Background : The microbiological diagnosis of bone and joint infections (BJI) currently relies on cultures, and the relevance of molecular methods is still debated. The aim of this study was to determine whether polymerase chain reaction (PCR) could improve the etiological diagnosis of BJI.

Methods : A prospective study was conducted during a 4-year period at Lariboisiere University Hospital (Paris, France), including patients with suspicion of infectious spondylodiscitis, septic arthritis, prosthetic joint infections, and respective noninfected groups. Clinical and radiological data were collected at inclusion and during follow-up. All samples were analyzed by conventional cultures and 16S ribosomal deoxyribonucleic acid (rDNA) gene (16S-PCR). Specific cultures and PCR targeting Mycobacterium tuberculosis were also performed for spondylodiscitis samples. Case records were subsequently analyzed by an independent expert committee to confirm or invalidate the suspicion of infection and definitively classify the patients in a case or control group. The sensitivity of the combination of culture and PCR was compared with culture alone.

Results : After expert committee analysis, 105 cases of BJI cases and 111 control patients were analyzed. The most common pathogens of BJI were staphylococci (30%), M tuberculosis (19%), and streptococci (14%). Adding PCR enhanced the sensitivity compared with culture alone (1) for the diagnosis of M tuberculosis spondylodiscitis (64.4% vs 42.2%; P < .01) and (2) for nonstaphylococci BJI (81.6% vs 71.3%; P < .01). It is interesting to note that 16S-PCR could detect BJI due to uncommon bacteria such as Mycoplasma and fastidious bacteria.

Conclusions : Our study showed the benefit of 16S-PCR and PCR targeting M tuberculosis as add-on tests in cases of suspected BJI.