Hematogenous osteomyelitis in children: Evaluation and diagnosis

INTRODUCTION — The evaluation and diagnosis of hematogenous osteomyelitis in children will be discussed here. The epidemiology, pathogenesis, microbiology, clinical features, complications, and management of osteomyelitis in children osteomyelitis are discussed separately:

●(See "Hematogenous osteomyelitis in children: Epidemiology, pathogenesis, and microbiology".)

●(See "Hematogenous osteomyelitis in children: Clinical features and complications".)

●(See "Hematogenous osteomyelitis in children: Management".)

DEFINITION AND PATHOGENESIS

Osteomyelitis is an infection localized to bone [1].

●Acute osteomyelitis – Bone infection diagnosed within two to four weeks after the onset of symptoms or signs in a previously uninfected bone [1]

●Chronic osteomyelitis – More protracted and generally indolent disease accompanied by development of one or more sequestra and/or relapse of infection in the same bone weeks to years after apparently successful treatment of initial infection [2]

Osteomyelitis is usually caused by microorganisms (predominantly bacteria) that enter the bone via the bloodstream (hematogenously).

Osteomyelitis also may result from direct inoculation of bone with bacteria in association with an open fracture or as a complication of a puncture wound or the placement of orthopedic hardware (such as pins or screws). (See "Infectious complications of puncture wounds".)

Less commonly, osteomyelitis may arise as an extension of a contiguous infection (eg, decubitus ulcer, deep wound infection, sinusitis, periodontal disease).

DIAGNOSTIC APPROACH — The diagnosis of osteomyelitis often is uncertain at the initial evaluation. The presentation may be delayed, and signs and symptoms may be nonspecific. A high index of suspicion and monitoring of the clinical course are essential to establishing the diagnosis. We consult an orthopedic surgeon or interventional radiologist as early as possible in the evaluation to assist in obtaining specimens from the site of infection and assessing the need for surgical intervention. (See 'Microbiology' below and "Hematogenous osteomyelitis in children: Management", section on 'Indications for surgery'.)

Clinical suspicion — Based upon the history and physical examination, acute hematogenous osteomyelitis should be suspected in infants and children with findings suggestive of bone infection, including:

●Constitutional symptoms (irritability, decreased appetite or activity), with or without fever

●Focal symptoms and signs of bone inflammation (eg, warmth, swelling, point tenderness) that typically progress over several days to a week; symptoms and signs in infants and small children may be poorly localized

●Limitation of function (limited use of an extremity; limp; refusal to walk, crawl, sit, or bear weight)

Additional clinical features of osteomyelitis in children are discussed separately. (See "Hematogenous osteomyelitis in children: Clinical features and complications", section on 'Clinical features'.)

Osteomyelitis also should also be considered in children who are found to have bacteremia or an abnormal imaging study of bone in the evaluation of trauma or nonspecific signs or symptoms (eg, irritability, fever without a source).

Risk factors that may raise suspicion for osteomyelitis include (see "Hematogenous osteomyelitis in children: Epidemiology, pathogenesis, and microbiology", section on 'Risk factors'):

●Bacteremia or sepsis

●Primary or acquired immune deficiency (eg, sickle cell disease, chronic granulomatous disease)

●Recent or current indwelling vascular catheter, including hemodialysis catheter

●For neonates – Skin infection; complicated pregnancy, labor, or delivery; urinary tract anomalies; late onset neonatal sepsis

We suggest that an orthopedic surgeon or interventional radiologist be consulted as early as possible to assist in obtaining specimens from the site of infection and assessing the need for surgical intervention. (See 'Microbiology' below and "Hematogenous osteomyelitis in children: Management", section on 'Indications for surgery'.)

Initial evaluation — The initial evaluation for children with suspected osteomyelitis includes blood tests and radiographs of the affected area(s).

Blood tests — Initial blood tests for children with suspected osteomyelitis include a complete blood count (CBC) with differential, or C-reactive protein (CRP), and possibly erythrocyte sedimentation rate (ESR) [1,3]. More recent clinical studies and guidelines emphasize measurements of CRP over ESR because of the accuracy of CRP measurements and the short half-life of the protein, which permits assessment of rapid changes in the degree of inflammation present in the body [1,4]. To avoid multiple venipunctures, at least one blood culture should be obtained at the same time as these studies. Blood cultures are discussed below. (See 'Microbiology' below.)

At the time of presentation, elevation of CRP (>10 to 20 mg/L [1 to 2 mg/dL]) and/or ESR (>20 mm/h) is sensitive (approximately 95 percent) in cases of culture-proven osteomyelitis in children, but CRP and ESR are nonspecific [1,4-7]. Elevated CRP and/or ESR support the diagnosis of osteomyelitis but do not exclude other conditions in the differential diagnosis. CRP/ESR may be normal early in the infection in some patients, and repeat testing may be warranted within 24 hours in children in whom osteomyelitis is strongly suspected. CRP and, to a lesser extent, ESR also are important markers for evaluating the response to therapy. (See "Hematogenous osteomyelitis in children: Management", section on 'Response to therapy'.)

Elevation of the white blood cell (WBC) count is neither a sensitive nor a specific indicator of osteomyelitis. In a 2012 systematic review of >12,000 patients, the WBC was elevated in only 36 percent [6]. However, the CBC and differential are helpful in evaluating other considerations in the differential diagnosis of children with bone pain (eg, vaso-occlusive crisis in sickle cell disease, leukemia). (See 'Differential diagnosis' below.)

Radiographs — Radiograph(s) of the affected region(s) should be performed as the initial imaging study to exclude other causes of pain (eg, bone tumors and fractures) [1,3,8-10]. In neonates and young infants, skeletal survey may be warranted because polyostotic disease is common. (See "Hematogenous osteomyelitis in children: Clinical features and complications", section on 'Birth to three months'.)

Radiographs are abnormal at the time of evaluation in most newborn and young infants who are ultimately found to have osteomyelitis [11-13]. In older children, radiographs are usually normal or inconclusive early in the course of osteomyelitis. For older children with suspected osteomyelitis and normal or inconclusive radiographs, magnetic resonance imaging (MRI) is usually indicated [1]. (See 'Advanced imaging' below and 'Differential diagnosis' below.)

Characteristic findings — Children with suspected osteomyelitis and characteristic abnormalities on initial radiographs are likely to have osteomyelitis (table 1). They should receive empiric antibiotic therapy pending further evaluation (microbiology and possibly additional imaging studies). (See "Hematogenous osteomyelitis in children: Management", section on 'Empiric parenteral therapy' and 'Microbiology' below and 'Advanced imaging' below.)

Radiographic findings that are compatible with osteomyelitis include [10]:

●Evidence of periosteal new bone formation (periosteal reaction) (image 1A-B)

●Periosteal elevation (suggestive of periosteal abscess) (image 1D)

●Lytic lesions (image 1C) or sclerosis, indicating subacute/chronic infection

However, these findings generally are not apparent at the onset of symptoms. The timing and typical sequence of radiographic changes of osteomyelitis varies depending on which bones are affected and the age of the patient.

●Long bones – The typical sequence of radiographic changes in long bones of children is as follows [14,15]:

•Approximately three days after symptom onset – Small area of localized, deep, soft-tissue swelling in the region of the metaphysis (image 2)

•Three to seven days after symptom onset – Obliteration of the translucent fat planes interposed within muscle by edema fluid

•Ten to 21 days after symptom onset (7 to 10 days in neonates) – Evidence of bone destruction (osteopenia, osteolytic lesions), periosteal reaction, cortical thickening, periosteal elevation (due to subcortical purulence (image 1A-D)); these lesions may require surgical debridement or drainage (see "Hematogenous osteomyelitis in children: Management", section on 'Indications for surgery')

•One month or longer – Lytic sclerosis

●Membranous, irregular bones – Bone destruction and periosteal elevation generally are apparent two to three weeks later than in long bones.

●Pelvic bones – Radiographs usually are not useful in the diagnostic evaluation of osteomyelitis of the pelvis; in one large series, they were abnormal in only 25 percent of patients [16].

●Vertebral osteomyelitis – In children with vertebral osteomyelitis, radiographs are normal at initial presentation in more than 50 percent of cases [17]. Initial abnormalities may include localized rarefaction of one vertebral plateau, followed by involvement of adjacent vertebrae. Marked destruction of bone, usually anteriorly, can occur, followed by anterior osteophytic reactions with bridging and bone sclerosis.

●Discitis – The radiographic changes of discitis are first noted several weeks after the onset of symptoms, with the following sequence:

•Narrowing of the disc space two to four weeks after the onset of symptoms

•Destruction of the adjacent cartilaginous vertebral end-plates

•Herniation of the disc into the vertebral body

In older children, anterior spontaneous fusion is common. Rarely, vertebral body compression or wedging is noted.

Despite the delay in radiographic changes, findings suggestive of discitis often are apparent at the time of presentation (because of the indolent course). In one series, as an example, radiographic changes were present on initial evaluation in more than 70 percent of cases [17].

Normal radiographs — A normal radiograph early in the course (eg, within seven days of symptom onset) does not exclude osteomyelitis. In systematic reviews, the sensitivity of radiographs at initial presentation ranges from 16 to 37 percent [1,6].

Children with suspected osteomyelitis (eg, localized bone pain, limited function, elevated ESR or CRP) and normal initial radiographs generally should receive empiric antibiotic therapy pending additional imaging and the results of microbiologic studies. However, school-age children who are afebrile and have equivocal physical examination findings, and normal radiographs may occasionally be followed closely without antimicrobial therapy pending additional evaluation. (See "Hematogenous osteomyelitis in children: Management", section on 'Empiric parenteral therapy' and 'Advanced imaging' below and 'Microbiology' below.)

Further imaging, usually with MRI or scintigraphy, should be performed as soon as possible in any child with suspected osteomyelitis whose initial radiographs are normal, regardless of whether antibiotic therapy is initiated (image 3). (See 'Advanced imaging' below.)

Advanced imaging — Most children with suspected osteomyelitis undergo additional imaging with MRI and/or possibly scintigraphy; computed tomography (CT), and/or ultrasonography may be useful in selected cases (table 2). Indications for these imaging studies include:

●Support of the diagnosis in children with normal initial radiographs

●Further evaluation of abnormalities identified on radiographs (eg, bone destruction)

●Evaluation of extension of infection (eg, growth plate, epiphysis, joint, adjacent soft tissues)

●Guidance for percutaneous diagnostic and therapeutic drainage procedures (eg, needle aspiration, abscess drainage)

Magnetic resonance imaging — If it is available, MRI is the imaging modality of choice when imaging other than radiography is needed to establish the diagnosis of osteomyelitis (eg, early in the course) or to delineate the location and extent of bone and soft tissue involvement (table 2) [1,3,8,9,18]. Intravenous contrast is not generally required but may help define intramedullary abscess or intramuscular abscesses [19]. Osteomyelitis is unlikely if MRI is negative.

MRI is particularly useful in identifying and/or distinguishing:

●Early changes in the bone marrow cavity (before changes in cortical bone are apparent on radiographs) [20]

●Pelvic osteomyelitis [21-23]

●Involvement of the vertebral body and the adjacent disc in children with vertebral osteomyelitis [24]

●Areas that may require surgical drainage (eg, intraosseous abscesses, subperiosteal or soft-tissue collections of pus (image 4); sinus tracts in cases of chronic osteomyelitis) (see "Hematogenous osteomyelitis in children: Management", section on 'Indications for surgery')

●Involvement of the growth plate (image 3) [25-27]

●Contiguous septic arthritis (especially in the evaluation of young children with possible septic arthritis of the hip or femoral osteomyelitis)

●Associated pyomyositis

●Evidence of venous thrombosis

The major advantages of MRI compared with other imaging modalities include accurate identification of subperiosteal or soft-tissue collections of pus and avoidance of exposure to ionizing radiation [28]. In addition, the efficacy of MRI does not appear to be affected by diagnostic or surgical intervention. However, repeat MRI may be of limited value after surgical drainage because it can be difficult to distinguish between postsurgical changes and recurrent or persisting infection. Repeat MRI seldom leads to management changes in patients with clinical improvement. In a retrospective review of 60 cases of acute osteomyelitis, only 11 of 104 repeat MRIs resulted in a change in treatment; in all 11 cases, the CRP was persistently elevated or rising [29].

Disadvantages of MRI include a longer scanning time than CT and the need for sedation or general anesthesia for an adequate study in most young children, which may be a limiting factor in some institutions. MRI is less useful when multiple sites of involvement are suspected or there are no localized clinical findings. Finally, MRI is not always readily available.

MRI demonstrates excellent anatomic detail and differentiation among soft tissue, bone marrow, and bone (image 5).

●Areas of active inflammation show a decreased signal in T1-weighted images and an increased signal in T2-weighted images [30]. Fat-suppression sequences, including short-tau inversion recovery (STIR), decrease the signal from fat and are more sensitive for the detection of bone-marrow edema.

●The penumbra sign (high-intensity-signal transition zone between abscess and sclerotic bone marrow in T1-weighted images (image 6)) is characteristic of Brodie abscesses and in one study was helpful in differentiating between indolent infections and neoplasms [31].

●The signal from infected bone marrow can be enhanced with intravenous gadolinium contrast [32], but this is seldom necessary for diagnostic purposes [19,33]. Gadolinium contrast clarifies the anatomy of complex cases (eg, chronic osteomyelitis with a sinus tract identified on examination and a sequestrum identified on plain radiographs). Because of the risk of nephrogenic systemic fibrosis, imaging with gadolinium should be avoided, if possible, in patients with moderate or advanced renal failure. (See "Nephrogenic systemic fibrosis/nephrogenic fibrosing dermopathy in advanced kidney disease".)

MRI abnormalities in discitis include reduction of disc space, increased T2 signal in the adjacent vertebral end plates, and signal intensity changes in the intervening discs [17,20].

The sensitivity and specificity for the detection of bone involvement in children with suspected osteomyelitis with MRI are high (81 to 100 percent and 67 to 94 percent, respectively, in a systematic review) [1]. Given the high sensitivity, osteomyelitis is unlikely if the MRI is negative [8], although rare patients lack MRI abnormalities at the time of presentation. False-positive results may occur in patients with adjacent soft tissue infection; adjacent soft tissue infection can cause edema of the bone that may be interpreted as "consistent with osteomyelitis" but represents sympathetic inflammation rather than bone infection. False-positive MRI results also may occur in children with vitamin C deficiency [34-37]. (See 'Other noninfectious conditions' below.)

Scintigraphy — We suggest scintigraphy (also known as radionuclide scanning or bone scan) when MRI is not available and imaging other than radiography is needed to establish the diagnosis of osteomyelitis (table 2). Scintigraphy also may be useful when the area of suspected infection cannot be localized or multiple areas of involvement are suspected.

Scintigraphy is helpful early in the course, readily available, relatively inexpensive, and requires sedation less frequently than MRI in young children. However, it does not provide information about the extent of purulent collections that may require drainage (eg, intramedullary abscess, muscular phlegmon) [8]. Scintigraphy is nonspecific; other causes of positive scintigraphy include fracture, bone tumor, cellulitis, and pyogenic arthritis [1]. Scintigraphy may be falsely negative if the blood supply to the periosteum is disrupted (eg, subperiosteal abscess) and during the transition between decreased and increased activity [38,39]. In addition, scintigraphy involves exposure to ionizing radiation [40].

The three-phase bone scan utilizing technetium 99m (99mTc) usually is performed as the initial nuclear medicine procedure in the evaluation of osteomyelitis. It consists of:

●A nuclear angiogram (the flow phase), obtained two to five seconds after injection

●The blood pool phase, obtained 5 to 10 minutes after injection

●The delayed phase, specific for bone uptake, obtained two to four hours after injection

Increased uptake of tracer in the first two phases can be caused by anything that increases blood flow and is accompanied by inflammation. Osteomyelitis causes focal uptake in the third phase; the intensity of the signal reflects the level of osteoblastic activity. Localization of a lesion near a growth plate can complicate interpretation.

The sensitivity and specificity for the detection of bone involvement in children with suspected osteomyelitis with scintigraphy generally are high, but the range is wider than MRI (53 to 91 percent and 47 to 84 percent, respectively in a systematic review) [1]. Sensitivity is variable in neonates but appears to be lower than in older children [11,41,42]. Scintigraphy has been shown to have poor sensitivity (53 percent) in cases of osteomyelitis caused by methicillin-resistant Staphylococcus aureus (MRSA) [8].

MRI generally is preferred if the results of the three-phase 99mTc bone scan are equivocal. Scintigraphy with inflammation imaging tracers (eg, gallium or indium) may be helpful but involve additional exposure to radiation.

Computed tomography — MRI usually is preferred to CT in the evaluation of suspected osteomyelitis. However, CT better delineates changes in bone and may be preferred when significant bone destruction is identified on radiographs [20]. CT findings of osteomyelitis include increased density of bone marrow, periosteal new bone formation, and periosteal purulence.

Other potential indications for CT may include:

●Delineation of the extent of bone injury in chronic osteomyelitis or planning the surgical approach to debridement of devitalized bone (sequestra) (image 7)

●Lack of availability of MRI or contraindications to MRI

CT is less time consuming than MRI, and young children usually do not require sedation. However, CT exposes children to ionizing radiation.

Ultrasonography — Ultrasonography usually is not helpful in the diagnosis of osteomyelitis. However, it can identify fluid collections associated with bone infections and may improve the success of percutaneous diagnostic and therapeutic drainage procedures [9,18,43].

Ultrasonography findings consistent with osteomyelitis include fluid collection adjacent to the bone without intervening soft tissue, elevation of the periosteum by more than 2 mm, and thickening of the periosteum. (See "Approach to imaging modalities in the setting of suspected nonvertebral osteomyelitis", section on 'Ultrasonography'.)

Microbiology

General principles — Isolation of a pathogen from bone, subperiosteal fluid collection, joint fluid, or blood or polymerase chain reaction (PCR) detection of pathogens in bone, subperiosteal fluid collection, or joint fluid establishes a diagnosis of confirmed or probable osteomyelitis in children with compatible clinical and/or radiologic findings and speciation and susceptibility testing are essential for planning treatment. (See "Hematogenous osteomyelitis in children: Management", section on 'Antimicrobial therapy'.)

With an increasing proportion of cases caused by antibiotic-resistant organisms (eg, community-associated MRSA) and fastidious pathogens (eg, Kingella kingae in young children), it is particularly important to collect specimens for culture from as many sites of infection as possible [1].

It is preferable to obtain microbiologic specimens before administration of antibiotics, but this must be weighed against the potential for additional complications of ongoing bacteremia by S. aureus or other pathogens. If bone culture cannot be obtained immediately, decisions regarding immediate or delayed administration of antibiotics generally are made in consultation with the orthopedic surgeon. For children with signs of systemic illness (eg, fever, tachycardia), we provide antibiotic therapy immediately following blood culture. In a retrospective reviews of children with osteomyelitis, the rates of bone culture positivity were similar (ranging from 70 to 84 percent) whether or not children received antibiotics before cultures were obtained [44-46].

Sites to culture — Specimens for culture should be obtained from blood and as many potential sites of infection as possible. An orthopedic surgeon and/or interventional radiologist should be consulted as early as possible in the evaluation of children with suspected osteomyelitis to assist in obtaining specimens for histopathology and/or culture and assessing the need for surgical intervention.

Culturing multiple sites increases the chance of identifying a pathogen and susceptibilities, enabling narrowly targeted therapy [44]. In a meta-analysis of 12 observational studies of children with confirmed osteomyelitis (published between 2005 and 2019), blood cultures were positive in 33 percent, whereas the combination of blood, bone, or tissue cultures were positive in 55 percent [1].

It is important to inform the microbiology laboratory if less common or fastidious organisms (eg, K. kingae) are suspected because they may require specific media, growth conditions, or prolonged culture time (table 3) [47].

●Blood cultures – We recommend at least one, and preferably two, blood cultures be obtained before administration of antibiotics to children with osteomyelitis. Although less frequently positive than cultures from bone or adjacent abscesses, blood cultures may be the only positive source of identification of the pathogen and may avoid bone aspiration that is pursued solely for microbiologic studies [1,6,44,47,48].

In a meta-analysis of 17 observational studies including 1422 children with confirmed osteomyelitis, the pooled positivity rate was 31 percent (95% CI 26-36) [1]. In a separate series limited to osteoarticular infections with S. aureus, blood cultures were positive in 54 percent.

●Bone culture – Bone samples for aerobic culture, Gram stain, and histopathology should be obtained whenever possible [44,48]. Cultures for rare pathogens (eg, anaerobic bacterial, fungus mycobacteria) are reserved for children with clinical suspicion or epidemiologic risk factors (eg, penetrating inoculation, immunocompromise (table 3)) [1,49]. Culture specimens may be obtained by percutaneous needle aspiration (which may be guided by ultrasonography, fluoroscopy, or other imaging modality [50]) or open biopsy (particularly if surgery is required for therapeutic drainage and debridement). (See "Hematogenous osteomyelitis in children: Management", section on 'Indications for surgery'.)

In a meta-analysis of 17 observational studies including 962 children with confirmed osteomyelitis, the average positivity rate of bone and tissue cultures was 65 percent (95% CI 56-75) [1].

●Other cultures – Subperiosteal exudate, joint fluid, and pus from adjoining sites of infection should be obtained and sent for Gram stain and culture whenever possible, as directed by imaging studies [44]. Specimens may be obtained by percutaneous needle aspiration (which may be guided by ultrasonography, fluoroscopy, or other imaging modality). Injection of bone aspirates or periosteal collections into blood culture bottles is recommended to enhance recovery of K. kingae [51,52]. (See "Hematogenous osteomyelitis in children: Epidemiology, pathogenesis, and microbiology", section on 'Kingella kingae'.)

Percutaneous needle aspiration is often successful in neonates with extensive soft-tissue and periosteal involvement and infants and young children with subperiosteal collections.

Percutaneous techniques are less successful in older children and adolescents, who may require bone aspiration or drilling to obtain culture specimens. In such cases, the risk of epiphyseal damage and further destruction of bone must be weighed against the benefit of obtaining a specimen.

Other microbiologic studies — Other microbiologic studies may be helpful in the detection of specific pathogens. PCR (if available) may identify a pathogen in children with negative cultures [1,53]. It can be particularly helpful for detecting K. kingae in purulent materials or bone specimens, particularly when Kingella-specific methods are used [54-56]. PCR for Streptococcus pyogenes or S. aureus may be positive when cultures are negative, especially in the pretreated patient.

Histopathology — The diagnosis of osteomyelitis is confirmed in children who have histopathologic evidence of inflammation in a surgical specimen of bone (picture 1A-C). An orthopedic surgeon and/or interventional radiologist should be consulted as early as possible in the evaluation of children with suspected osteomyelitis to assist in obtaining specimens for histopathology and/or culture and assessing the need for surgical intervention.

Histopathologic specimens may be obtained by percutaneous needle aspiration (which may be guided by ultrasonography, fluoroscopy, or other imaging modality) or open biopsy (particularly if surgery is required for therapeutic drainage and debridement). (See "Hematogenous osteomyelitis in children: Management", section on 'Indications for surgery'.)

Response to empiric therapy — Improvement in constitutional symptoms and localized inflammation (eg, erythema, point tenderness) with empiric antimicrobial therapy helps to support the diagnosis of osteomyelitis, particularly in children in whom no pathogen is isolated. (See "Hematogenous osteomyelitis in children: Management", section on 'Response to therapy' and 'Probable osteomyelitis' below.)

In children with normal MRI and/or scintigraphy, a lack of clinical response to empiric therapy and the results of blood cultures help to direct additional evaluation for other conditions in the differential diagnosis. (See 'Osteomyelitis unlikely (advanced imaging studies normal)' below and 'Differential diagnosis' below.)

DIAGNOSTIC INTERPRETATION

Supportive findings — The diagnosis of osteomyelitis is supported by a combination of [57]:

●Clinical features suggestive of bone infection (constitutional symptoms, focal symptoms and signs of bone inflammation including limitation of function) associated with elevated C-reactive protein (and/or erythrocyte sedimentation rate) (see 'Clinical suspicion' above and 'Blood tests' above)

●An imaging study with abnormalities characteristic of osteomyelitis (table 1) (see 'Radiographs' above and 'Advanced imaging' above)

●A positive microbiologic or histopathologic specimen (see 'Microbiology' above and 'Histopathology' above)

●Clinical improvement with empiric antimicrobial therapy (see 'Response to empiric therapy' above)

Categorization

Confirmed osteomyelitis — The diagnosis of osteomyelitis is confirmed by histopathologic evidence of inflammation in a surgical specimen of bone (picture 1A-C) (if obtained) or identification of a pathogen by culture or Gram stain in an aspirate or biopsy of bone or periosteal fluid collection [58].

Probable osteomyelitis — The diagnosis of osteomyelitis is probable in a child with compatible clinical, laboratory, and/or radiologic findings (table 1) and:

●Isolation of a pathogen from blood or joint fluid, or

●If cultures are negative, polymerase chain reaction (PCR) detection of S. aureus, S. pyogenes, Streptococcus pneumoniae, or K. kingae in bone aspirates, subperiosteal collections, or synovial fluid, or

●If cultures and PCR are negative, expected response to empiric antimicrobial therapy

Given the potential morbidity of delayed treatment, children with probable osteomyelitis should be managed in the same manner as children in whom infection has been confirmed by isolation of an organism from bone or blood [59]. (See "Hematogenous osteomyelitis in children: Clinical features and complications", section on 'Complications' and "Hematogenous osteomyelitis in children: Management", section on 'Empiric parenteral therapy'.)

The response to empiric antimicrobial therapy in children with culture-negative osteomyelitis directs the need for additional evaluation. (See "Hematogenous osteomyelitis in children: Management", section on 'Culture-negative osteomyelitis'.)

Osteomyelitis unlikely (advanced imaging studies normal) — Osteomyelitis is unlikely if advanced imaging studies (usually MRI or scintigraphy) are normal throughout the evaluation. The major considerations in the differential diagnosis and approach to continued evaluation and management depend upon the results of the blood culture and response to antimicrobial therapy:

●Positive blood culture, improvement with empiric therapy – A source of infection other than osteomyelitis must be sought (see 'Other infections' below)

●Positive blood culture, no improvement with appropriate therapy – Other sources of bacteremia and noninfectious conditions that may have predisposed the patient to bacteremia must be sought aggressively (see 'Differential diagnosis' below)

●Negative blood culture, improvement with empiric therapy – The child may have a more superficial source of infection (eg, cellulitis); a shorter course of antimicrobial therapy may be warranted (see 'Other infections' below)

●Negative blood culture, no improvement with appropriate therapy – A bacterial infection is unlikely; fungal and noninfectious causes of musculoskeletal pain should be sought; discontinuation of empiric antimicrobial therapy may be warranted (see 'Noninfectious conditions' below)

DIFFERENTIAL DIAGNOSIS — The differential diagnosis of osteomyelitis includes infectious conditions, noninfectious conditions, and radiographic mimics (table 4).

Other infections — Infections that do not involve bone may cause fever, pain, and tenderness overlying bone, mimicking hematogenous osteomyelitis. These infections usually are distinguished from osteomyelitis by imaging studies that lack the characteristic features of osteomyelitis (table 1). (See 'Advanced imaging' above.)

Other infections that may share features of osteomyelitis (and may complicate or be complicated by osteomyelitis) include:

●Septicemia. (See "Clinical features, evaluation, and diagnosis of sepsis in term and late preterm neonates", section on 'Clinical manifestations' and "Sepsis in children: Definitions, epidemiology, clinical manifestations, and diagnosis", section on 'Clinical manifestations'.)

●Cellulitis. (See "Cellulitis and skin abscess: Epidemiology, microbiology, clinical manifestations, and diagnosis", section on 'Cellulitis and erysipelas'.)

●Septic arthritis – Approximately one-third of cases of osteomyelitis extend to the contiguous joint (as many as 75 percent in neonates) [60-63]. (See "Bacterial arthritis: Clinical features and diagnosis in infants and children", section on 'Clinical features'.)

●Deep abscesses (eg, psoas abscess or obturator internus abscess [64]). (See "Psoas abscess", section on 'Clinical manifestations'.)

●Pyomyositis. (See "Primary pyomyositis", section on 'Clinical manifestations'.)

●Garré sclerosing osteomyelitis – Garré sclerosing osteomyelitis is characterized by rigid bony swelling at the periphery of the mandible and nonsuppurative sclerotic bone inflammation [65]. It has been reported at all ages [61]. Garré sclerosing osteomyelitis is thought to be triggered by odontogenic infection [66-68], but noninfectious causes of inflammation also may play a role.

●Congenital syphilis – Long bone abnormalities related to congenital syphilis may be associated with pain and limited movement of the involved extremity. (See "Congenital syphilis: Clinical manifestations, evaluation, and diagnosis", section on 'Clinical manifestations'.)

Noninfectious conditions

Chronic nonbacterial osteomyelitis — Chronic nonbacterial osteomyelitis (CNO; also called chronic recurrent multifocal osteomyelitis [CRMO]) is a chronic inflammatory bone disorder that primarily affects children. It is characterized by bone pain with insidious onset. The initial presentation is similar to that of osteomyelitis. Imaging may localize the areas of bony involvement and indicate the absence of features suggestive of chronic osteomyelitis, such as an abscess or sinus tract. Microbiologic and pathologic studies from bone biopsy specimens are generally necessary to differentiate CNO from infectious osteomyelitis. The evaluation and treatment of CNO are discussed in detail separately. (See "Chronic nonbacterial osteomyelitis (CNO)/chronic recurrent multifocal osteomyelitis (CRMO)".)

Other noninfectious conditions — Several noninfectious conditions have clinical features that overlap with osteomyelitis. These include:

●Malignancy – Tumor growth can cause bone pain, and some children with malignancies (particularly leukemia and Ewing sarcoma) have fever as part of their initial presentation. Unlike those with osteomyelitis, symptoms can be intermittent in children with cancer. Affected children also fail to respond to empiric antibiotic therapy. Osteomyelitis and cancer involving bone are usually differentiated with bone biopsy. (See "Overview of common presenting signs and symptoms of childhood cancer", section on 'Bone and joint pain' and "Overview of the clinical presentation and diagnosis of acute lymphoblastic leukemia/lymphoma in children", section on 'Presentation' and "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Signs and symptoms' and "Bone tumors: Diagnosis and biopsy techniques".)

●Bone infarction – Bone infarction secondary to hemoglobinopathy, especially in infants with dactylitis, can be particularly difficult to distinguish from osteomyelitis.

Radiographs, scintigraphy, and MRI all show similar results in both conditions. Unlike bone disease with hemoglobinopathy, osteomyelitis does not respond to hydration and other supportive measures. (See "Acute and chronic bone complications of sickle cell disease".)

●Vitamin C deficiency – Vitamin C deficiency (scurvy) may cause musculoskeletal pain and refusal to bear weight, particularly in children with autism spectrum disorder (the prevalence of which has increased since the early 2000s), intellectual disability, food aversion, or limited food preferences [34,35,37]. Additional findings of vitamin C deficiency, which do not always accompany the musculoskeletal findings, include petechiae, ecchymoses, bleeding gums, coiled hairs, and hyperkeratosis (picture 2). (See "Overview of water-soluble vitamins", section on 'Deficiency' and "Autism spectrum disorder (ASD) in children and adolescents: Terminology, epidemiology, and pathogenesis", section on 'Prevalence'.)

●Gaucher disease – Children with Gaucher disease can have painful bone crises similar to those that occur in patients with sickle cell disease. During bone pain crises, ischemia can be detected by technetium bone scan. Radiographs of the distal femur may demonstrate deformities caused by abnormal modeling of the metaphysis that are characteristic of Gaucher disease. However, the possibility of osteomyelitis should be considered in febrile patients with Gaucher disease who do not improve with hydration and other supportive measures. (See "Gaucher disease: Pathogenesis, clinical manifestations, and diagnosis" and "Gaucher disease: Treatment", section on 'Skeletal disease'.)

●Complex regional pain syndrome – Complex regional pain syndrome (CRPS) is an uncommon chronic condition that causes pain, swelling, and limited range of motion in the affected extremity. It frequently begins after an injury, surgery, or vascular event. Vasomotor instability and chronic skin changes also occur. CRPS may be distinguished from osteomyelitis by autonomic dysfunction and normal erythrocyte sedimentation rate /C-reactive protein. (See "Complex regional pain syndrome in children", section on 'Diagnosis'.)

●Neuropathic arthropathy (Charcot foot) – Neuropathic arthropathy is a progressive degeneration of the weightbearing joints of the foot that occurs in people with peripheral neuropathies (eg, related to spinal cord abnormality or injury) [69,70]. Clinical features include localized swelling, erythema, and warmth. Biopsy may be necessary to differentiate neuropathic arthropathy from osteomyelitis, particularly in the early/inflammatory stage when there are few or no radiologic abnormalities [71]. (See "Surgical management of neuropathic arthropathy (Charcot foot)", section on 'Anatomy and classification'.)

●Caffey disease – Caffey disease (infantile cortical hyperostosis) is an inherited disease that usually presents in early infancy with fever, subperiosteal bone hyperplasia, and swelling of overlying soft tissues. It is a rare disorder caused by a subset of mutations in the type 1 collagen gene COL191 [72] and is difficult to distinguish from osteomyelitis on initial presentation. Caffey disease can be distinguished from infectious osteomyelitis by bone biopsy and genetic testing. (See "Differential diagnosis of the orthopedic manifestations of child abuse", section on 'Infantile cortical hyperostosis (Caffey disease)'.)

Radiographic mimics — Benign and malignant bone tumors and tumors involving bone can have radiographic appearance similar to osteomyelitis. The acute clinical features and response to antibiotics in children with osteomyelitis usually distinguish osteomyelitis from these conditions. However, bone biopsy can be performed if necessary for histopathologic differentiation.

Bone lesions that can have radiographic appearance similar to osteomyelitis include (table 4):

●Fibrous dysplasia (see "Nonmalignant bone lesions in children and adolescents", section on 'Fibrous dysplasia')

●Osteoid osteoma and osteoblastoma (see "Nonmalignant bone lesions in children and adolescents", section on 'Bone-forming lesions')

●Chondroblastoma and chondromyxoid fibroma (see "Nonmalignant bone lesions in children and adolescents", section on 'Cartilage-forming tumors')

●Eosinophilic granuloma and other forms of histiocytosis (see "Clinical manifestations, pathologic features, and diagnosis of Langerhans cell histiocytosis", section on 'Lytic bone lesions')

●Osteosarcoma (see "Osteosarcoma: Epidemiology, pathology, clinical presentation, and diagnosis", section on 'Clinical presentation')

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Septic arthritis and osteomyelitis in children".)

SUMMARY AND RECOMMENDATIONS

●Diagnostic approach – The diagnosis of osteomyelitis often is uncertain at the initial evaluation. A high index of suspicion and monitoring of the clinical course are essential to establishing the diagnosis. We consult an orthopedic surgeon and/or interventional radiologist as early as possible in the evaluation to assist in obtaining specimens for culture and/or histopathology. (See 'Diagnostic approach' above.)

●Clinical suspicion – Findings suggestive of bone infection that raise suspicion for acute hematogenous osteomyelitis include (see 'Clinical suspicion' above):

•Constitutional symptoms (irritability, decreased appetite or activity), with or without fever

•Focal findings of bone inflammation (eg, warmth, swelling, point tenderness) that typically progress over several days to a week

•Limitation of function (eg, limited use of an extremity; limp; refusal to walk, crawl, sit, or bear weight)

•Detection of bacteremia or characteristic imaging findings during evaluation for other conditions (eg, fever, trauma)

●Initial evaluation – The initial evaluation of children with suspected osteomyelitis includes (see 'Initial evaluation' above):

•Complete blood count with differential

•C-reactive protein (CRP); erythrocyte sedimentation rate may be useful if CRP is normal

•Blood culture

•Radiograph(s) of the affected region

●Advanced imaging – If radiographs do not demonstrate characteristic findings (table 1), advanced imaging is necessary. If available, MRI is generally preferred to other advanced imaging modalities (scintigraphy, CT, ultrasonography) (table 2). Scintigraphy may be better for multifocal or poorly localized disease. (See 'Advanced imaging' above.)

●Microbiologic studies – Obtain specimens for Gram stain and culture from as many sites of infection as possible. Isolation of a pathogen from bone, periosteal collection, joint fluid, or blood is necessary for diagnosis. Speciation and susceptibility testing are essential for treatment planning. (See 'Microbiology' above.)

●Diagnostic interpretation – The diagnosis of osteomyelitis is supported by a combination of clinical features suggestive of bone infection, an imaging study with abnormalities characteristic of osteomyelitis (table 1), a positive microbiologic or histopathologic specimen, and/or a clinical improvement with empiric antimicrobial therapy. (See 'Supportive findings' above.)

The clinical features or combination of features determine the categorization of the diagnosis:

•Confirmed (see 'Confirmed osteomyelitis' above)

-Histopathologic evidence of inflammation in a surgical specimen of bone (picture 1A-C) (if obtained), or

-identification of a pathogen by culture or Gram stain in an aspirate or biopsy of bone or a periosteal fluid collection

•Probable (see 'Probable osteomyelitis' above):

Compatible clinical, laboratory, and/or radiologic findings (table 1) and:

-isolation of a pathogen from blood or joint fluid, or

-If cultures are negative, polymerase chain reaction (PCR) detection of Staphylococcus aureus, Streptococcus pyogenes, Streptococcus pneumoniae, or Kingella kingae in bone aspirates, subperiosteal collections, or synovial fluid, or

-If cultures and PCR are negative, expected response to empiric antimicrobial therapy

•Unlikely – Normal advanced imaging studies (usually MRI or scintigraphy) throughout the evaluation. (See 'Osteomyelitis unlikely (advanced imaging studies normal)' above.)

●Differential diagnosis – The differential diagnosis of osteomyelitis includes other infections, noninfectious conditions, and radiographic mimics (table 4). (See 'Differential diagnosis' above.)