of the USF clinic cohort"]},"engineering":{"fit":"Clinical decision-support scoring on existing concussion assessment pipelines for collegiate sports medicine.","architectureIntegration":"Export visit-level VOMS subscores and related fields; compute difference features; score in a controlled environment; return risk tiers to AT workflows with audit logs.","prerequisites":["Longitudinal two-visit concussion records with clearance labels","Standardized assessment timing","Clinical governance approval"],"constraints":["Source study used LOOCV only","Severe class imbalance and low specificity","Clinic demographics may differ from campus athlete population"],"security":"Restrict cognitive/mental-health features; no secondary model-training rights for vendors without athlete-facing consent review.","proofOfValue":["Local balanced accuracy and specificity on held-out seasons","Calibration within agreed error bands","Clinician override rate reviewed monthly"]},"delivery":{"work":"Historical validation, threshold setting with ATs, coach-access policy, and season-one monitoring dashboard.","dependencies":["Athletic training documentation quality","IT access to assessment and clearance systems","Team physician sign-off"],"ownership":"Sports medicine / head athletic trainer owns clinical use; AD sponsors policy.","skillsAdoption":"ATs learn score interpretation and specificity limits; coaches learn they do not receive raw cognitive detail without policy.","governanceCheckpoints":["IRB/clinical review before prospective use","Written prohibition on scholarship use","Quarterly drift review"],"acceptanceCriteria":["Documented local validation report","Signed access matrix","No unauthorized roster uses in pilot"],"risks":["Over-trust in 0.84 accuracy headline despite low specificity","Leak of symptom/cognitive scores to coaching staff","Distribution shift versus clinic training data"]}},"retrievedAt":"2026-09-21T03:09:15Z","enrichedAt":"2026-09-21T03:14:00Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Interpretation: ATs need plain-language cards that emphasize specificity limits so junior staff are not flooded by prolonged-class alerts.","procurementImplications":"Interpretation: buy validation and integration around existing assessment vendors; do not procure a black-box clearance AI that ignores VOMS field definitions.","operatingModelImplications":"Interpretation: keep medical clearance with sports medicine; models triage attention, they do not clear athletes.","sourceVerification":{"openedUrl":"https://pmc.ncbi.nlm.nih.gov/articles/PMC13191159/","referenceExcerpt":"XGBoost achieving the highest accuracy (0.84, +5% gain over Visit 1)","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}