Spinal Epidural Abscess — Surgical Decompression

SpineAdvancedCore Procedure

Spinal Epidural Abscess — Surgical Decompression

Surgical technique guide for emergency decompression of a spinal epidural abscess — posterior laminectomy for dorsal collections, anterior debridement for ventral pus with vertebral osteomyelitis, indications, instrumentation, complications and post-operative rehabilitation

High-yield overview

Emergency posterior laminectomy and abscess evacuation, or anterior debridement for ventral collection with osteomyelitis | advanced

Surgical Imaging

Critical Danger Structures and Exam Traps
Delay to Surgery — Irreversible Paralysis

The trap: Waiting for inflammatory markers to normalise, or for an MRI on the next available list, in a patient with new neurological deficit. Each hour of compression worsens the prognosis.

The fix: Any patient with new or worsening motor weakness, sensory level, or sphincter disturbance from a confirmed epidural abscess goes to theatre urgently — draw blood cultures then give the first antibiotic dose, but do not let antibiotics delay surgery. The neurosurgical or spinal surgery team must be contacted immediately.

Neurologically Intact — Over-Treatment

The trap: Taking a neurologically intact patient with a small dorsal epidural collection to theatre when close neurological monitoring and targeted antibiotics would suffice.

The fix: A patient who is neurologically intact with a small collection, no cord compression on MRI, and no progression may be managed non-operatively — provided an hourly neurological chart is instituted, antibiotics are culture-directed, and a repeat MRI is performed at 48-72 hours. Any new deficit mandates urgent decompression.

Ventral Collection — Wrong Approach

The trap: Performing a posterior laminectomy for a predominantly ventral epidural collection where the pus compresses the cord or cauda equina from the front, leaving the ventral thecal sac inadequately decompressed.

The fix: Ventral collections with associated vertebral osteomyelitis and discitis require an anterior approach (cervical: anterior cervical discectomy/corpectomy; thoracic: transthoracic or thoracoscopic; lumbar: anterior lumbar interbody) with debridement and reconstruction. Posterior stabilisation may also be required. Posterior-only approaches cannot adequately decompress ventral pus.

Dural Tear — CSF Leak and Meningitis

The trap: Inadvertent durotomy during laminectomy or during evacuation of an epidural collection densely adherent to the dura — particularly after previous spinal surgery or with chronic infection.

The fix: Identify the dura before decompressing. If a dural tear occurs, repair primarily with 6-0 Prolene sutures under magnification, reinforce with a dural substitute and fibrin glue, and extend post-operative antibiotics to cover for meningitis risk. Keep the patient flat for 24-48 hours post-operatively to reduce CSF leak pressure.

Instability After Multilevel Laminectomy

The trap: Performing a multilevel laminectomy (3 or more levels) for an extensive abscess without planned instrumented fusion, leading to post-operative deformity, progressive kyphosis, and chronic pain.

The fix: If the laminectomy extends 3 or more levels, or if facet joints are violated during the infection debridement, plan for concurrent posterior pedicle screw instrumentation. In the cervical and thoracic spine, the instability risk after multilevel laminectomy is higher and fusion is more strongly indicated. Anterior reconstruction with cage or strut graft may also be needed.

Diabetic and Immunocompromised — Wound Failure

The trap: Primary wound closure in a patient with poorly controlled diabetes, chronic steroids, or end-stage renal disease, resulting in wound dehiscence and persistent infection.

The fix: In high-risk patients (HbA1c greater than 9%, chronic steroid use greater than 10 mg prednisolone daily, dialysis-dependent), consider delayed primary closure or vacuum-assisted closure (VAC) dressing for 48-72 hours with planned re-look and closure. Optimise glycaemic control perioperatively. Meticulous haemostasis and layered closure reduce dead space.

Mnemonic

A.B.C.E.S.SABCESS — Spinal Epidural Abscess Presentation and Diagnosis

Mnemonic

D.E.C.O.M.PDECOMP — Intraoperative Principles

Surgical Indications

Absolute Indications

  • Progressive neurological deficit — any new motor weakness, sensory level, or sphincter disturbance attributable to the epidural abscess
  • Complete or incomplete paralysis — urgent decompression within hours; recovery correlates with speed of intervention
  • Sepsis or haemodynamic instability from the spinal infection not responding to antibiotics alone
  • Abscess with cord compression on MRI in a patient who is deteriorating clinically despite appropriate antibiotics
  • Ventral collection with vertebral osteomyelitis where anterior debridement and reconstruction are required to decompress the thecal sac and stabilise the spine

Relative Indications

  • Large epidural collection (greater than 75% of canal cross-sectional area) with significant thecal sac compression, even in a neurologically intact patient
  • Failure of antibiotic therapy — persistent fever, rising CRP, or expanding collection on repeat MRI after 48-72 hours of targeted treatment
  • Patient unable to cooperate with neurological monitoring — intensive care, altered mental state, or language barrier precluding reliable serial exams
  • Recurrent abscess after previous non-operative management

Contraindications to Non-Operative Management

  • Any new or worsening motor weakness, sensory deficit, or sphincter disturbance
  • Complete paralysis — surgical decompression within hours is the standard of care
  • Sepsis refractory to antibiotics
  • Ventral collection with unstable osteomyelitis requiring debridement

When Non-Operative Management Is Appropriate

  • Neurologically intact patient
  • Small epidural collection with no significant cord compression on MRI
  • No clinical deterioration over a minimum of 72 hours of close observation
  • Reliable patient who can communicate changes in neurological status
  • Identified organism with targeted antibiotic susceptibility
  • Negative TB and fungal cultures before committing to a long non-operative course

Clinical Presentation

The Heusner Staging Framework

The clinical course of a spinal epidural abscess follows a recognisable pattern:

  • Stage 1: Localised back pain and tenderness — the earliest symptom, present in over 90% of cases. Often misattributed to mechanical back pain in the first days.
  • Stage 2: Radicular pain — nerve root irritation from the expanding collection; may be bilateral or unilateral, and mimics disc herniation.
  • Stage 3: Motor weakness and sensory deficit — upper motor neuron signs (hyperreflexia, clonus, Hoffman sign) in cervical and thoracic abscesses; lower motor neuron signs (hyporeflexia, flaccid weakness) in cauda equina compression. Sensory level may be present.
  • Stage 4: Paralysis — complete motor and sensory loss below the level of the lesion; sphincter disturbance (urinary retention, loss of anal tone). This is the point of no return for many patients.
  • Stage 5: Death — from sepsis, respiratory failure (cervical), or complications of prolonged paralysis.

Exam relevance: The exam answer for when to operate is any patient who reaches stage 3 or beyond. The best neurological outcomes are achieved when decompression is performed in stages 1-2 before motor deficit develops — but most patients present at stage 2-3.

Common Presenting Features

  • Back pain (90%+): localised, progressive, worse at night, unresponsive to simple analgesia
  • Fever (50-70%): may be absent in immunocompromised patients and in those already on antibiotics
  • Neurological deficit (30-50% at first presentation): weakness, sensory changes, gait disturbance
  • Raised inflammatory markers: ESR commonly greater than 50 mm/hr, CRP commonly greater than 100 mg/L
  • Tenderness on spinal percussion at the affected level

Risk Factors

  • Diabetes mellitus (present in 20-40% of cases)
  • Intravenous drug use (especially Staph aureus bacteraemia)
  • Immunocompromised state: HIV, chemotherapy, chronic steroids, TNF-alpha inhibitors, organ transplant
  • Chronic renal failure and dialysis
  • Alcohol abuse
  • Recent spinal procedure or epidural injection
  • Endocarditis
  • Skin and soft tissue infection (cellulitis, furuncle) with haematogenous seeding
  • Urinary tract infection (Gram-negative organisms)
  • Liver disease and cirrhosis

Investigations

MRI with Gadolinium Contrast — Gold Standard

  • Sensitivity: 90%+ for detecting epidural abscess
  • Specificity: 90%+ when gadolinium is used
  • Protocol: T1-weighted, T2-weighted, and T1-weighted fat-suppressed post-gadolinium sequences through the entire clinically relevant spine
  • Findings: Iso- or hypointense collection on T1, hyperintense on T2, with peripheral rim enhancement post-gadolinium
  • Associated findings: vertebral osteomyelitis (endplate signal change, enhancement), discitis (loss of disc height, enhancement), paraspinal phlegmon
  • Extent: image the whole spine if clinically suspicious — skip lesions occur in up to 10% of cases

Blood Tests

  • ESR: almost always elevated; a normal ESR makes the diagnosis unlikely but does not exclude it
  • CRP: typically markedly elevated (commonly greater than 100 mg/L); serial CRP is the most useful marker for monitoring treatment response
  • Full blood count: leukocytosis common but not invariable
  • Blood cultures: two sets from separate sites BEFORE antibiotics — positive in 50-70% of cases
  • Renal function, liver function, HbA1c (for diabetic patients)

Other Investigations

  • Plain radiographs: may show vertebral endplate erosion, disc space narrowing, or paraspinal soft tissue swelling if osteomyelitis is present — but are normal in early isolated epidural abscess without bony involvement
  • CT with contrast: second-line if MRI is contraindicated (pacemaker, severe claustrophobia); less sensitive for the abscess itself but can show bony destruction and canal narrowing
  • CT-guided biopsy or aspiration: may be used to obtain a microbiological diagnosis in patients where blood cultures are negative and surgery is not yet indicated

Evidence for Surgical vs Non-Operative Management

Surgical Decompression — When and Why

The fundamental surgical principle is that the spinal cord and cauda equina tolerate compression poorly, and pus in the epidural space produces direct mechanical compression and ischaemia. Surgical evacuation relieves the mechanical component and allows direct culture and debridement.

  • Neurological deficit or progression = surgery: this is the clearest evidence-based indication. Multiple large case series and systematic reviews demonstrate that patients with preoperative neurological deficit who undergo early surgical decompression have better neurological outcomes than those managed non-operatively or those whose surgery is delayed.
  • Timing: recovery correlates with the duration of preoperative deficit. Decompression within 24 hours of symptom onset gives the best chance of recovery; beyond 36-48 hours the prognosis worsens markedly, particularly for complete paralysis.
  • Approach selection: posterior laminectomy for dorsal or posterolateral collections (the most common pattern); anterior debridement and reconstruction for ventral collections with osteomyelitis.

Non-Operative Management — Highly Selected Patients

  • Reserved for neurologically intact patients with small collections, identified organisms, and reliable serial neurological monitoring
  • Requires: hourly neurological checks by trained nursing staff, CRP monitoring, repeat MRI at 48-72 hours, and a low threshold to abandon conservative treatment
  • Antibiotics: minimum 6-8 weeks total; start IV and step down to oral based on culture and CRP
  • Failure rate: a subset of non-operatively managed patients will deteriorate and require delayed surgery — these patients have worse outcomes than those decompressed early
  • No randomised controlled trial has compared surgery with antibiotics alone for spinal epidural abscess — the evidence is from case series and retrospective studies

Surgical vs Non-Operative Management — Decision Framework


Key Evidence

Evidence

Bacterial spinal epidural abscess. Review of 43 cases and literature survey

Level III
Darouiche RO, Hamill RJ, Greenberg SB, Weathers SW, Musher DMMedicine (Baltimore)
Clinical implication: Established the association between early surgical decompression and improved neurological outcomes; highlighted that back pain is the universal presenting symptom and that delay is the single biggest modifiable risk factor for poor outcome.
Source: Medicine (Baltimore) 1992 Nov;71(6):369-85
Evidence

Spinal epidural abscess: a meta-analysis of 915 patients

Level III
Reihsaus E, Waldbaur H, Seeling WNeurosurg Rev
Clinical implication: Reinforced that mortality remains significant despite modern antibiotics and surgical techniques; underscores the importance of early MRI in any suspicious case and the primacy of clinical neurological status over laboratory values in guiding surgical timing.
Source: Neurosurg Rev 2000 Dec;23(4):175-204
Evidence

Spinal epidural abscess in clinical practice

Level III
Sendi P, Bregenzer T, Zimmerli WQJM
Clinical implication: Provides a comprehensive clinical framework for diagnosis and management; reinforces that absence of risk factors does not exclude the diagnosis and that MRI must not be delayed in clinically suspicious cases.
Source: QJM 2008 Jan;101(1):1-12
Evidence

Cervical epidural abscess after epidural steroid injection

Level III
Huang RC, Shapiro GS, Lim M, Sandhu HS, Lutz GE, Herzog RJSpine (Phila Pa 1976)
Clinical implication: Highlights that invasive spinal procedures (epidural steroid injection, epidural catheterisation) are a recognised risk factor and that cervical abscess in particular demands emergency decompression — a small collection in the narrow cervical canal produces severe cord compression far earlier than in the lumbar canal.
Source: Spine (Phila Pa 1976) 2004 Jan 1;29(1):E7-9
Evidence

When Do You Drain Epidural Abscesses of the Spine?

Level III
Pourtaheri S, Issa K, Stewart T, Patel Y, Sinha K, Hwang K, Emami ASurg Technol Int
Clinical implication: Provides a decision framework for operative timing — surgery is non-negotiable for any neurological deficit, but a highly selected neurologically intact subgroup can be managed non-operatively provided monitoring is rigorous and the threshold to abandon conservative care is low.
Source: Surg Technol Int 2016 Oct 26;29:374-378

Clinical Decision Scenarios

Practise clinical reasoning and management decisions out loud

Viva scenarioAdvanced
Clinical prompt

A 55-year-old man with insulin-dependent diabetes (HbA1c 9.8%) presents with a 5-day history of progressive mid-thoracic back pain, fever, and now difficulty walking. Examination reveals T10 sensory level, bilaterally hyperreflexic legs, upgoing plantars, and 3/5 power in the hip flexors and knee extensors. MRI shows a dorsal epidural collection from T8 to T11 with cord compression. Blood cultures have been sent but results are pending. What do you do?

Practical approach
This patient has a thoracic spinal epidural abscess with established cord compression (T10 sensory level, upper motor neuron signs in the legs, and 3/5 power in proximal lower limb muscle groups). He is Heusner stage 3 — motor deficit is present. He requires urgent surgical decompression. **Pre-operative plan**: Draw two sets of blood cultures (already sent — check results are pending and antibiotics have not been given yet). Start empiric IV vancomycin (target trough 15-20 mcg/mL for Staph aureus including MRSA) plus IV ceftriaxone 2 g daily. Confirm blood cultures are drawn before the first dose. **Surgical plan**: Urgent posterior laminectomy from T8 to T11 under general anaesthesia, prone position on Jackson table with SSEP and MEP monitoring. Wide laminectomy over the MRI-demonstrated collection extent, extending at least one level above and below (T7-T12 exposure). Evacuate the epidural pus, collect three separate specimens for culture (aerobic, anaerobic, TB, fungal), copious irrigation with at least 3 litres of warm saline. **Wound management**: Given his HbA1c of 9.8% (poorly controlled diabetes), I would consider either primary closure with a closed suction drain and careful post-operative wound monitoring, or a VAC dressing with planned re-look and delayed primary closure at 48-72 hours. I would optimise his glycaemic control with an insulin sliding scale perioperatively. **Stabilisation**: A 4-level laminectomy (T8-T11) carries a significant instability risk. I would place posterior pedicle screw instrumentation from T7 to T12 to provide stability and allow early mobilisation. **Post-operative**: ICU or high-dependency care. Hourly neurological observations for 48 hours. Continue IV antibiotics (vancomycin plus ceftriaxone empirically) for 6-8 weeks, adjusting to culture-directed therapy when results return. CRP twice weekly. TLSO brace for 12 weeks. Neurorehabilitation input from day 2-3. **Key message to the examiner**: the single most important decision is timing. This patient has cord compression with motor deficit and is deteriorating. He needs to be in theatre urgently — the neurosurgical or spinal team must be contacted immediately. Blood cultures first, then antibiotics, then straight to theatre.
Viva scenarioAdvanced
Clinical prompt

A 35-year-old intravenous drug user presents with severe lumbar back pain and a fever of 38.9 degrees Celsius. She has no motor weakness, no sensory deficit, and normal bladder and bowel function. MRI shows a posterolateral epidural collection at L3-L4 compressing the thecal sac to about 60% of its normal anteroposterior diameter. Blood cultures are growing MSSA. How do you manage her?

Practical approach
This patient has a spinal epidural abscess confirmed on MRI with MSSA bacteraemia, but she is neurologically intact — no motor weakness, no sensory deficit, and normal sphincter function. She is a candidate for careful non-operative management with close monitoring, provided several conditions are met. **Non-operative management criteria**: She is neurologically intact. The collection is posterolateral (dorsal approach is available if she deteriorates). The organism is known (MSSA) with known susceptibility. She can communicate changes in her condition. The collection causes significant compression (60% canal compromise) but she has no neurological signs — this is a borderline case and I would discuss it with the spinal team and infectious diseases. **My management plan**: Admit to hospital. Institute an hourly neurological observation chart — nursing staff must assess and document lower limb power, sensation, and sphincter function every hour and escalate immediately if any change occurs. Start targeted IV flucloxacillin 2 g every 6 hours (narrowed from vancomycin now that MSSA is confirmed; flucloxacillin has better bone penetration than vancomycin). Continue antibiotics for 6-8 weeks total. Serial CRP twice weekly. Repeat MRI at 48-72 hours to assess for interval change in collection size. **Threshold for surgery**: Any new motor weakness, sensory change, or sphincter disturbance — even if mild — mandates immediate cancellation of non-operative management and urgent posterior laminectomy. Failure of CRP to trend down over the first week, or expansion of the collection on repeat MRI, are also indications for surgery. **Special considerations**: She is an IV drug user — screen for hepatitis B, C, and HIV. Social circumstances and OPAT suitability must be assessed before discharge (infectious diseases and social work input). She must be counselled that continued intravenous drug use increases the recurrence risk significantly. **If she deteriorates**: Urgent posterior L3-L4 laminectomy with abscess evacuation and copious irrigation. Intraoperative cultures, post-operative antibiotics for 6-8 weeks. The fact that she was initially neurologically intact gives her a good prognosis if surgery is performed promptly at the first sign of deterioration.
Viva scenarioAdvanced
Clinical prompt

You are called to see a 70-year-old woman 4 days after a T12-L1 laminectomy and pedicle screw fixation for a dorsal epidural abscess. She was neurologically improving post-operatively, but today she has new fever (38.5 degrees Celsius), wound erythema, and a serosanguinous discharge from the wound. Her CRP was falling and is now rising again. What is your assessment and management plan?

Practical approach
This patient has a post-operative wound infection with early wound breakdown at the surgical site 4 days after spinal epidural abscess decompression and instrumentation. This is a serious complication because she has hardware in situ and the original pathology was infection — the combination of hardware and wound breakdown creates a high risk of persistent deep infection. **Assessment**: Wound erythema with serosanguinous discharge 4 days post-operatively, new fever, and a rising CRP trajectory after an initial decline are highly consistent with surgical site infection (deep, given the timing and the presence of hardware). I must distinguish between superficial wound infection and deep infection involving the hardware and epidural space. **Immediate investigations**: Wound swab and tissue cultures (aerobic, anaerobic, fungal). Blood cultures (two sets). CRP, ESR, FBC, renal function. Urgent MRI of the thoracolumbar spine with gadolinium to assess for epidural collection recurrence, paraspinal collection, and hardware loosening. Plain radiographs to check hardware position and alignment. **Management — if deep infection is confirmed (the most likely scenario)**: 1. Return to theatre for wound exploration, debridement, and copious irrigation. Open the wound, take deep tissue and hardware-tip cultures. 2. Decision about hardware: I would retain the pedicle screws at this stage if they appear solid, because removal would destabilise a recently fused thoracolumbar junction. Deep infection with retained hardware requires suppressive antibiotics until fusion is solid (typically 6-12 weeks), followed by hardware removal once fusion is confirmed on CT. 3. Wound management: I would not close primarily. Place a VAC dressing with planned re-look and washout in 48-72 hours. Continue serial debridements until the wound is clean and granulating, then close by secondary intention or delayed primary closure. 4. Antibiotics: adjust based on new culture results. If the original organism was MSSA and the wound culture shows the same organism, continue flucloxacillin (or switch back to vancomycin if MRSA). If a new organism is identified, target accordingly. Antibiotics continue for a total of 6-8 weeks minimum, possibly longer if hardware is retained. 5. Nutritional optimisation: this patient is 70 and likely has comorbidities that impair wound healing — check albumin, pre-albumin, zinc, and vitamin C; correct deficiencies; consider nutritional supplementation.
Exam day cheat sheet
Spinal Epidural Abscess Decompression — Exam Day Summary

References

  1. Darouiche RO, Hamill RJ, Greenberg SB, Weathers SW, Musher DM. (1992). Bacterial spinal epidural abscess: review of 43 cases and literature survey. Medicine (Baltimore) 1992 Nov;71(6):369-85. PMID: 1359381. — Landmark case series establishing Staphylococcus aureus as the dominant pathogen and demonstrating the association between early surgery and improved neurological outcomes.

  2. Reihsaus E, Waldbaur H, Seeling W. (2000). Spinal epidural abscess: a meta-analysis of 915 patients. Neurosurg Rev 2000 Dec;23(4):175-204. PMID: 11153548 (DOI: 10.1007/pl00011954). — Large meta-analysis of published cases confirming mortality of 5-30% and the primacy of early decompression in determining outcome.

  3. Sendi P, Bregenzer T, Zimmerli W. (2008). Spinal epidural abscess in clinical practice. QJM 2008 Jan;101(1):1-12. PMID: 17982180 (DOI: 10.1093/qjmed/hcm100). — Comprehensive clinical review describing the classical presentation, diagnostic workup, and treatment framework used in contemporary practice.

  4. Huang RC, Shapiro GS, Lim M, Sandhu HS, Lutz GE, Herzog RJ. (2004). Cervical epidural abscess after epidural steroid injection. Spine (Phila Pa 1976) 2004 Jan 1;29(1):E7-9. PMID: 14699291 (DOI: 10.1097/01.BRS.0000106764.40001.84). — Case report with literature review of post-injection cervical epidural abscess, emphasising the rapid neurological deterioration possible in cervical epidural abscess and the need for emergent decompression.

  5. Pourtaheri S, Issa K, Stewart T, Patel Y, Sinha K, Hwang K, Emami A. (2016). When do you drain epidural abscesses of the spine? Surg Technol Int 2016 Oct 26;29:374-378. PMID: 27608748. — Surgical review of indications and timing for drainage of spinal epidural abscess, summarising when neurological deficit, sepsis or failure of antibiotics mandates surgery and when highly selected neurologically intact patients can be managed non-operatively.

  6. Wang LP, Hauerberg J, Schmidt JF. (2001). Long-term outcome after neurosurgically treated spinal epidural abscess following epidural analgesia. Acta Anaesthesiol Scand 2001 Feb;45(2):233-9. PMID: 11167170 (DOI: 10.1034/j.1399-6576.2001.450215.x). — Single-centre case series with long-term neurological follow-up after surgical treatment of spinal epidural abscess arising as a complication of epidural analgesia, demonstrating the high rate of residual deficit in delayed presentations.

Editorially reviewed — transparent references and correction processPublished by OrthoVellum Medical Education TeamEditorial boardMethodologyReview policy
Educational disclosure

Educational content is reviewed for source visibility, editorial coherence, and correction readiness.

No individual clinician credential is claimed unless a named person is shown.

Verify before clinical use; this is not medical advice or a substitute for local guidance.