Learning outcomes
By the end of this lesson you should be able to: explain why acute haematogenous osteomyelitis occurs at the metaphysis of a long bone and name the organism; distinguish acute osteomyelitis from septic arthritis by the age of the patient, the joint findings and the blood tests; apply the Kocher criteria to decide which child with a painful hip needs an arthrotomy; describe the investigation and management of both conditions, including why antibiotics must follow cultures; distinguish acute from chronic osteomyelitis and describe the features and treatment of Brodie's abscess; and list the differentials of a single painful joint in a child and in an adult.
Prerequisites
Lessons 1-2 (fractures and complications) and Lesson 4 (lower limb injuries, including the slipped capital femoral epiphysis and the child with knee pain). From Gross Anatomy I: the cancellous-metaphyseal bone and the growth plate, and the joints of the hip and knee.
The big idea
One anatomical fact explains the biology. The metaphysis of a growing long bone is supplied by the nutrient artery, whose branches run in a hairpin loop towards the growth plate and then turn back — and in doing so, in the child, they empty into venous sinusoids in which the blood is slow and swirling rather than streaming. Anything carried in the bloodstream therefore tends to be dropped exactly there. Add the further fact that the area is relatively avascular on the far (epiphyseal) side of the plate, where the phagocytes cannot easily reach, and you have the reason for both the site and the tendency to become chronic.
NoteThe metaphyseal infection also explains a piece of paediatric anatomy that will be examined again and again: in infants under about 18 months, transmitsseal vessels cross the growth plate, so metaphyseal infection can spread into the epiphysis and the joint; after that age, the physis is an avascular barrier, so metaphyseal osteomyelitis stays out of the joint and the joint is only involved if the metaphysis is intracapsular — as in the hip, the shoulder and the elbow, where the metaphysis lies within the capsule.
Acute haematogenous osteomyelitis
| Feature | What you see |
|---|---|
| Patient | Usually a child aged 2-10, boys more than girls; a newborn or an immunocompromised adult in the less typical cases |
| Organism | Staphylococcus aureus in the great majority; Kingella kingae in the toddler; Salmonella in sickle cell disease; Mycobacterium tuberculosis anywhere tuberculosis is common |
| Route | Haematogenous — from a skin infection, a dental infection, or sometimes no identifiable source |
| Site | The metaphysis of a long bone, most often the distal femur, proximal tibia and proximal humerus; rarely the axial skeleton |
| Symptoms | Sudden onset of fever and malaise with a severe, deep, unrelenting pain in one limb; the child refuses to bear weight or move the limb (pseudoparalysis in the newborn) |
| Signs | Focal bony tenderness over the metaphysis, which is the earliest and most reliable sign; warmth and swelling develop late; the joint beside it is usually not irritable |
| Investigations | Raised CRP and ESR; a normal total white count in about half; blood cultures in about half; radiographs are normal in the first 10-14 days and then show periosteal reaction and lytic change. MRI is the imaging of choice for early marrow oedema |
| Treatment | IV antibiotics after blood cultures have been drawn (flucloxacillin or a cephalosporin; adjust to the organism); drainage of any subperiosteal or intraosseous abscess; splint for comfort; then oral antibiotics for a total of 4-6 weeks |
Watch outDo not wait for radiographic change before starting treatment. The plain film is normal for the first 10-14 days of the disease, by which time irreversible growth-plate and articular damage can have occurred. The diagnosis is clinical — fever, one focal point of bony tenderness, and raised inflammatory markers — and the imaging you need is MRI for early marrow change, not a plain film waiting to catch up.
The natural history explains the urgency. Infection confined within the medullary canal raises the intramedullary pressure, which compresses the vessels that supply the cortex from within and additionally lifts the periosteum so that the cortical bone is deprived from outside as well. The result, if untreated, is a segment of dead bone — a sequestrum — that no antibiotic can reach and no immune cell can find. The infection then becomes a chronic abscess, and the surgery has to be done to remove bone that is already dead.
Septic arthritis
| Acute osteomyelitis | Septic arthritis | |
|---|---|---|
| Age of onset | Children, rarely adults | Any age, with a peak in the young child and the elderly |
| Site | Metaphysis of a long bone | A large joint — most often the hip in the child, the knee in the adult |
| The joint | The joint beside the bone is spared, and the child will move it gently (unless the metaphysis is intracapsular) | The joint itself is the focus: held in a position of comfort, exquisitely tender, and rigid on attempted movement |
| Temperature | Usually febrile | Usually febrile, but a newborn or an immunosuppressed patient may not be |
| Organism | Staphylococcus aureus; Kingella in toddlers; Salmonella in sickle cell | Staphylococcus aureus in adults; the neonate with group B Streptococcus; Kingella in the child; Neisseria gonorrhoeae in a sexually active adult |
| The emergency | Drain an abscess | Drain the joint. Every hour counts |
In a child the differential diagnosis of the painful hip is the clinical problem that most often goes wrong, because the hip lies deep and the signs are indirect. Septic arthritis must be separated from transient synovitis, which is a benign post-viral inflammation that resolves in days, and from the slipped capital femoral epiphysis and Perthes' disease, which have their own management.
| Differential | How it is distinguished |
|---|---|
| Septic arthritis | Fever, inability to bear weight, systemic toxicity, raised CRP and WCC, an effusion on ultrasound; joint aspiration gives pus or organisms |
| Transient synovitis (irritable hip) | A mild or absent fever, a well child, a moderate CRP (< 2 mg/dL), a normal or low WCC, a small effusion, and symptoms that improve within days. The classic Kocher-discriminating picture: the child walks, but limps |
| Perthes' disease (Legg-Calvé-Perthes) | A child of 4-8 years with a painless or mildly painful limp, no fever, and avascular necrosis of the femoral head on the radiograph |
| Slipped capital femoral epiphysis (SCFE) | An obese adolescent with a limp, referred knee pain and obligatory external rotation on hip flexion; frog-leg radiographs are diagnostic (Lesson 4) |
| Juvenile idiopathic arthritis | A persistent effusion over weeks without fever; other joints involved; a raised ESR but a sterile aspirate |
Watch outA child who will not bear weight on a leg, refuses to move one hip, and has a fever is drained. Septic arthritis of the hip destroys the femoral head's articular cartilage within days, and the subcapital epiphyseal vessels are tamponaded by the intra-articular pressure, so the head also becomes avascular. The management is aspiration (or open arthrotomy), antibiotics, and follow-up radiographs for avascular necrosis — not observation, however well the child seems.
The Kocher criteria
| Number of Kocher criteria present | Predicted probability of septic arthritis | Action |
|---|---|---|
| 0 | Under 0.2% | Discharge with instructions and review; the picture is transient synovitis |
| 1 | Approximately 3% | Consider ultrasound-guided observation or aspirate |
| 2 | Approximately 40% | Aspirate the hip under image guidance |
| 3 | Approximately 93% | Aspirate, and prepare for surgical drainage |
| 4 | Essentially 100% | Immediate joint drainage and IV antibiotics |
NoteThe rule works because of two of the four criteria in particular: a history of fever and an inability to bear weight are the pair that separates septic arthritis from transient synovitis in the original study. When a child with an irritable hip walks into the room and is febrile, the ducks are already in a row. The CRP adds the strongest single discriminator in practice, and a CRP above 2 mg/dL with a high WCC pushes hard towards drainage.
A four-year-old has a fever of 39 °C, refuses to bear weight, has an ESR of 48 mm/h and a white cell count of 16 × 10⁹/L, and ultrasound confirms a hip effusion. What is the correct management?
Chronic osteomyelitis and Brodie's abscess
| Feature | Chronic osteomyelitis | Brodie's abscess |
|---|---|---|
| How it arises | Untreated or inadequately treated acute infection; contamination of an open fracture; infection of an implant | An acute infection that has been partially contained by a robust host response |
| Radiograph | Sequestrum (dense dead bone) within a cavity, periosteal new bone (involucrum), sclerosis, sometimes a sinus tract | A small, round, well-defined lytic lesion with a sclerotic rim |
| Treatment | Radical debridement of all dead bone and infected tissue, culture-directed antibiotics, and reconstruction of any bone defect. Antibiotics alone cannot sterilise dead bone | Curettage or drilling of the cavity plus antibiotics; usually it settles |
| Prognosis | The recurrence risk is high and the management is surgical by definition | Excellent, provided it is recognised as an infection and not as a tumour |
Watch outA Brodie's abscess is the trap in the other direction: it looks like an osteoid osteoma or a eosinophilic granuloma on the film and has been misdiagnosed as a malignancy. But the story of a painful, boring night pain, deep and dull, in a metaphysis, with a rim of sclerosis, should prompt antibiotics and curettage — and the tissue grown on culture. If you are considering a bone biopsy for suspected tumour, chronic and subacute infection must be on the differential list too.
Common mistakes
Watch outWaiting for a radiograph before treating osteomyelitis. The film is normal for 10-14 days in acute haematogenous osteomyelitis. The decision is made on fever, focal bony tenderness and the inflammatory markers; the imaging you want is an MRI showing marrow oedema. Starting antibiotics on the day of presentation is the difference between treating an infection and removing dead bone later.
Watch outStarting antibiotics before cultures are drawn. Once the first dose is in, the yield from blood and from joint aspirate falls dramatically and you are choosing antibiotics blind. Draw blood cultures (and aspirate the joint if the diagnosis is septic arthritis) first, then give the antibiotic — this is the standard 'two from the blood, then the drug' sequence.
Watch outTreating a chronically discharging osteomyelitis with courses of oral antibiotics. Chronic osteomyelitis is a surgical disease: dead bone cannot be sterilised by antibiotics, because it has no blood supply to deliver the drug. The treatment is debridement of sequestra and infected tissue, and antibiotics around the surgery.
Watch outForgetting that a sore joint in an adult with a neuropathy, or a chronically painful joint in an immunocompromised patient, may be infected. The elderly diabetic with a hot, swollen knee and no fever, or the patient with rheumatoid arthritis on a biologic, has septic arthritis until proven otherwise — the absence of a fever is not the absence of pus.
Check your understanding
Why does metastatic infection of bone lodge at the metaphysis of a long bone in a child?
Name the four Kocher criteria and state the probability of septic arthritis when three are present.
Why can chronic osteomyelitis not be eradicated with antibiotics alone, even with a prolonged course of culture-directed therapy?
Multiple choice
A four-year-old refuses to bear weight, has a temperature of 39 °C, ESR 48 mm/h and white cell count 16 × 10⁹/L; ultrasound confirms a hip effusion. What is the most appropriate immediate management?
A six-year-old has fever, focal metaphyseal tibial tenderness and refusal to bear weight with raised ESR and CRP; early radiographs are normal but MRI shows marrow oedema with a subperiosteal collection. Which organism and management pairing is most appropriate?
Which statement is correct about a Brodie's abscess compared with chronic osteomyelitis?