Osteomyelitis
Description
- Infection of bone and marrow -> inflammation, vascular compromise, bone necrosis (sequestrum) and new bone formation (involucrum)
- Sequestrum is avascular -> antibiotics do not reach it -> surgical debridement, not longer antibiotics, is what cures chronic disease
Classification - by route
| Features | |
|---|---|
| Haematogenous | Children: metaphysis of long bones (slow sinusoidal flow). Adults: vertebral body |
| Contiguous spread | Diabetic foot ulcer, pressure injury, dental, sinus, chronic wound |
| Direct inoculation | Open fracture, surgery, prosthetic joint, penetrating trauma |
By chronicity
- Acute - days to weeks; systemic features; no sequestrum yet -> antibiotics may cure without surgery
- Chronic - sequestrum, involucrum, sinus tract, recurrent flares; surgery is almost always required
Anatomical (Cierny-Mader)
- I medullary / II superficial / III localised (full-thickness, stable) / IV diffuse (unstable)
- Modified by host: A normal, B compromised (local or systemic), C treatment worse than the disease
Epidemiology
- Bimodal: children (haematogenous long bone) and adults >50 (vertebral, diabetic foot, prosthetic)
- Vertebral osteomyelitis incidence rising - ageing population, more instrumentation, more IV drug use, more bacteraemia
- Diabetic foot osteomyelitis complicates ~20% of moderate and ~60% of severe diabetic foot infections
- Staphylococcus aureus causes ~50-70% of all cases across every category
- High burden among Aboriginal and Torres Strait Islander people - diabetes, remoteness, delayed presentation
- Prosthetic joint infection: ~1-2% of primary arthroplasties
Aetiopathogenesis
- Bacterial adherence -> biofilm formation on necrotic bone and hardware
- Biofilm -> metabolically dormant organisms, 1000x antibiotic tolerance, immune evasion
- This is why implant retention often fails and why cure needs debridement or removal
- Inflammation -> raised intraosseous pressure -> thrombosis of nutrient vessels -> necrosis
- S. aureus invades and survives within osteoblasts and induces RANKL -> osteoclastic bone destruction
- Small colony variants -> persistence and relapse years later
Organisms by setting
| Setting | Organism |
|---|---|
| All settings, commonest | Staphylococcus aureus (MSSA and MRSA) |
| Prosthetic/hardware | Coagulase-negative staphylococci, Cutibacterium acnes (shoulder), S. aureus |
| Diabetic foot | Polymicrobial - S. aureus, streptococci, Enterobacterales, anaerobes, Pseudomonas |
| Sickle cell disease | Salmonella (and still S. aureus) |
| IV drug use | S. aureus, Pseudomonas, Candida; sternoclavicular and vertebral |
| Vertebral, indolent, endemic exposure | Mycobacterium tuberculosis (Pott disease), Brucella |
| Puncture through a shoe | Pseudomonas aeruginosa |
| Human/animal bite | Eikenella, Pasteurella, anaerobes |
| Neonatal | Group B streptococcus, E. coli, S. aureus |
Diagnosis
Clinical
- Acute: localised bone pain, fever, tenderness, overlying warmth/swelling, refusal to weight-bear
- Vertebral: insidious back pain worse at night and at rest, focal spinal tenderness, often afebrile
- *Any new back pain with raised inflammatory markers, or with S. aureus bacteraemia, is vertebral osteomyelitis until excluded*
- New neurological deficit -> epidural abscess -> urgent MRI and neurosurgical referral
- Chronic: sinus tract discharging onto the skin is pathognomonic, non-healing ulcer, deformity
- Diabetic foot: neuropathy blunts pain; the ulcer, not the bone, is often the only sign
- Positive probe-to-bone test - a sterile probe passes through the ulcer to gritty bone
- The most predictive bedside sign of underlying osteomyelitis; high PPV in a high-prevalence setting
- Ulcer >2 cm2, depth >3 mm, duration >1-2 weeks, "sausage toe" all raise probability
- Positive probe-to-bone test - a sterile probe passes through the ulcer to gritty bone
Bloods
- CRP and ESR - most useful for monitoring response, not for diagnosis (normal values do not exclude, especially in diabetes)
- FBE, UEC, glucose/HbA1c
- Blood cultures x2 before antibiotics - positive in ~50% of haematogenous and vertebral disease; a positive blood culture may remove the need for bone biopsy
Imaging
| Plain X-ray | Normal for the first 10-14 days (needs 30-50% bone loss). Later: periosteal reaction, lucency, sequestrum. Cheap, useful as a baseline and for Charcot |
| MRI - the imaging test of choice | Sensitivity ~90%, specificity ~80%; marrow oedema, abscess, epidural extension, soft tissue. Positive within days |
| CT | Best for sequestrum, cortical destruction and surgical planning |
| Nuclear medicine | Three-phase bone scan sensitive but non-specific; labelled leucocyte SPECT/CT or FDG-PET where MRI is contraindicated or hardware causes artefact |
Microbiology - the decisive step
- *Obtain bone for culture and histology before starting antibiotics wherever the patient is stable*
- Percutaneous image-guided or open biopsy; ideally off antibiotics for 2 weeks if already started
- *Superficial wound and sinus tract swabs do NOT reflect bone pathogens - except S. aureus, which correlates reasonably*
- Prosthetic joint infection: >=3-5 periprosthetic tissue samples, sonication of explanted hardware, prolonged incubation (Cutibacterium up to 14 days)
- Send for histopathology as well as culture - inflammatory cells plus necrosis confirms osteomyelitis when cultures are negative
- Consider mycobacterial and fungal culture in indolent disease or the immunosuppressed
The key differential in the diabetic foot
| Discriminator | |
|---|---|
| Charcot neuroarthropathy | Warm, swollen, deformed midfoot, usually no ulcer; MRI marrow oedema can look identical - distinguish by distribution, joint involvement, and clinical context |
| Soft tissue infection alone | No bone contact, ulcer <2 cm2, normal X-ray |
| Gout, fracture, malignancy | History, imaging |
Management
Three pillars: microbiological diagnosis, surgical source control, and targeted antibiotics of adequate duration. Vascular supply and glycaemia determine whether any of it works.
1. Get the diagnosis before treating
- Delay antibiotics until bone or blood cultures are obtained, unless septic or with rapid neurological deterioration
- Empirical broad-spectrum therapy in a stable patient with chronic osteomyelitis destroys the only chance of a targeted regimen
2. Surgery
- Debridement to healthy bleeding bone; excise sequestrum, sinus tracts and necrotic tissue
- Dead space management: antibiotic-loaded cement/beads, local flap, bone graft
- Stabilisation; remove infected hardware once the fracture has united
- Prosthetic joint infection: DAIR (debridement, antibiotics, implant retention) only if early (<3-4 weeks), stable implant, susceptible organism; otherwise one- or two-stage revision
- Vertebral osteomyelitis is usually managed non-operatively; operate for epidural abscess with neurological deficit, instability, deformity, or failure of medical therapy
- Diabetic foot: assess and revascularise before deciding on amputation; conservative surgery (resection of the infected bone) is often preferable to major amputation
3. Antibiotics
- Targeted to culture; high dose; bone-penetrating
| Organism | Agent |
|---|---|
| MSSA | Flucloxacillin 2 g IV 4-6 hourly (or cefazolin) |
| MRSA | Vancomycin (AUC-guided) or daptomycin; oral step-down: clindamycin, doxycycline, trimethoprim-sulfamethoxazole, linezolid |
| Streptococci | Benzylpenicillin/ceftriaxone |
| Enterobacterales | Ceftriaxone or ciprofloxacin (excellent bone penetration) |
| Pseudomonas | Ceftazidime/piperacillin-tazobactam, then ciprofloxacin |
| Prosthetic/staphylococcal biofilm | Add rifampicin (never as monotherapy - rapid resistance; check interactions) |
- Duration
- Native bone osteomyelitis: ~6 weeks from the last debridement (or from the start of effective therapy)
- Vertebral osteomyelitis: 6 weeks (12 weeks if undrained abscess, extensive disease or slow response)
- Diabetic foot osteomyelitis: 6 weeks if bone is NOT resected; ~3 weeks if residual infected bone only after minor amputation; ~2-5 days if all infected bone is removed
- Oral step-down is evidence-based - OVIVA showed oral therapy non-inferior to IV at 6 weeks, with fewer line complications and lower cost
- Switch at ~1-2 weeks once clinically improving, source controlled, organism known and a highly bioavailable oral agent is available
- Australian practice uses OPAT or oral step-down rather than prolonged inpatient IV therapy
4. Host optimisation - decides the outcome more than the antibiotic
- Revascularisation where there is peripheral arterial disease
- Glycaemic control, nutrition (albumin, protein, vitamin D), smoking cessation
- Offloading - total contact cast or removable device for the diabetic foot
- Pressure area care; treat oedema; MDT high-risk foot service
5. Monitoring
- Clinical response + CRP trend (CRP falling by ~50% in the first 1-2 weeks predicts response)
- Do not repeat MRI early - imaging lags clinical recovery by months
- Drug safety monitoring (FBE, UEC, LFT; vancomycin levels; linezolid - cytopenias, neuropathy, serotonin toxicity)
Associations
- Diabetes mellitus with peripheral neuropathy and peripheral arterial disease - the dominant Australian context
- Staphylococcus aureus bacteraemia - always ask where the metastatic focus is
- Injecting drug use, haemodialysis, indwelling vascular catheters
- Orthopaedic hardware, prosthetic joints, open fractures, spinal instrumentation
- Sickle cell disease (Salmonella), immunosuppression, HIV, malignancy
- Pressure injuries, spinal cord injury, peripheral vascular disease
- Endocarditis (consider a TTE/TOE in vertebral osteomyelitis with staphylococcal or streptococcal bacteraemia)
- Charcot neuroarthropathy - the diagnostic confounder
- Tuberculosis (Pott disease), brucellosis, melioidosis (northern Australia), fungal infection
- Chronic recurrent multifocal osteomyelitis (CRMO) - sterile, autoinflammatory, paediatric
Natural history & complications
- Acute haematogenous osteomyelitis in children: >90% cure with prompt antibiotics alone
- Adult chronic osteomyelitis: relapse in 20-30%, sometimes years later (small colony variants, biofilm, residual sequestrum)
- Cure requires adequate debridement - antibiotic duration cannot compensate for retained dead bone
- Diabetic foot osteomyelitis
- Remission ~60-80% with combined medical and surgical management and revascularisation
- Major amputation in 10-20%; 5-year mortality after a major amputation exceeds that of many cancers
- Recurrent ulceration in ~40% within a year - lifelong podiatry and footwear
- Vertebral osteomyelitis: most recover; residual back pain in ~30%; neurological deficit is usually permanent if established before treatment
- Prosthetic joint infection: success ~70-90% for two-stage revision, lower for DAIR
- Complications: epidural abscess and cord compression, pathological fracture, sepsis, growth arrest (paediatric physeal involvement), chronic sinus, amputation, secondary amyloidosis, Marjolin ulcer (SCC in a chronic sinus tract)
🔒
6 more sections, plus exam facts
Premium unlocks every note across every specialty, and the full exam fact library behind it.
Get premium access