Procalcitonin as a Predictor of Septic Knee Arthritis: A Retrospective Cohort Study

Abstract

Introduction: Differentiating septic arthritis from aseptic arthritis (AA) of the knee is difficult without arthrocentesis. Although procalcitonin (PCT) has shown diagnostic value in identifying bacterial infections, it has not been established as a reliable marker for identifying septic arthritis (SA). Recent studies have shown promise in the use of PCT as a useful systemic marker for identifying septic arthritis versus AA. This observational retrospective review compares PCT with routine inflammatory markers as a tool for differentiating septic arthritis versus AA in patients with acute, atraumatic knee pain.

Methods: Fifty-three consecutive patients (24 SA, 29 AA) were retrospectively reviewed at one institution with concern for SA. SA was diagnosed based on a physical examination, laboratory markers, and arthrocentesis. Laboratory indices were compared between the septic arthritis and AA groups. Data analysis was conducted to define sensitivity and specificity. Receiver operator characteristic curve analysis and regression were conducted to determine the best marker for acute SA of the knee.

Results: Using multiple logistic regression, bacteremia (OR 6.75 ± 5.75) was determined to be the greatest predictor of SA. On linear regression, concomitant bacteremia (coef 3.07 ± 0.87), SA (coef 2.18 ± 0.70), and the presence of pseudogout crystals (coef 1.80 ± 0.83) on microscopy predicted an increase in PCT. Using a PCT cutoff of 0.25 ng/mL yields a sensitivity of 91.7% and specificity of 55.2% for predicting SA; however, the ideal cutoff in our series was 0.32 ng/mL with a sensitivity of 79.2% and specificity of 72.4%. PCT was superior to the white blood cell count, erythrocyte sedimentation rate, and C-reactive protein in the area under the receiver-operating characteristic curve analysis.

Discussion: Procalcitonin seems to be the most sensitive and specific systemic marker in differentiating septic from AA.

Figure 1

Graph showing ROC analysis of laboratory values for all patients

Figure 2

Graph showing ROC analysis of laboratory values excluding those with another infection