Abstract
Sir,
Dr Chu and colleagues have provided us with an important and well-conducted randomized trial addressing the question of duration of antibiotic therapy for a diabetic foot infection (DFI). 1 In light of the potential for antibiotic therapy to cause adverse clinical effects, increase financial costs, and potentially induce antibiotic resistance, it would be best to discontinue therapy as soon as possible. But, with many diabetic patients having peripheral neuropathy, peripheral arterial disease, or the ill-defined immunological problems that may impair host responses to infection, it may be difficult to know what duration of therapy is sufficient.2,3 Not only might some patients with a DFI need more prolonged therapy, but it may also be difficult to know when their infection has resolved. Chu et al have provided useful data on at least one part of this complex mix of questions.
I think readers should note, however, that there are several limitations to this study, some of which have not been sufficiently emphasized and others not mentioned in the article. I also have several questions for the authors, the answers to which may put their data into better perspective.
You excluded patients requiring revascularization, yet you classified some of your patients as having “severe peripheral arterial disease” (defined as an ankle brachial index of <0.4). What criteria did you use to decide if the patient needed revascularization?
You stated how you defined diabetic foot osteomyelitis (based on clinical and radiographic findings), but did not state whether or not these patients were excluded. Since osteomyelitis requires a longer duration of antibiotic therapy than soft tissue infection, it is key to know if osteomyelitis was an exclusion criterion. Even if it was, osteomyelitis may be missed at the time a DFI is diagnosed, so repeat foot X-rays a few weeks later would be needed to ensure it was not present in some patients randomized to 1 of the 2 treatment groups.
It appears that all enrolled patients were hospitalized, including “most” of those with a mild infection, who would normally be treated as outpatients in centers that have this type of ambulatory treatment available. Since you use the term “most,” were any of your patients treated as outpatients? If so, was there a difference in the percentage hospitalized in the 2 treatment groups?
You used the recommendations in the Infectious Diseases Society of America (IDSA) guidelines 2 for defining the presence and severity of the DFI, but you used your own system for defining clinical failure or recurrence. This system included parameters not included in the IDSA infection definition, for example, presence of necrotic tissue, growth of granulation tissue, bacterial culture results, hs-CRP level. I am not aware of any study validating the use these parameters to define resolution or recurrence of infection, are you? Why not use the resolution of the findings that were used to define infection as evidence of it being cured, as has been done in several published studies?
I cannot find the duration (mean, median, standard deviation) of antibiotic therapy for patients in each of the 2 treatment groups. As this is a key issue that your study was designed to address, it is important to clearly provide these data. Was the difference in duration between those who continued versus discontinued antibiotic therapy statistically significant and clinically important?
Again, my congratulations to the authors and thanks for undertaking this important study. You have provided useful data for answering a key question. I hope your responses to my questions will clarify a few additional issues.