Guides › How legal nurse consultants build a medical chronology, step by step

How legal nurse consultants build a medical chronology, step by step

Legal nurse consultants build a medical chronology in 9 steps: scope the question, inventory the records, prep the pages, extract cited rows, flag gaps and conflicts, add clinical analysis, run a cite check, deliver, and supplement. Software can draft steps 2 to 5 in minutes. Steps 1, 6 and 7 stay with the nurse.

By Ahmed Jemaa, Co-Founder & CEO of Medrecords AI · Published 24 Oct 2026 · Updated 24 Oct 2026Summarize with ChatGPT, Claude, Perplexity or Google AI

How do legal nurse consultants build a medical chronology?

An LNC builds a medical chronology by working from the attorney's legal question back to the records: confirm what the case is about, list what records exist and which are missing, then build dated, cited rows and add clinical commentary on what each event means for the standard of care, causation or damages.

That is what separates an LNC chronology from a paralegal's date list. The rows are the evidence the attorney will put in front of an expert or a deponent; the commentary is the work product. The workflow below is our planning model for a 2,000-page personal injury file, with the time and the output each step should produce.

The manual times are planning estimates, not measurements. Steps 1 to 8 add up to about 28 to 49 hours by hand (0.5 + 1 + 2 + 15 + 2 + 4 + 2 + 1 at the low end; 1 + 2 + 4 + 25 + 4 + 8 + 3 + 2 at the high end), before supplements. 2 captured sources point the same way. RRR Health Tech, a chronology service, says a 600-page PI file usually takes 4 to 8 hours to chronologize (captured 2026-10-23), which is 75 to 150 pages an hour, or about 13 to 27 hours for 2,000 pages. DigitalOwl, a vendor, says a paralegal typically spends 40 to 60 hours on a 5,000-page PI chronology (vendor claim, captured 2026-10-23).

Step 1: scope the question before you open the records

Scoping takes 30 to 60 minutes and saves the most time of any step. Ask the attorney 4 things: what the claim is, what dates matter, what the opposing theory is likely to be, and what the deliverable is. Write the answers down and send them back for confirmation.

A med mal case about a missed sepsis diagnosis needs minute-level rows for 1 admission and a summary of everything else. A PI case about a disc injury needs every treatment visit, every gap over 30 days, and the prior history of the same body part. Same records, different chronology. Without a scope, LNCs build the second kind for the first case and lose a day.

Output: a 1-page scope with the issues, key dates, providers in and out of scope, and the format the attorney wants. If the attorney has no format, our medical chronology template is a starting point.

Step 2: what records do you have, and what is missing?

Inventory every provider, date range and page count before extracting anything. The record index tells the attorney what you have. The missing-records list tells them what to request next, and in many cases it is the most useful thing you deliver in week 1.

Read for references: a referral to an orthopedist whose notes never arrived, an MRI ordered with no report, a 5-day admission with nursing notes for 3. Each missing item should cite the page that implies it exists. Medrecords AI drafts this list with missing records identification, but the LNC decides which gaps matter and which records probably never existed.

Output: a record index by provider and date range, and a missing-records list with the citing page for each item.

Step 3: prep the pages

Prep means removing duplicates, making scanned pages searchable, applying Bates numbers and pulling out pages that belong to another patient. By hand it is 2 to 4 hours of tedious work. With software it takes minutes, and your job becomes checking what was removed.

Duplicates are the big one. The same ED note can arrive in the hospital chart, the billing packet and the subpoena response. Medrecords AI removes exact duplicates before building the timeline, logs the removals, and does not bill for them. A wrong-patient page should be quarantined, not deleted, so you can show the attorney what was set aside and why. In the Medrecords AI demo case (#IME-4812, 342 pages in 2 packets), 11 duplicates were removed and 1 wrong-patient page (p.140) was quarantined.

Output: a clean, deduplicated, Bates-numbered set, plus a log of removed and quarantined pages. See Bates numbering and deduplication.

Step 4: extract the rows

Extraction is the step that used to take most of the hours: 1 row per relevant event, with date and time, provider, source page and finding. Software now drafts it. Medrecords AI reads about 100 pages a minute, so a 2,000-page file produces a draft chronology in about 20 minutes, and every row cites its source page.

Compare that with a manual pace. At RRR Health Tech's stated 75 to 150 pages an hour, the same 2,000-page file is 13 to 27 hours of extraction. What changes is the LNC's role. You are no longer typing rows; you are reviewing them. That review should be targeted. Read every row in the scoped date range closely. Skim the rest for anything that contradicts the theory of the case.

The 5 fields every row needs

1. Date and time of service, not the date the document was printed.
2. Provider and facility.
3. Source page, as a Bates number or page reference that opens the page.
4. Finding, in the record's own clinical terms.
5. Significance, left blank in the draft and filled in at step 6.

The first 4 are facts the software can draft. The 5th is clinical judgment and belongs to you.

Step 5: flag gaps, conflicts and late entries

Flagging means marking what the attorney needs to know before reading the rows: gaps in treatment, conflicting entries, undated or unsigned pages, late-entry and addendum labels, and pre-existing conditions. Software drafts the flags with their pages. The LNC decides which ones matter.

Hospital entries must be dated, timed and authenticated under 42 CFR 482.24(c)(1), so an undated or unsigned entry is a flag with its page, not a gap to fill in. A 47-day gap in physical therapy is a fact. Whether it hurts the damages case, reflects an insurance denial, or means the client got better is an analysis question. Write the flag as a fact, then answer the question in step 6. For med mal files, late entries and charting gaps carry more weight; our guide to AI chronologies for med mal cases covers what software can and cannot see in an EHR audit trail.

Output: a flag list, each with its page citation and a 1-line note on why it may matter.

Step 6: add the clinical analysis

Analysis is the step that makes the chronology an LNC work product. Add significance to each key row, note where care departed from what you would expect, connect symptoms to treatment over time, and mark where an expert opinion is needed. No software should do this step for you, and Medrecords AI does not.

Good analysis is short and tied to the row. Here is an illustrative row, not from a real file: "Lactate 4.1 resulted 21:52, not acknowledged until 00:30 (p.412). Sepsis bundle timing is a question for the ED expert." That is more useful to an attorney than a paragraph. Leave standard-of-care conclusions to the testifying expert, and keep your notes clearly marked as the nurse's commentary.

Output: the commented chronology, plus a short list of questions for the expert. Our guide to legal nurse consultant reports covers how the chronology feeds the narrative report.

Step 7: how do you cite-check a medical chronology?

The cite check is where you protect your name. Click every row the analysis relies on and confirm the page says what the row says, with the right date, time and provider. Then sample 10% of the remaining rows. Any error in the sample means a full pass.

Those key rows are what the expert will be crossed on, so a line with no page behind it is a problem the attorney inherits.

Software that cites every row makes this step about half as long, because you click instead of search. It does not make the step optional. AI does the reading and helps you make the call. The last word is yours.

Step 8: deliver the summary and the export

Deliver 3 things: a 1 to 2-page executive summary that answers the attorney's scope questions, the commented chronology, and the missing-records list. Export in a format the attorney and the expert can annotate, with citations that still open the page outside your software.

Medrecords AI exports hyperlinked Word and HTML files, so a citation in the exported file opens the page, with no account or plugin on the reader's side. Attorneys rarely read every row; they read the summary, then jump to the rows behind each point. Write the summary so each claim points to a row number.

Output: executive summary, commented chronology, missing-records list.

Step 9: supplement when new records arrive

Records keep arriving after delivery. Add each batch, re-run deduplication against what you already have, and mark new rows so the attorney sees what changed. Software makes this a minutes-long job per batch. Re-check any analysis the new records touch.

Output: an updated chronology with a change log by date received. See supplemental record review.

Where does AI fit, and what does the LNC still own?

AI fits the high-volume, low-judgment steps: inventory, prep, first-pass rows and draft flags. The LNC owns scope, clinical significance, the cite check and anything that ends up in an opinion. That split is how LNCs take more files without lowering their standard.

The billing question follows. If extraction used to be about half your hours (15 to 25 of 28 to 49 in the table) and now takes 20 minutes plus review, hourly billing for page-turning stops making sense. An LNC can move to a per-file or per-deliverable price and pass the software cost through.

A worked per-deliverable price

Take the 2,000-page file. The software line is 2,000 x $0.10 = $200 on Medrecords AI Self-Service. The list rate is 10¢, with 5¢ per page at volume. Your time with software drafting runs about 7 to 13.5 hours: scope 0.5 to 1, review of drafted steps 2 to 5 about 1 to 1.5, analysis 4 to 8, cite check 1 to 2, delivery 0.5 to 1. At $125 to $150 an hour, the rate 21 of 23 independent consultants state in our AI-native legal nurse consultant guide, that is $875 to $2,025 of nurse time plus $200 of software, or $1,075 to $2,225 for the deliverable. Quote the attorney 1 number for the scope, and the attorney knows the cost before the records arrive. The same guide covers per-case pricing and positioning in more detail.

Which setup fits which LNC practice?

Solo LNCs get the most from software that drafts and cites rows they can edit, priced per page so a slow month costs nothing. LNC firms with several nurses need per-matter access and an activity log. Nurses who prefer to build every row by hand should pick a tool built for manual entry.

• Solo LNC, mixed PI and med mal: draft in software, keep steps 1, 6 and 7 yours, price per deliverable. Start at Medrecords AI for legal nurse consultants.
• LNC firm with 3 or more nurses: per-matter roles, MFA and exportable activity logs matter as much as drafting speed.
• LNC who wants full manual control: highlight-to-create tools fit better. Casefleet, for example, says you highlight text in the records to create entries (captured 2026-10-23). Our list of medical chronology software for attorneys compares both kinds.
• Format questions first: if the attorney's format is the open question, read how to write a medical chronology before building anything.

Bring us your ugliest case. Book a demo, then run your next file and compare it with your last manual chronology, row by row. First case is free on us. See pricing.

Frequently asked questions

Sources and method

Vendor prices and features below come from public pages captured on the dates shown. Vendors change pricing often, so confirm before you buy.

• DigitalOwl, AI medical record summaries for personal injury law firms (captured 2026-10-23)
• RRR Health Tech, medical record review cost for attorneys (600 pages in 4 to 8 hours; captured 2026-10-23)
• Casefleet, medical chronology software (captured 2026-10-23)
• 42 CFR 482.24, Condition of participation: medical record services (eCFR, read 2026-10-24)
• Medrecords AI, hyperlinked exports (Word and HTML, read 2026-10-24)
• Medrecords AI, legal nurse consultant reports guide
• Medrecords AI, medical chronology template
• Medrecords AI, pricing (read 2026-10-24)

Related reading