How long to treat with antibiotics following amputation in patients with diabetic foot infections? Are the 2012 IDSA DFI guidelines reasonable?
Correspondence: Steven W. Johnson, 1612 Barndale Glen Ct, Winston-Salem, NC 27106, USA. Tel.: 303-720-5978; fax: 919-681-7494; e-mail: email@example.com
What is known and Objective
To the best of our knowledge, there has been no published study designed to identify the most appropriate duration of antibiotic therapy in lower extremity skin and skin structure infections in diabetic patients [aka “diabetic foot infections” (DFI)] post-amputation. However, recent guidelines published by the Infectious Diseases Society of America (IDSA) provide recommendations for treatment duration in these patients. Therefore, our objective is to review the literature evaluating antibiotic treatment in DFI to determine if the IDSA guidelines are reasonable.
Evidence for the use of antibiotics after amputation comes largely from perioperative surgical prophylaxis studies evaluating the rate of infection after amputation. Three such studies were identified; 2 found a 5-day course of antibiotics post-amputation resulted in a reduction of infection rate, while 1 found no additional benefit. Comparative antibiotic studies in DFI also offers evidence for treatment duration, of which, 10 studies were identified. Five included patients who received amputations; however, only 1 reported treatment outcomes in a subset of diabetics requiring amputation. In this study, the authors concluded that antibiotic treatment is likely necessary after amputation.
What is new and Conclusion
Given the general lack of data, we recommend that post-operative treatment duration be individualized, and, until further studies are done, it seems reasonable to adhere to the recommendation provided by the 2012 IDSA DFI guidelines for a 2–5 day course of antibiotic therapy post-operatively when no residual infected tissue remains.
What is known and objective
Lower extremity skin and skin structure infections in diabetic patients, commonly known as diabetic foot infections (DFI) are common in persons with diabetes mellitus and can be associated with considerable morbidity and (less frequently) mortality.[1-4] Although foot ulcers are estimated to occur in 15% of US patients with diabetes, recent research by the American Diabetes Association (ADA) suggests this figure may be as high as 25%.[5-8] The foot is the most common site of infection in diabetics and represents the most frequent indication for hospital admission in these patients.[1-4] Diabetic patients have a 10-fold greater risk of hospitalization due to lower extremity infections when compared with non-diabetics.[6-8] Most hospitalizations occur in patients with poor glycaemic control or who have infections classified as moderate-to-severe.
In May 2012, the Infectious Diseases Society of America (IDSA) updated its clinical practice guidelines for the diagnosis and treatment of DFI from its original publication in 2004.[9, 10] These guidelines feature reviews of the pathophysiology, microbiology and the overall management of DFI. Most notably, the guidelines focus on the antibiotic treatment of DFI and provide recommendations for future research.
Despite the recent publication of the IDSA guidelines, the optimal management of DFI and its complications remain unclear and controversial. Initial management should consider the need for hospitalization and the role of surgical intervention. Surgical intervention can range from drainage and excision of infected tissue to revascularization of lower extremity and amputation. In the United States, almost 70% of non-traumatic lower extremity amputations occur in diabetics, the majority of which are preceded by an infected foot wound.[11-13] Even with optimal antibiotics, amputation is the only or best option for a cure of DFI and its complications in some patients.[14, 15] Amputations may be needed in those patients with necrotizing fasciitis, surgical management of osteomyelitis, gangrene, extensive soft tissue loss, evidence of compartment syndrome, or in those with critical limb ischaemia.[9, 14, 15] In these patients, cure most likely occurs by surgical removal of all infected bone, since leaving infected bone behind increases the possibility for recurrence. It remains unknown, however, how amputation impacts the needed antibiotic treatment duration, since the primary site of infection has been removed in these patients. For most infections, it is often governed by factors which include (but are not limited to) the site and severity of infection, the patient's immune status, and the rate of resolution of signs and symptoms of infection. In regards to lower extremity skin and skin structure infections, antibiotic therapy is generally discontinued when the patient has received a minimum course of therapy, is metabolically stable and the cardinal signs of infection have resolved. There are currently no data to support using antibiotics to treat wounds after evidence of infection resolution in an attempt to either hasten healing or to prevent recurrent infection.[2, 9] Unfortunately, evidence-based treatment recommendations for patients who require amputations are based on poorer quality of evidence than patients who just require antibiotic therapy. This is especially evident when trying to determine the duration of antibiotic treatment after amputation. Therefore, our objective is review the literature evaluating antibiotic treatment in DFI to determine if the IDSA guideline recommendations are reasonable.
The 2004 IDSA guidelines provide a framework for determining the total length of therapy required to treat DFI based on severity of soft tissue infection, presence/absence of boney involvement and (if bone involvement is present) the presence/absence of residual infection (see Table 1). These guidelines have remained unchanged in the 2012 update. In patients with a bone or joint infection who undergo amputation and have no evidence of residual infected tissue post-operatively, a 2–5 day course of antibiotics is currently recommended. Those with bone and joint infection with residual infected soft tissue should receive 2–4 weeks of antimicrobials, and those with residual (but viable) infected bone should receive 4–6 weeks.
Table 1. Suggested route, setting and durations of antibiotic therapy, by clinical syndrome (adapted from)
|Soft-tissue (mild)||Topical or oral||1–2 weeks, may extend up to 4 weeks if slow to resolve|
|Soft-tissue (moderate)||Oral or initial parenteral||2–4 weeks|
|Soft-tissue (severe)||Initial parenteral, then switch to oral when possible||2–4 weeks|
|Bone or joint (no residual infected tissue – e.g. amputation)||Parenteral or oral||2–5 days|
|Bone or joint (residual infected soft tissue – but not bone)||Parenteral or oral||2–4 weeks|
|Bone or joint (residual but viable infected bone)||Initial parenteral, then consider switch to oral||4–6 weeks|
|Bone or joint (no surgery, or residual dead bone post-operatively)||Initial parenteral, then consider switch to oral||> 3 months|
While the guidelines provide an evidence grade of ‘weak, low’ (uncertainty in the estimates of desirable effects provided by evidence from at least 1 critical outcome from observational studies, RCTs with serious flaws or indirect evidence, other alternatives may be equally reasonable. Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.) for the treatment durations provided, there has been no published study designed explicitly to identify the most appropriate duration of therapy in such patients. Evidence for the use of antibiotics after amputation comes from perioperative surgical prophylaxis studies evaluating the rate of infection after major limb amputation. In one such retrospective (non-randomized) evaluation in patients undergoing major lower limb amputation, the impact of a 5-day course of antibiotics vs. a 24-h course of antibiotics on stump infections was conducted. A consecutive series of 40 patients receiving a 24-h course of prophylactic antibiotics was followed by a prospective evaluation of 40 consecutive patients receiving a 5-day regimen of the same antibiotics [flucloxacillin 1 g IV or vancomycin 1 g IV (vancomycin was used in penicillin-allergic patients and where MRSA was present or suspected clinically) plus gentamicin 120 mg IV or ciprofloxacin 200 mg IV plus metronidazole 500 mg IV]. The groups were similar in terms of patient demographics, vascular risk factors and level/type of amputation. Significantly reduced infection rates (5% vs. 22·5%, P = 0·023) and length of hospital stay (22 vs. 34 days, P = 0·001) were observed in the 5-day antibiotic regimen compared to the 24-h regimen respectively. Revision rates were lower with the 5-day antibiotic regimen but did not reach statistical significance (2·5% vs. 10%, respectively, P = 0·36). However, none of the cases received amputation for a DFI and less than half of the subjects were diabetics, therefore, the usefulness of this trial in determining antibiotic treatment duration in diabetic amputees is questionable. A similar study examined the duration of antibiotic prophylaxis in 302 subjects receiving vascular reconstruction procedures. Subjects were prospectively randomized to a single dose or a multidose-regimen of ticarcillin/clavulante 3·1 g IV until all lines were removed (up to 5 days). The multidose group had a lower wound infection rate (10% vs. 18%, respectively, P = 0·04). In contrast, an earlier study had failed to find any benefit of a longer 5-day course of antibiotics. Neither of these studies looked specifically at the role of antibiotics in major lower limb amputation. Furthermore, neither study looked specifically at diabetic patients.
Comparative antibiotic studies in DFI also provide some evidence for treatment duration post-amputation. In one such study, ampicillin/sulbactam was compared with imipenem/cilastatin. Patients were evaluated at day-5 and at the end-of-treatment (EOT). Amputations were performed on 33 (69%) and 28 (58%) of the ampicillin/sulbactam and imipenem/cilastatin groups (respectively) as part of a standard protocol for osteomyelitis. Of the 61 amputations, 57 were limited to the infected digits and metatarsal heads and were considered foot-sparing procedures. Four patients underwent below-the-knee amputation (BKA). Amputation was undertaken before the day-5 assessment in 22 patients and in 38 by EOT. For all patients, EOT occurred at a mean (± SD) of 13 (± 6·5) days and 15 (± 8·6) days in the in the ampicillin/sulbactam and imipenem/cilastatin groups respectively. Among those who had amputations, pathogen eradication (independent of treatment group) was achieved by the day-5 and EOT assessment for 20 of 61 and 41 of 61 patients respectively. However, this is difficult to extrapolate to DFI amputation patients in general. Other than the day-5 and EOT evaluation time points, it is not possible to determine when the amputations were performed in the subset undergoing surgery, and thus ascertain the duration of post-operative antibiotics. In this study, where the authors noted that surgery rarely resulted in the total resection of all infected tissue, antibiotic treatment is likely necessary after amputation.
Since the publication of the IDSA guidelines in 2004, results of several key clinical trials evaluating the use of antimicrobials in this patient population have been published.[22-30] Of these, 9 randomized clinical trials that focus on antimicrobial treatment of DFI were identified (see Table 2). Of the 4 studies that included patients who received amputations, 1 study eliminated patients who received amputations from the clinical response endpoint and were deemed endpoint failures. Although the remaining 3 trials included patients who received amputations, outcomes in this subset were not differentiated from other subjects in the study. Consequently, of these studies, only 1 was identified that may aid to the current trials mentioned that have led to the 2012 IDSA guideline recommendations. In this study, piperacillin/tazobactam was compared with imipenem/cilastatin in 62 patients with DFI. Of these, 48 had associated osteomyelitis. Treatment was planned for 14 days. For osteomyelitis, treatment was administered for 28 days, counting from the time of debridement (if performed). The mean duration of treatment was 21 days for the piperacillin/tazobactam group and 24 days for the imipenem/cilastatin group. Successful clinical response was seen in 14 (46·7%) and 9 (28·1%) respectively. Relapses occurred in 0 and 2 patients respectively. Of those patients with positive cultures, 47 (96%) had a complete response which was similar between both groups. Amputation was performed in 18 (60%) and 22 (69%) of patients respectively. The amputations included minor digital amputations (47%), transmetatarsal (10%), tarsometatarsal [i.e. Chopart, (20%)] and BKA (22·5%). These amputations were performed for uncontrolled infection in 30 patients and ischaemia or non-infectious reasons in 10 patients. The authors did note that after total excision of the bone, 5 days of treatment were considered adequate depending on the clinical response. However, this study reported patients who had amputations despite completing appropriate therapy as failures and were not included as having clinical response. Therefore, it is not possible to determine if the 5 days of post-amputation treatment was justified or adequate. Furthermore, the authors did not mention if any patients required more than 5 days of treatment based on clinical response and patients who received amputations were not differentiated from other patients in the reporting of duration of treatments.
Table 2. Published randomized trials on diabetic lower extremity infections from 2004 to 2012
|Saltoglu et al. (2010)||OL||64 ||P/T vs. I/C||Included ||Included||14 (osteo for 14–28)||21 vs. 24|
|Lipsky et al. (2008)||DB||835 ||pexiganan vs. ofloxacin||Excluded||Excluded||14–28||23 (study 303) 22 (study 304) |
|Noel et al. (2008)||DB||257a||ceftobiprole vs. vanco plus ceftaz||Excluded||Excluded||7–14||9 vs. 9·1|
|Lipsky et al. (2007)||DB||127 ||moxi vs. P/T then A/C ||Excludedb||Included||7–14||6·7 (iv) vs. 6·3 (iv), plus 7·4 (po) vs. 7·9 (po)|
|Giordano et al. (2005)||DB||78a||moxi vs. P/T then A/C po||Excludedb||Included||7–14||6 (iv) vs. 6 (iv), plus po (mean not provided)|
|Harkless et al. (2005)||OL||314 ||P/T vs. A/S||Excluded||Excluded||4–14 (possibly extended to 21)||8 vs. 8·5 |
|Lipsky et al. (2005)||DB||586 ||ertapenem then A/C vs. P/T then A/C||Excludedb||Included||5–28||11·1 (iv) vs. 11·3 (iv), plus 9·7 (po) vs. 9·7 (po) |
|Lipsky & Stoutenburgh (2005)||SB||103||daptomycin vs. vanco or semi-synthetic pcn||Excluded||Excluded||7–14||Not provided|
|Lipsky et al. (2004b)||OL||361 ||linezolid vs. A/S or A/C||Included ||Not mentioned||7–28||7·8 (iv) vs. 10·4 (iv), plus 15·9 (po) vs. 15 (po) |
What is new and conclusion
The management of lower extremity infections in diabetic patients continues to challenge clinicians, especially in subjects that received amputation. Updated guidelines recently published May 2012 by the IDSA on the diagnosis and treatment of DFI can help guide clinicians when treating these patients, however, they do not add any additional information from the original 2004 guidelines on treatment duration. According to these guidelines, the recommended treatment duration of antibiotics is 2–5 days in subjects with no residual infected tissue remaining. However, there appears to be no good evidence to support this recommendation. The 2–5 day recommendation is extrapolated from studies evaluating the rate of infection after major limb amputation in uninfected patients receiving prophylactic antibiotics in both diabetic and non-diabetic patients. Only 1 study was identified that reported treatment outcomes in a subset of diabetics with lower extremity infections requiring amputation. However, this study was not specifically designed to evaluate the optimal duration. It was designed to evaluate the treatment of limb-threatening foot infections in diabetic patients in which a majority received amputation.
Given the lack of data, we recommend that post-operative treatment duration be individualized. Clinicians should consider the level of amputation (whether it was likely to successfully remove all clinically infected tissue). Determination whether there is residual infection still left in the tissues following amputation is very much a judgment call, with varying levels of certainty depending on the case. Other factors to consider include: evidence for residual infection post-operatively (local and systemic symptoms and clinical signs of inflammation) as well as personal history (history of previous amputations due to DFI). Furthermore, the severity of presenting infection and timing of ablative surgery should be considered. It is reasonable to assume patients with a severe and/or deep DFI are likely to undergo amputation early in their hospital course. Therefore, these patients are more likely to have retained infection post-operatively than those presenting with superficial/mild-to-moderate infections that have had ample exposure to IV antibiotics and time for infection localization prior to surgery. Patients at greater risk for continued infection after amputation (e.g. immunocompromized, history of recurrent infections, delayed wound healing) may benefit from receiving a longer course of antibiotics post-amputation. In patients with low risk for continued infection post-operatively and those without evidence of infection, shorter antibiotic treatment post-amputation may be adequate. Future studies should focus on determining the optimal antibiotic regimen, specifically antibiotic treatment duration. Until these are completed, it seems reasonable to adhere to the recommendations from the IDSA guidelines for 2–5 days of antibiotic therapy post-amputation when no residual infected tissue remains.