Traumatic brain injury lawsuits: causation, long-term impact and the evidence that survives cross

Strip away the medicine and a brain injury claim has 5 links. Each one needs its own evidence, and each is where one side or the other will push.

Links 1 and 2 belong to the first 72 hours, and we cover them in depth in the companion guide on TBI imaging, diffuse axonal injury and DTI. This guide starts where that one ends: a documented head injury, often labeled mild, and months later headaches, poor concentration, sleep trouble and trouble at work.

Why "mild" is the hardest category to litigate

Severe TBI with a hemorrhage on CT, a week in intensive care and an obvious deficit rarely produces a causation fight. The damages fight is about numbers. Mild TBI is different. The 2023 American Congress of Rehabilitation Medicine criteria, built by a 17-member working group and a 32-member expert panel, recorded that 30 of 32 panelists (93.8%) agreed "concussion" may be used interchangeably with "mild TBI" when neuroimaging is normal or not clinically indicated (Silverberg and colleagues, 2023). So the typical case has a negative CT, a short or disputed loss of consciousness, and symptoms that are real to the person but overlap with many other conditions. The TRACK-TBI study defined its mild group by an admission Glasgow Coma Scale of 13 to 15 and reported that most TBIs are classified mild on that basis (Nelson and colleagues, 2019).

No blood test proves the headache in month 9 came from the crash. There is only a pattern: who the person was before, what changed and when, and what else was going on. Every piece of it is a document.

Mild TBI or concussion

A brain injury from external force with transient disturbance of brain function. The ACRM 2023 criteria allow the terms to be used interchangeably when imaging is normal or not indicated.

Persistent post-concussive symptoms

Symptoms such as headache, dizziness, fatigue, irritability, poor concentration and sleep disturbance that last beyond the expected recovery window. Older sources say "post-concussion syndrome"; ICD and DSM terms have changed across editions.

GOSE

Glasgow Outcome Scale Extended: an 8-point functional outcome scale from 1 (dead) to 8 (upper good recovery). A score under 8 means some injury-related limitation.

Performance validity test (PVT)

A test built into a neuropsychological battery to check whether the person gave full effort. Failure does not prove malingering; it means the other scores cannot be read at face value.

Defense experts in brain injury cases use a short list of alternative explanations. None of them is illegitimate. Each is a real confounder in the medical literature, and each can also be overreached. We would print this table and put it at the front of the file.

The pattern across the table is simple. Every defense theory predicts something about timing: the symptom was there before, it came from something that started later, or it tracks something other than the injury. Every plaintiff answer is also about timing: the symptom is new, it started when the injury happened, and its course fits the injury. Underneath everything, a brain injury case is a chronology case.

Chronic pain is the attack with the best data behind it

Iverson and McCracken studied 170 chronic pain patients who had not had a head injury. Specific symptom endorsement ranged from 5% to 76.5%. Disturbed sleep, fatigue and irritability were reported by most patients, forgetfulness by 29%, and 80.6% endorsed 3 or more symptoms from the DSM-IV postconcussional disorder research criteria (Iverson and McCracken, 1997). Their conclusion was that postconcussive-like symptoms "are not unique sequelae of mild head injury."

The study cuts both ways. A symptom checklist cannot tell a brain injury from a neck injury, but that does not prove the plaintiff's symptoms come from pain. The record has to show which came first and whether cognitive complaints changed when pain did.

For the law of pre-existing conditions (the eggshell rule, aggravation and apportionment, and who carries the burden), see the eggshell plaintiff guide. Each doctrine turns on the same before and after records.

Both sides cite recovery statistics. Both usually cite them loosely. Here is what 2 common sources say, and what they do not.

The older view: most adults recover in 3 to 12 months

The WHO Collaborating Centre Task Force reviewed 428 prognosis studies and accepted 120 (28%) after critical review. It found consistent evidence that children recover quickly. For adults, it said the majority of studies report recovery for most people within 3 to 12 months, and that where symptoms persist, "compensation/litigation is a factor, but there is little consistent evidence for other predictors." It also warned that causal inferences in this literature "are often mistakenly drawn from cross-sectional studies" (Carroll and colleagues, 2004).

Defense experts quote the first and second sentences. Plaintiff experts quote the third. All 3 are in the same abstract.

The newer data: TRACK-TBI at 12 months

TRACK-TBI enrolled 1,453 patients at 11 US level I trauma centers within 24 hours of injury: 1,154 with mild TBI (admission GCS 13 to 15 with a clinical head CT) and 299 with orthopedic injuries as controls. It followed them for 12 months on the GOSE.

A jury will hear both readings:

Our view: TRACK-TBI fairly rebuts the claim that all mild TBI resolves in weeks, and it warns that injured people report limits for many reasons. Presenting the 53% as the chance this plaintiff has a permanent brain injury is the kind of leap Rule 702(d) invites a judge to catch (chapter 8).

Validity testing and the base rate of poor effort

A widely cited figure on symptom exaggeration comes from a 2002 survey of American Board of Clinical Neuropsychology members, covering an estimated 33,531 annual cases. Respondents reported probable malingering or symptom exaggeration in 29% of personal injury cases, and diagnostic impressions of probable malingering in 39% of mild head injury claims (Mittenberg and colleagues, 2002). Rates tracked the share of plaintiff versus defense referrals.

That figure is a survey of practitioners' impressions, not a measured prevalence, and it is more than 20 years old. It is a reason to insist that validity testing be done and reported, not evidence about any single plaintiff.

Long-term risk is where brain injury damages get large, and where expert testimony most often outruns its data. Each study below reports an association in a population, not a prediction for a plaintiff. We state them as the authors did.

Post-traumatic epilepsy

Annegers and colleagues followed 4,541 children and adults with TBI in Olmsted County, Minnesota, injured from 1935 through 1984. They graded severity by the record: mild meant loss of consciousness or amnesia under 30 minutes; moderate meant 30 minutes to 24 hours or a skull fracture; severe meant more than 24 hours, a subdural hematoma or a brain contusion. They compared unprovoked seizures with population rates (Annegers and colleagues, NEJM, 1998).

The significant risk factors in their multivariate analysis were brain contusion with subdural hematoma, skull fracture, loss of consciousness or amnesia for more than 1 day, and age 65 or older. A later Danish registry study of 1,605,216 children and young adults born from 1977 to 2002 found a relative risk of epilepsy of 2.22 after mild brain injury and 7.40 after severe injury, and found the risk still raised more than 10 years after mild injury (relative risk 1.51) and severe injury (4.29). Risk was notably higher in people with a family history of epilepsy (Christensen and colleagues, Lancet, 2009).

Note the disagreement: Minnesota saw no excess after 5 years for mild injury; Denmark saw one past 10 years. Either way, seizure risk scales with severity, so an epilepsy projection rests on the day 1 severity markers. A family history of epilepsy in the pre-injury chart matters to both sides.

Dementia

2 large cohort studies on dementia after TBI:

Sources: Fann and colleagues, 2018; Barnes and colleagues, 2018.

3 points keep dementia testimony honest. A hazard ratio is relative: 24% more of a modest baseline risk is a small absolute change, which is why Barnes's absolute figures (6.1% against 2.6%) matter. Both studies are observational and built from coded diagnoses, so they show association. And the spike in the first 6 months in the Danish data is worth asking your expert about: a question epidemiologists raise with this pattern is whether early, undiagnosed cognitive decline contributes to falls and head injuries in the first place. The abstract does not settle it; your expert should address it.

Chronic traumatic encephalopathy and traumatic encephalopathy syndrome

CTE is a neuropathological diagnosis: it is made from brain tissue. In 2021 a National Institute of Neurological Disorders and Stroke consensus panel published research diagnostic criteria for traumatic encephalopathy syndrome, the clinical disorder associated with CTE. Diagnosis of TES requires "substantial exposure to repetitive head impacts" from contact sports, military service or other causes, core clinical features of cognitive impairment or neurobehavioral dysregulation, a progressive course, and that the features are not fully accounted for by other conditions. The panel said the criteria were developed "with a primary goal of facilitating future CTE research" (Katz and colleagues, Neurology, 2021).

Under those research criteria, a single-crash case does not meet the exposure requirement. Repetitive-impact cases (contact sports, military blast exposure) are a different analysis; see sports concussion and CTE record review.

Whether increased future risk is compensable is a question of state damages law. The medicine sets the numbers; the law decides what they are worth.

In a mild TBI case the baseline decides more than the imaging does. A plaintiff promoted twice and then put on a performance plan 4 months after the crash has a case the defense must answer. A plaintiff whose chart shows 5 years of headaches and poor sleep has a problem the plaintiff must answer.

Our advice to both sides: request baseline records in the first wave, before the neuropsychological evaluation.

An order of requests that works

1. Intake the history first. Every school, employer, doctor, pharmacy and prior injury for 10 years before the incident. Gaps in the list are leads.
2. Request primary care and pharmacy first. They are fast, and they name the specialists, therapists and prior injuries you did not know about.
3. Request school and employment in the same wave. They take longest, and they are the objective baseline that neither side's expert created.
4. Follow every referral. "Referred to counseling" implies a record. Request it.
5. Close the loop in writing. Log each request, response and "no records" answer.

Reading the baseline without cherry-picking

The defense picks every prior headache; the plaintiff picks every "exceeds expectations." A baseline that survives cross shows frequency and function: how often a symptom was charted, what treatment it needed, and whether it affected school or work. Count, date and cite.

Watch templated text in the primary care chart. A review of systems that reads "denies headache, dizziness, memory loss" at 12 consecutive visits may be a real negative or a default template nobody changed. The same caution applies after the injury, when copy-forward text can make symptoms look more constant than they were.

This example is hypothetical: the person, records and page numbers are invented. The patterns are common ones.

D.M. is 41, a logistics coordinator at a regional distributor. On 03/14/2025 her car is rear-ended at a stoplight. She sues, claiming mild TBI with persistent cognitive symptoms, headaches and dizziness, lost promotion prospects and future care. The defense says any concussion resolved, and that her complaints track a prior concussion, an anxiety history and ordinary tension headaches.

The pre-injury record

The post-injury record

Before and after, side by side

What decides this case is not the negative CT. It is the job file, the missing pain and sleep records, and whether the loss of consciousness history holds up.

A life care plan projects lifetime medical and support needs and their cost. In a brain injury case it is often the biggest damages number, and the biggest target. Each line needs 2 supports: a record showing the condition and the need, and a medical opinion that the need will continue. A line without a record is a projection; a line without an opinion is a wish list.

For D.M., a plan built strictly from the record looks different from one built from the complaint:

What the defense checks in a TBI life care plan

• Record tie. Each line should cite a record or a physician statement. Lines that appear first in the plan are open to challenge.
• Severity match. Items typical of moderate or severe TBI (attendant care, residential programs, seizure care) in a mild TBI plan with a negative CT need specific support.
• Duration. A lifetime of treatment for a condition the treating record shows improving.
• Pre-existing care. Treatment already under way before the incident belongs to the baseline. The eggshell and apportionment rules in the pre-existing conditions guide decide what is recoverable.
• Method. A template copied between cases is weak ground under amended Rule 702(d).

The plaintiff side pre-empts most of these by building the plan from a cited chronology, with written physician sign-off on each item. See life care planners, the life care plan foundation deliverable, and catastrophic injury life care foundation for more severe cases.

Brain injury cases are expert cases, and in federal court every expert must pass Federal Rule of Evidence 702, which the Supreme Court amended by order of April 24, 2023, effective December 1, 2023 (Supreme Court order). The rule now reads:

"A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if the proponent demonstrates to the court that it is more likely than not that: (a) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (b) the testimony is based on sufficient facts or data; (c) the testimony is the product of reliable principles and methods; and (d) the expert's opinion reflects a reliable application of the principles and methods to the facts of the case."

What changed

2 things. First, the rule now says expressly that the proponent must show the court each requirement is met "more likely than not." The committee note explains why: many courts had held that whether an expert's basis was sufficient, and whether the method was applied reliably, were questions of weight for the jury. The note calls those rulings "an incorrect application of Rules 702 and 104(a)." Second, subpart (d) now says the opinion must "reflect" a reliable application, which the note says emphasizes "that each expert opinion must stay within the bounds of what can be concluded from a reliable application of the expert's basis and methodology."

The note also sets limits on the change. "Nothing in the amendment imposes any new, specific procedures." The standard "does not require perfection," though it does not let an expert make claims unsupported by the basis and method.

Many states follow Rule 702; others still apply the general acceptance test from Frye v. United States, 293 F. 1013 (D.C. Cir. 1923), and not every state adopted the 2023 wording. Check your forum.

Applying it to each brain injury expert

The rules around the report

3 other rules govern what the expert must disclose and may rely on:

• FRCP 26(a)(2)(B). A retained expert's report must give all opinions and their basis, facts or data considered, exhibits, qualifications with publications from the previous 10 years, testimony from the previous 4 years, and compensation. Drafts are protected under 26(b)(4)(B). Attorney and expert communications are protected under 26(b)(4)(C) except for compensation, facts or data the attorney provided and the expert considered, and assumptions the attorney provided and the expert relied on (FRCP 26). A chronology counsel sends the expert is likely to fall within that facts or data exception if the expert considers it.
• FRE 703. An expert may rely on facts or data that experts in the field would reasonably rely on, even if inadmissible. The underlying inadmissible facts reach the jury only if their probative value in helping evaluate the opinion substantially outweighs their prejudicial effect (FRE 703).
• FRCP 35. When a party's mental or physical condition is in controversy, the court may order an examination by a suitably licensed or certified examiner, on motion for good cause. The examined party can request the examiner's detailed report (FRCP 35).

Is your expert ready for a 702 challenge?

Documents

• Grades and test scores. A student whose grades fall after an injury, with attendance and teacher comments explaining why, has strong evidence. Get several years on each side of the injury date.
• Performance reviews. Reviews written before the claim are the strongest. Reviews written after it are still useful but can be challenged as written with knowledge of the lawsuit.
• Work product and accommodations. Error logs, productivity metrics and reduced-hours requests can show a change in function more precisely than any test.

Lay witnesses

Spouses, coworkers, supervisors and teachers are most credible when they are specific ("she used to plan the whole family reunion; now she writes everything on sticky notes") and when their account matches the documents. They are weakest when their testimony copies the complaint.

Histories and how they drift

Statements made for diagnosis or treatment that describe history, symptoms, their onset or general cause fall under the hearsay exception in FRE 803(4). So each note's history is evidence, and changes between notes are visible.

For the plaintiff, the ED's "amnestic to impact" is a contemporaneous marker of post-traumatic amnesia, and amnesia sits beside loss of consciousness in the severity definitions used by the Minnesota epilepsy study in chapter 4. For the defense, the growing duration and the late source are a credibility point and a reason to test every later history against the first 72 hours. Neither side should let an expert adopt the later history without saying where it came from.

Demonstratives: FRE 107 and FRE 1006

2 rules now govern the charts and timelines used at trial. FRE 107, effective December 1, 2024, covers illustrative aids: a court may allow one to help the jury understand evidence or argument, weighing its utility against unfair prejudice, confusion and delay. It is not evidence, and it "must not be provided to the jury during deliberations unless" all parties consent or the court orders it for good cause. It should be entered into the record when practicable. FRE 1006, amended the same day, covers summaries of voluminous admissible records, which are admissible whether or not the underlying records are introduced, provided the originals are available to the other side.

Once school, employment, pharmacy and 10 years of primary care records arrive, a mild TBI file can run to thousands of pages. The cross-referencing this guide describes can take a nurse consultant weeks, and it is the work AI medical record review tools are built for.

What the technology does

Most tools combine OCR (optical character recognition) to turn scanned pages into text, handwriting recognition for handwritten notes such as athletic trainer logs and older PCP charts, and a large language model (LLM) that reads the text, extracts dated events and drafts an AI medical chronology or summary. Older clinical natural language processing did narrower extraction, such as pulling ICD-10 codes and medication names. Generative AI can draft prose, which is useful and is also where the risk lives.

The comparison below shows where these tools help on a TBI file. Here is where they fail:

• Hallucination. An LLM can produce a fluent sentence that no record supports: a loss of consciousness that was never charted, a diagnosis carried from 1 patient to another in a merged production. On a causation issue that turns on a single line, that is unacceptable. The remedy is a page-level citation on every line, and a human who checks the citations that matter.
• Copied-forward and templated text. EHR text copied from visit to visit, and notes drafted by ambient AI scribes, can make a symptom look charted 20 times when it was assessed once. Counting mentions without spotting repeated text inflates the course.
• Judgment. Whether sleep explains a processing speed deficit is expert opinion, never a tool's finding.
• Bad scans. Faxed school records and handwritten notes defeat OCR more often; low-confidence pages need a human read.

The courts have already set the stakes for unchecked AI output. In Mata v. Avianca, Inc., 678 F. Supp. 3d 443 (S.D.N.Y. 2023), lawyers were sanctioned under Rule 11 after filing a brief that cited cases ChatGPT had invented. If an expert relies on a chronology with a fabricated fact, that fact is in the expert's basis, and Rule 702(b) puts it before the judge.

What to ask any vendor, in priority order

For a wider comparison of legal AI tools used by neuropsychology and IME evaluators, see AI tools for neuropsychology IME evaluators. For accuracy standards in court, see is AI accurate enough for court, and for privacy, HIPAA compliant AI medical record review.

3 templates you can copy. Edit the bracketed fields. They are working tools, not legal forms; adapt them to your jurisdiction and your case.

CASE: [CASE NAME]            INJURY DATE: [MM/DD/YYYY]
LOOKBACK: [10] years before injury to present

SCHOOL (FERPA: school must respond within 45 days)
[ ] K-12 transcripts, standardized test scores, attendance  [SCHOOL, YEARS]
[ ] Special education, IEP, 504 plan, gifted or ADHD testing
[ ] School nurse and athletic trainer records, concussion logs

EMPLOYMENT
[ ] Complete personnel file incl. job description  [EMPLOYER]
[ ] Performance reviews [YEAR] to present
[ ] Attendance, leave, FMLA and accommodation records
[ ] Payroll records and W-2s; tax transcripts [YEARS]

MEDICAL (HIPAA access: action within 30 days)
[ ] Primary care, [5+] years before injury  [PROVIDER]
[ ] Psychiatry, psychology, counseling  [PROVIDER]
    (psychotherapy notes are outside the HIPAA access right)
[ ] Pharmacy fill history  [PHARMACY / PBM]
[ ] Prior ED visits, imaging, neurology  [FACILITY]
[ ] Military service treatment records  [IF APPLICABLE]

PRIOR CLAIMS
[ ] Prior injury claims, workers' compensation, disability files

LOG: source | date requested | date received | custodian says missing

ALTERNATIVE        | PRE-INJURY RECORD (freq, treatment, function) | POST-INJURY RECORD | GAP / RECORD NEEDED | EXPERT WHO ADDRESSES IT
Prior concussion   | [date, return to activity, cite]              | [cite]             | [ ]                 | [ ]
Psychiatric hx     | [dx, meds, last status, cite]                 | [cite]             | [ ]                 | [ ]
Chronic pain       | [site, treatment, cite]                       | [cite]             | [ ]                 | [ ]
Sleep disorder     | [complaints, study?, cite]                    | [cite]             | [ ]                 | [ ]
Meds, substances   | [fills by date, screening, cite]              | [cite]             | [ ]                 | [ ]
Litigation effects | [n/a before claim]                            | [cite]             | [ ]                 | [ ]
Later event        | [n/a]                                         | [date, cite]       | [ ]                 | [ ]

HISTORIES OF THE EVENT (LOC / amnesia / GCS):
source | date | what it says | who gave the history | page

EXPERT: [NAME, SPECIALTY]       REPORT DUE: [DATE]

FILE SENT
[ ] Cited chronology, pre-injury and post-injury  [VERSION, DATE]
[ ] Baseline records: school [ ] employment [ ] PCP [ ] pharmacy [ ]
[ ] Every history of the event (EMS, ED, PCP, specialists)
[ ] Imaging reports and images, if within scope
[ ] Opposing expert reports and Rule 35 report
[ ] Missing-records log with request status

REPORT SHOULD
[ ] Cite a page for every factual statement
[ ] Address each alternative explanation in the causation matrix
[ ] State which history of the event it relies on, and why
[ ] Report performance validity results (neuropsychology)
[ ] Keep conclusions within the test data and literature
[ ] List facts, data and assumptions provided by counsel

Before you serve the report

• Baseline records receivedSchool, employment, 5 or more years of primary care, pharmacy
• First 72 hours documentedEMS, ED, imaging, GCS and every LOC or amnesia statement
• Missing records requestedReferrals and orders checked against the production
• Causation matrix complete7 rows, both columns, page cites
• Histories comparedDrift explained, source of each history named
• Validity testing reviewedResults reported in both neuropsychological reports
• Long-term risk stated with absolute numbersStudy design and comparison group named
• Life care plan lines sortedDocumented, pending, projection
• Before and after evidence gatheredReviews, grades, lay witnesses with specific accounts
• Exhibits classifiedFRE 1006 summary or FRE 107 illustrative aid
• Forum standard checkedRule 702 as amended, or your state's version or Frye
• AI output verifiedCitations checked on the lines that matter

0 of 12 checked

Medrecords AI is medical chronology software. It works on the records you upload; it does not request or retrieve records from providers, schools or employers. On a TBI file it does 6 things:

• Builds a cited chronology across pre-injury and post-injury records, with a citation on every line back to the source page.
• Extracts patient history and recovery factors: prior accidents (including head injuries), prior surgeries, pre-existing conditions such as a psychiatric diagnosis, and social, family and occupational history, keeping what the patient reported separate from what a clinician confirmed.
• Lays out a causation and apportionment graph: the chain from the incident through symptom onset, imaging and treatment, with prior conditions and alternative causes shown alongside, every node cited. It is a structured, cited view of the record, not a conclusion.
• Flags missing records the file implies, such as a pain management referral or a sleep study order with no result. Each flag cites the line that implies it.
• Tracks work capacity and functional restrictions over time, and loads CT and MRI studies as DICOM in a browser viewer on the timeline through imaging review.
• Drafts a life care plan foundation: diagnoses, course, limitations and recurring care, each cited, for the planner to cost, revise and sign.

What it does not do: it does not interpret imaging or neuropsychological test data, does not decide causation, does not give legal advice and does not score a case. Flags are signals, not verdicts. It runs under SOC 2 and HIPAA with a signed BAA (see security and HIPAA). Self-Service bills 10 cents a deduplicated page, down to 5 cents at volume, duplicates free; Enterprise On-Prem is an annual license. More on the use case: traumatic brain injury record review, for plaintiff and defense teams, legal nurse consultants and forensic psychological evaluators.

What has to be proven in a traumatic brain injury lawsuit?

That the incident caused a brain injury, that the injury caused the symptoms and limits the plaintiff reports, how long they will last, and what they cost. In mild TBI the fight is usually over causation and duration.

How common are lasting symptoms after a concussion?

In the TRACK-TBI study of patients seen at US level I trauma centers, 53% of mild TBI patients still reported functional limitations at 12 months, against 38% of orthopedic trauma controls.

Can the defense blame symptoms on a prior concussion or anxiety?

It can argue it, and the records decide how well the argument works. The pre-injury school, work and primary care records are the test.

Does a normal CT or MRI mean there was no brain injury?

No. Most mild TBI produces normal standard imaging, and the ACRM consensus panel agreed that "concussion" and "mild TBI" may be used interchangeably when neuroimaging is normal or not indicated.

Can a single concussion cause CTE?

The 2021 NINDS research criteria for traumatic encephalopathy syndrome require substantial exposure to repetitive head impacts, and CTE itself is diagnosed from brain tissue. A single-event CTE claim does not fit those criteria.

What changed in Federal Rule of Evidence 702 in December 2023?

The rule now states that the proponent must show the court it is more likely than not that each requirement is met, and that the opinion must reflect a reliable application of the method to the facts.

Can AI summarize brain injury records for a lawsuit?

Yes, with limits. AI medical record review tools can build a cited chronology across school, work and medical records, pull every loss of consciousness history, and flag missing records. They cannot decide causation, and every line should be cited to its source page and checked by a person before an expert relies on it.

Is it HIPAA compliant to upload medical records to AI software?

It can be, if the vendor signs a business associate agreement, maintains security controls such as SOC 2, and does not use your records to train its models. A consumer chatbot without a BAA is not an appropriate place for a client's medical records.

Can ChatGPT write a TBI medical chronology?

A general LLM can produce text that looks like a chronology, but without page-level citations you cannot check it, and it can hallucinate facts. In Mata v. Avianca a court sanctioned lawyers under Rule 11 for filing cases ChatGPT invented. Use a tool that cites every line, under a BAA.

Can software decide whether the accident caused the brain injury?

No. Software can organize the evidence on each side of the causation question and cite it. Whether the incident caused the symptoms is an opinion for qualified experts, and admissibility of that opinion is a question for the judge under Rule 702 or your state's standard.

Every rule and number here was checked against a primary source in September 2026. Claims we could not check were left out. The worked example and the diff are hypothetical.

Rules and regulations

• Federal Rule of Evidence 702, text and 2023 committee note.
• Supreme Court order of April 24, 2023 amending Rules 106, 615 and 702, effective December 1, 2023.
• US Courts, Federal Rules of Evidence: enacted January 2, 1975 (Pub. L. No. 93-595), last amended in 2024.
• FRE 107 (illustrative aids, effective December 1, 2024), FRE 703, FRE 803(4), FRE 1006.
• FRCP 26 and FRCP 35.
• 34 CFR 99.10 (FERPA, 45 days) and 45 CFR 164.524 (HIPAA access, 30 days, psychotherapy notes excluded).

Studies and consensus statements

• Nelson LD and colleagues. TRACK-TBI 12-month recovery. JAMA Neurology, 2019. PubMed 31157856; full text.
• Carroll LJ and colleagues. WHO Task Force on mild TBI prognosis. J Rehabil Med, 2004. PubMed 15083873.
• Iverson GL, McCracken LM. Brain Injury, 1997. PubMed 9354255.
• Mittenberg W and colleagues. J Clin Exp Neuropsychol, 2002. PubMed 12650234.
• Annegers JF and colleagues. Seizures after TBI. NEJM, 1998. PubMed 9414327.
• Christensen J and colleagues. Epilepsy after TBI. Lancet, 2009. PubMed 19233461.
• Fann JR and colleagues. Dementia after TBI in Denmark. Lancet Psychiatry, 2018. PubMed 29653873.
• Barnes DE and colleagues. Mild TBI and dementia in veterans. JAMA Neurology, 2018. PubMed 29801145; full text.
• Katz DI and colleagues. NINDS TES criteria. Neurology, 2021. PubMed 33722990.
• Silverberg ND and colleagues. ACRM mild TBI criteria. Arch Phys Med Rehabil, 2023. PubMed 37211140.

What we left out

National incidence figures, because the primary agency pages could not be retrieved for checking. Cumulative seizure probabilities by year, because the source abstract reports ratios only. Population-attributable dementia figures, for the same reason. Case law under the amended rule, which varies by circuit.

Related guides

• TBI imaging, diffuse axonal injury and DTI
• Eggshell plaintiffs and pre-existing conditions
• Blunt force trauma medical records
• Child abuse medical records
• How to get medical records for a lawsuit
• Altered medical records
• Is AI accurate enough for court