Will AI replace lawyers?

No. AI automates document-heavy work like research and review, but lawyers remain essential for judgment, strategy, and client relationships.

How accurate is AI at reviewing legal documents?

One industry study found AI reviewed legal documents with about 94% accuracy versus roughly 85% for human reviewers, and far faster.

How to Compare Medical Records With EUO & Deposition Testimony

All Posts

Insights

How to Compare Medical Records With EUO & Deposition Testimony

Last Updated:

Oct 6, 2026

Medical records and deposition transcripts almost never tell the exact same story.

You’ll have a plaintiff swear under oath they’ve never had back issues, only for you to dig up an old urgent care note treating a lumbar strain.

But pinning down a usable contradiction between the two is maddening. When you spot a mismatch, you're immediately stuck trying to figure out what actually happened.

Did the deponent genuinely forget a random 15-minute clinic visit from three years ago? Did a rushed doctor just copy-paste an inaccurate chart note? Or is it simply a phrasing issue, where the witness testified about "lower back pain" but the triage nurse wrote down "sciatica"?

If you try to weaponize every little discrepancy, opposing counsel might explain them away and make you look desperate. This simple guide is meant to help you sift through the noise, map the testimony against the chart, and figure out if what you've found is an actual contradiction in a medical deposition (or not).

Let’s begin.

TL;DR

To spot and confirm contradictions between medical records and depositions:

Build a medical chronology: Compare testimony and medical records side by side across identical timeframes and issues.
Verify original sources: Double-check original records and transcripts directly to confirm the conflict isn't just an indexing or transcription error.
Account for qualifiers: Retain specific phrasing, timeline qualifiers, and hedge words used by the witness (e.g., "to the best of my knowledge").
Investigate explanations: Rule out reasonable explanations like distinct injuries, misremembered dates, or pre-existing condition flare-ups.
Document the gap: Formally log confirmed conflicts alongside exact transcript page-and-line citations and medical record page numbers.

What counts as a contradiction between medical records and deposition testimony?

A contradiction exists when deposition testimony and a medical record make statements about the same fact that cannot both be true.

Here is a simple case.

A witness testifies, "I never saw anyone for my knee before the fall." A clinic note dated one year before the fall records a visit for right knee pain and a cortisone injection. If the note belongs to the witness and describes the same knee, both statements cannot be accurate.

However, before you call any difference a contradiction, these things have to match:

The record belongs to the witness and not to someone with a similar name.
Both sources describe the same body part or the same condition.
The record falls inside the time frame the testimony covers.
The question the witness answered actually covers what the record shows.

So where should you look first?

What types of contradictions show up between medical records and depositions?

When the chart and the transcript split apart, it usually happens in a few predictable places. Here are the most common types of contradictions you'll run into, along with a quick way to check each.

Prior injuries and treatment

The witness denies or downplays a problem the records show existed before the incident.

Someone testifies their neck was fine, but a chiropractor treated that person for neck pain 18 months earlier.
A man says he never hurt his knee, but an urgent care note shows a sprain two years ago.

Check first: Reread the question. "Ever" and "in the year before the accident" cover very different periods.

Symptom onset

The witness's account of when a symptom began doesn't match the first time it appears in the chart.

Someone says her headaches started the day of the crash, but the first record of them is five weeks later.
Another person says his back pain began at work, but an earlier note records back pain after moving furniture.

Check first: A symptom can start long before a doctor writes it down. It only conflicts if a note from that period records "denies headache" or something similar.

How the injury happened

The witness describes the accident one way, and the intake or triage note describes it another way.

A person testifies she slipped on a wet floor, but the triage note says she "tripped over a box."
They say they were rear-ended at a stoplight, but the ER note says "patient hit a parked car."

Check first: Find out who gave the history. A paramedic or family member may have supplied it, and nurses often paraphrase.

Treatment and follow-through

The witness's description of their care doesn't match what the providers recorded.

A woman says she went to every therapy session, but the clinic logged five missed visits.
A man says he took his pain medication daily, but the pharmacy shows one refill in three months.

Check first: Make sure the file is complete. Request the full records and billing ledger before treating a gap as a conflict.

Physical limitations

The witness claims limits that other records don't support.

A woman can't lift a gallon of milk, but a therapy note says she lifted bags of mulch that weekend.
A man can't stand for an hour, but his time sheets show full eight-hour shifts that month.

Check first: Compare the same time period, and weigh one-time effort against repeated activity. Lifting something heavy once is different from doing it every day at work.

The same five types apply to medical experts. If an expert's causation opinion assumes the patient had no earlier neck complaints, pre-accident records showing neck pain undercut the facts behind it.

Now let's look at how to check each one without overstating what you find.

How do you compare medical records with deposition testimony?

You compare medical records with deposition testimony by matching the exact testimony against the relevant medical records, using the case timeline to keep both in context.

Here’s the full process, step-by-step:

Step 1: Confirm you have the right patient and the right documents

Before you compare anything, make sure the records belong to the correct patient and cover the dates you need. Check identifying details and provider names against each production.

Use the correct transcript version and review any errata and exhibits. Look for medical-record amendments as well, and keep each correction next to the original wording so you can see what changed.

Next, note any gaps in the file. If a witness describes six therapy visits and your file includes only four notes, you have a records question to chase down.

Step 2: Build a medical chronology around the disputed issue

A medical chronology lists every relevant medical event in date order. It lets you see exactly where testimony and records line up and where they split apart.

Each entry should include:

The date of service and the provider
The relevant finding or statement
The source document and page number

Include care from before the incident whenever prior symptoms or aggravation of an old injury are part of the dispute.

Watch out for dates that look alike but mean different things. The date of service is the day the visit happened. The note may have been signed or entered days or weeks later. Suppose a witness testifies that she stopped chiropractic care on March 15. A note for a March 2 visit that the chiropractor signed on March 20 could make it look like she kept going. Log the date of service, and the false conflict disappears.

The same applies to symptoms. The day a symptom began and the day it first appears in a record are often different dates.

Step 3: Record the deposition question along with the answer

A deposition answer only covers what the question asked. If you record the answer without the question, your summary can claim more than the witness actually said.

Consider this exchange:

Q: Did you have physical therapy for your neck before the accident?

A: No.

Now suppose the records show chiropractic adjustments for the neck before the accident. It is tempting to write "Witness denied prior neck treatment." But the witness was asked only about physical therapy, and chiropractic care is a different kind of treatment.

The answer may be completely accurate. What you have found is a gap in the questioning, and it may call for a follow-up question at the next deposition.

To avoid this problem:

Write down the exact question and answer, with page and line numbers (for example, Dep. 42:7-15).
Keep every qualifier the witness used, like "as far as I remember," "after the accident," or "for that shoulder."
Read a few pages before and after the answer, since the witness may have clarified or corrected it.

Step 4: Find the medical records that address the same claim

Search the records for the same issue the testimony covers, using every relevant term and variation. Neck pain, for example, may appear in a chart as a cervical complaint. When the question involves treatment, check consultations and referrals separately, and review procedures as a separate category.

You also have to verify medical terms instead of assuming two labels mean the same thing. Search results show you where the evidence is, and interpreting it is still your job.

If an expert says the patient had no earlier complaints, check the available pre-incident history before you assess the conclusion built on that claim.

Step 5: Compare timing, scope, and source

With the testimony and the matching record in front of you, confirm they describe the same thing. A difference only matters if both sources cover the same body part and the same event during the same time period.

Next, look closely at what the witness meant. A witness who testifies she was "unable to work" may mean she could not return to her warehouse job. A note showing she worked part-time at a front desk may not conflict with that testimony at all.

Then identify the source of each statement in the record. Medical records contain two kinds of evidence:

Patient statements: A note that says "patient denies numbness" records what the patient told the doctor.
Clinical findings: An exam showing "sensation intact" records what the doctor observed.

If the witness testifies that her hand has been numb since the accident, a "denies numbness" entry from that period records her words and creates a direct conflict. A normal sensation exam conflicts less directly, because numbness can come and go.

Step 6: Check the original documents and rule out innocent explanations

Many apparent conflicts come from your working copies. A summary may shorten a quote and change its meaning, and scanning software (OCR) can misread a word or a number. Providers also copy text from old notes into new ones, which carries outdated information forward. A copied note might say "no joint pain" during a visit about joint pain.

Read around the entry, too. A nearby note may explain the change, show that the patient’s condition improved, or reveal that the record itself was corrected.

You also have to consider if the note is wrong. In a 2020 JAMIA study, researchers secretly recorded doctor visits by actors posing as patients and compared each recording with the doctor's note.

Of 105 notes, 90% had at least one error, and some recorded things that never happened. The study was small, but it shows why a witness who says "I never told the doctor that" deserves a careful check.

And if deciding whether two statements truly conflict requires medical judgment, bring in a qualified medical reviewer.

Step 7: Document the finding and decide your next move

If your checks confirm an inconsistency, record both passages with their citations. Note any context that might explain the difference. Then give the finding a clear status:

Potential conflict: You found a mismatch but haven't finished checking it.
Explained difference: You checked it and found an ordinary explanation.
Verified inconsistency: The mismatch survived every check in Steps 1 through 6.

Next, decide what the finding calls for. Some findings call for a missing record or a clinical review. Others call for a follow-up question that counsel can use.

Spend your time on findings that could affect causation or damages, including findings about treatment and physical limits. A dramatic inconsistency matters less when it has little to do with the issue in dispute.

The example below shows what this looks like in practice.

Example: The witness denies earlier treatment

Here is how the steps come together on one finding. This example is fictional.

A witness testifies that he never received treatment for neck pain before the collision. An orthopedic clinic note from 14 months earlier records a neck-pain visit and a trigger-point injection that same day.

Step 1: The name, date of birth, and medical record number on the note match the witness.
Step 2: The visit date falls 14 months before the collision, inside the period the question covers.
Step 3: The question was "Before the accident, did you ever get treatment for your neck?" The answer was "No," with no qualifiers, and the witness did not correct it later.
Step 4: The note describes "cervicalgia," which means neck pain.
Step 5: Both sources describe the same body part, and the note records a treatment the clinic performed.
Step 6: The original note, along with a matching billing entry, confirms the clinic performed the injection.

The finding survives every check, so you can record it as a verified inconsistency:

"The witness denied any pre-collision treatment for neck pain (Dep. 38:4-9). The orthopedic clinic record documents a neck-pain visit and trigger-point injection 14 months before the collision (Ortho Records, p. 112)."

This finding may justify follow-up questions. It does not, by itself, answer whether the collision made a pre-existing condition worse. That question still requires the full medical history and a clinical opinion.

You compare medical records with deposition testimony by matching the exact testimony against the relevant medical records, using the case timeline to keep both in context.

Here’s the full process, step-by-step:

Step 1: Confirm you have the right patient and the right documents

Before you compare anything, make sure the records belong to the correct patient and cover the dates you need. Check identifying details and provider names against each production.

Use the correct transcript version and review any errata and exhibits. Look for medical-record amendments as well, and keep each correction next to the original wording so you can see what changed.

Next, note any gaps in the file. If a witness describes six therapy visits and your file includes only four notes, you have a records question to chase down.

Step 2: Build a medical chronology around the disputed issue

A medical chronology lists every relevant medical event in date order. It lets you see exactly where testimony and records line up and where they split apart.

Each entry should include:

The date of service and the provider
The relevant finding or statement
The source document and page number

Include care from before the incident whenever prior symptoms or aggravation of an old injury are part of the dispute.

The same applies to symptoms. The day a symptom began and the day it first appears in a record are often different dates.

Step 3: Record the deposition question along with the answer

A deposition answer only covers what the question asked. If you record the answer without the question, your summary can claim more than the witness actually said.

Consider this exchange:

Q: Did you have physical therapy for your neck before the accident?

A: No.

The answer may be completely accurate. What you have found is a gap in the questioning, and it may call for a follow-up question at the next deposition.

To avoid this problem:

Write down the exact question and answer, with page and line numbers (for example, Dep. 42:7-15).
Keep every qualifier the witness used, like "as far as I remember," "after the accident," or "for that shoulder."
Read a few pages before and after the answer, since the witness may have clarified or corrected it.

Step 4: Find the medical records that address the same claim

You also have to verify medical terms instead of assuming two labels mean the same thing. Search results show you where the evidence is, and interpreting it is still your job.

If an expert says the patient had no earlier complaints, check the available pre-incident history before you assess the conclusion built on that claim.

Step 5: Compare timing, scope, and source

Then identify the source of each statement in the record. Medical records contain two kinds of evidence:

Patient statements: A note that says "patient denies numbness" records what the patient told the doctor.
Clinical findings: An exam showing "sensation intact" records what the doctor observed.

Step 6: Check the original documents and rule out innocent explanations

Read around the entry, too. A nearby note may explain the change, show that the patient’s condition improved, or reveal that the record itself was corrected.

You also have to consider if the note is wrong. In a 2020 JAMIA study, researchers secretly recorded doctor visits by actors posing as patients and compared each recording with the doctor's note.

And if deciding whether two statements truly conflict requires medical judgment, bring in a qualified medical reviewer.

Step 7: Document the finding and decide your next move

If your checks confirm an inconsistency, record both passages with their citations. Note any context that might explain the difference. Then give the finding a clear status:

Potential conflict: You found a mismatch but haven't finished checking it.
Explained difference: You checked it and found an ordinary explanation.
Verified inconsistency: The mismatch survived every check in Steps 1 through 6.

Next, decide what the finding calls for. Some findings call for a missing record or a clinical review. Others call for a follow-up question that counsel can use.

The example below shows what this looks like in practice.

Example: The witness denies earlier treatment

Here is how the steps come together on one finding. This example is fictional.

Step 1: The name, date of birth, and medical record number on the note match the witness.
Step 2: The visit date falls 14 months before the collision, inside the period the question covers.
Step 3: The question was "Before the accident, did you ever get treatment for your neck?" The answer was "No," with no qualifiers, and the witness did not correct it later.
Step 4: The note describes "cervicalgia," which means neck pain.
Step 5: Both sources describe the same body part, and the note records a treatment the clinic performed.
Step 6: The original note, along with a matching billing entry, confirms the clinic performed the injection.

The finding survives every check, so you can record it as a verified inconsistency:

How should you document contradictions in a comparison table?

Litigation teams usually track contradiction findings in an issue chart, sometimes called a contradiction chart. It is a table with one row per disputed fact, placing the testimony and the matching medical record side by side. Extra fields for context and review status let any teammate follow your reasoning and pick up the next step.

Field	What to record
Disputed issue	The point being tested
Testimony	Exact Q&A with page and line citations
Medical record	Relevant entry with date and source
Context	Qualifications, amendments, or missing records
Status	Potential, explained, or verified conflict
Next step	Records request, clinical review, or follow-up question

Missing information also deserves its own place in the analysis. A 2024 systematic review found some form of incompleteness in 47 of the 48 studies it reviewed. That does not mean every medical record is incomplete, but it is a useful reminder that silence in the file should not automatically be treated as proof.

Once the finding is verified and properly sourced, you can carry it into the next examination with much more confidence.

Field	What to record
Disputed issue	The point being tested
Testimony	Exact Q&A with page and line citations
Medical record	Relevant entry with date and source
Context	Qualifications, amendments, or missing records
Status	Potential, explained, or verified conflict
Next step	Records request, clinical review, or follow-up question

Once the finding is verified and properly sourced, you can carry it into the next examination with much more confidence.

How do you turn medical record contradictions into deposition follow-up questions?

Turn each verified finding into a focused question that clarifies one specific fact. Most follow-up questions aim to clarify the scope of an answer or identify who supplied a history. Others ask directly about an earlier treatment event.

For example:

“When you said you never received treatment, did you mean therapy or any neck treatment at all?”
“What care did you receive at the visit documented on this date?”
“Did you give this history to the clinician yourself?”

Keep each question tied to one factual point. That gives counsel a clean issue to work with.

Handling all of this by hand works for a small file. It gets much harder as the record grows.

How can AI software help compare medical records with depositions?

A single personal injury claim stacks up fast because there are a dozen medical providers, months of billing, and hundreds of pages of deposition testimony.

At that scale, manual review means checking each answer against a dozen open documents tabs on their screen, completely missing the line buried among hundreds of pages that contradicts it.

This is the kind of problem InFactIQ solves.

InFactIQ is an AI claims and litigation intelligence platform that reads the entire case file together and flags the points where testimony disagrees with the records and bills. Each finding links back to its source page, so your team can check the underlying record before deciding whether it is a genuine inconsistency.

For this type of review, InFactIQ can:

Compare testimony with medical records and bills: It flags conflicts across the file and can show deposition or Examination Under Oath (EUO) testimony alongside the medical record it conflicts with.

Keep the chronology connected: Its persistent timeline brings events from across the case into one view and can be filtered by date, record type, or billing incident.

Carry those findings into the examination: Before a deposition or EUO, it can draft an outline from the case file and the team’s playbook. It also tracks contradictions, admissions, corroborations, and unanswered lines as testimony develops.

Recheck the testimony afterward: Once the examination ends, InFactIQ summarizes the testimony and compares it again with the claim file and demand package.

Apply your team’s existing review process: Playbooks can turn existing templates or standards into cited work products such as chronologies, issue charts, and deposition outlines.

The software does not decide whether a difference is meaningful. Your team still needs to check the source, consider the context, and decide whether the finding deserves follow-up. What AI changes is how much of the file you can cross-check before you get to that judgment.

Turn every contradiction into a question you can support

Think back to that old back-pain note from the beginning. By itself, it's just one line in a chart. It becomes valuable once you find the testimony it contradicts and verify the treatment it describes. From there, you account for the context and shape a question that both sources back up.

Before you carry a contradiction forward, make sure another reader can follow the evidence to the same point and clearly see what remains unresolved.

Book an InFactIQ demo to see how your team can review testimony and medical records together on your next case.

FAQs

Does a missing symptom in a medical record contradict later deposition testimony?

Not automatically. The note may be incomplete or focused on a different complaint, or it may come from a partial production. Compare explicit statements about symptoms and check the surrounding records.

Can AI find every contradiction between medical records and testimony?

No review process can promise that. Missing documents and poor scans can limit what any tool finds, and so can ambiguous wording or clinical context. Use AI to surface passages worth comparing, then verify the sources and judge what they mean.

Does an inconsistency prove that a witness is lying?

No. Memory lapses and simple misunderstandings can produce differences between testimony and records. Documentation errors and changes in a condition over time can do the same.

Contents