Blog
October 6, 2026
par
Andrej Evtimov
Four checkpoints in the record that carry most of the weight in an mTBI file.
A mild traumatic brain injury claim is graded mild because of what was recorded in the first hours, not because of what the claimant is living with two years later. Four checkpoints in the record carry most of the weight in these files.
"Mild" is a description of the injury event, not of the claim
A mild traumatic brain injury (mTBI) claim is graded mild because of what was recorded in the first hours. It is not graded mild because of what the claimant is living with two years later.
That gap is why these files behave differently from other injury claims. The diagnosis rests largely on symptoms the claimant reports. The damages run for decades: lost earning capacity, cognitive rehabilitation, life care costs. And the evidence that decides the file is a set of documents, not an image.
Claims leaders do not need to adjudicate the neuroscience. They need to know which four things in the record carry weight, and whether the file contains them. This article is a records-reading guide. It does not support diagnosing a claimant, and the clinical questions belong to a qualified reviewer.
What the outcome data says about "mild"
It is tempting to treat "mild" as self-limiting. One large US cohort does not support that.
The TRACK-TBI study followed 1,154 patients with mTBI at 11 US level I trauma centers. Mild there means a Glasgow Coma Scale (GCS) score of 13 to 15. These were patients who reached a major trauma center, not every claimant who reports a concussion. At 12 months, "the percentage of mTBI participants reporting functional limitations was 53% (95% CI, 49%-56%)." Most such patients "report persistent, injury-related life difficulties at 1 year postinjury" (JAMA Neurology, 2019).
The comparison group in the same study is the number worth carrying into a reserve discussion. Among orthopedic trauma controls, patients with injuries but no brain injury, 38% also reported functional limitations at 12 months (JAMA Neurology, 2019). Persistent difficulty after a serious accident is common. Attributing all of it to the head injury is the contested step. That is an argument about attribution, not about whether the claimant is impaired.
What the record has to establish
The 2023 diagnostic criteria of the American Congress of Rehabilitation Medicine (ACRM) are the current reference point, and clinical guidelines have begun adopting them. One Ontario guideline is explicit. It adheres to the ACRM criteria "for concussion or uncomplicated mTBI" (Living Concussion Guidelines, 2023).
The structure of those criteria matters more than any single threshold. The guideline directs clinicians to assess clinical signs, acute symptoms and clinical findings, rather than a symptom report alone. What follows from that, for a claims file, is a reading order. This article uses four checkpoints.
- Glasgow Coma Scale at the scene and at the emergency department. Where it lives: EMS run sheet, ED triage and trauma notes. What it establishes: Severity grading at the time, and how the initial injury was classified.
- First appearance of symptoms. Where it lives: ED note, first primary care visit, first therapy note. What it establishes: Whether symptoms were contemporaneous or later in the timeline.
- Validity testing. Where it lives: Neuropsychological evaluation report. What it establishes: Whether the cognitive scores can bear the weight placed on them.
- Treating record against expert report. Where it lives: Treating notes versus the retained expert's history. What it establishes: Whether the expert's account matches the source documents.
Checkpoint 1: The Glasgow Coma Scale at the scene
The GCS scores eye opening, verbal response and motor response, and runs from 3 to 15. A score of 13 to 15 is the mild band, and it is what the TRACK-TBI cohort used as its inclusion criterion. EMS means the emergency medical service crew at the scene; ED means the emergency department.
Three practical cautions belong with the scene score.
First, it is a snapshot. A 15 at the roadside and a 15 at triage is a stronger record than one number on one form. Look for the score at each handoff: EMS, triage, trauma bay, and any repeat assessment.
Second, intoxication is common in this population. One prospective study of traumatic brain injury (TBI) admissions puts the figure high. Alcohol intoxication "has been present in 35%-50% of traumatic brain injury (TBI) patients admitted in the hospitals." That study compared intoxicated and sober patients on the GCS-Pupils score. It found a lower score in the intoxicated group "which was not statistically significant" (Chinese Journal of Traumatology, 2022). That is one study of all-severity admissions, not proof either way. It does mean an argument that the score was really the alcohol needs support from the reviewer, not assertion.
Third, a missing GCS is not a defense. Many claimants never reach a trauma center. Without a documented score, the severity grading has to come from the rest of the acute record. Loss of consciousness, disorientation, amnesia for the event, and what the responders wrote down.
Checkpoint 2: When the symptoms first appear in the record
This checkpoint is decided by reading order rather than by expertise. Build the date-ordered medical chronology first, then read the symptom timeline off it.
Four questions the chronology answers:
- Were cognitive complaints recorded at the emergency department, or only in later visits?
- What is the interval between the loss and the first mention of headache, memory difficulty, or concentration difficulty?
- Where do the dates of representation and of the first neuropsychological referral sit on the same timeline? Map the dates; do not infer motive from them.
- Do the symptoms follow a recovery curve, or do they change direction at some point in the record?
None of these questions answers itself. Late-appearing symptoms have ordinary explanations: people return to work, discover they cannot do the job, and seek care then. A gap is a question for the reviewer, not a conclusion.
Checkpoint 3: Validity testing in the neuropsychological report
Neuropsychological testing produces scores. Validity measures test whether those scores can be relied on. The report should state which validity measures were administered and how the claimant performed on them.
There are two kinds, and they are not interchangeable. Performance validity tests (PVTs) examine whether the claimant gave a valid performance on cognitive tasks. Symptom validity tests (SVTs) examine the self-reported symptom profile. A review of the field reports that the two load as separate factors. "Failure in one validity domain does not necessarily invalidate the other domain," and the two "should be evaluated separately" (Brain Injury, 2015).
What a validity failure does and does not mean
This is where files go wrong in both directions.
Bigler's review is explicit that a failed measure is not proof of intent. "Malingering requires intentionality," it states. It adds that many publications use the term loosely in association with a failed measure, "even though the only statistical proof of intentionality is when SVT performance is below chance" (Brain Injury, 2015). Below-chance performance is the narrow case. Scores below a cut-point but well above chance are the common case. The same review calls "the interpretative problems with SVT performance below the cut-point yet far above chance" substantial.
The same source records the base condition in its samples. "The majority of individuals with a history of TBI undergoing neuropsychological testing perform above cut-points" on these measures (Brain Injury, 2015). Most people tested pass.
So the usable position is narrow. An invalid profile raises a question about whether the cognitive scores measure ability. Whether they can still support the causal opinion built on them is for the neuropsychologist or physician reviewer to say. It does not establish that the claimant is feigning. A file that argues otherwise invites a credibility problem of its own.
What to check in the report
- Were validity measures administered at all? If none were, ask the reviewer what that omission means for the weight of the report.
- How many, and of which type? A single measure supports far less than a battery.
- Does the narrative discuss the validity results, or does it state a diagnosis and move on?
- If a measure was failed, does the report say what follows from that, or does it interpret the invalid scores anyway?
Checkpoint 4: Does the expert report match the treating record?
The retained expert usually writes a history section. It draws on the claimant interview and on documents the file already contains, so much of it can be checked line by line.
What to compare:
- Reported duration of loss of consciousness. Check against: EMS run sheet and ED note.
- "Immediate onset" of cognitive symptoms. Check against: The first three clinical encounters.
- Pre-accident functioning described as unimpaired. Check against: Prior primary care records, prior psychiatric or headache history, school or work records.
- Current functional limitations. Check against: Treating therapy notes and any return-to-work documentation.
- Mechanism of injury. Check against: The claimant's own account at the first visit.
A mismatch is not a finding on its own. Histories drift as they pass through hands, and a claimant's description at month 18 will not match the triage note. Two things matter. Whether the expert's opinion rests on a version of the facts the source documents do not support. And whether the treating clinicians, who were treating rather than reporting, recorded something different.
Base rates: the symptoms are common in people with no injury at all
Post-concussion symptom questionnaires are part of most mTBI files. Their scores need a comparison point.
A study collected Rivermead Post-Concussion Symptoms Questionnaire responses from 11,759 general-population respondents in the United Kingdom, Italy and the Netherlands. None of those adults had a head injury. Among them, 41.4% reported fatigue, 35.8% headaches, 35.3% sleeping problems and 34.3% irritability (Journal of Clinical Medicine, 2022).
Two conclusions follow, and they point in opposite directions. A symptom checklist with several items endorsed is not, by itself, evidence of a brain injury. And a claimant who endorses those items is not, by itself, exaggerating, because a third of the uninjured population endorses them too. The questionnaire is a starting point for the reviewer, not an exhibit.
Building the file before the number moves
mTBI claims are long-tail by construction. The diagnosis is made early, the damages model is built late, and the reserve is usually set before anyone has read the whole record. That sequence is what produces development.
Three things move the file earlier.
Read the acute record on day one. The GCS entries, the ED note and the first week of care are short documents. They shape everything afterwards, and they are hardest to obtain once litigation hardens. Early review of the acute set is the same discipline that drives reserve accuracy from day one.
Ask for the validity data before the deposition. The neuropsychological report is the hinge document. Knowing which measures were used, and what the report says about them, changes the questions asked of both the claimant and the expert.
Put the chronology and the expert history side by side. This is document comparison across several hundred pages, and it is slow by hand. On auto liability claim files, that is the work AI-assisted extraction is aimed at. The condition is provenance. Every extracted date and quotation must link back to its source page. The adjuster and the physician reviewer read the source, and the reviewer answers the clinical question.
The position that holds
A defensible mTBI position rarely rests on denying that the claimant has symptoms. It rests on four document-level statements that the file can support.
The severity recorded at the time was X. The symptoms first appear in the record on date Y. The cognitive scores are, or are not, supported by validity measures. The expert's history does, or does not, match the treating record.
Each of those is checkable. Each of them narrows the question the physician reviewer has to answer. Each of them also survives cross-examination better than an argument that a brain injury graded mild should have resolved by now. The TRACK-TBI outcome data does not support that argument.
Key takeaways
- "Mild" grades the injury event, not the claim. In the TRACK-TBI cohort, 53% of mild TBI patients still reported functional limitations at 12 months, against 38% of orthopedic trauma controls.
- Checkpoint 1 is the Glasgow Coma Scale at the scene and at the emergency department. Read it at every handoff, and treat a missing score as a gap in the record rather than a defense.
- Checkpoint 2 is when cognitive symptoms first appear on a date-ordered chronology. A gap is a question for the reviewer, not a conclusion.
- Checkpoint 3 is validity testing in the neuropsychological report. A failed measure is not proof of intent, and most people tested pass; below-chance performance is the narrow exception.
- Checkpoint 4 is whether the retained expert's history matches the treating record. Post-concussion symptoms are also common in uninjured adults, so a symptom checklist is a starting point for the reviewer, not an exhibit.
