# HIE birth injury cases: the medical records that decide them

> The maternal and neonatal records that decide an HIE birth injury case: ACOG criteria, NICHD tracing categories, missing fetal strips and AI review.

Canonical page: https://medrecords.ai/guides/hie-birth-injury-medical-records/

---
New — **Missing Records Detection:** flags every visit, provider, and date missing from the file. [See how →](https://medrecords.ai/product/missing-records-identification/)
Field guide, September 2026. 14 chapters.

## Hypoxic-ischemic encephalopathy (HIE): the records that decide a birth injury case

For birth injury plaintiff attorneys, defense counsel and legal nurse consultants with labor and delivery experience. You walk away with the ACOG criteria mapped to the records that prove them, a decision path for incomplete fetal monitoring strips, a records request checklist and a minute-by-minute timeline worksheet.

By [Ahmed Jemaa](https://medrecords.ai/authors/ahmed-jemaa/), Co-Founder and CEO, Medrecords AI. Published 25 September 2026.

HIE birth injury medical records are the maternal and neonatal documents used to reconstruct what happened around birth: prenatal records, the labor nursing flowsheet, the full electronic fetal monitoring tracing, oxytocin records, delivery and resuscitation notes, cord blood gases, Apgar scores, NICU and cooling records, EEG, neonatal MRI and placental pathology. Experts weigh them against the 2014 ACOG and AAP neonatal encephalopathy criteria.

An HIE case is decided by about 6 hours of records: the last stretch of labor, the delivery, the first hour of resuscitation and the decision to cool. Those hours are split across 2 patients, 4 or 5 electronic systems and at least 3 clocks, and the most important document, the fetal monitoring tracing, is the 1 most likely to be missing from the production.

This guide is built from the American College of Obstetricians and Gynecologists and American Academy of Pediatrics report *Neonatal Encephalopathy and Neurologic Outcome*, second edition (2014, reaffirmed 2019), including its executive summary, its chapters on intrapartum assessment, neonatal assessment and placental pathology, and its review of the cooling trials; the 2008 NICHD, ACOG and Society for Maternal-Fetal Medicine workshop on electronic fetal monitoring as summarized in that report; the HIPAA access rule at 45 CFR 164.524; the Virginia Birth-Related Neurological Injury Compensation Act (Va. Code 38.2-5000 and following); and the Florida Birth-Related Neurological Injury Compensation Plan (Fla. Stat. 766.301 to 766.316). Every case example is a labeled hypothetical. Nothing here is legal or medical advice. Whether care met the standard, and what caused a child's injury, are questions for qualified experts in your jurisdiction, and procedure and expert rules vary by state.

9 numbers

### HIE records in 9 numbers

Under 5
Apgar score at both 5 and 10 minutes, a neonatal sign consistent with an acute intrapartum event
ACOG and AAP, 2014, executive summary II.A
pH under 7.0
umbilical artery pH, or base deficit of 12 mmol/L or more, or both, raises the probability of an intrapartum hypoxic component
ACOG and AAP, 2014, II.B
35 weeks
gestational age floor in the report's case definition of neonatal encephalopathy
ACOG and AAP, 2014, I
24 to 96 hours
of life: the MRI window the report calls most useful for timing an injury
ACOG and AAP, 2014, III.C
6 hours
from birth: randomization window in all but 1 of the large cooling trials
ACOG and AAP, 2014, chapter 11
72 hours
of cooling in the trials, target rectal temperature 33.5 C, rewarming at 0.5 C an hour
ACOG and AAP, 2014, chapter 11
4.5 mmol/L
rise in umbilical artery base deficit when a clamped cord segment sits 60 minutes before analysis
ACOG and AAP, 2014, chapter 6
11.2%
of placentas examined in 1 series of infants with neonatal encephalopathy, though all met criteria
ACOG and AAP, 2014, chapter 4
65%
of emergency primary cesareans in an 11,481-delivery review began within 30 minutes of the decision
ACOG and AAP, 2014, chapter 6
Chapter 1 Everyone

### HIE and neonatal encephalopathy in plain terms

Neonatal encephalopathy is abnormal brain function in a newborn in the first days of life. The 2014 ACOG and AAP report defines it as "a clinically defined syndrome of disturbed neurologic function in the earliest days of life in an infant born at or beyond 35 weeks of gestation, manifested by a subnormal level of consciousness or seizures." The baby often also has trouble starting and keeping up breathing, and low tone and reflexes.

Hypoxic-ischemic encephalopathy, HIE, is the subset where the cause is too little oxygen (hypoxia) and too little blood flow (ischemia). That word "cause" is the whole lawsuit. The report is blunt about the label: if a full workup of causes is not possible, the term HIE "should best be replaced by neonatal encephalopathy," because neither hypoxia nor ischemia can be assumed to be the only mechanism. When you see "HIE" written in a NICU admission note at hour 2, treat it as a working diagnosis written before anyone had the placenta report, the MRI or the genetics. It is a clue, and it is also a fact the defense will attack.

#### Why the records carry so much weight

Nobody watches a fetal brain being injured. Every side reconstructs it from a tracing, nursing entries, a cord blood pH, a resuscitation form and an MRI taken days later. The report itself calls for "a comprehensive multidimensional assessment" of neonatal status and all potential contributing factors, including maternal medical history, obstetric antecedents, intrapartum factors (fetal heart rate monitoring and the delivery) and placental pathology. Read that list as a records request. Each item names a document.

The report also warns that many injuries are not intrapartum at all. It notes that epidemiologic studies have suggested 70% of causation is related to chronic antenatal factors, while MRI studies show most term injuries look acute. It resolves that tension by pointing out that imaging in the first 2 to 3 weeks cannot tell whether an injury happened during labor or in the days before. For a plaintiff, that means the prenatal chart and the placenta are defense exhibits waiting to happen, and you want them before your expert signs anything. For the defense, it means the intrapartum records are rarely the whole story, and a gap in them cuts both ways.

Neonatal encephalopathy (NE)Disturbed neurologic function in a newborn at or beyond 35 weeks: reduced consciousness or seizures, often with breathing difficulty and low tone. A description, not a cause.
Hypoxic-ischemic encephalopathy (HIE)Neonatal encephalopathy attributed to hypoxia and ischemia around birth. A causal label, which is why its first appearance in the chart is worth dating.
Sentinel eventAn acute event immediately before or during labor that can cut off fetal oxygen: uterine rupture, severe placental abruption, cord prolapse, amniotic fluid embolus with severe maternal hypotension, maternal cardiovascular collapse, or fetal exsanguination.
Base deficitA calculated measure of metabolic acid buildup. The report treats 12 mmol/L or more in umbilical artery blood as pathologic acidemia.
Therapeutic hypothermiaControlled cooling of the newborn for a set period to reduce secondary brain injury. It creates its own flowsheet, device log and eligibility record.
EFMElectronic fetal monitoring: the continuous fetal heart rate and contraction tracing, often stored outside the main chart.

If you remember one thing
"HIE" in the chart is a working diagnosis, not a finding of cause. The case turns on whether the maternal, intrapartum, neonatal and placental records together support it, so request all 4 groups before any expert commits.

Chapter 2 Everyone

### The ACOG and AAP framework and the record behind each criterion

[*Neonatal Encephalopathy and Neurologic Outcome*, second edition](https://www.acog.org/-/media/project/acog/acogorg/clinical/files/task-force-report/articles/2014/neonatal-encephalopathy-and-neurologic-outcome.pdf), is the 2014 report of ACOG's Task Force on Neonatal Encephalopathy, endorsed by the American Academy of Pediatrics and [published in *Pediatrics*](https://publications.aap.org/pediatrics/article/133/5/e1482/32738/Neonatal-Encephalopathy-and-Neurologic-Outcome). ACOG reaffirmed it in 2019. Its first edition in 2003 was titled *Neonatal Encephalopathy and Cerebral Palsy*; the rename signals a wider range of outcomes than cerebral palsy.

The 2014 report does not work as a pass or fail checklist. It lists neonatal signs and contributing factors, and says that the more elements are met, "it becomes increasingly more likely" that intrapartum hypoxia-ischemia played a role. Experts on both sides use it that way, as a set of weights. Your job is to make sure every weight has a document under it.

"The presence of metabolic acidemia does not define the timing of the onset of a hypoxic-ischemic event."

**The 2014 criteria and where each one lives in the file** — ACOG and AAP, 2014 — Executive summary, sections I to IV
| Criterion (as the report states it) | What the report says it means | Record that proves or disproves it |
| --- | --- | --- |
| Case definition: encephalopathy in an infant at or beyond 35 weeks | First mandatory step: confirm the infant meets the definition | Prenatal dating ultrasound, delivery record gestational age, NICU admission neuro exam, early seizure documentation |
| Apgar score under 5 at 5 minutes and 10 minutes | Increased relative risk of cerebral palsy. If the 5-minute Apgar is 7 or more, a major role for peripartum hypoxia-ischemia is unlikely | Delivery record, expanded Apgar form, neonatal resuscitation record, newborn admission note. Check that all 3 agree |
| Umbilical artery pH under 7.0, base deficit 12 mmol/L or more, or both | Increases the probability of an intrapartum hypoxic component. Arterial pH above 7.20 makes it unlikely | Cord gas lab report with sample type labels, collection and result times; delivery note stating which vessel was sampled |
| MRI or MR spectroscopy consistent with hypoxia-ischemia | Deep nuclear gray matter or watershed patterns are typical; other patterns point elsewhere. MRI obtained after 24 hours with no injury makes a significant intrapartum injury unlikely | MRI reports and the DICOM images for every study, with the hour of life at each scan |
| Multisystem organ failure | Renal, hepatic, hematologic, cardiac, metabolic or gastrointestinal injury, alone or combined | NICU labs (creatinine, liver enzymes, coagulation), urine output, echocardiogram, pressor records |
| Sentinel hypoxic or ischemic event | Uterine rupture, severe abruption, cord prolapse, amniotic fluid embolus with severe prolonged maternal hypotension and hypoxemia, maternal cardiovascular collapse, fetal exsanguination | Operative note, nursing notes, anesthesia record, placental pathology, maternal labs |
| Fetal heart rate patterns consistent with an acute event | A tracing that starts Category I and converts to Category III suggests an event; a Category II tracing present from admission with minimal or absent variability for 60 minutes or more suggests earlier compromise | The complete EFM tracing from admission to delivery, plus nursing category assessments |
| No other proximal or distal factors | Abnormal fetal growth, maternal infection, fetomaternal hemorrhage, neonatal sepsis and chronic placental lesions make an acute intrapartum event as the sole cause much less likely | Prenatal growth scans, maternal temperature and labs, Kleihauer-Betke or similar testing if done, blood cultures, placental pathology, genetic workup |
| Outcome: spastic quadriplegia or dyskinetic cerebral palsy | Other subtypes are less likely to be linked to acute intrapartum events | Pediatric neurology records, developmental evaluations, therapy records through the present |

Every row in the framework maps to at least 1 document, and 3 of them (MRI, cord gas, tracing) are routinely produced incomplete.

2 points in this table decide more cases than they should. First, the report says acute intrapartum asphyxia severe enough to cause encephalopathy will most often injure other organs, and "the absence of evidence of such injury should call into question the diagnosis of significant intrapartum asphyxia." Without the NICU labs, nobody can test that. Second, imaging can time an injury only "within days rather than hours or minutes." An MRI does not prove the injury happened at 03:40; the tracing and the timeline carry that argument.

MythA cord pH under 7.0 proves the injury happened during labor.
RuleThe report says acidemia raises the probability but "does not define the timing" of the event, and most significantly acidemic newborns are neurologically normal.
MythA good 5-minute Apgar ends the case.
RuleA 5-minute Apgar of 7 or more makes a major hypoxic role "unlikely," which is a weight, not a bar. Check who scored it and whether the resuscitation record agrees.
MythThe MRI report is enough.
RuleThe report calls for interpretation by a trained neuroradiologist and an expert opinion when expertise is limited. Your expert needs the images and the scan times, not a 1-page impression.
If you remember one thing
Build the file around the 2014 report's rows. For each criterion, name the document that proves it, the document that could disprove it, and whether you have both.

Chapter 3 Everyone

### The fetal heart rate vocabulary: NICHD Categories I, II and III

In 2008 the Eunice Kennedy Shriver National Institute of Child Health and Human Development, ACOG and the Society for Maternal-Fetal Medicine jointly published updated guidelines for fetal heart rate monitoring ([Macones and colleagues, *Obstetrics and Gynecology*, 2008](https://pubmed.ncbi.nlm.nih.gov/18757666/) ). They set the 3-tier terminology that labor nurses and obstetricians now chart every shift. The 2014 ACOG and AAP report calls it "the accepted standard in the United States." The definitions below are the ones that report restates.

**The 3 NICHD tracing categories** — 2008 workshop — As restated in ACOG and AAP, 2014, chapter 6
Category I — **Baseline 110 to 160 bpm, moderate variability, no late or variable decelerations** — Strongly predictive of normal fetal acid-base status at the time of observation.
Category II — **Everything that is not Category I or Category III** — Not indicative of abnormal fetal acid-base status. The largest and most argued-over group.
Category III — **Absent variability with recurrent late decelerations, recurrent variable decelerations or bradycardia; or a sinusoidal pattern** — Identifies fetuses at increased risk of abnormal acid-base status at the time of observation.
The categories describe the tracing "at the time of observation." A category charted at 14:00 says nothing about 14:20.

That last phrase is the reason the strip beats the flowsheet. A labor nurse charts a category at intervals set by unit policy and the phase of labor. Between those entries, the tracing kept running. When a flowsheet says "Category II" at 3 consecutive checks, that is 3 snapshots of a continuous record. Your expert has to see the film, not the snapshots.

#### What the 2014 report says about tracing patterns and timing

The report treats the tracing as timing evidence in both directions, which is why both sides fight over it:

- A Category I or II tracing together with a 5-minute Apgar of 7 or more, normal umbilical artery gases, or both, "is not consistent with an acute hypoxic-ischemic event."
- A tracing that is Category I on presentation and converts to Category III "is suggestive of a hypoxic-ischemic event."
- After a Category I start, tachycardia with recurrent decelerations, or persistent minimal variability with recurrent decelerations, may also suggest intrapartum timing.
- A Category II tracing lasting 60 minutes or more from the time of presentation, with persistently minimal or absent variability and no accelerations, even without decelerations, "is suggestive of a previously compromised or injured fetus."

For the defense, a flat tracing from admission points to injury before the mother arrived. For the plaintiff, those first 60 minutes, and what the team did with them, can become the case. Both readings need the admission segment of the strip, the part most often printed short or not at all.

#### Oxytocin, contractions and the category

The report defines tachysystole as more than 5 contractions in 10 minutes, averaged over a 30-minute window. It says that during induction or augmentation, a decrease in the oxytocin dose "should be considered" if tachysystole occurs with a Category I tracing, and that with a Category II or III tracing, oxytocin "should be reduced or stopped" and intrauterine resuscitation started. Whether that happened in a given labor is a standard-of-care question for an expert. Whether the record lets anyone answer it is a records question, and it needs 3 things lined up on 1 clock: the contraction channel of the tracing, every oxytocin rate change from the pump or medication record, and the charted category.

##### What the flowsheet gives you

Charted by the nurse

- Category at each assessment
- Baseline, variability, decelerations in words or picklists
- Oxytocin rate entries as charted
- Notification of the provider, when charted
- Interventions: position change, fluids, oxygen

##### What only the tracing gives you

Recorded by the monitor

- Every minute between assessments
- The contraction pattern and frequency
- When a change actually began, not when it was charted
- Signal loss, maternal heart rate capture, lead changes
- Event marks and annotations stored in the archive

If you remember one thing
NICHD categories describe the tracing at the moment someone looked. The case is usually about the minutes between looks, and only the complete tracing covers them.

Chapter 4 Building

### The HIE record set: what to request and where it lives

An HIE file is 2 charts, not 1. The mother and the baby have separate medical record numbers, and often separate hospitals when the baby is transferred for cooling. A request for 1 does not produce the other, and each needs its own authorization, index and Bates range. The table below is the list we would start from. It is long on purpose: every row feeds a criterion from chapter 2 or an entry in the timeline in chapter 8.

| Record | What it shows | Where it usually lives | Production trap |
| --- | --- | --- | --- |
| Prenatal records | Dating, risk factors, growth scans, Dopplers, prior pregnancies, infections, prior stillbirth | Outpatient OB clinic, maternal-fetal medicine, ultrasound unit | A separate practice from the hospital. Request directly, including ultrasound reports and images |
| Triage and admission | Presenting complaint, initial exam, first tracing interpretation, time of admission to L&D | Mother's hospital chart | The triage record can sit in an ED or OB triage module that prints separately |
| L&D nursing flowsheet | Timed vitals, cervical exams, FHR category and features, contraction assessment, notifications, interventions | Mother's EHR, perinatal module | Printed views vary. Ask for every flowsheet row with entry and documentation times |
| EFM tracing | Continuous fetal heart rate and uterine activity from admission to delivery | Perinatal surveillance or fetal monitoring archive, often a separate system | Produced as printed segments, or not at all. Chapter 5 covers this |
| Oxytocin and medication records | Start time, every rate change, stops, other labor drugs | MAR, infusion pump data, flowsheet | Rate changes charted late or rounded. Ask whether pump logs exist |
| Physician and midwife notes | Labor progress, tracing reviews, decision to deliver | Mother's EHR | Late entries and addenda after the outcome. Get version history |
| Anesthesia record | Epidural, maternal blood pressure, time into the OR, incision time | Anesthesia information system | A separate system with its own clock, often the most precise times in the file |
| Delivery or operative note | Mode of delivery, indication, findings (rupture, abruption, cord), birth time | Mother's EHR | Compare its times against nursing and anesthesia |
| Cord blood gas report | Arterial and venous pH, pCO2, pO2, base deficit, collection and result times | Lab system, either chart | Samples unlabeled or swapped; result time far after birth |
| Neonatal resuscitation record and Apgars | Timed steps: ventilation, intubation, compressions, epinephrine, team arrival, Apgars at 1, 5, 10 minutes and later | Paper form scanned into the baby's chart, or a code sheet | Written after the event, rounded times, scanned at poor quality |
| NICU admission and course | Neuro exam, encephalopathy staging, seizures, organ labs, ventilation, pressors | Baby's chart at the birth or receiving hospital | Transfer splits it across 2 hospitals |
| Transport record | Referral call time, team arrival, passive or active cooling in transit, temperatures | Transport team chart | Owned by the receiving hospital or a transport service; rarely produced unasked |
| Cooling log | Eligibility screen, start time, target and actual temperatures, duration, rewarming | NICU flowsheet and the cooling device's own data | Device data is not in the chart; ask whether it was saved |
| EEG and amplitude-integrated EEG | Background, seizures, timing | Neurophysiology system | Reports produced, raw recordings not |
| Neonatal MRI and MRS | Pattern and extent of injury | Radiology and PACS | Reports without images; hour of life at each scan missing |
| Placental pathology | Infection, inflammation, vascular lesions, cord findings | Pathology department | Placenta never sent, or slides and blocks not preserved |
| Genetic and metabolic workup | Alternative causes | Baby's chart, genetics clinic, outside labs | Results arrive months later from outside labs |

3 rows usually sit outside the core chart: the tracing, the device data (monitor, pumps, cooling machine) and the images. Name each in its own paragraph. A generic "complete medical record" request gets whatever the release-of-information print profile includes, and that profile was built for continuity of care, not litigation.

#### Put the request on the right legal footing

For a patient's own request, HIPAA gives a right of access to protected health information in the designated record set, and [45 CFR 164.524(b)(2)](https://www.law.cornell.edu/cfr/text/45/164.524) requires the covered entity to act "no later than 30 days after receipt of the request," with 1 extension allowed. Section 164.524(c)(2)(ii) adds that where the information is kept electronically and the person asks for an electronic copy, the entity must provide it "in the electronic form and format requested by the individual, if it is readily producible in such form and format," or in an agreed readable electronic format if not. That is your footing for asking for the tracing as data rather than paper. In litigation, subpoenas and requests for production do the same job under your jurisdiction's rules. For the full request workflow, see [how to get medical records for a lawsuit](https://medrecords.ai/guides/how-to-get-medical-records-for-a-lawsuit/).

- **Mother's chart and baby's chart requested separately** *Separate record numbers, separate authorizations*
- **Prenatal records from the outpatient practice** *Growth, infection and dating evidence often lives only there*
- **Complete EFM tracing in native electronic format** *Admission to delivery, with event marks and annotations*
- **Flowsheet rows with entry times** *Separates when care happened from when it was charted*
- **Anesthesia record** *Often the tightest clock for OR entry and incision*
- **Cord gas lab report with sample labels and times** *Arterial versus venous, collected versus resulted*
- **Resuscitation record and expanded Apgar form** *The original scan, not a retyped summary*
- **Transport record** *Referral time and cooling in transit*
- **Cooling flowsheet and device data** *Start time and temperature curve*
- **EEG and aEEG reports and recordings** *Seizure timing*
- **MRI and MRS images in DICOM with reports** *Every study, with the hour of life*
- **Placental pathology report, slides and blocks** *Or written confirmation it was never sent*

**0** of 12 checked

If you remember one thing
Request the tracing, the device data and the images by name, in separate paragraphs, in the first request. "Complete medical record" gets you the print profile, and the print profile leaves out the 3 things the experts need most.

Chapter 5 Building

### The missing strip problem

The strip arrives as a few printed pages covering the last 40 minutes, or a scan of a scan with an unreadable time axis, or not at all, with a letter saying the tracing is not part of the legal medical record. Each version leaves the experts arguing about a labor they cannot see.

The cause is structural. In many hospitals the tracing is captured and stored by a perinatal surveillance system that feeds the EHR only summaries, or nothing. The release-of-information office prints what its EHR print profile includes. Unless someone asks the perinatal system's owner, often nursing informatics or clinical engineering, for an export, the tracing never enters the production.

#### What a complete tracing request covers

1. **The whole period.** From the first minute of monitoring in triage to delivery, including any antepartum testing (nonstress tests) done in the same admission or earlier visits.
2. **The native electronic format.** The archive's own export, or a complete export that preserves the time axis, both channels and the system's time stamps. Printed segments are a fallback, not a substitute.
3. **Annotations and event marks.** Many systems let nurses mark events (position change, exam, medication) on the tracing itself. Those marks are stored with the tracing, not the flowsheet.
4. **Signal source.** Whether the heart rate came from an external ultrasound transducer or a fetal scalp electrode, and when that changed. A maternal heart rate recorded as fetal is a known failure of external monitoring, and the source channel is how an expert checks.
5. **A gap statement.** A written account, signed by the custodian, of every period with no data, and why: disconnection, transfer to the OR, system downtime, archive failure.
6. **Retention and audit data.** When the tracing was archived, whether it was ever edited or annotated after delivery, and who viewed or printed it. See [EHR audit trails in medical malpractice](https://medrecords.ai/guides/ehr-audit-trail-medical-malpractice/) for how to ask for access logs.

#### When the strip comes back incomplete

Treat an incomplete strip as a fact to document, not a dead end. The decision path below is how we would work it. It does not decide what the gap means. It decides what to ask for next and what to record.

**Incomplete EFM tracing: what to do next** — workflow — Illustration
1
**Does the production cover admission to delivery with no gaps in time?**
YesLog the time range, format and page count. Check the time axis against the flowsheet clock (chapter 6) and move on to the timeline.
NoMap exactly which minutes are missing before asking anyone about them. Go to 2.
2
**Was it produced as printed segments or a scan, not an electronic export?**
YesAsk for the native export by name, citing the electronic form and format rule at 45 CFR 164.524(c)(2)(ii) for a patient request, or your discovery rules. Ask who administers the perinatal system.
NoAn electronic export with gaps: go to 3.
3
**Do other records show monitoring during the missing minutes?**
YesFlowsheet entries describing the tracing, event marks, or notes saying "strip reviewed" during the gap show data existed. List each 1 with its page cite and ask the custodian to explain the gap in writing.
NoNothing refers to monitoring then: ask whether the patient was off the monitor (transfer, ambulation, OR move) and where that is documented.
4
**Does the custodian say the data was lost or never stored?**
YesAsk for the retention policy in force on the delivery date, downtime logs, and the date of loss. Preservation letters and spoliation remedies vary by state; counsel decides the next step.
NoKeep the request open with a date. Record every response in the missing-records log.
Every branch ends in a documented request or a documented answer. A gap nobody wrote down helps no one.

#### Virginia wrote the rule into its statute

At least 1 state has settled whether a tracing is part of the medical record in its birth injury program. Virginia's compensation act says:

"As used in this chapter, fetal monitoring strips, whether printed or in electronic format, shall be deemed to constitute part of the medical records relating to an infant who allegedly suffered a birth-related neurological injury."

The same section requires a hospital, on written notice that a legal representative intends to file a petition, to "promptly deliver to such person all available medical records relating to the infant" ([38.2-5004(E)](https://law.lis.virginia.gov/vacode/title38.2/chapter50/section38.2-5004/) ). The rule applies inside that chapter, and chapter 9 covers what the program is. Outside Virginia's program, whether a tracing is part of the "medical record" for a given request depends on the hospital's legal health record definition, state records law and the discovery rules of your court.

**Do not let a printed strip become the only copy.** If the hospital offers to print the tracing for you, ask first whether the electronic version still exists and ask that it be preserved. A printout is a derivative. The archive is the evidence.

If you remember one thing
Map the missing minutes before you complain about them. A list of gaps, each tied to a flowsheet line that shows monitoring was running, turns "the strip is incomplete" into a specific question the custodian has to answer in writing.

Chapter 6 Building

### Clocks, cord gases, Apgars and the resuscitation record

The first hour of life produces the densest records in the file and the least reliable times, because much of it is written afterward from memory or a scribe's paper notes. 3 of the 2014 criteria (Apgars, cord gases, organ injury onset) come from this hour.

#### Clocks first

Before you trust any interval, list every clock in the file. The fetal monitor has a clock. The EHR stamps entries with the time the nurse chose and the time she saved. The anesthesia system logs its own times. The lab analyzer stamps the cord gas. The resuscitation form carries whatever time someone read off a wall clock or an Apgar timer. These can differ by minutes, and nobody documents the offset.

Find an anchor event that appears in 2 or more systems, usually birth itself: the tracing ends, the anesthesia record logs delivery, the nursing record logs birth time, the newborn chart logs birth time. Compare them. If the monitor is 4 minutes ahead of the EHR, write that down with the cites, and adjust nothing silently. Every entry in your timeline should carry its source and its source clock.

Anchor: time of birth — Time shown — Source
Operative note — 04:12 — hypothetical
Anesthesia record, delivery event — 04:11 — hypothetical
L&D nursing flowsheet, birth time — 04:12 — hypothetical
Last data on EFM tracing — 03:58 — hypothetical
Resuscitation record, time of birth — 04:10 — hypothetical
Spread across sources — 2 minutes — Tracing ends 14 minutes before birth: log it as a gap

#### Cord gases: read the lab report, not the summary

The 2014 report says umbilical artery pH and base deficit "should be obtained in cases for which fetal metabolic status is in question," and that both vessels "should be sampled primarily to ensure that the artery has been sampled." The artery reflects the fetus before placental exchange; the vein reflects blood after it. A venous value reported as arterial makes the baby look better than the fetus was. Arterial and venous values that are nearly identical can mean the same vessel was sampled twice.

Time also changes the numbers. In clamped cord segments, the report cites data that pH, pO2 and pCO2 stay essentially unchanged for up to 60 minutes, but base deficit does not: umbilical artery base deficit rose by 1.2 mmol/L after 20 minutes in room air and by 4.5 mmol/L after 60 minutes, and lactate rose by 44% at 20 minutes and 245% at 60. The report says to double clamp the cord promptly and analyze quickly, and that when analysis is delayed, base deficit and lactate "should be interpreted with caution." So your timeline needs 3 cord gas times: birth, collection and result. A result stamped 70 minutes after birth is a question for the experts, not a footnote.

#### Apgars and the resuscitation record

The report says the Apgar score for an infant undergoing resuscitation "should be recorded on an expanded Apgar score form," which captures the resuscitative measures in use at each scoring time. Ask for that form by name. Then check the 5-minute and 10-minute scores against 3 places: the delivery record, the resuscitation record and the newborn admission note. They disagree more often than you would hope, because each was written by a different person at a different time.

For the resuscitation itself, list each step with its time: positive-pressure ventilation, intubation, chest compressions, epinephrine doses, first heart rate over 100, team arrivals. Round numbers (every event at :00 or :05) suggest a reconstruction after the fact. That is not wrongdoing, but it lowers the weight of those times against the anesthesia log and the monitor.

If you remember one thing
Write down every clock and the offset between them before you compute a single interval. Then give the cord gas 3 times, not 1: birth, collection and result.

Chapter 7 Building

### Cooling, EEG, MRI and the placenta: the records that date an injury

The neonatal records after the first hour do 2 jobs. They document treatment, and they carry most of the evidence both sides use to argue when the injury happened. Request them as carefully as the labor records.

#### Therapeutic hypothermia

Cooling is the treatment for moderate to severe HIE that the 2014 report calls "the first efficacious neural rescue therapy for neonatal encephalopathy." The large trials it reviews shared a design: infants at or near term, randomization within 6 hours of birth in all but 1 study, 72 hours of cooling, then slow rewarming at 0.5 C per hour. Trials that used whole-body cooling targeted a rectal temperature of 33.5 C. The report's box of clinical trial criteria for whole-body cooling lists gestational age above 36 weeks, NICU admission at less than 6 hours of age, and a cord or first-hour gas with pH of 7.0 or less or base deficit of 16 mmol/L or more, with additional criteria when the gas is borderline or unavailable.

Those are trial criteria, not a rule for every NICU today; local protocols differ. They tell a records team which data points the cooling decision rests on. The cooling record set:

- The eligibility screen: the form or note documenting the gas, the neurologic exam, the encephalopathy staging and the decision, with its time.
- The referral call and transport record when the baby was born at a hospital without cooling, including any passive cooling or temperature readings in transit.
- The cooling start time as charted, and the device's own log if it kept 1.
- The temperature flowsheet across 72 hours and rewarming.
- Any documented reason cooling was not offered, delayed or stopped early.

The interval from birth to cooling start is 1 of the entries both sides will compute. It depends on the birth time you anchored in chapter 6, which is another reason to settle the clocks first.

#### EEG and amplitude-integrated EEG

Several cooling trials used amplitude-integrated EEG or EEG to grade encephalopathy for enrollment. Request the EEG reports, ask whether recordings are retained, and put seizure onset times on the timeline with their cites.

#### MRI and spectroscopy

The report says early MRI between 24 and 96 hours of life "may be more sensitive for the delineation of the timing of perinatal cerebral injury," while an MRI "optimally at 10 days of life (with an acceptable window between 7 days and 21 days of life)" best shows the full extent. It names deep nuclear gray matter and watershed cortical injury as typical hypoxic-ischemic patterns, and lists patterns such as focal arterial infarction, venous infarction and isolated hemorrhage as suggesting that peripartum hypoxia-ischemia did not play a role. It also says that even when the pattern fits, imaging "cannot determine the etiology of the hypoxia-ischemia, such as placental insufficiency or interruption of umbilical cord blood flow."

For the file: every study, the hour of life at each scan (from your anchored birth time), the sequences performed, the report and the DICOM images. The reading belongs to a pediatric neuroradiologist; your job is to give them every study and its exact time.

#### The placenta

The placenta is the record most often destroyed before anyone asks for it. The report states that gross and microscopic examination of the placenta, "if still available, should be performed in cases of neonatal encephalopathy." It also reports how rarely that happens: placental examination is performed in fewer than half of deliveries that meet the consensus indications, only 11.2% of placentas were examined in 1 series of infants with neonatal encephalopathy, and only 24% in a study of children with cerebral palsy.

About 1 of 10 — 11.2%
Placentas examined in 1 series of infants with neonatal encephalopathy, all of which met criteria for examination. ACOG and AAP, 2014, chapter 4.

If there is a pathology report, request the slides and blocks too, so a second pathologist can read them. If there is none, get a written answer: was the placenta sent, held or discarded, and under what policy. Both sides' experts will be asked about it, and "we don't know" is the worst answer.

If you remember one thing
Give every neonatal record an hour of life, computed from 1 anchored birth time. Cooling start, seizure onset and each MRI are timing evidence only when their hours are right.

Chapter 8 Building

### Worked example: a minute-by-minute intrapartum timeline

Standard chronologies go day by day. An HIE timeline has to go minute by minute across the last hours of labor, pulling from every source at once, and it has to show what is missing as clearly as what is there. The example below shows the format. Every fact in it is invented.

**Hypothetical.** A 39-week induction with oxytocin. The mother's chart runs 1,412 pages; the baby's birth hospital chart runs 388 pages and the receiving NICU chart 2,960 pages. The hospital produced the EFM tracing as 64 printed pages. Times below are as recorded in each source, with the source clock noted. Nothing here is drawn from a real case or a customer file.

**Hypothetical intrapartum and neonatal timeline** — hypothetical — Illustration
1. 21:30
**Oxytocin started at 2 mU/min**
Flowsheet shows Category I at start. Pump log not produced.

Mother, MAR p. 612; flowsheet p. 488
2. 01:10
**Oxytocin at 14 mU/min, 6 contractions in 10 minutes charted**
Nurse charts Category II, recurrent variable decelerations, provider notified. No rate change charted.

Flowsheet p. 503; EFM printout p. 38
3. 01:40 to 02:55
**No tracing produced**
Printed pages jump from 01:40 to 02:55. Flowsheet entries at 02:00 and 02:30 describe the tracing, so the monitor was running.

EFM printout pp. 44 to 45; flowsheet pp. 507, 509
4. 03:05
**Oxytocin stopped; minimal variability, recurrent late decelerations charted**
Category III charted at 03:15. Obstetrician paged per nursing note.

Flowsheet p. 511; nursing note p. 540
5. 03:32
**Decision for cesarean delivery**
Physician note (entered 06:48, labeled late entry) records decision at 03:32. Nursing note records "MD at bedside, plan for C/S" at 03:35.

Physician note p. 552; nursing note p. 541
6. 03:49
**In operating room**
Anesthesia record logs patient in room. EFM printout ends at 03:58 (monitor clock).

Anesthesia record p. 590; EFM printout p. 64
7. 04:07
**Incision**
Anesthesia record 04:07; operative note 04:08. Decision-to-incision interval 35 minutes on the physician's decision time, 32 minutes on the nursing note.

Anesthesia record p. 591; operative note p. 560
8. 04:11
**Birth**
Apgars 1, 2 and 3 at 1, 5 and 10 minutes on the resuscitation record; delivery record shows 1, 3, 4.

Baby, resuscitation record p. 12; delivery record p. 8
9. 04:14
**Cord gases collected**
Arterial pH 6.94, base deficit 15; venous pH 7.08. Resulted 05:24, 70 minutes after collection.

Baby, lab report p. 31
10. 06:25
**Transport team arrives; passive cooling started**
Referral call logged at 05:02.

Transport record pp. 2 to 3
11. 08:40
**Active whole-body cooling started at receiving NICU**
4 hours 29 minutes after birth. Device log not produced.

NICU flowsheet p. 144
12. Day 1 to 5
**Placental pathology: none in either chart**
Delivery record says "placenta to pathology." No report produced.

Mother, delivery record p. 562

The timeline produces 3 intervals both sides will argue about (decision to incision, collection to result, birth to cooling) and 2 gaps to chase (75 minutes of tracing, the placenta).

#### What the example teaches

**The decision time has 2 sources, and the better-looking 1 is a late entry.** The physician's decision time comes from a note written more than 3 hours later. That is not improper by itself, but the timeline should show both times and cite both. See [altered medical records](https://medrecords.ai/guides/altered-medical-records/) for how to tell a labeled late entry from a problem.

**The interval is a fact; its meaning is not.** The 2014 report reviews the so-called 30-minute rule, the capability to begin a cesarean within 30 minutes of the decision. It calls the limit arbitrary, says some conditions such as cord prolapse or uterine rupture need delivery as soon as possible, and cites a review of 11,481 primary cesareans at 13 NICHD network centers in which only 65% of emergency cases began within 30 minutes, with no significant difference in HIE between the under and over 30-minute groups (0.7% versus 0.5%). Put the interval in the timeline, cite both ends, and leave its weight to the experts.

**The gap is documented, not guessed.** The 75 minutes of missing tracing is flagged with the 2 flowsheet entries that prove the monitor was running. That list goes straight into the next request (chapter 5).

If you remember one thing
Every timeline entry carries its time, its source clock and its page cite, and every interval is shown with both ends. When 2 records disagree, show both. The disagreement is often the most useful line in the timeline.

Chapter 9 Deciding

### Virginia and Florida: no-fault birth injury programs

Virginia and Florida run no-fault compensation programs for a defined class of birth-related neurological injuries. Whether a case falls inside a program, and what that means for a civil claim, are legal questions with their own case law; ask counsel in that state. For a records team, both statutes do something unusual: they write the medical record requirements into the claim itself.

##### Virginia

StatuteBirth-Related Neurological Injury Compensation Act, [Va. Code 38.2-5000 and following](https://law.lis.virginia.gov/vacode/title38.2/chapter50/section38.2-5001/)
Injury definedBrain or spinal cord injury from oxygen deprivation or mechanical injury in labor, delivery or resuscitation in a hospital, leaving the infant permanently motorically disabled and developmentally or cognitively disabled, and permanently needing assistance in all activities of daily living. Excludes genetic or congenital abnormality, degenerative neurological disease and maternal substance abuse
Where filedElectronic petition to the Virginia Workers' Compensation Commission, $15 filing fee
Records dutyPetition must include all available relevant medical records and identify unavailable records and why. Hospitals must promptly deliver records on written notice. Fetal monitoring strips, printed or electronic, are deemed medical records (38.2-5004)

##### Florida

StatuteBirth-Related Neurological Injury Compensation Plan, [Fla. Stat. 766.301 to 766.316](https://www.flsenate.gov/Laws/Statutes/2025/766.302)
Injury definedBrain or spinal cord injury from oxygen deprivation or mechanical injury in labor, delivery or resuscitation in the immediate postdelivery period in a hospital, leaving the infant permanently and substantially mentally and physically impaired. Birth weight at least 2,500 grams for a single gestation, 2,000 grams for each infant in a multiple gestation. Excludes genetic or congenital abnormality
Where filedPetition to the Division of Administrative Hearings, $15 filing fee
Records dutyWithin 10 days after the petition, the claimant furnishes the association all available relevant medical records and a list of unavailable records with reasons. The association has 45 days from service of a complete claim to respond (766.305)
Read those records duties together and 1 point stands out. Both statutes expect a list of what is missing and why, filed with the petition. That is the same missing-records log this guide recommends building anyway. A team that keeps a cited gap list from the first production is ready to file 1; a team that discovers gaps at the end is not.

The definitions turn on records too. Florida's birth weight floor comes from the delivery record, both statutes exclude genetic or congenital causes, and both require injury in labor, delivery or resuscitation: a timing question the same timeline answers.

If you remember one thing
In Virginia and Florida, the program statute requires a list of unavailable records with reasons. Keep that list, with page cites, from the first day of review.

Chapter 10 Everyone

### Where AI helps and where it fails on labor and delivery records

An HIE production suits AI medical record review in 1 way and not in another. It suits it because thousands of pages across 2 or 3 charts are timed flowsheet rows, labs and notes that a large language model (LLM) can read, sort and cite faster than a person. It does not because the deciding evidence is a tracing, an MRI and a placenta slide, and none of those is text.

By hand
Build the timeline — Days of indexing 2 charts, retyping times
Find conflicting times — Only if the reviewer happens to compare
Missing records — Noticed when an expert asks
Cite each fact — Manual page references, often dropped
With AI and a human reviewer
Build the timeline — Draft AI medical chronology across both charts, then human review
Find conflicting times — Same event with different times flagged for a person to resolve
Missing records — Referenced but unproduced records flagged with the line that implies them
Cite each fact — Page-level citation on every line, checked by the reviewer

#### Where it fails

- **Tracings and images.** An LLM reading a scanned strip sees pixels and printed labels. It cannot categorize a tracing, and nobody should let it try. The same goes for MRI. Those are expert tasks.
- **OCR on scanned forms.** Resuscitation records and expanded Apgar forms are often handwritten paper scanned in. Optical character recognition (OCR) and handwriting recognition misread 1s and 7s and 03:15 versus 08:15. Low-confidence pages should be flagged, not smoothed over.
- **Copy-forward and templated text.** Labor notes copied forward, and notes drafted by ambient AI scribes, can carry a time or finding from an earlier entry. A model that trusts the text will put the wrong time on the timeline.
- **Hallucination.** Generative AI can state a fact no page supports. In a timeline where 3 minutes matter, 1 invented time is enough to wreck an expert's credibility.

The answers are the same ones courts now expect. Every line grounded in a page-level citation so a person can check it in 1 click. A human-in-the-loop reviewer who checks the cites before anything leaves the firm. Lawyers have been sanctioned under Rule 11 for filing fabricated AI-generated citations (*Mata v. Avianca*, S.D.N.Y. 2023); a fabricated clinical time in an expert disclosure is the same failure in a different document.

#### Vendor checklist for HIPAA compliant AI on birth injury files

1 —

##### A citation on every line

Each timeline entry links to the source page. No citation, no use.

2 —

##### A signed BAA, SOC 2, no training on your data

Birth records are PHI for 2 patients. Get the business associate agreement before the first upload.

3 —

##### Handles 2 charts in 1 matter

Mother and baby as separate sources with their own page ranges, cited together in 1 timeline.

4 —

##### Flags low-confidence OCR

Scanned resuscitation forms must be marked for a human read.

5 —

##### Says what it will not do

No tracing interpretation, no causation, no standard-of-care verdicts. A tool that claims those is a liability in a deposition.

If you remember one thing
Use AI for the text: the flowsheets, notes, labs and the gap list. Keep the tracing, the images and every judgment about cause with the experts, and check every cited line before it leaves the office.

Chapter 11 Building

### Templates: the records request checklist and the timeline worksheet

Copy these into your own forms. Adjust the legal basis line to your jurisdiction and the type of request.

##### 1. HIE records request checklist

Use it for the first request to each facility, 1 copy per patient (mother, baby) per hospital.

PATIENT: [MOTHER / INFANT] [NAME], DOB [DATE], MRN [IF KNOWN]
FACILITY: [HOSPITAL] DATES: [ADMISSION] to [DISCHARGE OR TRANSFER]
BASIS: [HIPAA ACCESS REQUEST, 45 CFR 164.524 / AUTHORIZATION / SUBPOENA]

Please produce each item below. For any item not produced, state in
writing whether it exists, where it is kept and why it is not produced.

[ ] Prenatal records received from outside providers
[ ] Triage and L&D admission records
[ ] L&D nursing flowsheet, all rows, with entry and documentation times
[ ] Complete EFM tracing, first monitoring to delivery, in native
 electronic format with event marks, annotations, signal source,
 and a written statement of any gaps
[ ] Oxytocin and all labor medications: MAR, infusion pump data
[ ] Physician and midwife notes, with version history and addenda
[ ] Anesthesia record
[ ] Delivery or operative note
[ ] Cord blood gas report: vessel labels, collection and result times
[ ] Neonatal resuscitation record and expanded Apgar form (original scan)
[ ] Newborn admission, NICU records, transport record
[ ] Cooling eligibility screen, cooling flowsheet, device data
[ ] EEG and aEEG reports and recordings
[ ] MRI and MRS reports and DICOM images, every study
[ ] Placental pathology report, slides and blocks, or written
 confirmation of disposition and the policy applied
[ ] Genetic and metabolic test results
[ ] Retention policy for fetal monitoring data in force on [DATE]

##### 2. Intrapartum timeline worksheet

1 row per event. Fill the clock column before computing any interval.

CLOCK ANCHOR: birth time per [SOURCE, PAGE] = [HH:MM]
KNOWN OFFSETS: EFM [+/- MIN] EHR [+/- MIN] ANESTHESIA [+/- MIN] LAB [+/- MIN]

TIME | EVENT | SOURCE + PAGE | SOURCE CLOCK | ENTRY TIME IF LATE | CONFLICT? | GAP?
-----|-------|---------------|--------------|--------------------|-----------|-----
[ ] | Oxytocin start / each rate change / stop
[ ] | Each FHR category charted (and tracing page)
[ ] | Each provider notification and response
[ ] | Decision to deliver (every source that states it)
[ ] | In OR / incision / birth
[ ] | Apgars 1, 5, 10 min (every source)
[ ] | Resuscitation steps
[ ] | Cord gas collected / resulted (artery, vein)
[ ] | Referral call / transport arrival / cooling start
[ ] | Seizure onset / EEG / each MRI (hour of life)

INTERVALS (show both ends and both sources):
Decision to incision: [ ] min Collection to result: [ ] min
Birth to cooling start: [ ] h [ ] min

MISSING-RECORDS LOG: item | evidence it exists (page) | date requested | answer
If you remember one thing
The worksheet's conflict and gap columns are the ones experts read first. Fill them in as you go, not at the end.

Chapter 12 Publisher

### What Medrecords AI does with an HIE file

Medrecords AI is medical chronology software. It works on the records you upload; it does not request or retrieve records from hospitals. On an HIE file it does 4 things:

- Builds a [cited chronology](https://medrecords.ai/product/chronology/) across the mother's and the baby's charts, with [a citation on every line](https://medrecords.ai/product/citations/) back to the source page, so the timeline in chapter 8 starts as a draft instead of a blank sheet.
- [Flags missing records](https://medrecords.ai/product/missing-records-identification/) the file implies, such as a delivery record that says "placenta to pathology" with no report, or flowsheet lines that describe a tracing for minutes the production lacks. Each flag cites the line that implies it.
- Routes scanned and handwritten pages through [OCR](https://medrecords.ai/product/ocr/) and flags low-confidence pages for a human read.
- Loads MRI studies as DICOM in a browser viewer and places them on the timeline through [imaging review](https://medrecords.ai/product/imaging/), so your neuroradiologist can cite a slice like a page.

What it does not do: it does not interpret fetal heart rate tracings or MRI images, does not decide causation or standard of care, and does not score a case. Flags are signals, not verdicts. It runs under SOC 2 and HIPAA with a signed BAA (see [security](https://medrecords.ai/security/) and [HIPAA](https://medrecords.ai/hipaa/) ). Self-Service bills 10 cents a deduplicated page, down to 5 cents at volume, duplicates free; Enterprise On-Prem is an annual license. More on the use case: [birth injury record review](https://medrecords.ai/solutions/birth-injury-record-review/).

The offer

#### See a cited labor and delivery timeline from your own file.

Book a demo on an HIE 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.

[Book a demo](https://medrecords.ai/demo/) [See missing records identification](https://medrecords.ai/product/missing-records-identification/)
Scheduling only. No records move from a public page.

Chapter 13 Everyone

### Frequently asked questions

What records are needed for an HIE birth injury case?Prenatal records, triage and L&D nursing flowsheets, the complete fetal monitoring tracing, oxytocin records, physician, anesthesia and delivery notes, cord blood gases, the resuscitation record and Apgars, NICU, transport and cooling records, EEG, MRI reports and images, placental pathology and any genetic workup. Mother and baby have separate charts.
Why is the fetal monitoring strip often missing from the chart?Many hospitals store the tracing in a separate perinatal system, so a standard chart request does not produce it. Ask for it by name, for the whole labor, in native electronic format.
What are the ACOG criteria for HIE?The 2014 ACOG and AAP report lists signs that make an intrapartum hypoxic-ischemic contribution more likely: Apgar under 5 at 5 and 10 minutes, umbilical artery pH under 7.0 or base deficit of 12 mmol/L or more, a typical MRI pattern, multisystem organ failure, a sentinel event and certain tracing patterns. They weigh probability; they do not prove timing or cause.
What is the NICHD 3-tier fetal heart rate system?The 2008 NICHD, ACOG and SMFM classification. Category I is normal, Category III is abnormal (absent variability with recurrent decelerations or bradycardia, or a sinusoidal pattern), and Category II is everything else.
How soon must cooling start?The large cooling trials reviewed in the 2014 report randomized infants within 6 hours of birth in all but 1 study and cooled for 72 hours. Current protocols vary by NICU; whether timing met the standard is an expert question.
Can AI read a fetal heart rate tracing?Not for litigation. AI tools can index the pages and cite what nurses charted about the tracing, but categorizing the tracing is a task for a qualified obstetric or nursing expert.
Is it HIPAA compliant to upload birth records to an AI tool?It can be, with a vendor that signs a business associate agreement, holds SOC 2, does not train on your data and logs access. Consumer chatbots without a BAA are the wrong place for 2 patients' PHI.
Can software decide whether HIE was caused by the delivery?No. Software can build a cited timeline and flag missing records. Causation and standard of care belong to medical experts, under the expert rules of your state.
Can ChatGPT build a birth injury timeline?A general chatbot can summarize text you paste, but it gives no page citations and can invent times. For a timeline an expert will rely on, use a tool that cites every line to its source page and check each cite.

Chapter 14 Everyone

### Sources and method

Clinical criteria and figures were checked against the full text of the ACOG and AAP report in September 2026; quotes are verbatim except that dashes in compound terms are rendered as hyphens. Virginia and Florida statutes were read on the legislatures' sites. The timeline, anchor table and chart sizes are hypothetical. Nothing here is legal or medical advice.

- [ACOG and AAP, *Neonatal Encephalopathy and Neurologic Outcome*, 2nd ed.](https://www.acog.org/-/media/project/acog/acogorg/clinical/files/task-force-report/articles/2014/neonatal-encephalopathy-and-neurologic-outcome.pdf) (2014, reaffirmed 2019): executive summary; chapter 4 (placenta); chapter 6 (FHR categories, tachysystole, cord gases, 30-minute rule); chapter 7 (expanded Apgar); chapter 10 (imaging); chapter 11 (cooling trials).
- [Pediatrics 133(5):e1482](https://publications.aap.org/pediatrics/article/133/5/e1482/32738/Neonatal-Encephalopathy-and-Neurologic-Outcome), AAP publication of the report.
- [Macones and colleagues, 2008 NICHD workshop report](https://pubmed.ncbi.nlm.nih.gov/18757666/), Obstetrics and Gynecology; definitions as restated in the ACOG report.
- [45 CFR 164.524](https://www.law.cornell.edu/cfr/text/45/164.524), HIPAA right of access: (b)(2) and (c)(2)(ii).
- [Va. Code 38.2-5001](https://law.lis.virginia.gov/vacode/title38.2/chapter50/section38.2-5001/) and [38.2-5004](https://law.lis.virginia.gov/vacode/title38.2/chapter50/section38.2-5004/).
- [Fla. Stat. 766.302](https://www.flsenate.gov/Laws/Statutes/2025/766.302) and [766.305](https://www.flsenate.gov/Laws/Statutes/2025/766.305) (2025).
- *Mata v. Avianca, Inc.*, 678 F. Supp. 3d 443 (S.D.N.Y. 2023): Rule 11 sanctions for fabricated AI-generated citations.

Related guides: [EHR audit trails in medical malpractice](https://medrecords.ai/guides/ehr-audit-trail-medical-malpractice/), [altered medical records](https://medrecords.ai/guides/altered-medical-records/), [clinical practice guidelines and the standard of care](https://medrecords.ai/guides/clinical-practice-guidelines-standard-of-care/), and [record review for legal nurse consultants](https://medrecords.ai/solutions/legal-nurse-consultants/).

### More guides

- [**In-house vs outsourced medical record review** — In-house vs outsourced medical record review compared on cost, turnaround, data residency, and control, with a…](https://medrecords.ai/guides/in-house-vs-outsourced-medical-record-review/)
- [**Informed consent documentation: how to review it in the medical record** — How to review informed consent documentation in the chart: the 2 disclosure standards, CMS form rules, consent…](https://medrecords.ai/guides/informed-consent-medical-records/)
- [**Is an AI medical chronology defensible in court?** — Whether an AI medical chronology holds up in court comes down to verifiability, not the tool. What makes a…](https://medrecords.ai/guides/is-ai-accurate-enough-for-court/)
- [**Legal nurse consultant reports: what attorneys expect you to deliver** — The 8 reports a legal nurse consultant is asked for, from merit review to chronology, what each holds, how long…](https://medrecords.ai/guides/legal-nurse-consultant-reports/)
- [**Medical abbreviations for legal professionals: how to read the shorthand without misreading the case** — Read medical abbreviations in records for litigation: the Joint Commission Do Not Use list, the ISMP list, 150…](https://medrecords.ai/guides/medical-abbreviations-for-legal-professionals/)
- [**Medical billing review: how to catch overcharges and duplicates** — What medical billing review is, the errors reviewers look for, why bills must be checked against the records…](https://medrecords.ai/guides/medical-billing-review/)