Serum inflammatory markers for the screening and diagnosis of periprosthetic joint infection: a systematic review and meta-analysis

Abstract

Aim: As part of a multi-society effort to derive a unified consensus definition of periprosthetic joint infection (PJI), a systematic review of serum inflammatory marker diagnostic performance for hip, knee, and shoulder PJI was performed. Methods: PubMed (MEDLINE) and EMBASE were searched for studies reporting the diagnostic performance of C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), white blood cell count (WBC), fibrinogen, interleukin-6 (IL-6), or D-dimer for PJI. From these, each markers’ pooled sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV), and area under the summary receiver operating characteristic curve (AUSROC) were calculated using a random-effects model. Results: A total of 89 studies reported all diagnostic performance measures for at least one marker. CRP (84 studies, 22 351 patients) demonstrated a pooled sensitivity, specificity, PPV, NPV, and AUSROC of 79.4 % (95 %CI: 78.5–80.3), 77.7 % (77.1–78.3), 67.0 % (63.3–70.7), 86.6 % (84.5–88.7), and 0.872 (SE 0.01), respectively. Corresponding performance estimates for fibrinogen (14 studies, 3433 patients) were 70.9 % (68.3–73.3), 85.9 % (84.3–87.3), 77.2 % (71.8–82.6), 82.1 % (77.1–87.2), and 0.889 (0.02), respectively, and those for IL-6 (20 studies, 2318 patients) were 76.3 % (73.4–79.0), 85.8 % (83.8–87.6), 74.5 % (69.0–80.0), 86.0 % (80.6–91.3), and 0.900 (0.01), respectively. ESR, D-dimer, and WBC did not offer greater predictive values than these markers. Conclusion: Although serum CRP, fibrinogen, and IL-6 demonstrated the best performance among all analysed parameters, their diagnostic accuracy remains insufficient to reliably confirm or exclude PJI. Elevated serum markers should be re-evaluated as a diagnostic criterion in future PJI definitions. Level of evidence: The level of evidence was Level III.