Background Diagnosing infectious spondylodiscitis is greatly facilitated by improved access to MRI testing. However, MRI testing increases the possibility of false positives. The efficiency of bacteriological testing is highly variable depending on the series performed, and this efficiency is not improving.
Objectives The objective of this study is to evaluate the efficiency of such testing in the last decade and to compare it with the efficiency determined in the preceding decade by the same department.
Methods This is a retrospective monocentre study of records of patients hospitalised in the rheumatology department of the CHU de Clermont Ferrand university hospital from 2003 to 2012 for infectious spondylodiscitis. These patients were compared with a historical cohort of 90 patients from the department, including all spondylodiscitis cases from 1992 to 2002 selected using the same criteria.
Results From 2003 to 2013, 99 patients were hospitalised for infectious spondylodiscitis and were compared with the cohort of 90 patients hospitalised from 1992 to 2002.
A pathogen was identified in 78% (77/99) of the cases, which is not significantly different from the preceding decade: 68% (61/90). Blood cultures were positive respectively in 37% (37/99, 37/95 excluding tuberculosis) and 39% (35/90, 35/88 excluding tuberculosis) (NS). In the last decade, pathogens were more often isolated through fine needle aspiration biopsy (FNAB) of the vertebral disc: 47/99 (47%) vs. 28/90 (31%) p=0.02. In the last decade, the efficiency of vertebral disc FNAB was 66% (47/71); eight patients underwent two vertebral disc FNAB and the second biopsy was only positive once. The efficiency was 77% (72/93) for primitive non-tubercular spondylodiscitis in the last decade.
Over both periods, the proportion of staphylococci (49%), streptococci and enterococci (23% and 26%), gram negative bacilli (13 and 20%) and tuberculosis (5 and 3%) were not significantly different. However, amongst staphylococci, there was an observed decrease in the proportion of S. aureus in the last decade (47% vs. 70%, p=0.03) and increase in the proportion of coagulase negative staphylococci.
Patients of the last decade were older (66 vs. 61 years of age; p=0.01) and more often had infection risk factors (51% vs. 33%; p=0.03), and diabetes in particular (27% vs. 9%; p=0.001).
Conclusions In our recruitment, the efficiency of bacteriological testing for spondylodiscitis has not declined in 20 years, and has even slightly improved. However, for one quarter of spondylodiscitis cases presumed to be infectious, the pathogen remains unknown.
Disclosure of Interest None declared