Can C-Reactive Protein Be Used To Predict Acute Native Joint Septic Arthritis in the Adult Population?

Background:

The hot swollen joint is a common presentation to the emergency department. A wide range of potential differential diagnoses exist. There is evidence to support CRP as a predictor of septic arthritis(SA) in children.  Limited evidence exists in adults.  The primary aim of this study was to establish whether C-reactive protein (CRP) can predict native joint SA in the adult population.

Method:

Retrospective cohort study. All patients who underwent native joint aspiration by the orthopaedic team over a 4 year period were identified from the lab microbiology records. Exclusion criteria applied.  Patients were divided into three groups for analysis: patients with SA, crystal arthropathy and those with normal/arthritic joints.

Results:

190 patients were identified. 15 (8%) were deemed to have SA giving a local incidence of 1 per 100,000.18 had a crystal arthopathy (10%) and 157 (82%) had normal or osteo/inflammatory arthritis.  We identified a significant difference in mean CRP between the SA group and the other two groups (p<0.001).  Receiver operative curves produced a CRP of 90mg/L as the cut off to differentiate between non-infective causes and SA (Sensitivity 100%, Specificity 96%).  We showed a significant difference in the aspirate WCC in SA compared with normal joints. There was no significant difference when comparing infected and crystal arthropathy.  However we identified 20,000 as the cut off to effectively exclude SA (Sensitivity 27%, Specificity 100%).   We identified a low sensitivity of aspiration gram stain (27%).

Conclusion:

Gram stain has a low sensitivity.  It is therefore useful to identify markers predictive of SA.  CRP is a reliable independent marker to differentiate cause of a hot swollen joint.  A threshold of 90mg/L is a very sensitive tool to diagnose septic arthritis.  This, combined with a WCC of <20,000 is very sensitive and specific way to exclude SA.