Septic Arthritis Imaging

Updated: Sep 19, 2019
Author: Lourdes Nunez-Atahualpa, MD; Chief Editor: Felix S Chew, MD, MBA, MEd 

Practice Essentials

Septic arthritis (SA) in the United States has an annual incidence of 10 per 100,000.[1]  It is more prevalent in patients who are elderly (80 years or older), have prosthetic joints, have undergone joint surgery, have immunocompromised states such as HIV, have skin infections, or have autoimmune diseases.[2, 3]   Septic arthritis is a medical emergency associated with high rates of morbidity and mortality, especially when the diagnosis is delayed or the treatment is suboptimal. The large joints of the hip and the knee are most commonly affected (in more than half of the cases), but any joint can be involved. The sternoclavicular joint (SCJ) is another common site of septic arthritis, involved in 0.5-1.0% of all joint infections.[4]  

In general, infectious arthritis is classified as pyogenic (septic) or nonpyogenic. Pyogenic septic arthritis is caused by Staphylococcus aureus in up to 80% of cases,[5, 6]  followed by other pathogens such as staphylococci, streptococci, Gonococcus species, Escherichia coli, Haemophilus, Klebsiella, Pseudomonas, and Candida. Infection can lead to rapid and severe joint destruction.

Methicillin-resistant Staphylococcus aureus (MRSA) has become a major cause of septic arthritis in the United States. MRSA joint infection seems to be associated with worse outcomes. This may be due to its greater virulence, delays in initiation of appropriate antibiotics, and the older age of many affected patients.[7]

Nonpyogenic infective arthritis tends to be less aggressive and have a more chronic course, with causative organisms including Mycobacterium tuberculosis, fungi, and spirochetes.[8, 2, 3]  

Septic arthritis can be acquired through several routes of transmission. The most common cause is acquisition through hematogenous spread from a distant source. Direct seeding can occur through trauma, surgery, or spread from a contiguous infection such as osteomyelitis or cellulitis.

(Septic arthritis images are provided below.)

A 30-year-old man who was taking steroids presente A 30-year-old man who was taking steroids presented with a joint effusion and knee pain. Anteroposterior view of the knee demonstrates patchy demineralization of the tibia and femur and joint-space narrowing. This was caused by tuberculoid infection of the joint.
Coronal short-tau inversion recovery MRI of the pu Coronal short-tau inversion recovery MRI of the pubic symphysis demonstrates a hyperintense joint effusion and increased signal intensity in the bone marrow of the pubic rami. Abnormal high signal intensity is also present in the bilateral hip adductor muscles. The diagnosis was septic arthritis with associated osteomyelitis and inflammatory changes in the soft tissues.
Septic arthritis. Anteroposterior view of the shou Septic arthritis. Anteroposterior view of the shoulder demonstrates subchondral erosions and sclerosis in the humeral head. These are relatively late findings of septic arthritis. Periosteal reaction due to coincident osteomyelitis is present adjacent to the surgical neck of the humerus.
Coronal T2-weighted fat-saturated MRI of the shoul Coronal T2-weighted fat-saturated MRI of the shoulder demonstrates a joint effusion, bone marrow edema, and marked adjacent soft tissue inflammation with a fluid collection in the infraspinatus muscle. This is an example of septic arthritis with associated soft tissue abscess.

Preferred examination

Septic arthritis is an emergency condition that needs to be diagnosed early by clinical examination and by radiologic imaging and laboratory workup. Early diagnosis can prevent complications that could ultimately lead to patient disability. The imaging workup for suspected septic arthritis includes radiography or ultrasound before surgical intervention to detect the effusion and to see whether an associated osteomyelitis is present.[9, 10, 11, 12, 13, 14, 15]

Ultrasound may provide assistance in determining the presence of intra-articular effusion and locating the site of optimal aspiration. However, if further imaging is required, MRI is the most sensitive and specific technique, although it is of low utility for acute diagnosis.[16]  Scintigraphy, CT, and FDG-PET are also used, although to a lesser extent.[8, 17, 18, 9, 19, 20, 21, 22]

Imaging is not the primary means of diagnosing septic arthritis. Clinical evaluation and joint fluid aspiration are the key to the diagnosis.[5] Samples should be obtained in all suspected cases of septic arthritis. This sampling can usually be achieved with fine-needle aspiration performed either blindly or by being guided though imaging methods, such as fluoroscopy, ultrasound, MRI, or CT, depending on the location. Surgical exploration may be necessary.

Fluid should be sent for Gram staining, culturing, glucose testing, leukocyte count, and differential determination. White blood cell counts are usually 50,000-60,000/µL, with more than 80% neutrophils. Synovial fluid glucose levels are decreased. Gram stain results are positive in 75% of patients with gram-positive cocci. Gram staining is less sensitive in cases of gonococcal infection. Only 25% of cultures of gonococcal synovial fluid are positive. In 9% of cases, blood cultures are the only source of pathogen identification and should be obtained before antibiotic treatment.[23]

Conventional radiographic findings of septic arthritis may include soft tissue swelling and edema, capsular distention, or displacement of the articular structures, which may suggest the presence of joint effusion, but it has low sensitivity and specificity for acute infection. Later stages of septic arthritis may reveal chronic bony changes and calcium deposits.[9, 10]

Ultrasound can help detect hip joint effusion by detecting elevation of the joint capsule with anechoic or complex fluid effusion, as well as the presence of synovial thickening and hyperemia. A difference of 2 mm between the 2 hip joints is considered an abnormal finding.[10]

The imaging workup of either septic arthritis or osteomyelitis should include MRI. In cases of overlap in clinical presentation between septic arthritis and osteomyelitis, MRI provides better evaluation of the location and extent of disease.[10]

American College of Radiology Appropriateness Criteria provide specific approaches for septic arthritis in children, particularly children who present with a limp and monoarticular pain, with description of the roles of radiography, ultrasound, nuclear medicine, CT, and MRI. According to the ACR guidelines, MRI, because of its sensitivity to soft tissue and bone marrow pathology, has high accuracy in diagnosing infection, including septic arthritis, osteomyelitis, pyomyositis, and discitis, and could be considered as the initial imaging study.[12, 15, 24]



The earliest plain film radiographic findings of septic arthritis are soft tissue swelling around the joint and a widened joint space from joint effusion; however, uniform narrowing of the joint has also been described.[25]

With progression of disease, plain films reveal joint-space narrowing as articular cartilage is destroyed, loss of continuity of the white cortical line as bone destruction begins, and development of marginal erosions when the bone is further destroyed. Bone debris within the joint space may also be seen in chronic untreated disease, which in adults may be mistaken as a neuropathic hip.[25]

Findings of superimposed osteomyelitis may also be identified, such as periosteal reaction, bone destruction, and sequestrum formation.

(See the images below.)

During the progression of infectious arthritis of During the progression of infectious arthritis of the hip, this image was obtained early in the disease and shows only concentric joint-space loss.
During the progression of infectious arthritis of During the progression of infectious arthritis of the hip, subchondral erosions and sclerosis of the femoral head are present.
During the progression of infectious arthritis of During the progression of infectious arthritis of the hip, subchondral erosions and sclerosis of the femoral head are present.

Mycobacterial infection presents as unilateral, focal, or diffusely decreased bone density, with eventual loss of the joint space that can evolve to ankylosis. Further studies, such as ultrasonography, can identify small to moderate joint effusion, accompanied by synovial thickening and enhancement on MR imaging.[26]

A triad of radiographic abnormalities known as the Phemister triad is characteristic of tuberculous arthritis: peripherally located bony erosions, juxta-articular osteoporosis, and gradual narrowing of the joint space.[6]

Gas within the joint or adjacent soft tissues can sometimes be seen in infection secondary to gas-forming organisms, such as E coli or Clostridium perfringens. However, gas within the joint is usually secondary to prior aspiration or the vacuum phenomenon, which can occur secondary to limb traction during positioning while under examination or a recent dislocation.

Plain radiographic findings in the infant hip include obliteration of soft tissue planes, swelling, displacement of the fat pads, and juxta-articular osteoporosis. Hefke and Turner also described the obturator sign, which consists of swelling of the obturator internus in septic arthritis of the hip.[27]  Subluxation or dislocation of the femoral head secondary to intra-articular fluid can occur. However, this can be difficult to identify if the femoral head is not ossified.

In children, lateral displacement of the femoral epiphysis relative to the contralateral hip signifies a joint effusion.[28] As little as 2 mm of asymmetry in the distance measured from teardrop of the acetabulum to the medial metaphysis of the femoral neck is considered pathologic.

Inadequately treated SA can show osteomyelitis, osteoarthritis, ankylosis, periarticular calcifications, subchondral bone loss, and sclerosis.

Fluoroscopic imaging can help in the guidance of aspirations and to identify sinus tracts.[29]

Degree of confidence

Initial plain radiographic findings are frequently normal. In a prospective study that evaluated 123 patients with suspected hip joint effusion, plain films had only a sensitivity of 27.8% as compared to a 100% sensitivity from ultrasonography.[30]

The characteristic findings are somewhat nonspecific by themselves, but they can be nearly diagnostic when correlated clinically.

Poorly defined bony erosions are characteristic of septic arthritis and a helpful feature in differentiating septic arthritis from other diseases in the differential diagnosis. Osseous erosions in gout, rheumatoid arthritis, seronegative spondyloarthropathies, pigmented villonodular synovitis, hemophilia, and synovial osteochondromatosis tend to be sharply marginated.



Ultrasonography (US) is a low-cost, widely available, noninvasive technique that allows side-to-side characterization of joints, differentiation between intra- and extra-articular disease,[31]  and guided relief of tension due to effusion in a painful joint.[32]  The use of ultrasound for diagnosis of joint effusions has been described since 1987. It is an extremely sensitive method[33]  that can detect as little as 1-2 ml of fluid in a joint. Capsular distention in the hip, noted as convexity of the anterior recess when compared to the contralateral, can be easily identified.[32]

Effusions in septic arthritis tend to be hypoechoic, rather than echo-free,[30, 31]  and it has been suggested that an ultrasound that fails to show a fluid collection may virtually exclude the diagnosis.[30]

In the hip joint, the presence of joint asymmetry, the presence of fluid, and increased thickness of the articular capsule have been described as diagnostic criteria.[30]

Synovial hypertrophy has been observed in patients with tuberculosis, brucellosis, lyme disease, and fungal infections.[31]

US has been referred as the “orthopedic stethoscope“[34]  and is considered the modality of choice in children because of the ease of image acquisition,[32]  obviation of sedation, and radiation-free safety profile. In pediatric patients, it can also suggest the need for further studies, such as an MRI, or the need for arthrotomy, irrigation, and/or debridement in patients with an irritable hip.[35]

It has also been suggested that ultrasound can help eliminate the need for diagnostic joint aspiration in suspected septic arthritis while detecting the extent of the pathology.[33]

Besides B-mode ultrasonography, studies have been performed with other modes, such as a power Doppler, for which it has been suggested that an increased vascularity of the affected joint could exclude the diagnosis with certainty, but that has not been successful.[36]

US could potentially be used to estimate the size of joint effusions and abscesses and allow monitoring of treatment, although it has been described by MRI.[37]

Degree of confidence

In a retrospective study that evaluated 158 patients for joint effusion of the knee, US showed an 81.3% sensitivity and a 100% specificity, with a positive predictive value of 100% and a negative predictive value of 77.5% when compared to MRI.[38]  In the hip, US showed a specificity of 80%, sensitivity at 96.3%, positive predictive value of 94%, and negative predictive value of 85.7% when compared to surgical results and further follow-up. This prospective study further suggested that US is superior to clinical, laboratory, and plain radiography parameters for the detection of septic arthritis in patients with a mean age of 45 days.[39]

Known caveats of this imaging method is its operator dependability. The effusion characteristics can be difficult to differentiate from chronic inflammatory conditions such as rheumatoid arthritis.[31]


Computed Tomography

Although not commonly used for the evaluation of joint infections, CT scanning can be accurate in the evaluation of septic arthritis.[40, 41]  It is very helpful when evaluating the sternoclavicular joint and the sacroiliac joint, which may be difficult to evaluate using plain films. CT scanning can depict early findings in septic arthritis, such as synovial thickening or joint effusion. Periarticular abscesses or fluid collections can also be identified.[40]

CT findings are otherwise similar to those on plain radiographs and include joint effusion, joint-space narrowing, bone and cartilage erosions, gas within the joint, and soft tissue swelling.

Moreover, real-time CT scanning can be used to guide joint aspirations in uncommon or difficult sites such as the sacroiliac or sternoclavicular joints.

Some authors have reported that the sensitivity of CT is similar to that of MRI. However, joint aspiration should be performed in all cases of suspected septic arthritis, and it is the primary means of diagnosis.

Although associated findings of osteomyelitis or soft tissue abscess increase the specificity of CT, distinguishing septic arthritis from other diseases in the differential diagnosis may be difficult.


Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) has been increasingly used to evaluate musculoskeletal infections, including septic joints. MRI is a sensitive and relatively specific imaging modality.[42] A combination of T1-weighted, T2-weighted, short-tau inversion recovery, and postcontrast T1-weighted fat-suppressed series are most helpful.[6]

(See the images below.)

Coronal short-tau inversion recovery MRI of the pu Coronal short-tau inversion recovery MRI of the pubic symphysis demonstrates a hyperintense joint effusion and increased signal intensity in the bone marrow of the pubic rami. Abnormal high signal intensity is also present in the bilateral hip adductor muscles. The diagnosis was septic arthritis with associated osteomyelitis and inflammatory changes in the soft tissues.
Coronal T2-weighted fat-saturated MRI of the shoul Coronal T2-weighted fat-saturated MRI of the shoulder demonstrates a joint effusion, bone marrow edema, and marked adjacent soft tissue inflammation with a fluid collection in the infraspinatus muscle. This is an example of septic arthritis with associated soft tissue abscess.

Synovial enhancement and the presence of a joint effusion have been reported to have the highest correlation with the clinical diagnosis of a septic joint. Bone erosions, bone marrow edema, cartilage loss, and erosion enhancement have also been identified, particularly in the setting of bacterial infections.[6]

Patients with nonpyogenic arthritides, such as lyme disease, may present with massive recurrent joint effusions. In the knee joint, an association with uniform cartilage loss, enthesitis, popliteal fossa lymphadenopathy, and popliteal muscle myositis has been described.[6]  Patients with tuberculous arthritis may have more bone erosions and less marrow-signal abnormality on MRI than patients with pyogenic arthritis. Use of intravenous gadolinium contrast is also very helpful in patients with a suspected septic joint to distinguish a periarticular abscess from surrounding myositis and to evaluate the degree of synovial inflammation.[2]

It has been suggested that MRI should be used in the first 12 hours to diagnose concurrent infections in newborns and in adolescents, as well as involvement of the shoulder joint, symptomatic arthritis of more than 6 days, and infection by methicillin-susceptible Staphylococcus aureus (MSSA) or methicillin-resistant S aureus (MRSA).[11]

When dynamic contrast-enhanced MRI is used, the signal intensity at 3.5 minutes (2.7-4.3 min) could be considered as a fair way to help differentiate SA from transient synovitis in the hip joint.[43]

Decreased intensity is expected when compared to the contralateral joint.[43]

In vivo MR studies marking macrophages with nanoparticles, such as ultrasmall supermagnetic iron oxide, and with later follow-up are being performed to detect early synovial inflammation and to monitor treatment.[44, 45]

In a study by Kang et al of 65 children who underwent MRI for acute hip pain and were suspected of having a septic condition (ie, presenting with fever or increased inflammatory markers), 28 (43.1%) were diagnosed with septic arthritis. Whereas a high-grade joint effusion was a significant MRI finding differentiating septic arthritis from transient synovitis for the entire study group, the presence of contralateral joint effusion in the short-term group (2 days or less of symptom duration) and the absence of a change or enhancement of the signal intensity of soft tissue in the long-term subgroup (3 days or longer of symptom duration) were significant predictors of transient synovitis.[46]

Degree of confidence

A clinically consistent history and the extra-articular findings of bone marrow edema or soft tissue inflammation are important in increasing the specificity of MRI for septic arthritis. The sensitivity and specificity of MRI is increased in the setting of associated osteomyelitis. MRI is as sensitive as technetium-99m methylene diphosphonate (MDP) scintigraphy in detecting osteomyelitis, and it is more sensitive than other scintigraphic techniques. It is excellent for the evaluation of soft tissues because of its high spatial resolution and multiplanar capability[47]

Infected and noninfected joint effusions have the same signal-intensity characteristics and cannot be distinguished by using MRI.

On T2-weighted images, high signal intensity in the adjacent bone marrow helps differentiate septic arthritis from synovitis. However, increased signal intensity does not necessarily indicate osteomyelitis. It can be secondary to hyperemia due to nearby infection or other etiologies. Thus, MRI is sensitive but is not always specific.[19]

Bone marrow edema and bone erosions caused by the infection of the joint can persist after successful treatment of septic arthritis [37] . 

Gadolinium-based contrast agents have been linked to the development of nephrogenic systemic fibrosis (NSF) or nephrogenic fibrosing dermopathy (NFD). For more information, see Nephrogenic Fibrosing Dermopathy. The disease has occurred in patients with moderate to end-stage renal disease after being given a gadolinium-based contrast agent to enhance MRI or MRA scans.


Nuclear Imaging

Early-phase (blood flow) and later (blood pool) images show increased activity at the joint and on both sides of the affected area. Delayed images obtained at 4-6 hours should demonstrate continued increased activity in the bone with associated osteomyelitis.

Decreased uptake in the femoral head can be seen with decreased perfusion related to high intra-articular pressures from the joint effusion or occlusion of blood vessels by bacteria.[48]

Scintigraphy with 99mTc-MDP is extremely sensitive, though not specific, for septic arthritis. Nevertheless, early stages of the disease may yield normal findings.[49]  Three-phase bone scanning has a reported sensitivity of 90-100% and a specificity of 73-79% for osteomyelitis. These values are likely decreased in septic arthritis without associated osteomyelitis.

Newer agents such as indium (In)-labeled autologous white blood cells, leukocytes labeled with 99mTc-hexamethyl propylamine oxime (HMPAO), 99mTc-labeled antigranulocyte monoclonal antibodies, and gallium-67 citrate have also been used to evaluate septic arthritis and osteomyelitis with increased specificity. However, 99mTc-MDP remains the mainstay of scintigraphic imaging.[20, 21]

A positively labeled white cell scan is specific. A positive 3-phase bone scan is specific if no other factors that could cause increased bone turnover are present. Thus, bone scanning is most useful if radiographic results are normal. If other factors (eg, trauma, arthritis) that could cause a false-positive bone scan are present, results should be confirmed with another study, such as white blood cell scanning.

Any process that results in increased bone turnover (eg, trauma, nonseptic arthritis) can result in a false-positive finding. Reflex sympathetic dystrophy could cause uptake on both sides of a joint, but it can be differentiated by assessing the clinical history and by finding involvement of all the joints in the affected extremity rather than a single joint.


Questions & Answers