Fat Embolism Syndrome — The 12–72h Triad, Gurd's Criteria & Lung-Protective Support

Overview & definition

The fat embolism syndrome (FES) is a multisystem disorder characterised by the triad of respiratory insufficiency, neurological dysfunction and a petechial rash that declares itself 12 to 72 hours after the entry of medullary fat into the venous circulation — most often from a long-bone (femur, tibia) or pelvic fracture, but also from orthopaedic surgery, liposuction, pancreatitis, sickle-cell crisis, bone-marrow transplant and parenteral lipid infusion.[7][11]

The cardinal points for the exam: (1) it is a clinical diagnosis — no single laboratory test confirms it; (2) the timing is the discriminant — respiratory failure in the first 1–2 hours after injury is contusion, aspiration or haemorrhage, NOT FES; (3) marrow fat embolises to EVERY major trauma patient, but only a minority (1–5% of isolated long-bone fractures, up to 10–20% of multiple fractures) develop the syndrome; and (4) management is supportive with an ARDS lung-protective ventilatory strategy, and the most effective treatment is prevention by early fracture fixation.[2][5][9]

Epidemiology & incidence — fat emboli versus the fat embolism syndrome

The single most important epidemiological distinction is between fat embolism (the histological/radiological presence of marrow fat in the circulation — present to some degree in almost every major trauma patient) and the fat embolism syndrome (the clinical expression in a minority). Confusing the two leads to over-diagnosis.[7][11]

• Incidence of the syndrome: ~1–5% of isolated long-bone fractures; 10–20% with multiple long-bone fractures or a femur + pelvic combination; up to 35% historically in unplated major lower-limb trauma before early fixation was standard. Mortality 5–15%; full recovery in the majority.[2][9]
• Risk is proportional to the energy and marrow burden: closed fractures > open (the open fracture decompresses the medullary canal); femur > tibia > humerus; polytrauma and high ISS; age 10–40 (the marrow is fatty and the bones are high-energy); male sex.[8][11]
• latrogenic / non-traumatic triggers — recognise these because the patient may have NO fracture: intramedullary nailing, hip/knee arthroplasty (cemented, pressurisation), Kuntscher rod insertion, liposuction, medullary reaming, bone-marrow harvest/transplant, parenteral lipid emulsion, sickle-cell crisis (marrow infarction), acute pancreatitis, decompression sickness, and high-dose corticosteroid withdrawal.[7][14]

Triggers — the #1 is the long bone fracture

[7] [8] [13]

Pathophysiology — the two-hit mechanical + biochemical model

FES is best understood as a two-mechanism model — a mechanical phase and a biochemical phase acting in sequence and amplifying each other. Understanding this model explains why the lungs dominate the picture, why the brain is involved despite all emboli being venous, and why the syndrome looks like ARDS.[8][11]

The mechanical phase — fat globules embolise to the pulmonary capillaries

The marrow sinusoids are torn by the fracture; the sudden rise in intramedullary pressure forces liquid fat globules (30–100 µm, too large to traverse the pulmonary capillary bed) into the draining veins. The globules travel via the right heart to the pulmonary microvasculature, where they mechanically obstruct precapillary arterioles and capillaries, producing an acute rise in pulmonary vascular resistance, a degree of right-heart strain, and an early, often mild, ventilation–perfusion mismatch. This mechanical phase accounts for the immediate (minutes to hours) hypoxaemia and the transient pulmonary hypertension seen during intramedullary nailing.[8][14]

In some patients a fraction of the fat bypasses the lung: through a patent foramen ovale (present in ~25% of adults), through pulmonary precapillary arteriovenous anastomoses, or by simple distension of capillaries under high pressure — delivering fat to the systemic (cerebral and retinal) circulation and explaining the cerebral and petechial manifestations.[13]

The biochemical phase — free fatty acids → inflammation → endothelial damage → ARDS

The embolic triglyceride is itself relatively inert. The damage is done by its hydrolysis. Pulmonary lipase (and tissue lipase) cleaves the triglyceride to free fatty acids (FFAs) — toxic, detergent-like molecules that:[8][11]

1. Directly injure the pulmonary capillary endothelium → increased permeability, protein-rich oedema, haemorrhage into the alveoli — the histological picture is a haemorrhagic alveolitis.
2. Activate the inflammatory cascade — complement (C5a), neutrophils, platelets, cytokines (TNF-α, IL-1, IL-6, IL-8), thromboxane and leukotrienes — producing a systemic inflammatory response and amplifying endothelial injury.
3. Generate a coagulopathy — platelet activation and consumption (thrombocytopenia is a Gurd minor criterion), microvascular thrombi, and a clinical overlap with early DIC. [1]

The end result is a permeability-type pulmonary oedema that is clinically and radiologically indistinguishable from ARDS — bilateral diffuse alveolar infiltrates, refractory hypoxaemia, reduced compliance. The time lag between the fracture and the onset (12–72 h) reflects the time required for lipase hydrolysis and the inflammatory cascade to develop — this is the biological basis of the delayed presentation.[6][11]

The brain — systemic embolism, not primary lung failure

Cerebral involvement is from systemic fat emboli (paradoxical or via pulmonary AV shunts), NOT simply hypoxia. Autopsy studies show microvascular fat and petechial haemorrhages distributed in the white matter and brainstem — a pattern distinct from anoxic encephalopathy. This is why neurological signs can precede the respiratory failure, and why they can be focal or bizarre (not the symmetric picture of hypoxia).[13]

The petechial rash — the pathognomonic clue

The petechial rash is the embolisation of platelet-fibrin-fat microthrombi to the dermal capillaries, with thrombocytopenia amplifying it. Its distribution over the upper chest, axillae, neck, conjunctivae and the oral mucosa reflects dependent and gravity-independent zones with abundant capillaries — and its transient nature (resolves in 24–48 h) is the reason the diagnosis is missed if the skin is not examined early and serially.[2][9]

[8] [11]

The clinical triad & timing

The triad, in order of frequency and importance:[2][9]

1. Respiratory insufficiency (75–85%, the commonest and the leading reason for ICU admission): the earliest and most reliable sign is hypoxaemia out of proportion to the mechanism — a falling SpO₂, tachypnoea, increasing work of breathing, then frank ARDS (bilateral diffuse alveolar infiltrates on the CXR developing over hours, refractory hypoxaemia, reduced compliance). Almost every patient with FES has some degree of respiratory involvement.
2. Neurological dysfunction (~80%, often second): ranges from subtle confusion and agitation to obtundation, seizures and coma. The picture is non-focal and fluctuating (a toxic-metabolic encephalopathy pattern), though focal deficits and bizarre behaviour occur from discrete cerebral microemboli. Neurological signs may precede the respiratory signs and are the source of the most exam-worthy confusion with traumatic brain injury.[13]
3. Petechial rash (50–60%, the pathognomonic but transient feature): a 1–2 mm non-blanching eruption over the conjunctivae, oral mucosa, neck, upper chest, axillae (the dependent, gravity-independent skin of the upper body). It appears 24–36 h after injury, lasts 24–48 h, and may be the only sign that clinches the diagnosis — examine the conjunctivae, oral mucosa and axillae of every polytrauma patient at 24 h.[2]

Timing — the discriminant against early diagnoses

The 12–72 hour window is the single most useful clinical discriminant:[7][11]

• Immediate (0–2 h) respiratory failure after chest trauma = pulmonary contusion, tension pneumothorax, haemothorax, aspiration — NOT FES. FES is excluded by being too early.
• 2–12 h = the latent period; the patient often looks well; this is when the biochemical cascade is building.
• 12–72 h = the classical window; a previously stable polytrauma patient who now hyperventilates, desaturates and becomes confused has FES until proven otherwise.
• Beyond 72 h is unusual but recognised; late or stuttering presentations occur, particularly with delayed or repeated surgical manipulation of the fracture. [1]

Gurd's criteria — the diagnostic framework

Gurd and Wilson (1970–1974) formalised the diagnosis. The classic construct requires at least one major criterion plus four minor criteria, plus the presence of macroglobulinaemia (fat macroglobulinuria). In modern practice the criteria are used more flexibly — the triad of respiratory + CNS + petechial rash in the right temporal window is itself diagnostic, and Gurd's criteria are the structured way to defend the diagnosis in a vivva.[1][2]

[1] [2] [7]

Differential diagnosis — what FES is NOT

The polytrauma patient at 24–72 hours can deteriorate from many causes; FES is one of several, and the wrong diagnosis delays the right treatment. The petechial rash and the timing are the discriminants, but several mimics must be actively excluded.[7][11]

[7] [9] [13]

Investigations — supportive, not definitive

No single test confirms FES; the diagnosis is clinical (Gurd's criteria in the right temporal context). Investigations serve to support the diagnosis, grade the severity, and exclude the mimics.[7][11]

• Arterial blood gas: the earliest and most useful single test — hypoxaemia (often PaO₂/FiO₂ <300, meeting ARDS criteria) with a respiratory alkalosis from hyperventilation, progressing to a metabolic acidosis as shock and organ failure develop.
• Chest X-ray: typically normal initially, then bilateral diffuse alveolar infiltrates develop over 12–24 h (the picture of ARDS); the radiograph lags behind the hypoxaemia (the CXR may be near-normal while the PaO₂ is dire). Cardiomegaly and effusions are absent.
• Full blood count: a falling platelet count (thrombocytopenia) and a falling haemoglobin (haemodilution/haemolysis) are useful early clues; leucocytosis is common.
• Coagulation: prolonged INR/aPTT, raised D-dimer, low fibrinogen in the more severe cases (overlap with DIC); a falling platelet count plus coagulopathy supports the diagnosis.
• Biochemistry: hypocalcaemia (saponification of free fatty acids), mild hyperbilirubinaemia, raised LDH (haemolysis), and a rising creatinine with urinary fat and protein. Serum lipase may be elevated.
• Fundoscopy: retinal petechiae, cotton-wool spots or visible fat globules (Purtscher-like retinopathy) — a positive finding is highly specific though insensitive.
• Urine and sputum cytology for fat: fat globulinuria / fat in the urine (the Gurd laboratory criterion) supports the diagnosis but is insensitive and not routinely performed.
• CT head: to exclude TBI; cerebral fat embolism may show the "starfield" pattern on MRI (scattered bright dots on diffusion-weighted imaging) — highly suggestive but usually a delayed finding.
• Bronchoalveolar lavage: staining for fat (Sudan) in alveolar macrophages supports the diagnosis but is invasive and not routinely required; positive in a majority of FES patients but also in major trauma without the syndrome (low specificity).
• Echocardiography (transoesophageal, intraoperative): during nailing, shows the "snowstorm" of echogenic material in the right heart; in established FES, look for right-heart strain, pulmonary hypertension and a patent foramen ovale.[8]
• ECG: non-specific ST/T changes, arrhythmia — from myocardial microemboli and hypoxaemia; exclude ischaemia.

Management — supportive, with an ARDS lung-protective strategy

There is no specific therapy for FES. Management is supportive, focused on the respiratory failure (the usual reason for ICU admission) and the prevention of further fat release. The principles are: oxygenate and ventilate with a lung-protective strategy, circulate cautiously, treat the fracture definitively (early fixation), and avoid secondary brain and lung injury.[1][7][11]

Respiratory support — escalate along the oxygen ladder to lung-protective ventilation

• Supplemental oxygen and high-flow nasal cannula (HFNC) or non-invasive ventilation (NIV) for the mild–moderate case with alert, cooperative neurology and PaO₂/FiO₂ >200. Close monitoring — escalation to intubation should not be delayed if work of breathing rises, neurology deteriorates, or gas exchange worsens.
• Invasive mechanical ventilation for the severe case (refractory hypoxaemia, falling GCS, exhaustion, shock). Use a lung-protective ARDS strategy: tidal volume 6 mL/kg predicted body weight, plateau pressure <30 cmH₂O, moderate PEEP titrated to oxygenation, and permissive hypercapnia (pH >7.20). The ARDS Network trial (lower tidal volumes) established the mortality benefit that underpins this approach in the ARDS-like lung of FES.[6]
• Prone positioning for moderate–severe ARDS (PaO₂/FiO₂ <150): the PROSEVA trial showed a 28-day mortality benefit of sustained prone ventilation (16 h/day) in severe ARDS — applicable to the severe FES lung.[12]
• Neuromuscular blockade (48 h cisatracurium) for severe early ARDS improves oxygenation and may reduce mortality; inhaled pulmonary vasodilators (nitric oxide, prostacyclin) are rescue therapies for refractory hypoxaemia.
• ECMO (venovenous) as a last resort for refractory hypoxaemia unresponsive to prone ventilation — a bridge through the biochemical storm, which is self-limiting.

Circulation and fluids — cautious, lung-sparing

• Cautious crystalloid resuscitation — the FES lung is a permeability-type oedema; aggressive fluid deepens the ARDS. Use a balanced crystalloid, target euvolaemia, and consider early colloid or albumin and vasopressors (noradrenaline) to maintain perfusion rather than flooding the lung.
• Vasopressors for shock; inotropes if right-heart failure develops from pulmonary hypertension (sildenafil, inhaled pulmonary vasodilators in selected cases).
• Transfusion for anaemia — keep haemoglobin adequate for oxygen delivery; avoid over-transfusion. [1]

Definitive fracture management — the single most effective treatment

• Early surgical stabilisation of the long-bone or pelvic fracture (within 24 h) is the single most effective intervention to prevent and reduce the severity of FES, by stopping ongoing fat release. The Bone et al. prospective randomised study established that early fixation of femoral fractures reduced pulmonary complications (including FES and ARDS) compared with delayed fixation, especially in polytrauma.[5]
• Reamed versus unreamed intramedullary nailing — reaming pressurises the canal more and liberates more fat; in the patient with established or impending FES or severe chest injury, unreamed nailing, plating, or external fixation may be preferred. The choice is individualised by the orthopaedic and trauma team, but the principle — stabilise early, stop the fat release — is universal.
• Avoid repeated manipulation of the fracture; splint and stabilise in the ED.

Corticosteroids — prophylaxis reduces incidence but remains controversial

• High-dose methylprednisolone prophylaxis (various regimens, e.g. 1.5 mg/kg q8h for six doses, or 7.5 mg/kg/kg at admission and at 24 h) given to high-risk patients (femoral shaft fracture, multiple long-bone fractures) within 24 h of injury has been shown in several prospective studies (Schonfeld 1983; Lindeque 1987) and a meta-analysis (Bederman 2009) to reduce the incidence of FES.[3][4][10]
• However, steroids are NOT routine: the studies are heterogeneous, the definition of FES inconsistent, and there is concern about infection, wound healing and the masking of sepsis in polytrauma. Most contemporary trauma and ICU guidelines reserve steroids for the highest-risk patient on an individual basis; they are not a treatment for established FES (no benefit once the syndrome is declared).[7][10]
• The exam answer: "corticosteroid prophylaxis reduces the incidence of FES in high-risk patients in meta-analysis but is not used routinely because of methodological concerns and the risk of sepsis; it has no role once the syndrome is established."

Supportive and general ICU measures

• DVT and stress-ulcer prophylaxis (FES patients are immobilised, critically ill and often coagulopathic — balance the risk).
• Nutrition — early enteral feeding; the FES patient is catabolic.
• Venous thromboembolism prophylaxis with mechanical and pharmacological means, balancing the coagulopathy.
• Avoid secondary brain injury in the encephalopathic patient — keep PaO₂ >60 mmHg, avoid hypotension, control seizures (the cerebral fat embolism patient is at risk).
• Normoglycaemia, normothermia, infection surveillance (FES mimics sepsis; cultures are negative initially but nosocomial infection is common). [1]

Cerebral fat embolism — the neurology that confuses the viva

Cerebral fat embolism (CFE) is the neurological expression of FES, occurring in up to 80% of FES patients (mild in most). It deserves separate attention because it is the most common source of diagnostic confusion with traumatic brain injury and because the MRI finding — the "starfield" pattern — is an exam-favourite.[13]

• Clinical spectrum: mild confusion and agitation (the most common) → obtundation, decorticate/decerebrate posturing → focal deficits (hemiparesis, visual disturbance, aphasia) → seizures and coma. The picture is non-focal and fluctuating in most, but discrete focal deficits occur from discrete microemboli.
• Mechanism: systemic (paradoxical, via PFO, or via pulmonary AV shunts) fat emboli to the cerebral microcirculation → microvascular obstruction, petechial haemorrhage in the white matter and brainstem, and a vasogenic oedema. Histology shows a "purpuric" pattern of multiple petechial haemorrhages — distinct from the confluent contusion of TBI.
• Imaging: CT is often normal early or shows non-specific oedema (it is performed to exclude haemorrhage). MRI is the modality of choice: on diffusion-weighted imaging, the "starfield" pattern — scattered, discrete, bright-restricted-diffusion dots on a dark background (vasogenic + cytotoxic oedema from microemboli) — is highly suggestive of CFE. Additional findings: multiple small T2/FLAIR hyperintensities ("raindrop"), susceptibility-weighted imaging microhaemorrhages.[13]
• Management: as for FES (supportive, lung-protective ventilation, avoid secondary brain injury — normoxia, normotension, normoglycaemia, normocapnia, seizure control, head-up 30°). There is no specific therapy; most neurological deficits recover completely over days to weeks, though a minority have residual cognitive or focal deficits.

Prevention — the most effective treatment is to not let it happen

Prevention is the cornerstone of FES management, because no specific therapy reverses the syndrome once declared. The hierarchy:[5][8][10]

1. Early (within 24 h) surgical stabilisation of long-bone and pelvic fractures — the single most effective intervention. Stabilise the fracture, stop ongoing fat release, mobilise the patient. The Bone et al. study showed fewer pulmonary complications with early versus delayed fixation.[5]
2. Careful operative technique — unreamed or limited reaming, venting of the medullary canal, reduced pressurisation, external or plating alternatives in the highest-risk patient (severe chest injury, established FES). Intraoperative transoesophageal echo and prompt action (pause, reduce pressurisation, vasopressor) at the first sign of the snowstorm.
3. Avoid repeated fracture manipulation; splint and handle gently in the ED and during transfers.
4. Corticosteroid prophylaxis for the highest-risk patient on an individual basis (see Management) — reduces incidence but not routine.[3][10]
5. Vigilance — monitor every polytrauma patient with a long-bone or pelvic fracture with serial oxygen saturation, respiratory rate, GCS, platelet count and a daily skin/fundi examination through the 12–72 h window.

Prognosis & outcomes

• Mortality 5–15% overall; deaths are usually from multi-organ failure (refractory ARDS, sepsis, MODS) rather than FES itself. With modern lung-protective ventilation and early fixation, mortality has fallen.[9][14]
• Recovery in the majority over 1–2 weeks; the respiratory failure resolves as the biochemical cascade exhausts itself and the endothelium recovers. The CXR clears over 7–14 days.
• Neurological recovery is usually complete, but a minority (~10%) have residual cognitive or focal deficits, particularly after a severe cerebral fat embolism.[13]
• Predictors of severity: high ISS, multiple long-bone fractures, femur + pelvic combination, delayed fixation, older age, pre-existing lung disease, and the depth of the initial hypoxaemia.

Red flags

Exam practice

[1] [2] [5] [6] [10] [13]

Trial cards

[1] [1] [1] [1] [1]

Clinical pearls

References

1. [1]Gurd AR Fat embolism: an aid to diagnosis J Bone Joint Surg Br, 1970.PMID 5487573
2. [2]Gurd AR, Wilson RI The fat embolism syndrome J Bone Joint Surg Br, 1974.PMID 4547466
3. [3]Schonfeld SA, Ploysongsang Y, DiLisio R, et al. Fat embolism prophylaxis with corticosteroids. A prospective study in high-risk patients Ann Intern Med, 1983.PMID 6354030
4. [4]Lindeque BG, Schoeman HS, Dommisse GF, Boeyens HL, Vlok AL Fat embolism and the fat embolism syndrome. A double-blind therapeutic study J Bone Joint Surg Br, 1987.PMID 3818718
5. [5]Bone LB, Johnson KD, Weigelt J, Scheinberg R Early versus delayed stabilization of femoral fractures: a prospective randomized study. 1989 Clin Orthop Relat Res, 2004.PMID 15187827
6. [6]The Acute Respiratory Distress Syndrome Network (Brower RG, Matthay MA, Morris A, et al.) Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome N Engl J Med, 2000.PMID 10793162
7. [7]Mellor A, Soni N Fat embolism Anaesthesia, 2001.PMID 11167474
8. [8]Hofmann S, Huemer G, Salzer M [Pathophysiology of fat embolisms in orthopedics and traumatology] Orthopade, 1995.PMID 7753543
9. [9]Taviloglu K, Yanar H Fat embolism syndrome Surg Today, 2007.PMID 17186337
10. [10]Bederman SS, Bhandari M, McKee MD, Schemitsch EH Do corticosteroids reduce the risk of fat embolism syndrome in patients with long-bone fractures? A meta-analysis Can J Surg, 2009.PMID 19865573
11. [11]Akhtar S Fat embolism Anesthesiol Clin, 2009.PMID 19825491
12. [12]Guérin C, Reignier J, Richard JC, et al. (PROSEVA trial) Prone positioning in severe acute respiratory distress syndrome N Engl J Med, 2013.PMID 23688302
13. [13]Alrasheed AS, Elazrag AM, et al. Pathophysiology, clinical manifestations, and prognostic insights of cerebral fat embolism: a literature review Front Neurol, 2025.PMID 41603011
14. [14]Miyake T, Okada H, et al. Advances and uncertainties in fat embolism syndrome: a review Trauma Surg Acute Care Open, 2026.PMID 42375760