This page is for the person who actually builds the chronology: the paralegal handed a banker's box of records on a Friday, or the legal nurse consultant who bills for the work and has to defend it later. It gives you the column layout, a filled-in sample built from a hypothetical record, the order of operations for doing it by hand, the four ways chronologies most often go wrong, and a template you can download and start typing into. It does not assume you have any software. Where software helps, it says so, and where it does not, it says that too.
A medical chronology is a table. Each row is one event in a patient's care, taken from the medical records, placed in strict date and time order across every provider who touched the patient, and cited to the page of the record it came from. That last part is what separates a chronology from a summary. A summary tells the reader what happened. A chronology shows the reader where to look.
The people who rely on chronologies rarely build them, which is why the format matters more than it seems. The reviewing attorney reads it to decide whether the case has a theory and to plan depositions. The retained expert reads it before opening the records, and often bills fewer hours because of it. The claims professional or defense counsel reads it to test the plaintiff's timeline against the chart. The mediator reads it because nobody at the table will read 3,000 pages. In each case the reader wants the same thing: to be able to pick any row, open the record to the cited page, and see the sentence that supports it.
Chronologies are used in medical malpractice, personal injury, workers' compensation, long-term disability, nursing home and product liability matters, on both the plaintiff and the defense side. The format does not change with the side. What changes is which rows get flagged, and that belongs in the last two columns, where it can be separated from the facts.
Chronologies get attacked in two ways. Opposing counsel asks the author, or the expert relying on it, "Where in the record does it say that?" and "Is that what the record says, or what you think it means?" A column layout that answers both questions on every row makes the chronology defensible. A layout that mixes fact with commentary, or that cites a document without a page, does not.
These eight columns answer both questions. The first six are facts from the record. The last two are the reviewer's analysis, kept separate on purpose.
| Column | What goes in it | Why it survives cross-examination |
|---|---|---|
| 1. Date/time | Date of service, and the clock time whenever the record has one. Use one format throughout (MM/DD/YYYY 24:00 is the least ambiguous). When a document is dated differently from the event it describes (a late entry, a dictated report), record the event time here and put the authoring time in the Event or Flag column. | Sequence is the whole point. A time that cannot be traced to the record, or a date format that sorts wrong, breaks the sequence. |
| 2. Provider | The person who did or documented the thing: name and credential as they appear in the record (RN, PA-C, MD, resident). If the author is unclear, write "unsigned" rather than guessing. | The question at deposition is who knew what, when. "The hospital" did not know anything; a named nurse at 21:20 did. |
| 3. Facility / department | Where the care happened, down to the unit when it matters (ED, ICU, OR, PACU, outpatient clinic, lab, radiology, EMS). | Handoffs between departments are where cases live. You cannot see a handoff failure if the column is missing. |
| 4. Event | What happened, stated the way the record states it: complaint, exam finding, order, result with reference range, medication given, consult, procedure, discharge. Objective, one event per row, no adjectives. | This is the column the expert relies on. If it contains opinion, the expert's opinion inherits the reviewer's opinion, and both are attacked together. |
| 5. Source document (Bates page) | Document type and the Bates number of the page. Not "ED records"; "ED Provider Note, SAMPLE-000015." For a multi-page document, cite the page the fact is on. | An uncited row is an assertion. A cited row is evidence anyone can check in ten seconds. |
| 6. Quote | The exact words from the page, short, in quotation marks, including abbreviations and errors as written. Use "[sic]" sparingly and "..." for omissions. | A quote settles the "is that what it says" question before it is asked. It also preserves language the record uses that a paraphrase would lose. |
| 7. Significance | Why the row is in the chronology: an interval, a value against its reference range, a result with no documented action, a contradiction with another row. Plain, factual, and clearly the reviewer's note. | Because it is in its own column, a reader can ignore it, redact it, or disagree with it without touching the facts. |
| 8. Flag | A short controlled label so the table can be sorted and filtered: BASELINE, NOTIFY, DELAY, GAP, LATE ENTRY, CONFLICT, OUTCOME, DUPLICATE, or blank. Agree on the list before you start and do not add to it mid-case. | Flags let the attorney pull every DELAY row into a deposition outline in one filter. Free-text notes do not. |
Two columns people sometimes add, and why they are optional: a "record type" column is useful for filtering (nursing note, lab, imaging) but it duplicates the front half of the Source column, so it is left out here; and a "page count" or "document range" column belongs in the records inventory, not the chronology.
| Date/time | Provider | Facility / dept. | Event | Source (Bates page) | Quote | Significance | Flag |
|---|---|---|---|---|---|---|---|
| 04/02/2025 19:48 | M. Ortiz, RN | Riverside Community Hospital, ED triage | Triage. Abdominal pain 2 days, chills. T 38.9 C, HR 118, BP 102/64, RR 22. ESI 3 assigned. | ED Triage Note, SAMPLE-000012 | "Pt reports RLQ pain x2 days, chills. ESI 3." | Arrival vitals. Temperature and heart rate both abnormal at presentation. | BASELINE |
| 04/02/2025 20:35 | K. Lund, PA-C | Riverside ED | Initial provider exam. RLQ tenderness, no rebound. Orders: CBC, BMP, lactate, CT abdomen/pelvis with contrast. | ED Provider Note, SAMPLE-000015 | "Abd soft, RLQ tenderness, no rebound. Will obtain labs and CT." | First provider contact 47 minutes after triage. | |
| 04/02/2025 21:12 | Laboratory (analyzer result); called to M. Ortiz, RN | Riverside, clinical lab | CBC and lactate resulted. WBC 17.4 (ref 4.0–11.0). Lactate 3.1 mmol/L (ref 0.5–2.0), critical. | Lab Report, SAMPLE-000041 | "LACTATE 3.1 H CRITICAL. Called to RN Ortiz 2120, read back." | Critical value documented as called to the nurse at 21:20. No documented notification of the PA or attending. | NOTIFY |
| 04/02/2025 22:05 | R. Feld, MD (radiology) | Riverside, radiology | CT abdomen/pelvis final read. | Radiology Report, SAMPLE-000058 | "Acute appendicitis with periappendiceal fluid. Early perforation cannot be excluded." | Result final at 22:05. Next documented provider action is 23:50 (row 6). | DELAY |
| 04/02/2025 22:05 to 23:50 | None documented | Riverside ED | No provider or nursing entries for 1 hour 45 minutes after the CT read. | Absence: SAMPLE-000015 to 000020 (ED notes); SAMPLE-000080 to 000083 (nursing flowsheet) | (none) | Interval between a CT read suggesting perforation and the first documented response. Confirm against the audit trail before relying on it. | GAP |
| 04/02/2025 23:50 | K. Lund, PA-C | Riverside ED | Provider reassessment. Surgery consult paged. No antibiotic order at this time. | ED Provider Note (addendum), SAMPLE-000016 | "CT c/w appendicitis. Surgery paged 2350." | First documented action on the CT (1 h 45 min) and on the critical lactate (2 h 38 min). | DELAY |
| 04/03/2025 00:42 | A. Nwosu, MD (surgery resident) | Riverside, general surgery consult | Consult. Recommends IV antibiotics now and operative management in the morning. | Consult Note, SAMPLE-000071 | "Rec: pip/tazo now, NPO, OR in AM. Discussed with Dr. Halvorsen." | Antibiotic recommended at 00:42. See row 8 for the administration time. | |
| 04/03/2025 01:31 | M. Ortiz, RN | Riverside ED | Piperacillin-tazobactam 3.375 g IV, first dose administered. | Medication Administration Record, SAMPLE-000094 | "PIP/TAZO 3.375 G IV. Given 0131. ORTIZ M." | First antibiotic 5 h 43 min after arrival, 4 h 19 min after the critical lactate, 49 min after the recommendation. | DELAY |
| 04/03/2025 06:15 | P. Halvorsen, MD (attending surgeon) | Riverside, OR 2 | Laparoscopic appendectomy. Perforated appendix with purulent fluid in the right lower quadrant. Washout performed. | Operative Report, SAMPLE-000112 | "Perforated appendix with gross contamination of the RLQ." | Perforation confirmed intraoperatively. | OUTCOME |
| 04/02/2025 21:30 (event as stated); authored 04/09/2025 09:02 | K. Lund, PA-C | Riverside ED | Late entry to the 04/02 provider note stating that the lactate was reviewed and discussed with the patient at 21:30. | ED Provider Note (late entry), SAMPLE-000017 | "LATE ENTRY 4/9/25 0902: lactate 3.1 reviewed 2130, discussed w/ pt, plan surg consult." | Authored seven days after the visit and after the operation. Not reflected in any contemporaneous entry. Compare against the audit trail. | LATE ENTRY |
Four things to copy from the sample. Times are given to the minute wherever the record has them, and intervals are computed in the Significance column rather than left for the reader. Lab values carry their reference range in the same cell. A gap is a row of its own (row 5), with a citation to the pages that would have contained an entry and do not. And the late entry (row 10) is placed at the time of the event it describes, with the authoring date shown beside it, so it sorts where the defense will say it belongs while making plain that it was written a week later.
The order matters more than any individual step. Most of the rework in chronology projects comes from doing step 4 before step 2, or step 5 as an afterthought.
Before reading a page, list every producing source (each hospital, practice, EMS agency, pharmacy, imaging center, lab, insurer) with the date range covered, the page count, and the date received. Then read the records you have for references to records you do not: a discharge summary that names a cardiology consult, a nursing note that mentions a transfer, a primary-care note that refers to an outside MRI. Each becomes a line on a "requested / outstanding" list. Do this first because a chronology built on an incomplete production will have gaps that are artifacts of the production, not of the care, and you will not be able to tell the two apart later. The inventory is also where the page counts and document ranges live, so they do not clutter the chronology.
Assign a stable page number to every page of every document before extraction starts. If the production is already stamped by the other side, keep their numbers; do not re-stamp, because every filing in the case will use theirs. If it is not stamped, stamp it now with a short prefix per source and a zero-padded sequence (SAMPLE-000001). Do not skip this because the set is small, and do not use PDF viewer page numbers, which change the moment a page is inserted. The Source column in the chronology is only as durable as the page numbers behind it.
Records arrive grouped by who produced them. Chronologies are read by when things happened. Before extracting, arrange the documents into one date-of-service sequence across all sources, keeping each document's Bates range intact. In practice this means building a document index (Bates start, Bates end, source, document type, date of service) and sorting that index, rather than physically reshuffling pages. Watch for documents whose printed date is the print date or the signature date rather than the service date; EHR exports do this often.
Read in the sorted order and create a row for each clinically or legally significant event. Significant means anything that changes what the next provider knew or should have done: a complaint, an abnormal or normal result, an order, a medication given or held, a consult, a change in condition, a notification, a procedure, a discharge instruction. Keep the Event column to what the record says. Do not summarize a day into one row; a day in an ICU can be forty rows, and the intervals between them are the point. Do not skip nursing flowsheets and medication administration records because they are voluminous; they hold the clock times that physician notes lack.
Fill the Source and Quote columns as you extract, not afterward. A row without a page citation is not finished, and going back to find the page later takes longer than writing it down the first time. The quote should be short and exact, including the record's abbreviations. If the same fact appears in several documents (the lab value in the lab report, the nursing note, and the discharge summary), cite the primary source, the one closest in time to the event, and note the others in Significance only if they disagree.
Only after the facts are down, go back through the table and fill Significance and Flag. Compute intervals between a result and the documented response, a recommendation and the action, a change in condition and the notification. Add a GAP row wherever a stretch of time has no entries and the patient's condition suggests there should be some, citing the pages that would have held them. Mark every late entry, addendum and amended note with LATE ENTRY and record the authoring time. Mark CONFLICT where two documents disagree about the same event. Keep the language in Significance factual ("no documented notification of a provider between 21:20 and 23:50"), and leave the conclusion ("this was below the standard of care") to the expert.
Have a second person, or yourself on a different day, open a random sample of rows and check each against the cited page: is the page right, is the quote exact, is the time the event time and not the print time. Sort the table by Date/time and look for rows that fall out of order, which usually means a date-format error or a late entry placed by authoring date. Check that every source in the inventory appears in the chronology at least once; a source with zero rows was either irrelevant or not read. Then reconcile the outstanding-records list: any gap that coincides with missing records gets an annotation saying so.
The free chronology demo builds this same table from a synthetic record, with every row linked to the page it came from. No login.
Open the demo →The same four problems account for most of the chronologies that get taken apart in deposition or quietly discarded by the expert.
An entry with no page citation, or with a citation to a document rather than a page, cannot be verified without re-reading the record, which defeats the purpose. It is also the first thing opposing counsel looks for. The usual cause is extracting first and planning to cite later. The fix is procedural: no row is complete without a Bates page, and the QA step samples for it. If a fact genuinely comes from outside the record (a deposition, a client interview), it does not belong in the chronology; put it in a separate table with its own source column.
Electronic records let a provider add to a note days or weeks later. The late entry is usually labeled, but not always prominently, and a reviewer reading a printed export may see only the final text. If the chronology places the late entry at the time it describes without noting when it was written, the table now asserts that the lactate was reviewed at 21:30, and the fact that this was written after the operation disappears. Every addendum, late entry and amendment needs its authoring time captured and a LATE ENTRY flag. When the timing of a note matters to the case, request the EHR audit trail; the audit-trail guide for LNCs explains what it contains, and the audit trail request builder drafts the request.
Productions contain the same page many times: a lab report in the lab section, again in the nursing section, again in the discharge packet, again in the records the specialist received. A chronology that creates a row for each copy inflates the event count and can turn one result into three apparent events on different dates (the date the lab resulted, the date it was faxed, the date it was scanned into the outside chart). Identify duplicates during the sort step, keep the primary source, and note in the inventory which Bates ranges are copies. A duplicate that differs from the original in content is not a duplicate; it is a CONFLICT row.
Not every clock in a record is the same clock. Laboratory analyzers, imaging equipment, cardiac monitors, EMS run sheets, the pharmacy system and the EHR itself may be minutes apart, and a facility near a timezone line, or a transfer between facilities, can be an hour apart. An EMS record in one timezone and an ED record in another can make a patient appear to arrive before leaving. A medication may show a scan time and a documented "given" time that differ. The chronology cannot fix the clocks, but it must not hide them. Record times as the source states them, name the source in the Provider or Source column when it is a device or system, and put the discrepancy in Significance. When an interval in the case is measured in minutes, say which two clocks it was measured between.
The template below is the eight-column layout from this page. The CSV opens in Excel, Google Sheets, Numbers or any spreadsheet and is the format most firms end up with. The printable version is for reviewing on paper or for a case where the records themselves are paper.
Tip: in a spreadsheet, format the Date/time column as text before pasting, so that "04/02/2025 19:48" is not silently converted, and freeze the header row. Keep one workbook per case and one sheet per chronology version.
| Date/time | Provider | Facility / dept. | Event | Source (Bates page) | Quote | Significance | Flag |
|---|---|---|---|---|---|---|---|
Flag values: BASELINE · NOTIFY · DELAY · GAP · LATE ENTRY · CONFLICT · OUTCOME · DUPLICATE. Facts in columns 1–6; reviewer's notes in 7–8.
The honest answer is that it scales with the page count and the condition of the records. A common planning figure among LNCs and paralegals is on the order of a hundred pages an hour for a fully cited chronology from clean electronic records, and materially slower for scanned paper, handwriting, or productions with heavy duplication. On that basis a 3,000-page record is several working days of extraction before the quality check, and a 10,000-page record is weeks. Whatever figure you use, it is a planning number for your records, not a promise, and the only way to calibrate it is to time your own first 500 pages.
Where the time goes is instructive: most of it is steps 3 through 5 (sorting, extracting, citing), which are mechanical, and comparatively little is step 6 (judgment), which is where the value is. That is the case for software. Extraction and sorting are what software does well; it reads every page at the same pace, keeps the citation attached to the fact, and does not get tired on page 2,400. What it does not do is decide which rows matter to this case, or whether a two-hour interval was reasonable. Those columns stay with you.
Three ways to use software with the template on this page, depending on what you need:
Whichever route you take, apply the same two tests you would apply to a chronology a colleague handed you. First, can you open the cited page for any row and find the quote. Second, does the number of dated events look right for the size of the record. In our own testing in September 2026, on a synthetic record of roughly 3,500 pages containing 3,533 dated entries, a single pass through a general-purpose AI model returned 74 events, while processing the same record in sized segments returned 1,946; the single-pass output looked complete and was not. A tidy table is not evidence that the whole chart was read. The timing write-up has the full numbers, and this piece on defensible chronologies covers what happens when the table is put to a witness.
Upload a record set, get the dated, page-cited chronology in this layout, then finish the judgment columns yourself. Free cases to start, no card.
Start free →A medical chronology is a date-ordered table of the clinically and legally significant events in a patient's medical records, with every row cited to the page of the record it came from. It is built from the records themselves, not from the patient's account, and it is used by attorneys, legal nurse consultants, paralegals, experts and claims professionals to see what was known, what was done, and when.
Eight: date and time, provider, facility or department, event, source document with the Bates page, a short quote from the record, significance, and a flag. The first six are facts from the record. Significance and flag are the reviewer's analysis and should stay in their own columns so the factual part of the chronology can be defended on its own.
It scales with page count and record quality. A common planning figure among LNCs and paralegals is on the order of 100 pages an hour for a fully cited chronology from clean electronic records, and slower for scanned, handwritten or duplicated productions. On that basis a 3,000-page record is several working days, before quality checking. Treat any per-page figure as a planning number for your own records, not a promise.
A chronology is a table, one row per event, each row cited to a page, in strict date order across all providers. A summary is narrative prose that condenses the care into paragraphs and may or may not cite pages. Experts and attorneys generally want the chronology first, because they can verify any row against the record; a summary is written from the chronology, not instead of it.
Yes, or a stable page-numbering scheme that everyone on the case will keep using. If the pages are stamped after the chronology is written, every citation has to be redone. Stamp first, then read. If the production arrives already stamped by the other side, use their numbers and do not re-stamp.
Software can do the extraction and sorting, and it does that part in minutes rather than days. Whether the result is usable in deposition depends on two things you can check: does every row carry a page citation you can open, and does the number of dated events look right for the size of the record. A chronology that cannot be traced to pages, or that returns a few dozen events from thousands of pages, is a draft to be finished, not a finished work product. The judgment columns, significance and flag, remain the reviewer's work either way.
Related reading: how paralegals organize medical records (Bates, sorting and the inventory in more detail), the case timeline template (the attorney-facing document that is written from the chronology), and the paralegal and legal nurse consultant pages.
Questions about the template? [email protected]