Free account 2:35Medical chronology
Every medical event on its date, in order, tied to the page it came from. Filter it, search it and export it as PDF or Word.
The emergency department records that prove or defeat the claim
An EMTALA violation is a Medicare-participating hospital's failure to screen, stabilize or properly transfer an emergency patient. Timestamped emergency department records prove or defeat the claim.
No private action: 42 USC 1395dd(d)(2)(A) runs against the hospital; Medrecords AI cites records.
An EMTALA violation is a Medicare-participating hospital's failure, when someone comes to its emergency department, to give an appropriate medical screening examination, to stabilize an emergency medical condition it finds, or to transfer an unstable patient only as the statute allows. Under 42 USC 1395dd(d)(2), a harmed patient can sue the hospital, not the physicians, within 2 years. Timestamped ED records prove or defeat the claim.
An EMTALA case is won or lost on a clock. The question is rarely whether the emergency department did good medicine; it is whether the hospital screened this patient the way it screens everyone with the same complaint, whether it stabilized what it found, and whether the transfer paperwork existed before the ambulance doors closed.
For plaintiff med-mal and PI attorneys, legal nurse consultants and ED nurse experts working an EMTALA claim. You walk away with the rules, the record set to request, a timestamp method, and 2 templates.
Congress passed the Emergency Medical Treatment and Labor Act in April 1986 as part of COBRA, aimed at "patient dumping": hospitals turning away or shipping out emergency patients who could not pay. It is section 1867 of the Social Security Act, codified at 42 USC 1395dd. It binds every hospital that has a Medicare provider agreement and an emergency department, and the definition of "hospital" includes critical access hospitals and rural emergency hospitals.
The statute is short. Its first duty is 1 sentence, and you should read it once in full before reading any chart:
"In the case of a hospital that has a hospital emergency department, if any individual (whether or not eligible for benefits under this subchapter) comes to the emergency department and a request is made on the individual's behalf for examination or treatment for a medical condition, the hospital must provide for an appropriate medical screening examination within the capability of the hospital's emergency department, including ancillary services routinely available to the emergency department, to determine whether or not an emergency medical condition (within the meaning of subsection (e)(1)) exists."
Everything else follows from that sentence. If the screening finds an emergency medical condition, subsection (b) requires the hospital to stabilize it within the staff and facilities available, or to transfer the patient under subsection (c). Subsection (c) limits transfers of unstabilized patients. Subsection (h) bars delaying screening or stabilizing treatment to ask about payment or insurance. The regulation goes further and bars the hospital from seeking insurer authorization until after the screening is done and stabilizing treatment has started (42 CFR 489.24(d)(4)(ii)).
The table below is the one to print. For each duty it names the record that proves compliance or breach, and the defense you should expect the hospital to raise from its own chart.
| Duty | Source | Record that proves or defeats it | Common defense |
|---|---|---|---|
| Appropriate medical screening examination (MSE) | 1395dd(a); 489.24(a) | Triage note with time and ESI level; the provider note showing who examined, when, and what was ordered; results with times; the hospital's written screening policy or protocol for this complaint; the central log entry | "We gave her the same screening we give every patient with this complaint. The diagnosis was missed, which is a malpractice question, not an EMTALA one." |
| Stabilizing treatment | 1395dd(b); 489.24(d) | Orders and medication administration times; serial vital signs; reassessment notes; the admission order if the patient was admitted; the informed refusal form if treatment was declined | "No emergency medical condition was found," or "we admitted him in good faith as an inpatient, which ended our EMTALA duty," or "she refused treatment after being told the risks." |
| Appropriate transfer | 1395dd(c); 489.24(e) | Physician certification with its risk and benefit summary, or the patient's written request; receiving hospital acceptance (name, time, accepting physician); records sent; EMS run sheet; the receiving hospital's arrival vitals | "The patient was stable at transfer," or "the patient requested transfer in writing," or "the on-call specialist refused to come in, so transfer was the better choice." |
| No delay for payment | 1395dd(h); 489.24(d)(4) | Registration timestamps against triage and MSE times; financial counselor notes; insurance verification or authorization call logs | "Registration ran in parallel with care and did not delay the exam." |
Notice what the table leaves out: the standard of care. EMTALA does not ask whether the emergency physician was right. A hospital can follow its screening protocol to the letter, miss an aortic dissection, and have no EMTALA exposure at all; that is a negligence case. Most complaints plead both counts; the records overlap, the elements do not. For the negligence side, the emergency department failure-to-diagnose review page covers what that record set looks like.
Many EMTALA claims die at the threshold, before anyone reads the physician note. The defense argues the patient never "came to the emergency department," that the condition was not an "emergency medical condition," or that the patient had become an inpatient. Each of those is a defined term, and the definitions live in the regulation, 42 CFR 489.24(b), more than in the statute.
The inpatient exception depends on good faith, so an admission order written minutes before a transfer, or an admission that never produced inpatient care, invites the argument that it was paper cover.
The decision path below is the usual order of the questions. Each "yes" and "no" is a fact dispute litigated with records.
Each box is a fact question the records answer. The path organizes the analysis; it does not decide whether EMTALA was violated.
A note on question 4. The stabilization duty in 1395dd(b)(1) is triggered when "the hospital determines that the individual has an emergency medical condition." Hospitals read that as a knowledge requirement: no finding in the chart, no duty to stabilize. Plaintiffs answer with the chart itself, pointing to abnormal vitals, results and nursing notes the hospital had in hand. How much knowledge a court requires, and how it is proven, is circuit case law; research yours before you plead it.
The statute never defines "appropriate medical screening examination." The courts filled the gap, and they filled it the same way in most circuits: appropriate means the screening the hospital would give any other patient with the same signs and symptoms, not the screening a careful physician should have given. The Fifth Circuit collected the cases in Marshall v. East Carroll Parish Hospital Service District (1998). It agreed with the other circuits that EMTALA "was not intended to be used as a federal malpractice statute," and held that a screening is judged "by whether it was performed equitably in comparison to other patients with similar symptoms," not by whether it produced the right diagnosis. The First Circuit put it in 1 line, quoted there: "The essence of this requirement is that there be some screening procedure, and that it be administered even-handedly."
CMS reads it the same way. The interpretive guidelines surveyors work from say:
"The MSE must be the same MSE that the hospital would perform on any individual coming to the hospital's dedicated emergency department with those signs and symptoms, regardless of the individual's ability to pay for medical care."
Appendix V adds 2 points that plaintiffs use constantly. The MSE "is not an isolated event. It is an ongoing process that begins, but typically does not end, with triage." And "it is not appropriate to merely 'log in' an individual and not provide a MSE." Triage decides the order in which people are seen; the MSE decides whether an emergency medical condition exists. A triage note with vitals and a chief complaint, followed by a 4-hour wait and a departure, is a triage record and nothing more.
Disparate screening is a comparison, so you need 2 things: the standard, and the chart that departed from it. The standard is usually 1 of 4 documents, and I would ask for all 4 in the first request:
Then the chart. The departures that tend to matter are concrete: a protocol ECG never done, a pregnancy screen skipped, a psychiatric patient with no suicide risk assessment (Appendix V says the record should show "an assessment of suicide or homicide attempt or risk" for psychiatric presentations), a woman in labor with no fetal heart tones or cervical exam charted, or a screening that stopped when registration flagged the patient as uninsured. For obstetric screening claims, the birth injury record review page covers the rest of that record set.
Once the hospital determines that an emergency medical condition exists, it has 2 choices under 1395dd(b): stabilize within the staff and facilities available, or transfer under subsection (c). A third path is built into the statute. If the hospital offers the treatment or the transfer, explains the risks and benefits, and the patient refuses, the hospital is "deemed to meet" the duty, and it "shall take all reasonable steps to secure" a written informed refusal. The regulation adds that the medical record "must contain a description of the examination, treatment, or both if applicable, that was refused" (42 CFR 489.24(d)(3)).
That makes the AMA form 1 of the most important pages in the chart. A blank AMA form signed by the patient, with no description of what was refused and no risk discussion, is weak. A nurse note saying "pt refused to sign" can be enough; Appendix V tells surveyors that an unsigned form "is not, however, automatically a violation," provided staff documented the refusal and the hospital used best efforts to get a signature. Check who signed, at what time, and whether the time fits the rest of the timeline. An AMA form timed after the patient's departure in the central log is a finding.
An unstabilized patient may be transferred only with 1 of 3 authorizations: the patient's informed written request; a physician certification that "based upon the information available at the time of transfer," the expected benefits of treatment elsewhere outweigh the increased risks; or a qualified medical person's certification, countersigned later by a physician. Every certification "shall include a summary of the risks and benefits upon which the certification is based." The transfer must also be "appropriate" under 1395dd(c)(2): treatment within the sender's capacity to minimize the risks, a receiving facility with space and qualified staff that has agreed to accept, all available records sent (including the name and address of any on-call physician who refused or failed to appear), and qualified personnel and equipment for the trip.
Here is what a transfer certification looks like when it is read the way an EMTALA reviewer reads it. The form is a mock-up; field names vary by hospital.
Every field on the form maps to a statutory element. Read it field by field against the timeline, not as a signature page.
2 more rules sit on the receiving side. A hospital with specialized capabilities, such as a burn unit, shock-trauma unit or NICU, "may not refuse to accept" an appropriate transfer of a patient who needs those capabilities if it has the capacity (489.24(f)). And Appendix V tells a receiving hospital to report a suspected improper transfer to CMS or the state survey agency, with notification expected "within 72 hours of the occurrence." That report is often how a CMS investigation starts, and the receiving hospital's arrival record is often the best evidence of the patient's condition at the moment of transfer.
The private right of action is in 1 subsection, and it is narrower than many first complaints assume.
"Any individual who suffers personal harm as a direct result of a participating hospital's violation of a requirement of this section may, in a civil action against the participating hospital, obtain those damages available for personal injury under the law of the State in which the hospital is located, and such equitable relief as is appropriate."
Read it for what it leaves out. The defendant is "the participating hospital." The statute gives no private civil action against the emergency physician, the on-call specialist or the physician group; claims against them run under state malpractice law. The damages are those "available for personal injury under the law of the State," so state damage caps and pre-suit rules become live arguments. Causation is built in: the harm must be "a direct result" of the violation. A plaintiff who was screened late but whose outcome was fixed before arrival has a causation problem, and the timeline is how both sides argue it.
Subsection (d)(2)(B) gives a second claim that is easy to forget: a medical facility that "suffers a financial loss as a direct result" of a hospital's violation can sue the hospital too. It gives a receiving hospital that absorbed the cost of an improper transfer its own claim against the sender.
Subsection (d)(2)(C) is blunt: "No action may be brought under this paragraph more than 2 years after the date of the violation with respect to which the action is brought." The clock runs from the violation, not from discovery of the harm and not from the end of treatment. It can be shorter or longer than the state malpractice period that governs the companion negligence claim, and the 2 periods may start on different days. Calendar both on intake. I would also calendar the EMTALA date from the earliest arguable violation, which is often the door time on the first visit, not the transfer hours later.
Separate from the civil action, the government enforces EMTALA 2 ways. CMS can terminate the hospital's Medicare provider agreement (489.24(g)). And the HHS Office of Inspector General can impose civil money penalties under 1395dd(d)(1) on the hospital and on a responsible physician, including an on-call physician, for negligent violations, such as certifying a transfer when the physician "knew or should have known that the benefits did not outweigh the risks." Gross and flagrant or repeated violations can bring physician exclusion.
HHS adjusts the statutory maximums for inflation every year in the penalty table at 45 CFR 102.3, which points to the OIG's EMTALA penalty rule at 42 CFR 1003.510. Quote the current year's table, not the statute, when you cite a figure. Before CMS imposes a penalty, the statute requires it to ask the Quality Improvement Organization for a physician review of whether the patient had an unstabilized emergency condition, with at least 60 days for that review unless delay would jeopardize health or safety (1395dd(d)(3)).
Penalties are not damages and do not go to the patient. They matter to a civil case because the investigation behind them produces documents: the survey findings, the statement of deficiencies, and sometimes the QIO physician review. Chapter 6 covers how to get and use them.
An EMTALA complaint to CMS or the state survey agency can start an on-site investigation, and the surveyors' playbook is Appendix V of the State Operations Manual. Read it before you draft discovery. It is the federal government's own list of the documents that show whether a hospital complied, and hospitals know surveyors ask for them, so they are hard to call burdensome.
At the entrance conference, Appendix V tells surveyors to ask for the dedicated ED logs for the past 6 to 12 months, the ED policy and procedures manual (triage, labor assessment, transfers), consent forms for unstable transfers, ED committee and medical staff minutes, staffing schedules (physicians for 3 months, nurses for 4 weeks), the medical staff bylaws and roster, and physician on-call lists for the past 6 months.
Appendix V says "a single occurrence is considered a violation," but surveyors still sample to find patterns: at least 20 records reviewed in depth (at least 10 on a revisit), chosen by the surveyor, never by hospital staff. The selection criteria double as a plaintiff's log review plan: transfers to other facilities; "gaps, return cases, or non-sequential entries in the log"; refusals of examination, treatment or transfer; patients who left against medical advice or without being seen; and patients who returned to the ED within 48 hours. Surveyors review the central log for at least 6 months and "check for completeness, gaps in entries or missing information." A physician, usually through the QIO, then decides whether the screening, stabilization and transfer were appropriate.
Each finding is written up under a tag number. Knowing the tags lets you read a statement of deficiencies quickly and match each citation to a duty.
| Tag | Rule | What the finding is about |
|---|---|---|
| A/C-2400 | 489.20 | Policies and procedures addressing the anti-dumping provisions |
| A/C-2401 | 489.20(m) | Receiving hospital failed to report a suspected improper transfer |
| A/C-2402 | 489.20(q) | EMTALA rights signage |
| A/C-2403 | 489.20(r)(1) | Transfer records kept 5 years |
| A/C-2404 | 489.20(r)(2); 489.24(j) | On-call physician list and on-call policies |
| A/C-2405 | 489.20(r)(3) | Central log |
| A/C-2406 | 489.24(a), (c) | Appropriate medical screening examination |
| A/C-2407 | 489.24(d) | Stabilizing treatment |
| A/C-2408 | 489.24(d)(4), (5) | Delay in examination or treatment to inquire about payment |
| A/C-2409 | 489.24(e)(1), (2) | Appropriate transfer |
| A/C-2410 | 489.24(e)(3) | Whistleblower protections |
| A/C-2411 | 489.24(f) | Recipient hospital responsibilities |
Findings go on Form CMS-2567, "Statement of Deficiencies and Plan of Correction." CMS began posting 2567s for short-term acute care hospitals and critical access hospitals from complaint surveys in March 2013, according to its survey memo S&C 13-21. The posted versions are redacted and carry no plan of correction; for the plan, and for the full survey file, send a Freedom of Information Act request to CMS or a public records request to the state survey agency, and follow with a subpoena if you need the surveyor's worksheets.
Admissibility at trial depends on your court's evidence rules. Even where a 2567 never reaches a jury, it does 3 jobs well. Appendix V tells surveyors to support every deficiency with evidence from record reviews and interviews, so the 2567 shows which other charts were pulled and that comparators exist. It quotes staff interviews, which become deposition outlines. And it often states what the hospital's own policy required, in the surveyor's words, which is the yardstick chapter 3 says you need.
A standard "complete medical records" request gets the chart as the HIM department prints it, which is about half of an EMTALA file. The central log, the on-call list, the policies and the order-set configuration are hospital business records, not part of any 1 patient's chart, and they rarely arrive unless you name them. A patient's HIPAA access request (the covered entity has 30 days to act under 45 CFR 164.524) is a fast way to get the chart itself; the rest comes through discovery or subpoena. The general mechanics are in how to get medical records for a lawsuit.
Retention helps you here. 42 CFR 489.20(r) requires both the sending and receiving hospital to keep "medical and other records related to individuals transferred to or from the hospital for a period of 5 years from the date of the transfer," to keep an on-call list, and to keep a central log. Quote the rule in your preservation letter so nobody can claim routine deletion.
| Record | Who holds it | What it proves | Watch for |
|---|---|---|---|
| Registration and arrival record | Sending hospital (patient access) | Door time, mode of arrival, whether insurance questions came before triage | Registration time earlier than triage by a wide margin; insurance verification before the MSE |
| Triage note | Sending hospital (EHR) | Triage time, chief complaint, first vitals, ESI level, pain score | ESI level inconsistent with the vitals; "triage time" that is really the note's filing time |
| Nursing notes and vitals flowsheet | Sending hospital | Reassessments, trends, the last vitals before departure | Long gaps between reassessments; vitals charted in a batch after the fact |
| Physician or APP note | Sending hospital | Who did the MSE, when, what was examined, the EMC conclusion | Note signed hours after the encounter; exam time missing; templated negatives |
| Orders and results with times | Sending hospital (EHR, lab and imaging systems) | Ordered, collected, resulted and reviewed times for each test | Protocol tests never ordered; results that came back after discharge and were never seen |
| AMA, LWBS and refusal forms | Sending hospital | Informed refusal: what was refused, risks explained, who signed | Blank risk section; time after departure; no witness |
| Transfer certification and consent | Sending hospital | Authorization for an unstable transfer and the risk and benefit summary | Boilerplate reasons; missing time; signed by a non-physician with no countersignature |
| Receiving hospital acceptance | Both hospitals (transfer center) | Who accepted, when, and whether capacity was confirmed | Acceptance after departure; name missing; recorded calls not preserved |
| EMS run sheet (patient care report) | EMS agency | Condition at pickup, vitals en route, crew level, times | Vitals worse at pickup than the "stable" note; crew level below what the patient needed |
| Receiving hospital ED record | Receiving hospital | Condition and vitals on arrival: the best test of "stabilized" | Resuscitation on arrival; a receiving hospital report to CMS |
| Central log, 489.20(r)(3) | Sending hospital | That the patient came to the ED, and the disposition recorded | Disposition that contradicts the chart; gaps and non-sequential entries around the visit |
| On-call list, 489.20(r)(2), and paging logs | Sending hospital (medical staff office, operator) | Which specialist was on call, when paged, and whether they answered or refused | No specialist listed for a service the hospital offers; paging logs with short retention |
| Policies, protocols, order sets | Sending hospital | The screening yardstick for disparate screening | Today's version produced instead of the version in force on the visit date |
| EHR audit trail for the visit | Sending hospital (EHR vendor data) | When each note was opened, written, signed and amended | Late entries and addenda written after a complaint; see the EHR audit trail guide |
A word on the ESI level. The Emergency Severity Index is a 5-level triage algorithm, from 1 (most urgent) to 5 (least urgent), that sorts patients by acuity and expected resource needs. The AHRQ ESI page describes it; the Emergency Nurses Association has owned it since 2019. An ESI 2 who waited 3 hours is a different story from an ESI 4 who waited 3 hours, and an ED nurse expert can say whether the level fit the vitals.
Attach it to a subpoena, a request for production, or a preservation letter. Edit the bracketed fields and delete items that do not apply. Rules and procedure vary by jurisdiction.
RE: [PATIENT NAME], DOB [DATE]; ED visit(s) [DATE(S)]; [HOSPITAL NAME]
Time period: [DOOR DATE/TIME] through [DISPOSITION DATE/TIME + 24 HOURS]
A. Patient chart (produce in native EHR print format with all timestamps)
1. Registration and arrival record, including arrival mode and insurance
verification or authorization entries, with times
2. Triage note(s), including chief complaint, vital signs, pain score and
ESI or other acuity level, with the time each field was documented
3. All nursing notes, reassessments and the vital signs flowsheet
4. All physician, APP and resident notes, including the medical screening
examination, with created, signed and amended times
5. All orders, with ordered, collected, resulted and reviewed times
6. Laboratory and imaging results, including imaging acquisition times
7. Medication administration record
8. Against-medical-advice, left-without-being-seen and informed refusal
forms, and any note documenting a refusal to sign
9. Transfer certification, patient transfer request or consent, and any
qualified medical person certification with physician countersignature
10. Records of the receiving facility's acceptance: name of accepting
physician, time, transfer center reference number
11. List or copy of the records sent with the patient
12. Discharge instructions and discharge time
13. Monitor, telemetry or fetal monitoring strips for the visit
B. Hospital records outside the chart
14. Central log entries under 42 CFR 489.20(r)(3) for [DATE], [DATE - 1]
and [DATE + 1], de-identified for other patients as the court orders
15. On-call list under 42 CFR 489.20(r)(2) for [SPECIALTY/SPECIALTIES]
for [DATE]
16. Paging, operator and phone logs showing calls to on-call physicians
and to the receiving facility, and any recordings
17. ED triage, medical screening, transfer and on-call policies, and
[COMPLAINT]-specific protocols and order sets, as in force on [DATE]
18. EHR audit trail for this encounter (access, create, modify, sign)
19. Any report to CMS or the state survey agency about this visit, any
Form CMS-2567 and plan of correction arising from it
20. Diversion status logs for [DATE]
C. Third parties
21. EMS patient care report(s) from [AGENCY]
22. Receiving hospital ED and transfer center records from [HOSPITAL]
Preservation: records related to transfers must be kept 5 years under
42 CFR 489.20(r)(1). Please suspend routine deletion of the items above,
including phone recordings and paging logs.
EMTALA disputes are about intervals: how long from the door to triage, from triage to a provider, from a critical result to someone acting on it, from the transfer decision to acceptance to wheels rolling. The chart rarely hands you those intervals. You build them from every system that carries a clock, and you show your work.
1 row per event. Paste into a spreadsheet. Keep event time and documentation time in separate columns, and never fill a blank with an estimate without saying so.
CASE: [NAME] VISIT: [DATE] HOSPITAL: [NAME] PREPARED BY: [NAME] Clock offsets noted: [SYSTEM A vs SYSTEM B: +/- MINUTES, BASIS] # | Event | Event time | Documented time | Source (doc, page/Bates) | Author/role | Duty (MSE/Stab/Transfer/Delay) | Interval from prior anchor | Protocol limit (cite policy) | Conflict with other record? | Note 1 | Arrival / door | | | | | MSE | -- | | | 2 | Registration / insurance entry | | | | | Delay | | | | 3 | Triage, ESI level [ ] | | | | | MSE | | | | 4 | First provider contact (MSE start) | | | | | MSE | | | | 5 | Tests ordered [list] | | | | | MSE | | | | 6 | Results available / reviewed | | | | | MSE | | | | 7 | Reassessments [each one] | | | | | MSE/Stab | | | | 8 | EMC determination (yes/no, by whom) | | | | | Stab | | | | 9 | Stabilizing treatment [each] | | | | | Stab | | | | 10 | On-call paged / responded / refused | | | | | Stab/Transfer | | | | 11 | Disposition decision | | | | | Stab/Transfer | | | | 12 | Refusal / AMA form signed | | | | | Stab | | | | 13 | Transfer certification signed | | | | | Transfer | | | | 14 | Receiving facility accepted (name) | | | | | Transfer | | | | 15 | Records sent (list) | | | | | Transfer | | | | 16 | Departure / EMS pickup, vitals | | | | | Transfer | | | | 17 | Arrival at receiving ED, vitals | | | | | Transfer | | | | 18 | Central log disposition entry | | | | | All | | | | GAPS (no record where one is expected): [LIST, with what implies the record] CONFLICTS (2 sources, 2 times): [LIST, both citations]
Walk-in. Registration record shows arrival 21:02 and "self-pay" entered at 21:06, before triage.
Registration, HOSP 000002Chest pain to back, BP 182/104, HR 108, pain 9 of 10. No ECG ordered.
Triage note, HOSP 000004Policy calls for an ECG by 21:12. First ECG is at 22:31, 89 minutes late.
Chest pain protocol, HOSP 000301; ECG, HOSP 000019Flowsheet has no vitals between triage and the provider exam; protocol calls for 2 reassessments in that window.
Vitals flowsheet, HOSP 000011Exam time stated in the note as 22:25; note created 23:48 and signed 01:12. 2 times, both cited.
ED physician note, HOSP 000006 to 000009Ordered 22:50, acquired 23:18, resulted 23:34: type A dissection.
Orders, HOSP 000014; CT report, HOSP 000022Physician addendum: "dissection, needs CT surgery, not available here." Labetalol ordered 23:41, given 23:52.
Addendum, HOSP 000009; MAR, HOSP 000024Receiving transfer center logs acceptance by a named cardiothoracic surgeon at 23:58.
Receiving transfer log, RCV 000003Transfer certification signed and dated, time field empty. Risk summary reads "benefits outweigh risks."
Transfer certification, HOSP 000027BP 150/90, HR 96. Nurse note: "stable for transfer."
Vitals flowsheet, HOSP 000012Run sheet at patient contact: BP 88/50, diaphoretic, 4 minutes after the "stable" note. Departs 00:20.
EMS patient care report, EMS 000002BP 80/46, taken straight to the OR.
Receiving ED note, RCV 000011Log records "discharged," not "stabilized and transferred" or "transferred."
Central log, HOSP 0003505 marked gaps and 3 conflicting times, each with a citation. None of them decides the case; together they tell an expert and a lawyer where to look.
Screening, transfer and delay theories
Screening, stabilization and causation defenses
My view: the missing certification time and the EMS vitals are the strongest facts, because they go to the transfer rules, where the statute is specific and no disparity comparison is needed. The protocol gap is a good screening theory only if the comparator charts show insured patients got their ECG on time. If they show the protocol is missed for everyone, the hospital has a bad ED rather than a disparity, and the claim looks more like malpractice than EMTALA. Whether any of this is an EMTALA violation is for the court and the experts.
An EMTALA file is small by litigation standards, often a few hundred pages across 2 hospitals and an EMS agency, but dense with times. Generative AI built on a large language model (LLM) handles the tedious part of AI medical record review: pulling every timestamp from triage notes, flowsheets, order logs and run sheets, putting them in order, and noticing a note created 83 minutes after the exam it describes. Older clinical NLP (natural language processing) tools extracted terms; an AI medical chronology drafts the timeline itself. A newer wrinkle: ambient AI scribes now draft some ED notes from recorded conversation, so a note's text, draft time and signature time can be 3 different things. And the software can be wrong in ways that cost more here than in most claims.
The fixes are known. Every line should carry a grounded, page-level citation, so a reviewer clicks from "EMS BP 88/50 at 00:14" to the run sheet page. Retrieval-augmented generation (RAG) that answers only from the uploaded pages reduces invented facts but does not remove them. OCR should flag low-confidence pages, especially faxes and handwritten notes, so a human reads them directly. And human-in-the-loop review is required: the ED nurse expert or LNC checks every time that matters against its page. Courts have sanctioned lawyers under Rule 11 for filing AI-invented case citations, Mata v. Avianca (S.D.N.Y. 2023) being the best known, and a timeline with an invented minute is the same risk in a different document.
Download this guide as a worksheet (CSV), with Done and Notes columns.
0 of 8 checked
Medrecords AI is medical record review software. You upload the records you have; it drafts a cited medical chronology of arrival, triage, provider contact, orders, results, treatment and transfer, with a citation on every line linked to its source page. OCR routes typed, handwritten and table pages to the right engine and flags low-confidence pages. Missing records identification flags documents the file implies but does not contain, such as a transfer mentioned in a note with no certification produced. Undated document flagging catches the forms with no time. When the receiving hospital's records arrive later, supplemental record review shows where they agree with, conflict with or add to the file.
What it does not do: it does not retrieve records from hospitals or EMS agencies, does not decide whether EMTALA applied or was violated, and does not give legal or medical opinions. Flags are signals, not verdicts. You review, you revise, you sign.
Book a demo on an EMTALA file you are working, then run your first case free on us. Every line comes back cited to its source page. You review, you revise, you sign.
Scheduling only. No records move from a public page.
3 short guides from our help center, recorded on a demo case. They play inside the app: sign in or create a free account, and the video opens at the chapter you picked. First case is free on us.
Free account 2:35Every medical event on its date, in order, tied to the page it came from. Filter it, search it and export it as PDF or Word.
Free account 2:30Every record the case documents mention that the file does not contain, grouped by provider, with the page that mentions it.
Free account 2:12Ask a question in your own words. The answer comes back with numbered links, and each one opens the page it came from.
Statutes, regulations and CMS guidance were checked against primary sources in September 2026, and quoted text is verbatim. Circuit case law was read through the Fifth Circuit's opinion in Marshall, which quotes the other circuits; those quotations are cited as Marshall reports them. The hospital, patient and times in chapter 9 are hypothetical. Product facts come from this site's product pages. Nothing here is legal or medical advice.
Related guides: EHR audit trails in medical malpractice, altered medical records, how to write a medical chronology, and medical malpractice record review.