Abstract
Purpose of the study: Few data are available concerning the proper management of patients with a periprosthetic fracture of the hip who presents biological signs of inflammation (increased WBC, sedimentation rate, or C-reactive protein). The purpose of this work was to determine the prevalence of elevated biological markers in this type of patient and to determine the reliability of such markers for the diagnosis of periprosthetic infection.
Material and methods: A periprosthetic hip fracture was diagnosed in 204 patients from 2000 to 2006. The WBC count, the sedimentation rate and the serum CRP level were noted at admission to the emergency ward. The diagnosis of infection was confirmed by at least two positive bacteriological samples of tissue biopsy and/or joint fluid collected at surgery. A statistical analysis was conducted to determine the prevalence of elevated biological markers of inflammation, the sensitivity, their specificity and their positive predictive value for deep infection.
Results: Twenty-one patients (11.6%) developed a periprosthetic infection. Among the 204 patients, the WBC count increased in 16.2%, sedimentation rate in 33.3% and CRP in 50.5%. The sensitivity was 24% (WBC), 50% (sedimentation rate) and 83% (CRP). The specificity was 85% (WBC), 69% (sedimentation rate) and 56% (CRP). The positive predictive value was low (18, 21 and 29% respectively).
Discussion: Markers of inflammation are frequently ordered before surgery to search for infection but can be elevated for various reasons. Most often, these markers are elevated because of the patient’s general status and are thus related to other co-morbid conditions and/or reaction to the new fracture. In this population, the WBC count did not contribute to the diagnosis of infection as only 24% of the infected patients had a high count. CRP and sedimentation rate and the WBC count had low positive predictive values.
Conclusion: This study shows that an isolated elevation of biological markers of inflammation in a patient with a periprosthetic fracture is not a good indicator of infection.
Correspondence should be addressed to Ghislaine Patte at sofcot@sofcot.fr