The following guidance aims to help healthcare professionals make decisions about antibiotic agents for the treatment of the infected diabetic foot in order to improve patient outcomes.
This is a consensus document based on limited available clinical trial evidence, review of international guidelines and expert opinion. There may be circumstances where alternative courses of action are appropriate. Guidance about antibiotic choice is dependent on local microbiological epidemiology and susceptibility patterns. However, the consensus group felt that the pathogens causing various diabetic foot infections in Scotland are unlikely to vary substantially within Scotland. Therefore, merit was seen in providing broad, practical guidance on antibiotic choice, subject to local adaptation when necessary.
General approach to diabetic foot ulcer management
The multidisciplinary team
Diabetic foot ulcers should be treated by a multidisciplinary footcare team as this managment strategy has been shown to reduce amputation rates. In addition, attention to aggressive treatment of macrovascular risk factors in people with diabetic foot ulcers has been shown to prolong survival.
Re-ulceration
Previous ulceration is the strongest predictor for recurrent ulceration and preventative measures need to be addressed following healing. Re-ulceration rates of up to 70% at 5 years have been reported.
Specimens for culture
There is some debate over when a culture is necessary. Clinically uninfected ulcers rarely need to be cultured. An acutely infected wound of mild or moderate severity in a person who has not recently been treated with antibiotics does not need to be cultured. Other wounds should almost always be cultured. If a specimen is not taken at presentation of clinical signs, then cultures should be taken if there is clinical failure of empirical antibiotics.
Aspiration of purulent secretions, curettage of the post-debridement wound base, punch biopsy and extruded or biopsied bone are the best specimens for culture.
Diagnosing bone infection
Inability to touch bone when probing a wound with a sterile metal probe makes osteomyelitis unlikely, with a negative predictive value of approximately 90% (Jeffcoate and Lipsky, 2004). The positive predictive value of a positive probe-to-bone test is around 50%, meaning that half of all ulcers that probe to bone do not have osteomyelitis (Jeffcoate and Lipsky, 2004).
If there is clinical suspicion of osteomyelitis plain X-ray is the usual initial investigation of choice. However, it can take 2 weeks before any changes of acute osteomyelitis are seen on plain radiograph and thus serial X-rays may be required to rule out osteomyelitis. If there is ongoing concern of osteomyelitis and it cannot be diagnosed using X-ray, secondary investigations of choice are (in order of preference):
- Magnetic resonance imaging (MRI).
- Isotope white cell scan.
- Triple phase bone scan (highly sensitive at diagnosing osteomyelitis, but is not specific, and can remain positive for >1 year).
Imaging options may be dictated by the local availability of imaging equipment.
Loose bone
Loose bone extruded from an ulcer, or any bone debrided, should be sent for bone culture and microbiological assessment. The extrusion of a bone fragment (sequestrum) is highly suggestive of underlying osteomyelitis, although the infection may have arrested coincident to the passage of the sequestrum.
“Sausage” digit
The presence of a red, swollen “sausage”-shaped digit is suggestive of osteomyelitis, but can be the result of other foot problems (e.g. fracture).
Differential diagnosis
Differentiating between osteomyelitis and Charcot foot can be difficult. Diagnosis is based on a good history and physical examination, and is assisted by obtaining supplementary investigations such as X-ray, MRI, and possibly isotope white cell and triple phase bone scans. It is important to note that osteomyelitis and Charcot foot frequently occur simultaneously. Osteomyelitis most often affects the forefoot and heel, while Charcot neuro-arthropathy usually affects the forefoot or ankle.
Prophylactic antibiotic use
Antibiotics should be used only in those who have clinical signs of infection (i.e. “mild”, “moderate” or “severe” in the Infectious Disease Society of America [IDSA] infection grading system; see Table 1).
Foot ulcer classification
Various ulcer classification schemes are used. The primary ones are the Wagner score (Wagner, 1981), the University of Texas system (Lavery et al, 1996), the SINBAD (Site, Ischaemia, Neuropathy, Bacterial Infection, and Depth) score (Ince et al, 2008), the PEDIS (Perfusion, Extent/size, Depth/tissue loss, Infection, Sensation) score (Schaper, 2004). The Scottish Care Information – Diabetes Collaboration electronic ulcer management programme is based on the Texas classification system.
Classification of infection
It is recommended that the presence and severity of infection be classified according to the IDSA system (Table 1) or the PEDIS system developed by the International Working Group for the Diabetic Foot.
General principles of antibiotic use
Antibiotic choice is primarily dependent on causative pathogens and epidemiology. However, treatment with antibiotics often needs to be commenced before culture and sensitivity results are available. Thus initial therapy is usually empirical, and based on the local epidemiological information and local susceptibility data. As the pathogens in diabetic foot infections do not vary significantly in different parts of Scotland, the authors offer practical guidance on antibiotic use. These recommendations are, however, subject to circumstances related to local epidemiology and prescribing policy. Direct contact with local specialists may be necessary for advice on specialised use of these, or other, antibiotics.
Initial antibiotic treatment is frequently empirical, based on the presumed pathogen (Table 2). This guidance is of value until microbiological investigations and clinical response shed further light on the nature of the infection, where available. In light of international concern over Clostridium difficile infection associated with certain antibiotics (especially clindamycin, co-amoxiclav, cephalosporins and quinolones) and the risk of meticillin-resistant Staphylococcus aureus (MRSA) infection (associated with co-amoxiclav, cephalosporins, quinolones and macrolides), narrow-spectrum antibiotic therapy should be used wherever possible. C. difficile is a particular risk for people aged >65 years and for inpatients. Adjustment of therapy based on microbiology results, when available and clinical response to empirical therapy is important in the management of these risks.
The choice of antibiotic and the route of delivery should reflect the severity of infection (Table 1). Duration of antibiotic treatment should be adjusted according to the severity of infection and should be guided by monitoring clinical improvement. In general, the duration of antibiotic therapy should be kept to a minimum. Antibiotic therapy is used only to treat evidence of infection, not to heal a wound, which typically takes much longer.
Allergies to antibiotics include skin rashes and anaphylaxis, but do not include minor side-effects such as nausea.
Enterococci, Pseudomonas and anaerobes are frequently isolated from diabetic foot wounds, but often represent colonising, rather than infecting, organisms. If, however, there is a chronic, persisting infection, or they are the predominant organisms, they may represent pathogens and need targeted treatment.
Specific antibiotic guidance
Specific clinical symptoms identified during careful examination shed light on the likely microbiology of a diabetic foot ulcer. From the likely pathology, initial antibiotic therapy can be decided on with some degree of confidence – although there will on occasion be circumstances suggesting different selections. We emphasise the importance of getting good quality specimens (see page 64 for a guidance on specimens for culture) for microbiological investigation, and close liaison with local infection specialists.
This guidance is categorised by the severity of infection, and by whether the person is, or is not, antibiotic-naïve. The primary choice of antibiotic, and alternatives, for use in people with diabetic foot infections are provided and discussed. A summary of the guidance is provided in Table 3.
Mild infection (IDSA) or PEDIS 2 in an antibiotic-naïve person
Likely pathogens
- S. aureus (and sometimes coagulase-negative staphylococci) or beta-haemolytic streptococci.
- If the person has recently been treated with antibiotics, enterococci and gram-negative rods are more likely to be present.
Note
Take a microbiological culture if the presence of an unusual organism is suspected, or if initial treatment fails.
Antibiotics
Primary
- Oral flucloxacillin 1 g four times a day (qds).
– Assuming that the first course of flucloxacillin was given in high dose and for a full 5–7 days, a second course is rarely effective if the first was unsuccessful.
– There is an increased prevalence of resistant organisms after the first exposure to flucloxacillin.
Oral alternatives
- Doxycyline 100 mg twice a day (bd), or
- Clindamycin 300–450 mg qds, if the person is allergic to, or intolerant of, flucloxacillin.
Treatment duration
Treat for 5–7 days. Adjust therapy in light of clinical response and microbiological culture and sensitivity results.
Moderate infection (IDSA) or PEDIS 3 in an antibiotic-naïve person
Likely pathogens
- S. aureus or beta-haemolytic streptococci.
- Obligate anaerobes are often associated with limb ischaemia, gangrene, necrosis or wound odour and must also be addressed.
Note
A good quality microbiological culture can be particularly helpful in this circumstance. See page 63 for guidance on specimens for culture.
Antibiotics
Primary
- Oral flucloxacillin 1 g qds, or
- Intravenous (IV) flucloxacillin 1 g qds.
Oral alternatives
- Co-trimoxazole 960 mg bd, or
- Co-amoxiclav 625 mg three times a day (tds).
Allergic to penicillins
- Clindamycin (oral, 300–450 mg qds; IV, 450–600 mg qds), or
- Co-trimoxazole 960 mg bd.
Addition
- Add oral metronidazole 400 mg tds if anaerobes are suspected (unless using clindamycin, which has anaerobic cover).
Moderate infection (IDSA) or PEDIS 3 in a person who is not antibiotic-naïve
Likely pathogens
- People with chronic infections, especially if they have received antibiotics previously, often have polymicrobial infections, including aerobic Gram-negative bacilli among the flora.
Antibiotics
Primary
- IV co-amoxiclav 1.2 g tds (especially when anaerobes or coliforms are suspected).
Oral switch
- Co-amoxiclav 625 mg tds, or
- Co-trimoxazole 960 mg bd.
Allergic to penicillins
- IV ciprofloxacin 400 mg tds and IV metronidazole 500 mg tds, or
- IV gentamicin1 (monitor serum concentration) and IV metronidazole 500 mg tds.
– Add vancomycin (monitor serum concentration) to either of the above if MRSA infection is suspected. - Choice will depend on the relative risks of C. difficle infection versus those of renal impairment.
Oral switch options if allergic to penicillins
- Oral ciprofloxacin 500–750 mg bd with either:
– Oral metronidazole 400 mg tds, or
– Clindamycin 300–450 mg qds.
Treatment duration
Treat for 5–7 days. Adjust therapy in light of clinical response and microbiological culture and sensitivity results.
Severe infection (IDSA) or PEDIS 4 in an antibiotic-naïve person
Likely pathogens
- S. aureus or beta-haemolytic streptococci.
- Anaerobes, enterobacteriaceae and Pseudomonas aeruginosa may also need to be treated. P. aeruginosa is usually a coloniser rather than being the infecting organism.
Notes
A good quality microbiological culture should virtually always be taken (see page 63 for guidance).
Caution: As Severe or PEDIS 4 infection implies systemic toxicity, it is generally advised that people at this level of infection be admitted to hospital for the initial phase of management.
Antibiotics
Primary
- IV co-amoxiclav 1.2 g tds.
– If necessary, add gentamicin1 5–7 mg/kg once daily or as per local practice.
Allergic to penicillins or concern about renal function
- IV ciprofloxacin 400 mg bd and IV metronidazole 500 mg tds.
– Add vancomycin (monitor serum concentration) if MRSA infection is suspected.
Treatment duration
Treat for a minimum of 10–14 days. Adjust therapy in light of clinical response and microbiological culture and sensitivity results.
Severe infection (IDSA) or PEDIS 4
The following guidance is for the treatment of severe or PEDIS 4 diabetic foot infections in number of circumstances:
- People who have recently received antibiotic therapy (i.e. those who have received antibiotic treatment within the preceding 90 days).
- People with a proven drug-resistant infection.
- Those at risk of a drug-resistant infections (e.g. those who have previous had an MRSA colonisation).
- Those infected with an extended-spectrum beta-lactamase-producing (ESBL) Escherichia coli or Klebsiella spp. If infection with an ESBL pathogen is proven, seek specialist infection advice.
Notes
A good quality microbiological culture should always be taken (see page 63 for guidance).
Caution: As Severe or PEDIS 4 infection implies systemic toxicity, it is generally advised that people at this level of infection be admitted to hospital for the initial phase of management.
Antibiotics
Primary
- IV piperacillin/tazobactam 4.5 g tds,
– Add vancomycin if MRSA infection is suspected (consult pharmacy for dose, but aim for a trough vancomycin level of 15–20 mg/L).
Penicillin allergy
- IV ciprofloxacin 400 mg bd and IV metronidazole 500 mg tds.
Oral switch
- Be guided by microbiology where possible.
– Otherwise try oral ciprofloxacin 500–750 mg bd and metronidazole 400 mg tds.
Treatment duration
Treat for minimum of 10–14 days. Adjust therapy in light of clinical response and microbiological culture and sensitivity results.
MRSA
If continuing MRSA cover is required, and the person can safely be discharged from hospital, the authors suggest either:
- Outpatient or home parenteral antimicrobial therapy (OHPAT) if available (usually
- V teicoplanin).
- Oral linezolid 600 mg bd. Note that treatment beyond 2 weeks’ duration with this agent needs to be monitored closely as it can be associated with bone marrow toxicity (particularly thrombocytopenia or leucopenia) and lactic acidosis, which are usually reversible on discontinuation of the drug.
- Rifampicin 300 mg bd with either:
– Oral doxycycline 100 mg bd,
– Fusidic acid 500 mg tds, or
– Trimethoprim 200 mg bd.
All these agents can be used to treat people allergic to penicillins.
Osteomyelitis
Early, local surgery to excise infected and necrotic bone may help accelerate healing and reduce the length of time that treatment with antibiotics are required in cases of osteomyelitis (Berendt et al, 2008). The exact role of local surgery in treating osteomyelitis remains controversial.
Published studies have shown that antibiotic therapy without surgery can lead to resolution of infection in up to 80% of cases of osteomyelitis (Game and Jeffcoate, 2008). The evidence base for antibiotic choice for these infections is poor. However, the group recommends that, in those people who show evidence of osteomyelitis, and in all cases where infected bone is not resected, the treatments outlined above should be continued for at least 4–6 weeks, or longer if the clinical response is poor.
There is no evidence that IV antibiotic therapy is superior to oral therapy in the treatment of osteomyelitis, although in certain infections, like those caused by MRSA, IV therapy delivered in the OHPAT setting enhances compliance and reduces the duration of hospitalisation. For MRSA-related osteomyelitis there is some evidence that adding rifampicin 600 mg bd or sodium fusidate 500 mg tds to the usual therapies can be beneficial. Rationalisation of therapy should be discussed with an infection specialist.
Tolerability of oral antibiotics during osteomyelitis is a significant issue and therapy may need to be tailored accordingly. Various antibiotics require monitoring of liver function tests, full blood counts and or serum levels. If considering linezolid for the treatment of osteomyelitis, the prescriber must be aware of its unlicensed status for osteomyelitis, and of the risk of bone marrow toxicity, peripheral or optic neuropathy and lactic acidosis. Measurement of serial C-reactive protein and white blood cell counts may assist in determining the course of the osteomyelitis, but should not be taken more than once per week.
