Paediatric Pelvic Osteomyelitis & Sacroiliac Joint Infection

A vague presentation. Pelvic osteomyelitis and septic sacroiliitis present vaguely: a child with a limp or refusal to weight-bear, fever, and pain that may be referred to the hip, buttock, groin, thigh or low back. The presentation is non-localising, so the diagnosis is frequently missed or made late, and the infection is a great mimic: it is mistaken for a septic hip, transient synovitis, discitis or another spinal problem, or even an abdominal cause such as appendicitis.

Keep it in the differential. Think of pelvic or sacroiliac infection in the limping, febrile child, especially when the hip examination is relatively unremarkable, the pain localises to the buttock or back, or a 'septic hip' does not quite fit. Obtain an early MRI rather than waiting for plain films to change, which early on they often do not. Timely diagnosis and antibiotic-led treatment usually give an excellent outcome; delay risks abscess, chronic infection and growth disturbance.

Why the pelvis. Haematogenous osteomyelitis seeds metaphyseal equivalents: portions of flat or irregular bone adjacent to cartilage, with a sluggish vascular anatomy similar to that of a long-bone metaphysis. The pelvis has several. The sacroiliac joint region is the most frequently involved, and the ilium, pubis, ischium and triradiate cartilage are also affected, which is why pelvic infection often straddles bone and the adjacent joint as septic sacroiliitis.

View larger

The organisms. Staphylococcus aureus predominates. Consider Kingella kingae in younger children and Salmonella in sickle-cell disease, and seek a tissue or blood diagnosis before treatment where possible.

Kingella has to be actively looked for. It is the leading organism in children aged roughly 6 to 48 months when PCR is used, and the qualifier matters because it grows poorly on routine plates: a child with genuine Kingella infection frequently has negative conventional cultures. Inoculating samples into blood-culture bottles and requesting PCR, including oropharyngeal PCR, is what converts a culture-negative case into a diagnosis. The clinical picture is often milder, with lower fever and a lower CRP than staphylococcal disease, so the diagnosis gets missed twice over.

Not every destructive lesion is infection. A subacute pelvic focus may be a Brodie's abscess, which mimics, and is mimicked by, tuberculosis and bone tumours. Do not assume a destructive pelvic lesion is infective: when the picture is atypical obtain tissue, and biopsy or curettage is sometimes needed to confirm the diagnosis and exclude tumour.

Antibiotics lead. Most children respond to appropriately chosen antibiotics alone. The course follows the same principles as acute haematogenous osteomyelitis elsewhere, and its figures are taken from Paediatric Acute Osteomyelitis so that one set of numbers applies wherever in the skeleton the infection sits.

Drain when needed. Surgical drainage is reserved for:

• A drainable abscess
• Sacroiliac joint sepsis that is not responding
• Failure to improve on antibiotics

Delay risks growth disturbance, and in the pelvis that phrase has a specific meaning, because it determines how long the child is followed.

The triradiate cartilage is the acetabulum's growth plate. It is the Y-shaped physis where the ilium, ischium and pubis meet, and it expands the acetabulum to keep pace with the enlarging femoral head until it closes in early adolescence. It is also one of the pelvic metaphyseal equivalents, so infection reaches it.

Damage does not declare itself at the time. The infection is treated and the child recovers; the consequence appears over the following years as the acetabulum fails to grow. The result is a progressively shallow, dysplastic acetabulum, with lateralisation and subluxation of the femoral head and the secondary problems of an incongruent hip. A premature partial closure can instead produce asymmetrical growth and acetabular deformity rather than simple shallowness.

The follow-up outlasts the infection. Where imaging showed involvement of the triradiate cartilage or the acetabulum, the child needs radiographic surveillance of the hip through skeletal maturity, not discharge once the CRP has normalised, because the treatable window for developing acetabular dysplasia is long before the hip becomes symptomatic. Tell the family at the outset, so that years of appointments after an apparently cured infection make sense to them.

The same logic applies, less commonly, to infection crossing into the proximal femoral physis from a septic hip, covered in Septic Arthritis of the Paediatric Hip.

Why the hip misleads. The Kocher criteria (fever, non-weight-bearing, raised ESR, raised WCC) were derived to separate a septic hip from transient synovitis. They are hip-specific and do not capture pelvic or sacroiliac infection. A child who does not fit Kocher, or whose hip is comparatively comfortable, can still have pelvic sepsis, and that is exactly how the diagnosis is missed.

Signs that point to the pelvis. A hip that examines relatively well in a febrile, limping child is reassuring only if the hip is the thing wrong with them. Look instead for:

• Hip rotation and log-roll relatively preserved and not agonising, unlike the held, irritable hip of septic arthritis
• Direct tenderness over the sacroiliac joint, posterior pelvis or pubis
• Positive sacroiliac provocation tests that reproduce buttock or SI pain: FABER (Patrick's), pelvic compression and distraction, Gaenslen's, and the flamingo or single-leg stand

An irritable hip does not clear the pelvis. In the Kuckelman multicentre series of sacroiliac joint infection (see the evidence below), hip motion was irritable in 93%: the hip examination was abnormal even though the joint was innocent. A comfortable hip is a clue when it is present, but an irritable one does not prove the hip is the source, and the sacroiliac signs still need to be sought.