Arthrocentesis and Arthrography

Arthrography is performed in shoulder arthroplasty patients for the following indications:

  • Infection
  • Evaluation of rotator cuff integrity
  • Provocative anesthetic injection
  • Evaluation of polyethylene glenoid component


Arthrocentesis is usually performed under fluoroscopic guidance with digital subtraction. The surgical scar should be examined for areas of dehiscence or leakage. If present, markers should be placed near these areas. An anterior to posterior approach is used. Using sterile technique, a 20-gauge spinal needle is advanced to the medial edge of the humeral head portion of the prosthesis.

Aspirated fluid is sent for gram stain, aerobic and anaerobic cultures, and cell count and differential. Other cultures such as fungal and AFB may be sent as well.

Dry taps can occur for the following reasons:

  • Dry joint
  • Large periarticular bursa acting as sump for fluid
  • Sinus track allowing for continuous drainage of joint

In these instances, a wash is performed with non-bacteriostatic normal saline. Washed fluid is sent for the usual labs. Cell count and differential however is not performed on joint washes.

After aspiration or wash, contrast is injected slowly into the joint and digitally subtracted images are obtained in the anterior projection. Some of this contrast is aspirated and sent for aerobic and anaerobic cultures labeled knee wash. The needle is then withdrawn and additional spot films may be obtained in various projections.


Arthrocentesis is performed when infected joint is suspected. It is considered the gold standard in determining whether infection is the cause of a painful shoulder arthroplasty. In infected total knee replacements, Barrack et al12 reported arthrocentesis sensitivity of 75%, specificity of 96%, and accuracy of 90%. They found false positive results to be less frequent than in arthrocentesis of hip replacements, however false negative results to be more common. This was attributed to the frequent use of antibiotics in patients with painful total knee replacements, and they recommend discontinuing antibiotics and waiting at least 2 weeks before aspiration is performed.

Sinus-track formation to skin wound outlined by arthrography constitutes definitive evidence of infection. Detection of contrast between bone-cement or bone-prosthesis interfaces is suggestive of loosening or infection, but in no ways is definitive. Likewise, failure of contrast to enter abnormally widened interfaces does not exclude loosening or infection.

Patient with remote shoulder arthroplasty presented with leakage at surgical scar. EKG lead was placed just inferior to leak and digital subtracted arthrogram demonstrates sinus-track connection. No fluid was aspirated from the joint due to sinus-track decompression.

Evolution of rotator cuff integrity:

During arthroplasty, rotator cuff tears are repaired. Unless there is a large chronic rotator cuff tear at time of surgery, the post-operative shoulder should have an intact rotator cuff. Arthrography can be performed to evaluate cuff integrity. Like the native shoulder, contrast injected into the glenohumeral joint should not connect to the subacromial-subdeltoid bursa.

Intact rotator cuff after hemiarthroplasty: contrast surrounds humeral head portion of the prothesis in a normal manner. There is no contrast in the subacromial-subdeltoid bursa.

Full thickness rotator cuff tear in patient with total shoulder arthroplasty. Contrast injected into glenohumeral joint traverses tear and fills the subacromial-subdeltoid bursa (arrows).

Provocative anesthetic injection:

Injection of long-acting anesthetics such as bupivicaine can help to localize painful symptoms to the neocapsule. Pain relief after injection indicates the source is from the joint. Lack of pain relief suggests pain is coming from surrounding musculature, cervical spine, etc.

Chronic pain after shoulder hemiarthroplasty with frozen shoulder. Contrast injection shows markedly small joint space and there was marked resistance to injection, consistent with adhesive capsulitis. Subsequent injection with bupivicaine relieved pain.

Evaluation of intraarticular pathology:

Intraarticular contrast better defines joint spaces and thus localize pathology.

Sudden onset of shoulder pain and decreased range of motion. Arthrogram spot films and CT demonstrate linear filling defect in the neocapsule between the humeral head and polyethylene glenoid components (arrows). Entrapped synovium was found at surgery.

Interior dislocated polyethylene glenoid component. An aspiration/arthrogram was performed preoperatively to rule out infection. Scout film demonstrates dislocated component in the axillary pouch of the joint. Arthrogram images are confirmatory. Aspirate was sterile.

Posteriorly dislocated polyethylene glenoid component. Arthrogram spot films and CT demonstrate posterior dislocation polyethylene glenoid components (arrows).

Next Page