Artikel
Diagnosing deep sternal osteomyelitis: correlation between CT imaging, histopathology/bacteriological results and infectious parameters
Suche in Medline nach
Autoren
Veröffentlicht: | 10. September 2013 |
---|
Gliederung
Text
Introduction: Deep sternal osteomyelitis has an incidence of about 8%. CT or MRI imaging is described as gold standard for diagnosing deep sternal osteomyelitis non-invasively. So far, however, there are very few papers on correlation between CT imaging results, serum levels of white blood cell count (WBC) and C-reactive protein (CRP) showing infection, and histopathological and bacteriological findings.
Aim: In this retrospective study we evaluated the correlation between diagnosis of deep sternal osteomyelitis based on CT imaging studies and case by case WBC, CRP, histopathological and/or bacteriological results.
Patients and Methods: 72 patients operated at our institution from 2009 to 2010 were evaluated in this study. Inclusion criteria were patients being operated at our institution, existing CT imaging and WBC/CRP before our first operation and histopathological and bacteriological results existing after the first operation at our institution. Acute infection was defined if WBC was ≥10,000 cells/ml, chronic infection if CRP was ≥10 mg/l. Correlation analysis was done using SPSS program for calculating the Pearson correlation coefficient.
Results: 43 patients were included in this study. There were 25 male (58.1%) and17 female (41.9%) patients in this study. Age ranged from 44 to 84 years (median 70.2). All 43 patients had positive bacteriology. 28 (65.1%) patients had positive CT scan signs for deep osteomyelitis, 31 (72.1%) patients had positive histopathological results. 13 (30.2%) patients had a WBC >10,000/ml, 40 (93%) patients had a CRP level > 10 mg/l. In 10 cases (23.3%) and 12 cases (27.9%) (51.1% total) CT diagnosis was concordant with histopathological, bacteriological results, and elevated WBC levels or histopathological, bacteriological results, and elevated CRP levels, respectively. In 21 cases (48.8%) CT imaging was concordant with histopathological and bacteriological results only. For 11 patients (25.6%) and in 26 cases (60.5%) (83.7% total) CT imaging and WBC, and CT imaging and CRP levels showed deep sternal osteomyelitis, respectively. 9 patients (20.9%) had concordant histopathological/bacteriological and WBC results, and 27 patients (62.8%) (83.7% total) concordant histopathological/bacteriological, and CRP results for deep sternal osteomyelitis. Pearson correlation coefficient was 0.5 for CT imaging and histopathology, WBC and CRP. For CT imaging, WBC results and CRP levels Pearson correlation coefficient was 0.79. The correlation coefficient between histopathology/bacteriology and WBC and CRP levels was 0.77. (Figure 1 [Fig. 1]).
Discussion: Although results for CT imaging alone are not encouraging they must be seen critically. The number of patients included in this study is rather small; therefore a larger study is desirable. However, even in this small number the correlation between CT diagnosis of deep sternal osteomyelitis and WBC results and CRP levels is rather robust. Second drawback is that histopathologic results can only show osteomyelitis, if the operating surgeon has taken a representative sample. Combining laboratory levels of infection combined with non-invasive CT imaging seems to be the best solution.