Newborn · HIE Assessment
Thompson HIE Score
Alternative HIE severity score (0-22) to Sarnat staging. 9-item clinical exam; tracks daily; Day 4 score predictive of outcome. Thompson 1997.
Last reviewed 2 June 2026
Thompson HIE Score
9-item neonatal encephalopathy severity
Score all 9 items (0-3) to see total.
Educational tool only — not medical advice. Thompson HIE is an alternative to Sarnat for neonatal encephalopathy severity. Advantage: bedside-only, no equipment. Many UK cooling protocols use Thompson > 7 at 3-5 hours of life as cooling eligibility, alongside other criteria. Score serially through first 5-6 hours; peak score matters most for outcome prediction.
What does this mean?
The Thompson HIE score (Thompson 1997, Acta Paediatr) is a simple bedside alternative to Sarnat staging — 9 items, each 0–3, total 0–27. Many UK and South African cooling protocols use Thompson > 7 at 3–5 hours of life as the encephalopathy threshold for therapeutic hypothermia eligibility, alongside the standard gas/Apgar criteria. Peak score ≥ 15 over the first day has ~92 % PPV for abnormal neurodevelopmental outcome at follow- up. Scores typically peak at 24–72 h and improve over the first week if recovery is going to happen. The score is performed serially through the first 5–6 hours of life because the picture evolves rapidly — what is mild at one hour can become severe by three. Use alongside aEEG, cord/early gas, and clinical history; no single tool is sufficient on its own.
What is the Thompson HIE score?
Alternative severity score for newborn hypoxic-ischaemic encephalopathy. 9 clinical signs scored, total 0-22. Higher = worse. Thompson 1997.
How is it different from Sarnat?
- Sarnat: 3 categorical stages (mild/moderate/severe).
- Thompson: 0-22 numerical score; tracks daily.
Both validated. Thompson preferred for serial monitoring + research; Sarnat for initial cooling decision usually.
Severity bands
- ≤9: mild.
- 10-14: moderate.
- ≥15: severe.
The 9 items
- Tone (0-3).
- Level of consciousness (0-3).
- Fits (0-2).
- Posture (0-3).
- Moro reflex (0-2).
- Grasp (0-2).
- Suck (0-2).
- Respiration (0-3).
- Fontanelle (0-2).
Prognosis by score
- Day 1 <10: usually good outcome.
- Day 1 >15: poor outcome.
- Day 4 <10: predicts good outcome.
- Day 4 >15: predicts severe disability or death.
- Improving trajectory reassuring.
- Rising scores concerning.
Combined with MRI
MRI more predictive than Thompson alone. Both together = best prognostication. Thompson useful in resource-limited settings or before MRI.
Frequency
Daily during cooling + first week. Trained nurse / doctor. <5 min once familiar. Trended over time.
Investigations during HIE
- Cord blood gas, serial lactate.
- Glucose, electrolytes, calcium, magnesium.
- Renal + liver function.
- Coagulation (DIC risk).
- Sepsis screen.
- Cranial ultrasound day 1-2.
- MRI 4-10 days.
- EEG / aEEG continuous.
- Hearing screen.
- Cardiac echo.
Cooling alternatives (research)
- Erythropoietin — mixed evidence.
- Stem cell therapy — experimental.
- Xenon gas — TOBY-Xe trial; not standard.
- Melatonin — small trials.
Cooling remains gold standard.
Rehabilitation pathway
- Neonatal physiotherapy from NICU.
- Developmental paediatrician follow-up.
- Physio / OT / SALT at home.
- Hearing aid / glasses if needed.
- Seizure management.
- Education support (SEN, EHCP).
- Peer support: BLISS, Holdens Foundation UK.
Different scenarios
Scenario 1: Day 1 Thompson 6, mild HIE
Likely good outcome. Continued monitoring. May not need cooling depending on Sarnat.
Scenario 2: Day 1 Thompson 14, Day 4 dropped to 8
Improving trajectory. Reassuring. Good prognosis likely.
Scenario 3: Day 4 Thompson 18, MRI shows severe BGT injury
Concerning. Counselling about likely disability. Early intervention planning.
Scenario 4: Sarnat 2 + Thompson 11 + cooling completed
Standard pathway. Follow-up MRI + hearing + developmental.
Scenario 5: Normal Thompson + normal MRI + discharged
Excellent prognosis. Routine follow-up.
Care guidance — Thompson HIE
- Daily tracking provides trajectory.
- Day 4 score most predictive.
- Combine with MRI + EEG + clinical.
- Improving = reassuring.
- Rising = investigate (seizure, sepsis, metabolic).
- BLISS UK support for families.
Sources
- Thompson CM, et al. The value of a scoring system for hypoxic ischaemic encephalopathy. Acta Paediatr 1997.
- NICE IPG347. Therapeutic hypothermia for HIE.
- BLISS UK. bliss.org.uk.
Recommended for this calculator
Frequently asked questions
What is the Thompson HIE score?
ALTERNATIVE SEVERITY SCORE for newborn HYPOXIC-ISCHAEMIC ENCEPHALOPATHY. 9 clinical signs scored 0-3 each (some 0-2). TOTAL 0-22. HIGHER = WORSE. Developed by Thompson 1997 (South Africa). 9 ITEMS: tone, level of consciousness, fits, posture, Moro reflex, grasp, suck, respiration, fontanelle. ADVANTAGES: more granular than Sarnat 3 categories; tracks change over time; correlates with outcomes.
How is it different from Sarnat?
(1) SARNAT — 3 categorical stages (mild/moderate/severe); quick; categorical. (2) THOMPSON — 0-22 numerical score; tracks daily; predictive at day 1, 4, 7. BOTH validated. THOMPSON sometimes preferred in research/serial monitoring; SARNAT often used for initial cooling decision. SOME UNITS use both; sometimes either alone. INTERPRETATION: Thompson ≥15 = severe; 10-14 = moderate; ≤9 mild.
When is Thompson used?
(1) MONITORING progress during cooling + after; (2) Predicting outcomes — Day 1 + Day 4 scores have predictive value; (3) RESEARCH protocols; (4) ALTERNATIVE to Sarnat. AT BEDSIDE: clinical exam; can be repeated daily — useful for trajectory tracking; UNUSUAL in low/middle-income countries where access to MRI / EEG limited.
What about each Thompson item?
(1) TONE: 0 normal, 1 hyper-tone, 2 hypo-tone, 3 flaccid. (2) LEVEL OF CONSCIOUSNESS: 0 normal, 1 hyperalert, 2 lethargic, 3 comatose. (3) FITS: 0 none, 1 less than 3/day, 2 more than 2/day. (4) POSTURE: 0 normal, 1 fisting, 2 strong distal flexion, 3 decerebrate. (5) MORO REFLEX: 0 normal, 1 partial, 2 absent. (6) GRASP: 0 normal, 1 weak, 2 absent. (7) SUCK: 0 normal, 1 weak, 2 absent/bites. (8) RESPIRATION: 0 normal, 1 hyperventilation, 2 brief apnoea, 3 prolonged apnoea/IPPV. (9) FONTANELLE: 0 normal, 1 full not tense, 2 tense.
Does my baby need cooling based on Thompson?
USUALLY combined with Sarnat + clinical + biochemical criteria. THOMPSON ≥7-10 (varies by protocol) supports cooling decision when other criteria met. SARNAT typically primary indicator. SHARED clinical judgment + biochemical (cord pH <7.0, base deficit ≥12), gestational age ≥36 weeks, age <6 hours. SOME centres use Thompson as primary criterion; UK mainly Sarnat-based pathway.
What's the prognosis by Thompson score?
(1) DAY 1 SCORE: low scores (<10) — usually good outcome; very high (>15) — poor outcome; middle range — variable. (2) DAY 4 SCORE: more predictive than day 1. >15 by day 4 — strongly predicts severe disability or death. <10 by day 4 — predicts good outcome. (3) PEAK SCORE: peak score often predictive. (4) TRAJECTORY: improving scores reassuring; rising scores concerning. NOT ABSOLUTE — combined with MRI, EEG, clinical.
What about MRI findings + Thompson?
MRI more predictive than Thompson alone. THOMPSON useful in resource-limited settings or when MRI not yet done. BOTH together: best prognostication. NORMAL MRI in first week + low Thompson — excellent outlook. SEVERE MRI + high Thompson — predicts severe disability. DISCREPANCIES: investigate; sometimes MRI worse than clinical or vice versa.
How often is Thompson done?
DAILY during cooling + first week. CAN BE done by trained nurse / doctor. TAKES <5 minutes once familiar. SCORES RECORDED in notes; trended over time. PARENTS often shown — provides objective marker of progress. SCORES TYPICALLY: high at start; drop with cooling response; rise concerning. STOP doing routinely once baby stable + discharged from NICU.
Will my baby's score change?
YES — scores fluctuate. EARLY HOURS: peak severity often. DURING COOLING: gradual improvement in most. POST-REWARMING: stable or improving. PERSISTENT high score: concerning prognosis. SUDDEN deterioration: new event (seizure, sepsis, metabolic) — needs investigation. PARENTAL UPDATES: daily ward rounds explain changes.
Are there cooling alternatives?
THERAPEUTIC HYPOTHERMIA standard. NOTHING else proven superior. RESEARCH areas: (1) ERYTHROPOIETIN (EPO) — neuroprotection; mixed evidence. (2) STEM CELL therapy — experimental. (3) XENON gas (anaesthetic with neuroprotective properties) — TOBY-Xe trial; not standard. (4) ANTI-INFLAMMATORY agents. (5) MELATONIN — small trials. WAIT FOR EVIDENCE before adopting. COOLING remains gold standard. ALL ADJUNCTS: ensure good supportive care (BP, glucose, electrolytes, oxygenation) — protects brain too.
What investigations during HIE?
(1) BLOOD GAS at birth + serial — cord pH; base deficit; lactate trajectory. (2) GLUCOSE, electrolytes, calcium, magnesium. (3) RENAL + LIVER function. (4) COAGULATION (DIC risk). (5) CRP, cultures (sepsis screen). (6) CRANIAL ULTRASOUND (bedside) day 1-2. (7) FORMAL MRI 4-10 days. (8) EEG / aEEG continuous. (9) FORMAL EEG if seizures. (10) HEARING SCREEN. (11) ECHO cardiac.
What's the rehabilitation pathway?
(1) NEONATAL PHYSIOTHERAPY from NICU; (2) DEVELOPMENTAL paediatrician follow-up; (3) AT-HOME PHYSIO + OCCUPATIONAL THERAPY; (4) SPEECH AND LANGUAGE THERAPY; (5) HEARING aid / glasses if needed; (6) SEIZURE management with neurology if applicable; (7) FEEDING support; (8) EDUCATION SUPPORT (SEN coordinators, EHCP) as child grows; (9) PEER SUPPORT — BLISS, HIE-specific groups (Holdens Foundation in UK).
Will my baby develop cerebral palsy?
DEPENDS on severity + injury pattern. MILD HIE: low CP risk. MODERATE HIE: ~10-20% CP risk with cooling. SEVERE HIE: 30-60% CP risk. EARLY SIGNS: persistent abnormal tone, delayed motor milestones, asymmetric movements. CP DOES NOT MEAN NO QUALITY OF LIFE — varies vastly from mild to severe; many children with CP live full lives. EARLY INTERVENTION makes huge difference; physio + adaptive equipment + supportive education.
How does this relate to other calculators on BumpBites?
Companion: /calculators/sarnat-hie main alternative score; /calculators/neonatal-cooling; /calculators/apgar-score; /calculators/nrp-algorithm; /calculators/silverman-andersen; /calculators/eos-sepsis (sepsis can cause HIE); /calculators/new-ballard.