Response to Comment on Lázaro-Martínez et al. Antibiotics Versus Conservative Surgery for Treating Diabetic Foot Osteomyelitis: A Randomized Comparative Trial. Diabetes Care 2014;37:789–795

We appreciate very much the comments on our article (1) from Landman et al. (2). It could be extracted that every patient in whom osteomyelitis is suspected should undergo a bone biopsy, but this is not based on evidence. Based on this thought, the current tendency of treating diabetic foot osteomyelitis exclusively with antibiotics could not be supported. Most studies in which the patients were treated with antibiotics did not have microbiologic and/or histopathologic confirmation of bone infection. The only modern series confirming osteomyelitis by histopathology and microbiology are probably surgical. We could conclude that most patients in whom antibiotics were successful did not in fact have osteomyelitis. Patients were included in most modern series of treatment of diabetic foot osteomyelitis after a clinical suspicion by means of the probe-to-bone (PTB) test, simple X-ray examination, and/or advanced imaging. The role of percutaneous bone biopsy was established in a retrospective study in which bone biopsy with microbiologic, but not histopathologic, purposes was used in some centers (3). The material extracted from small bone biopsies was probably not enough for histopathologic analysis. The possibility that contamination occurs in cases of exclusive soft-tissue infections must also be taken into account when percutaneous bone biopsy is taken. Despite the fact that bone biopsies are performed through apparently intact skin, the existence of subcutaneous infection in patients in which the inflammatory response is attenuated cannot totally be ruled out. Landman et al. (2) highlight the fact that simple X-rays and the PTB tests on their own show poor performance for diagnosing diabetic foot osteomyelitis. However, a diagnostic sequence based on the combination of these two tests showed a positive likelihood ratio of 12.8 and a negative likelihood ratio of 0.02 (4). Only 6.6% of patients with negative results on both diagnostic studies had osteomyelitis (4). In that study, osteomyelitis was confirmed by means of histopathology and microbiology.

On the other hand, it has been demonstrated that a positive microbiologic and a negative histologic result from a bone biopsy are just as likely as a negative microbiologic and a positive histologic result (5).

Furthermore, in our trial every patient who underwent surgery had histopathologic confirmation of osteomyelitis including patients in which Staphylococcus epidermidis was isolated. It seems logical to think that if the same diagnostic sequence was applied in the arm of patients treated with antibiotics, the same rate of patients would have osteomyelitis. Logically, we cannot be sure and this has been addressed in the limitations of the study. In the same way that Landman et al. (2) argue that there is a discrepancy between swab ulcers and bone cultures, other authors have demonstrated that deep-wound cultures, as were performed in our study, correlate well with bone cultures (6).

In conclusion, every patient included in the arm of surgical treatment had osteomyelitis confirmed by histopathology. The limitation of the absence of bone biopsy in the antibiotic arm of our study is a limitation we had previously recognized in our article.

• © 2014 by the American Diabetes Association.