Diagnostic Accuracy for Periprosthetic Joint Infection Does Not Improve by a Combined Use of Glucose and Leukocyte Esterase Strip Reading as Diagnostic Parameters.

Abstract

The diagnosis of periprosthetic infections (PJI) can be challenging in clinical practice because the clinical presentations of aseptic loosening and low-grade infections are similar. Semiquantitative evaluation of leukocyte esterase (LE) in synovial fluid using a urine strip test has already established itself as a diagnostic method over the past decade. The analysis of LE in synovial fluid leads to a high number of false-positive test results. In the present study, the value of a combined semiquantitative determination of glucose and LE in synovial fluid to improve the diagnosis of PJI was investigated. Over a 4-year period, 145 synovial samples were collected from patients who developed joint effusion after arthroplasty. LE and glucose test strips were considered as an index test for the diagnosis of PJI. A ++ or +++ LE and a negative glucose test strip reading were considered as positive test results. Modified diagnostic criteria for PJI as recommended by the Musculoskeletal Infection Society (MSIS) served as the reference test, except that intraoperative findings were excluded. Forty-six out of 145 samples were classified as septic complication according to the reference test. In regard to PJI, our data showed that combined use of LE and glucose strip test reading displayed a 98.0% specificity (95% confidence interval (CI): 95.2% to 100%), a 50% sensitivity (95% CI: 35.6% to 64.4%), a 92% positive predictive value (95% CI: 81.4% to 100.0%), and an 80.3% negative predictive value (95% CI: 73.2% to 87.4%). In contrast, the exclusive analysis of LE on the urine strip to diagnose PJI demonstrated a 90.9% specificity (95% CI: 85.2% to 96.6%), a 67.4% sensitivity (95% CI: 53.8% to 80.9%), a 77.5% positive predictive value (95% CI: 64.6% to 90.4%), and an 85.7% negative predictive value (5% CI: 79.0% to 92.4%). A combination of LE and glucose test pad reading is considered superior as a potential “rule-in” method for the diagnosis of PJI compared with LE test pad analysis alone. However, combined LE and glucose synovial fluid testing also demonstrated lower test sensitivity and thus diagnostic accuracy compared with LE analysis alone. Therefore, combined glucose and LE test pad analysis does not represent a sufficient diagnostic standard to exclude PJI with certainty.