Converting between high- and low-sensitivity C-reactive protein in the assessment of periprosthetic joint infection.

Abstract

Although low-sensitivity CRP (Ls-CRP) is an important tool for evaluating infected orthopedic prostheses, no clinical studies have evaluated whether Hs-CRP is a suitable surrogate for Ls-CRP or other traditional infection biomarkers. The laboratory data of 98 arthroplasty patients with suspected prosthetic infection were reviewed. Hs-CRP was highly correlated with Ls-CRP (R = 0.93). ROC analysis generated 100% sensitivity and 97% specificity for both Hs-CRP and Ls-CRP at optimal cutoffs of 28.6 and 2.6 mg/dL, respectively. Both CRP tests were more accurate than serum erythrocyte sedimentation rate, neutrophil differential, and white blood cell count. Hs-CRP was no different from Ls-CRP after unit conversion, and regression analyses suggested conversion factors that approximated 10. Hs-CRP and Ls-CRP have equivalent utility in the diagnosis of infected joint arthroplasty.