01Core principlesThe concepts and mechanisms needed to understand the subject.
A spinal epidural abscess is infected material in the epidural space that can compress the spinal cord, conus or cauda equina and can coexist with discitis, vertebral osteomyelitis or paraspinal infection. Haematogenous spread is common in native infection, often from Staphylococcus aureus bacteraemia, but direct inoculation after spinal procedures and contiguous spread also occur. Postoperative and implant-associated disease has different microbiology, sampling and source-control problems from native vertebral infection.
Presentation is variable. Severe focal spinal pain is common; fever and neurological deficit may emerge later, so requiring the classic triad misses early disease. Radicular pain may precede weakness. Once cord or cauda equina dysfunction develops, progression can be rapid and pre-treatment deficit strongly affects outcome. Examine the whole neuraxis relevant to symptoms and repeat the examination because a normal first assessment has a short shelf life.
Risk context changes probability rather than defining disease. Diabetes, injection drug use, immunosuppression, dialysis, indwelling vascular access, recent bacteraemia, endocarditis and spinal intervention all matter. Ask about skin and soft-tissue infection, urinary infection, dental and cardiac symptoms, tuberculosis exposure and travel or residence in endemic areas. Absence of fever is unsurprising, especially after antipyretics or in immune compromise.
Contrast-enhanced MRI is the key anatomical investigation. It can show an enhancing epidural collection, neural compression, disc-endplate involvement and paraspinal extension. The ACR rates MRI of the spine area of interest without and with contrast as usually appropriate for suspected spine infection. If MRI cannot be performed, CT and nuclear techniques may contribute, but negative or non-diagnostic alternatives do not automatically end a high-suspicion pathway.
Microbiology guides definitive therapy. Take two blood-culture sets before antibiotics when safe and obtain operative or image-guided samples when appropriate. The IDSA native vertebral osteomyelitis guideline supports withholding empiric antibiotics in stable adults until microbiological diagnosis, but explicitly excludes isolated epidural abscess, postoperative infection and implant infection. Therefore, its sampling sequence and 6-week “most bacterial NVO” statement must not be universalised to every SEA.
Treatment integrates neural decompression, drainage or debridement, stability and antimicrobials. Neurological compromise, progression, sepsis, instability, a large compressive collection or failure of medical therapy pushes toward urgent source-control assessment. Antibiotics alone may be selected in carefully monitored patients without neurological deficit or with prohibitive operative risk, but failure can occur; the plan needs named ownership, frequent examination, inflammatory-marker trends and a low threshold for repeat imaging and surgery.
Key points
- Think of SEA in severe localised spinal pain with infection risk, raised ESR/CRP, bacteraemia or neurological change; do not wait for the full pain-fever-deficit triad.
- MRI of the relevant spine without and with contrast is usually appropriate and defines epidural, disc, vertebral and paraspinal involvement; image a wider extent when symptoms or findings are multifocal.
- Before antibiotics, obtain two sets of aerobic and anaerobic blood cultures and baseline ESR/CRP when this does not delay emergency treatment.
- If neurological compromise, impending sepsis or haemodynamic instability is present, start empiric antimicrobials and obtain immediate surgical assessment; do not delay for image-guided biopsy.
- In a haemodynamically and neurologically stable adult with associated native vertebral osteomyelitis, microbiological diagnosis before empiric antibiotics can improve targeted therapy—but this is not permission to wait in isolated or progressive SEA.
- No single universal antibiotic regimen or duration fits native, postoperative, device-associated, immunocompromised, tuberculous, fungal and paediatric infection. Use cultures, anatomy, source control and current local antimicrobial guidance.
02Mechanisms and patternsImportant relationships and how to distinguish them.
New severe focal pain, night pain, percussion tenderness or radicular pain may be the only early clue. Fever and deficit are neither required nor reliably simultaneous.
Record segmental power, tone, reflexes, sensory level, saddle sensation, gait and sphincter function. A new or progressive deficit requires immediate surgical and antimicrobial action.
Recent S. aureus bacteraemia, endocarditis, vascular access, skin infection or injection drug use supports haematogenous native infection and should trigger cultures and spinal imaging when pain or neurology appears.
Recent spinal surgery, epidural injection or implanted material changes organisms, anatomy and need for source control. Escalate to the operating or specialist spinal service early.
Immune suppression widens bacterial and fungal possibilities; tuberculosis and Brucella risks depend on exposure and geography. Alert microbiology before sampling when special culture or containment is needed.
Metastasis, epidural haematoma, degenerative compression and inflammatory disease can resemble infection; cancer and anticoagulation do not exclude simultaneous SEA.
03Interpreting evidenceInformation, measurements and their limitations.
Consider the information, its meaning and its limitations before deciding what follows.
- 01
MRI spine without and with contrast - Why
- Define the epidural collection, neural compression and associated disc, vertebral or paraspinal infection.
- Interpretation and limitations
- Urgency follows neurological and septic state. Image the symptomatic region promptly and extend coverage when multifocal symptoms, bacteraemia or discontinuous infection is plausible. Gadolinium best shows epidural extension in NVO.
- 02
Two sets of blood cultures - Why
- Identify bloodstream pathogen before antimicrobials and sometimes avoid a separate biopsy in compatible native infection.
- Interpretation and limitations
- Take aerobic and anaerobic sets before treatment if this does not delay rescue. A negative result does not exclude SEA; prior antibiotics and intermittent bacteraemia reduce yield.
- 03
CRP, ESR and full blood count - Why
- Support suspicion and establish a trend for response.
- Interpretation and limitations
- Inflammatory markers are sensitive context but nonspecific; normal or modest results cannot overrule progressive neurological findings or convincing MRI.
- 04
Operative or image-guided specimens - Why
- Obtain culture and histopathology for targeted treatment and distinguish infection from malignancy or inflammatory disease.
- Interpretation and limitations
- Prioritise operative sampling during urgent decompression. In stable associated NVO, image-guided sampling before antibiotics may improve yield; do not delay treatment for neurological compromise or sepsis.
- 05
Source and complication assessment - Why
- Find endocarditis, vascular-catheter, skin, urinary or other sources and evaluate sepsis organ dysfunction.
- Interpretation and limitations
- Use echocardiography, repeat cultures and source-directed tests according to organism and clinical context. Investigation proceeds with, not instead of, spinal source control.
- 06
CT spine or nuclear imaging - Why
- Assess bone, guide biopsy or provide alternatives when MRI is impossible.
- Interpretation and limitations
- CT has limited sensitivity for epidural soft tissue compared with contrast MRI. Select alternatives with radiology and spinal teams and maintain escalation if high suspicion persists.
04Applied reasoningWorked examples connecting principles to decisions.
01Worked case: deteriorating infectionFever, back pain and new weaknessA febrile patient with severe thoracic pain develops rapidly progressive leg weakness and hypotension.+
- 1Recognise possible SEA with neural compression and sepsis; activate spinal surgery, infection/microbiology, anaesthesia and critical care immediately.
- 2Obtain two blood-culture sets and baseline tests if they can be taken without delay, then start empiric intravenous antimicrobials using the current local severe-spinal-infection pathway.
- 3Arrange urgent contrast-enhanced MRI while resuscitating. Do not postpone antimicrobials or surgical assessment for a planned percutaneous biopsy.
- 4Proceed to urgent decompression/drainage and operative sampling when indicated by the spinal team; address stability and any additional infective source.
- 5Narrow treatment to culture and susceptibility results and set duration through infection specialists according to compartment, organism, source control, bone/disc involvement, host and response.
02Stable diagnostic pathwayPossible native vertebral infection without deficitA stable adult has persistent focal pain, raised CRP and MRI showing disc-endplate infection with a small epidural component but no neurological deficit.+
- 1Perform and document a complete neurological examination, obtain two blood-culture sets and involve spinal and infection specialists early.
- 2If haemodynamics and neurology remain stable, pursue microbiological diagnosis before empiric therapy when feasible, following the NVO pathway and local governance.
- 3Define monitoring frequency and failure criteria explicitly: any new deficit, sepsis, enlarging collection, instability, persistent bacteraemia or clinical/laboratory failure triggers reassessment for urgent source control.
03Scope pathwayChoose the correct infection compartmentMRI has confirmed an epidural collection.+
- 1Determine whether infection is isolated epidural, accompanies native discitis/vertebral osteomyelitis, follows a procedure, or involves an implant.
- 2Identify immune status, organism risks and geography; send the appropriate bacterial, mycobacterial, fungal and histological specimens with laboratory notification.
- 3Select antimicrobial route, spectrum and duration with infection specialists and local guidance rather than importing a single NVO duration into every compartment.
05Checking understandingVerify the reasoning, revisit uncertainties and apply feedback.
- Repeat a documented neurological examination frequently during diagnostic and medical management; new weakness, sensory level or sphincter change is an immediate escalation.
- Trend haemodynamics, lactate and organ function in sepsis, and provide critical-care support as needed.
- Follow pain, temperature and CRP/ESR trajectory, but do not define success from inflammatory markers alone.
- Review culture and histology results with microbiology/infection specialists and narrow antimicrobials promptly when safe.
- Repeat MRI for neurological change, suspected treatment failure or inadequate clinical/laboratory improvement; routine timing depends on compartment and treatment plan.
06Special situationsVariants, exceptions and circumstances that change the usual approach.
The triad is late
Pain, fever and deficit often do not coexist at first contact. Severe focal pain plus bacteraemia or risk context is enough to investigate.
Two sequences, two states
Cultures before antibiotics improves yield when safe; neurological compromise or sepsis reverses the priority—treat and obtain urgent source control.
NVO is not all SEA
IDSA’s guideline is authoritative for adult native vertebral osteomyelitis but explicitly excludes isolated epidural, postoperative and implant infection.
Anatomy changes consequence
The same collection volume can be tolerated differently in lumbar versus cervical or thoracic canal; clinical deficit and compression guide urgency.
Medical treatment is active treatment
A nonoperative plan needs named specialists, culture-directed drugs, serial neurology and explicit failure criteria—not passive observation.
07Common pitfallsFrequent interpretation and management errors.
- 01
Waiting for fever and neurological deficit to join back pain before ordering MRI.
- 02
Withholding antimicrobials for biopsy in a patient with neurological compromise, sepsis or haemodynamic instability.
- 03
Starting antibiotics in a stable patient without first taking blood cultures.
- 04
Applying the IDSA six-week NVO statement to isolated SEA, postoperative infection, implant infection, tuberculosis or fungal disease.
- 05
Calling a normal white-cell count or absence of fever a negative test.
- 06
Choosing antibiotics without accounting for acquisition setting, immune status, cultures, source control and current local resistance guidance.