gms | German Medical Science

Objectives: The diagnosis of periprosthetic joint infection is a challenging task profounding implications for further treatment. Although clinical findings such as redness, swelling, pain and sinus tract may be sufficient in acute infections to detect a PJI, low-grade infections are often a diagnostic challenge. Beside clinical findings, C-reactive protein (CRP) measurement in serum is one of the most commonly used tests for periprosthetic joint infection. While MSIS guidelines include these markers as minor criteria in diagnosing PJI, the AAOS guidelines do not recommend further diagnostic tools such as joint aspiration in cases with normal CRP and ESR. But the role of CRP in the definitive diagnosis of PJI is still controversial. The aim of the current study is to determine the role of CRP in the diagnosis of PJI and evaluate the influence of the infecting organism type on the CRP level.

Methods: A retrospective analysis of all hip and knee PJI diagnosed in our institution from 2013 through 2017 was performed. Demographic data such as age and gender, type of infection (acute postoperative, acute haematogenous or chronic PJI), clinical manifestation, laboratory values including serum CRP and blood leukocytes and microbiological and pathological results of revisions were recorded for all patients. PJI was diagnosed according to proposed European Bone and Joint Infection Society criteria.

Results and conclusion: 204 patients were included, 107 hip and 97 knee prosthesis. Preoperative CRP levels were normal (lower than 10 mg/l) in 73 patients, representing 36% of the study group with positive cultures, distributed equally in hip and knee patients (34 of 97 knee, vs. 39 of 107 hip patients, p = 0.88). The mean CRP level of the normal group was 4.9 mg/l (SD 2.8). A total of 1,344 cultures were taken in all patients and only 63% (840) were positive. In 26% (53 of 204) of patients a polymicrobial PJI was identified. Interestingly, patients with normal preoperative CRP had significantly lower rate of positive cultures compared to patients with elevated CRP (199/461 (43%) vs. 643/883 (73%), p < 0.001). Furthermore, intraoperative taken cultures tend to isolate high virulent organisms more frequently compared to less-virulent ones. So 87% of the cultures were positive in patients with a PJI caused by Staphylococcus aureus compared to 45% by CNS and only 38% by Propionibacterium spp. Low-virulent microorganisms represented 86% of all PJI with normal CRP levels.

In conclusion, patients with seronegative PJI represent a remarkable and clinically challenging subset. Treating physicians must beware that CRP may yield a false-negative result, especially in the setting of low-virulent microorganisms. It is important to lower our threshold to utilize further diagnostic tools, especially joint aspiration, in chronic painful prosthesis.