Abstract
We investigated the role of Plasma Viscosity (PV), C-reactive protein (CRP) and Frozen Section (FS) in diagnosing prosthetic joint infection. We compared these results with microbiological diagnosis of infection of the tissue samples (three or more samples grown same organisms in culture).
53 patients, average age 67 years (37 – 89) underwent joint revision surgery. 34 patients had hip and 19 patients had knee joint revision arthroplasty, this includes single and multiple stage revision surgeries and excision arthroplasty. Nine (17%) patients had microbiologically proven joint infection. PV had sensitivity of 100%, specificity of 43% and negative predictive value of 100%. CRP had sensitivity of 89 %, specificity of 75% and negative predictive value of 97%. FS (presence of infection being more than 5 neutrophils/hpf) had sensitivity of 56% and specificity of 84%.
We recommend PV and CRP to be used in the investigation of prosthetic joint infection. If both CRP and PV are normal the chance of infection is very low (negative predictive value of 100%). In our series an elevated PV and CRP represented a 50% chance of having a joint infection. The role of frozen section does not appear to be beneficial in the diagnosis of joint
