The treatment of diabetic foot infections

Andrew S. Powlson and Anthony P. Coll*

Wolfson Diabetes and Endocrine Clinic, Institute of Metabolic Science, Box 281, Addenbrooke’s Hospital, Hills Road,Cambridge CB2 0QQ, UK

*Corresponding author. Tel: +44-1223-769041; Fax: +44-1223-330598; E-mail: apc36@cam.ac.uk

Successful treatment of infection in the feet of patients with diabetes mellitus remains a challenge. Althoughthe diagnosis of infection remains a clinical decision, presentation in feet rendered insensate from diabeticneuropathy plus co-existing vascular insufficiency means presentation is often atypical. Wounds frequentlyyield polymicrobial growth and differentiating commensal from pathogenic organisms can be difficult;isolates from diabetic foot wounds are often multidrug resistant. Affected patients often have many otherco-morbidities, which not only affect the choice of appropriate antimicrobial regimen but also impedehealing. Further, much contention surrounds the management of osteomyelitis, with the merits and role ofsurgery still undecided. In this review we briefly consider the epidemiology and pathogenesis of diabetic footdisease, before discussing emerging best microbiological practice and how this fits with the multidisciplinaryapproach required to tackle this difficult clinical problem.

Keywords: diabetes, microbiology, osteomyelitis

Diabetic foot disease: a costly problemDiabetes mellitus (DM) is a complex metabolic disorder with anever-increasing prevalence.1 An estimated 2.6 million peopleare thought to have DM in the UK at present, with this likely toreach close to 4 million by 2025.2 Among the many compli-cations associated with diabetes, issues related to foot diseaserepresent a significant and often challenging clinical problem.

The proportion of patients with DM who develop foot infec-tions is �2% per year in the UK, with the annual prevalence ofsuch infections in this population reaching 5%.3 Patients withdiabetic foot disease are over-represented in hospital inpatientpopulations and, despite often prolonged4 and repeated admis-sion, many come to amputation.5 Indeed, diabetes is the mostcommon cause of non-traumatic limb amputation, with a footulcer being an antecedent event in the majority of cases.6

Pathogenesis of diabetic foot diseaseThe development of a foot ulcer invariably involves the conver-gence of several pathological mechanisms.7 Diabetic peripheralsensorimotor neuropathy is a key factor in the majority ofcases. As a result of damage to sensory nerves, minor trauma(from something as seemingly trivial as an ill-fitting pairof shoes or walking barefoot on unfamiliar terrain) can go un-noticed. Neuropathy can also deform the architecture of thefoot to such a degree that joints and digits are placed inmechanically unfavourable positions, making them highly vul-nerable to injury.

Once the skin has been breached, continued mobilization on abroken area impairs the healing process. Inevitably, direct

contiguous spread of microbes on the skin follows on, with colo-nization and infection of superficial and then deeper tissues likelyif the process is allowed to proceed unchecked. Both the healingprocess and the response to infection are further compromisedby vascular insufficiency, which is commonly present in patientsburdened with complications of diabetes. Each pathologicaldriver—inappropriate weight-bearing due to neuropathy, ischae-mia and infection—needs tackling for resolution of the ulcer butthe major focus of this review will be the infection burden.

Presentation and diagnosis of infectionin the diabetic foot

Clinical diagnosis

One major problem facing the management and study of dia-betic foot disease is that there is no unifying standard thatdefines infection. Classic features such as purulence, erythema,pain, swelling and systemic upset can all be present but noneis pathognomonic of infection. Indeed, as a result of peripheralneuropathy, limb-threatening lesions can present entirelywithout pain and the concomitant ischaemia present in manylimbs also means that the florid inflammatory response typicallyseen in a severe soft tissue infection may be greatly muted.

Despite these challenges, a variety of scoring systems for footlesions have been used in the literature. For example, theWagner8 score classifies a wound from grade 0 (pre- orpost-ulcerative lesion) through a range of increasing depth andseverity to grade 5 (whole-foot gangrene). It does not,however, distinguish aetiology. The University of Texas system9

also uses a stepwise progression in grading (0–3) depending

# The Author 2010. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy.

J Antimicrob Chemother 2010; 65 Suppl 3: iii3–9doi:10.1093/jac/dkq299

iii3

by guest on August 12, 2013

http://jac.oxfordjournals.org/D

ownloaded from

upon depth of wound and, additionally, takes account of thepresence of infection and/or ischaemia. In neither classification,however, is there the ability to record infective burden. The classi-fication developed by the Infectious Diseases Society of America(IDSA)10 is a simple, validated system that ranks infectedwounds as mild, moderate or severe (Table 1) and appears tobe a reliable and useful tool for predicting relevant clinical out-comes, such as the need to be admitted to hospital and ampu-tation.11 It is noteworthy that the ‘severe’ category—thatassociated with the highest percentage rate of amputation—isreliant upon assessing systemic clinical (blood pressure, pulse,temperature) and laboratory (acidosis, glucose levels, white cellcount) parameters, indicating that an assessment of thepatient as a whole is as critical as examination of the wound.

Microbiological diagnosis

Once a clinical diagnosis of infection has been made, the nextchallenge is to determine the aetiology so that rational andappropriate therapy can be instigated. Important clues as topotential pathogens can be garnered from the clinical history.Infection in a very recently acquired superficial ulcer is likely tobe due to aerobic Gram-positive cocci (for example staphylo-cocci) while a long duration of ulceration, increased depth/sever-ity and antecedent therapy are likely to increase the chances ofthe wound yielding both polymicrobial growth (often comprisingaerobic and anaerobic Gram-positive and Gram-negative bac-teria) and resistant organisms (Figure 1).12

Because the skin has an indigenous microbial flora, great caremust be taken not only in obtaining a sample for microbiologicalanalysis but also in interpreting the result. Invariably, a positivelaboratory culture can only suggest the presence of infectionand must be correlated with clinical findings. A simple superficialswab across the ulcer runs the great risk of isolating commensalorganisms unrelated to the infection, leading to misdirectedtreatment against an innocent bystander rather than focusingon the pathogenic culprit. When used, swabs should always betaken from the base of a lesion after debridement of superficialslough and necrosis. Ideally, biopsy of tissue from a debridedulcer base or, if applicable, extruded bony fragments, offersmore useful material for microbiological analysis.

Even when such best practice is observed, there remains adegree of subjectivity in analysis. Preliminary data employing

molecular biology are beginning to emerge that suggest it maybecome possible not only to differentiate more objectively infec-tion from non-infection but also to determine the virulence ofinfecting pathogens and hence predict clinical outcome. Forexample, Sotto et al.13 used a PCR-based approach to detectthe presence of 31 of the most prevalent virulence-associatedgenes in Staphylococcus aureus isolated from diabetic footulcers. Logistic regression showed that five markers (cap8, sea,sei, lukE and hlgv) had utility in distinguishing non-infectedfrom infected ulcers and were able to predict wound status atfollow-up. Although this approach has some way to go beforefinding its way into routine clinical practice, it does open upthe possibility that a relatively easy multiplex PCR might beable to identify at an early stage ulcers that have the potentialto become clinically difficult and highlight the need for a moreaggressive approach to be taken early on.

Responding to the laboratory result

There are no data to validate the treatment of culture-positivewounds when the clinical appearance is not one of infection.Indeed, responding solely to laboratory findings without refer-ence to the clinical scenario puts patients at unnecessary riskof antibiotic-associated side effects and is likely to increase theprevalence of drug-resistant organisms. Current best practice14

advocates immediate treatment in the face of clinical infec-tion, with initial therapy directed against the ‘most likely’pathogens, aiming for revision of this empirical choice on thebasis of culture results and clinical progress. Thus Gram-positive cocci are invariably involved in mild and moderateinfection and should respond to antistaphylococcal therapiessuch as semi-synthetic penicillins (e.g. flucloxacillin), whilerecent antecedent therapy increases the likelihood of Gram-negative bacilli (GNB) also being involved, necessitating theaddition of b-lactam/b-lactamase inhibitor combinations orfluoroquinolones.

Oral agents are adequate for mild and moderate infections.However, severe infections invariably require intravenoustherapy covering Gram-positive cocci [with a low threshold touse a glycopeptide to cover the possibility of infection withmethicillin-resistant S. aureus (MRSA)], GNB and anaerobes.Moreover, it is our view that a severe infection with clinical evi-dence of systemic involvement, such as spiking temperature or

Table 1. IDSA classification of wound infection (modified from reference 11)

Mild Moderate Severe

More than two of: Patient is systemically well andmetabolically stable but has any of:

Infection in a patient with systemictoxicity or metabolic instability

† Purulence, erythema, pain, tenderness, warmth orinduration

† Any cellulitis/erythema extending ≤2 cm around ulcer† Infection is limited to skin/superficial subcutaneous tissues† No local complications or systemic illness

† Cellulitis extending .2 cm† Lymphangitis† Spread beneath fascia† Deep tissue abscess† Gangrene† Muscle, tendon, joint or bone

involved

Powlson and Coll

iii4

by guest on August 12, 2013

http://jac.oxfordjournals.org/D

ownloaded from

rigors, should trigger evaluation by an experienced surgeon. Acollection within the foot may be best served by surgical inter-vention to limit further spread and tissue loss, reduce theseptic burden on the patient and allow collection of appropriatematerial for microbiological investigation, which should optimizethe likelihood of pathogen detection and initiation of correctlytargeted treatment at the earliest possible juncture.

Drug-resistant organisms and diabeticfoot infectionsDrug-resistant organisms are over-represented in samplesobtained from diabetic foot ulcers. A study from a secondarycare multidisciplinary diabetic foot clinic in Melbourne, Australiais typical of many such units,15 in that patients attending hada high prevalence of MRSA. Of 653 specimens from 379 patients,MRSA was isolated from 23%. Chronicity of ulcer, previousin-patient care and chronic kidney disease (CKD) each indepen-dently predisposed patients to MRSA infection.

But does isolation of a drug-resistant organism matter?A French study in 200816 looked to determine the impact ofmultidrug-resistant (MDR) organisms on healing rates 1 yearafter discharge in 188 patients admitted for an infected footulcer. Organisms identified included MRSA, Enterobacteriaceaeresistant to third-generation cephalosporins, Pseudomonasaeruginosa resistant to two antibiotics from among ticarcillin,ciprofloxacin, ceftazidime and imipenem, and Enterococcus spp.resistant to glycopeptides. The complexity of the ulcers in thestudy can be judged by the fact that two-thirds were graded asmoderate or severe, .70% were judged to be neuroischaemiculcers and around a quarter of the lesions yielded resistant organ-isms. The presence of MDR organisms was associated with ahigher incidence of lower-limb amputation (35.6% versus 11.2%in non-MDR infection), with the majority (87.5%) of these amputa-tions being minor. After adjustment for other factors, however,multivariate analysis indicated that the presence of MDR bacteria

did not affect healing time. This is in keeping with a similar studydirected by investigators in Nottingham UK, where Game et al.17

found no evidence that the presence of MRSA was associatedwith a significant difference in healing time or percentage ofhealed wounds. In the case of the French study,16 this may be adirect consequence of prescribing practice in the unit in questionbeing vigilant to the high prevalence of MDR organisms seenlocally, with the first-line antibiotic regimen frequently includingan agent active against MRSA given immediately after obtainingspecimens for culture, to be reassessed thereafter according tothe microbiological results.

Guidelines for therapyOne of the major difficulties in developing guidelines for therapyis the lack of randomized controlled trial evidence to supportthem.18 The lack of a standardized objective definition of infec-tion and its subsequent eradication, the heterogeneous influ-ences from variation in vascular supply, position and durationof ulceration, the variability in strategies to prevent weightbearing through the affected tissue and the diverse multipleco-morbidities invariably found in patients with diabetic footdisease are just some of the many factors that conspire tomake study design challenging. Industry is also unlikely to offerfinancial support to evaluate the comparative efficacy ofgeneric antibiotic agents long-since off patent. In addition, anyguidelines relating to the use of antimicrobial therapy mustfactor-in local susceptibility and resistance data. Thus, in theface of an increasing clinical problem, there remains unsatisfac-tory uncertainty about the optimal route and effective durationof antimicrobial treatment. At present, the situation is one ofconsensus agreement from multidisciplinary panels of expertsassimilating other published guidelines and the limited availablesupport from published studies, rather than guidance backed byrigorous trial data. One such example is that of the Scottishworking group19 who contend that local differences across the

(a) (b)

Figure 1. Case History. A 38-year-old man with long-standing poorly controlled diabetes presented with a 4 month history of a painless red swollenfoot. The problem first appeared 3 months after the appearance of a small neuropathic ulcer under the second metatarsal head. Repeated courses ofantibiotics in the community reduced the severity of the swelling but this always recurred on cessation of therapy. (a) Initial plain X-ray showed loss ofthe head of the second metatarsal and periosteal reaction around the proximal phalanx of the second toe (arrows). (b) An MRI showed a 26×17 mmcollection on the dorsum of the foot (arrows) with some destruction of the second proximal phalanx. At operation the head of the second metatarsalwas found to be purulent and was removed along with the phalanges of the second toe. Analysis of biopsied tissue showed Citrobacter spp. andcoagulase-negative staphylococci, resistant to flucloxacillin. The patient received 4 weeks of intravenous vancomycin and oral ciprofloxacin postoperatively with full resolution of clinical signs.

Diabetic foot infections

iii5

JAC

by guest on August 12, 2013

http://jac.oxfordjournals.org/D

ownloaded from

country in causative organisms and their susceptibility to specificagents are insignificant enough to make such a document validin most instances. This approach has much merit, emphasizingas it does the importance of a multidisciplinary approach tomanagement, the categorization of infection on clinicalgrounds and the acknowledgement that, although culture-supported specific targeted therapy is the gold standard, theremust be a role for initial empirical treatment.

OsteomyelitisThe significance of diagnosing osteomyelitis in diabetic foot infec-tions is clear; it increases refractoriness to antibiotic treatmentand is associated with an elevated risk that lower-limb amputa-tion, either minor or major, will subsequently be required.20,21

The correct determination of the presence or absence of osteo-myelitis therefore carries high clinical importance. In theory, thediagnosis should be an easy one to make since, put simply, osteo-myelitis is infection of bone. In practice, however, it remains achallenging condition. The gold standard for diagnosis is theculture of pathogenic organisms from an isolated bone culturein the presence of characteristic histological changes suggestingnecrosis, inflammation and reparative processes. However, suchbiopsies are infrequently taken and are not infallible: falsenegatives can result from poor sampling or samples being takenafter antibiotic therapy; false-positive microbiology can be intro-duced by contamination and other inflammatory conditions canmimic the histological findings. A diagnosis is more usuallymade on clinical suspicion (refractory infections, bone visible orpalpable on probing the ulcer, failure to respond to antibiotics)in conjunction with one or more imaging modalities.

Plain radiographs are readily available in most centres andrepresent the first imaging step if osteomyelitis is suspected.The films, ideally weight-bearing and in two projections, maydemonstrate the discrete areas of radiolucency, cortical destruc-tion and loss of trabecular patterning found in early osteomyel-itis. Periosteal reaction, sclerosis and osteogenesis may follow(Figure 1). However, radiographic features only typically start toappear after at least 2 weeks and can be mimicked by other pro-cesses such as Charcot neuroarthropathy or trauma.10 If theplain radiograph findings are equivocal, magnetic resonanceimaging (MRI) should be considered as it is the most sensitiveand specific imaging modality in diabetic foot infections for diag-nosing and assessing the extent of both bone and soft tissueinvolvement. Osteomyelitis is seen as areas of decreased signalon T1 imaging/high signal on T2 imaging and MRI can alsodemonstrate intraosseous abscess formation and breaching ofthe cortex. Soft tissue involvement can be further evaluatedwith the use of gadolinium enhancement. Diagnostic yield isreliant on experienced radiological reporting but provides a sen-sitivity and specificity of �90% and �80%, respectively.22 Assuch, MRI can prove extremely valuable in informing diagnosis,extent and thus management of osteomyelitis in diabetic footinfections. CT can also prove superior to plain radiography indemonstrating the described bony radiological features of osteo-myelitis. However, MRI gives superior imaging of the soft tissuesand is the preferred modality. Radionuclide bone scanning andwhite cell scanning also have their place when the diagnosis isunclear but a full review of their role lies outside the remit ofthis review.23

Consensus criteria for diagnosisof osteomyelitisAn established consensus view on the diagnosis of osteomyelitis isyet to emerge but is an important aim both for informing treat-ment and to enable meaningful comparison of managementstrategies in determining best practice. Recent work under theauspices of the International Working Group on the DiabeticFoot (IWGDF) has attempted to formulate a criteria-basedapproach to diagnosis in the manner already accepted for simi-larly diagnostically challenging conditions such as infective endo-carditis or various rheumatological conditions.24 This aims toclassify osteomyelitis as definite, probable, possible or unlikely,based on a combination of clinical, microbiological and radiologi-cal features as well as response to antimicrobial therapy(Figure 2). Each criterion is in turn weighted as indicating eitherpossible or probable infection. The decision to treat for osteomyel-itis is then based on this probability. This approach has not beenvalidated in practice but represents a quantum leap forwardtowards standardizing diagnosis and the decision to treat.

Management of osteomyelitisOnce the diagnosis has been made and the decision to treattaken, there is currently no evidence-based consensus view onthe best way to manage osteomyelitis in the diabetic foot.Broadly, the approach is either conservative or surgical. Thechoice of treatment is all too often based on local prejudicesand facilities with surprisingly little evidence to drive the develop-ment of robust guidelines. For many years, harking back to thedays of limited antibiotic availability and effectiveness, the surgi-cal removal of infected bone was held to be paramount. Morerecently, a primarily medical approach has been advocated,with the view that appropriate antibiotic and adjunctive treat-ments can obviate the need for surgery. So who is correct? Norandomized controlled trials addressing this question exist. Thegroup responsible for postulating the consensus criteriaapproach to diagnosis described above went on to undertake asystematic literature review of the effectiveness of osteomyelitistreatment in the diabetic foot and Fran Game of the NottinghamFoot Ulcer Trials Unit also recently evaluated this question. Bothdrew similar conclusions: there is little evidence to support thecase for early surgical intervention but, equally, there are norobust comparative data to conclude that there is a significantdifference in outcome after either primarily surgical or conserva-tive management.24,25

Role of surgery

A handful of limited observational studies do support the casefor surgery. ‘Successful’ treatment after surgery is reported in�60%–90% of cases though how this is defined is somewhatnebulous.24 One retrospective series analysis suggested thatsurgical intervention within the first 3 days reduced the sub-sequent need for lower-limb amputation (10/77 patients)when compared with initial therapy with antibiotics alone (24/87).26 Henke et al.27 also report that delaying surgery leadsto an increased risk of major amputation. Others indicate thatlimited surgical intervention improves the degree of healing

Powlson and Coll

iii6

by guest on August 12, 2013

http://jac.oxfordjournals.org/D

ownloaded from

and limits the required degree of antibiotic therapy usage whencompared with conservative management only.28 Aragon-Sanchez et al.29 advocate the use of early minimal debridementin conjunction with antibiotics and subsequent minor or majoramputation as required. All of these studies are limited obser-vational series and the inclusion criteria, indications forsurgery and endpoint measures are ill defined and vary con-siderably. A pragmatic view would consider that there will becases where surgery is absolutely required (e.g. a large collec-tion of pus in a patient with systemic signs of sepsis) butthere is not convincing evidence to mandate early surgery inall cases or sufficient clarity in the literature to adequatelydefine when surgery becomes a necessity. As a minimum,consideration is needed to determine whether an interventionhas an acceptable chance of resulting in a sound, weight-bearing unit with a low chance of ulceration in the future. Wealso believe antibiotic therapy alongside surgical treatmentremains obligatory.

Medical treatment

Conservative management, based primarily on antibiotictherapy alone, was found to eradicate diabetic foot osteomyel-itis in .60% of cases when 500 case reports were reviewed.30

A series of patients treated this way in a single centre31 showedarrest of the infection in 80% of cases, though 37% of their

total of 147 cases of osteomyelitis did in fact undergo earlysurgery, suggesting that there remains a clinical judgementcall to be made in all presentations. Major amputation ratesand relapse rates have been shown to be equivalent betweenprimarily conservative management and where initial minorsurgery is used.29,31 Medical management is not without itsproblems. The mean duration of antibiotic therapy in theGame series was 61 days while others have suggested thattreatment for .6 months is required in up to 50% of cases.32

Such prolonged courses of antibiotics potentiate the sideeffects, discourage compliance and encourage the develop-ment of drug-resistant organisms. It is, therefore, a priority tooptimize antibiotic therapy. Surgery may be one route to redu-cing the need for antibiotics as suggested above. There is littleevidence, however, to inform choice of antibiotic or route ofadministration. Antistaphylococcal agents and broad-spectrumantibiotics have been shown to be equally efficacious and,despite the frequent preference for parenteral administrationin osteomyelitis, there is little evidence to support its superiorityover oral therapy.

In the absence of high-quality comparative data, we suggestthat each presentation must be evaluated on its own clinicalmerits and that the need for surgery should be judged through-out the course of treatment by an experienced clinical team, inthe knowledge that antibiotic therapy alone may prove effectivein some cases.

DEFINITE

• positive bone culture and positive histology

• pus in bone at surgery

• atraumatically detached bone fragment removed from ulcer

• intraosseous abscess on MRI

PROBABLE

• Visible cancellous bone in ulcer

• MRI showing bone oedema with other signs of osteomyelitis

• Bone sample with positive culture but negative/absent histology

• Bone sample with positive histology but negative/absent culture

POSSIBLE

• Plain X-rays show cortical destruction

• MRI shows bone oedema OR cloaca

• Probe to bone positive

• Visible cortical bone

• ESR >70 mm/h with no other plausible explanation

• Non-healing wound despite adequate off-loading, and perfusion for >6 weeks OR ulcer of >2 weeks

duration with clinical evidence of infection

SCORE

1 definite Treat

Consider treating but may need

further investigation

Further investigation advised

before treatment

• 2 probable OR

• 1 probable + 2 possible OR

• 4 possible

2 possible

PROBABILITY MANAGEMENT

>90%

50%–90%

10%–50%

Figure 2. Diagnostic criteria to guide the management of osteomyelitis in patients with diabetic foot disease (modified from Berendt et al.24).

Diabetic foot infections

iii7

JAC

by guest on August 12, 2013

http://jac.oxfordjournals.org/D

ownloaded from

Beyond antimicrobial therapy in diabeticfoot infectionsSuccessful treatment of infection in a diabetic foot often requiresinterventions beyond the prescription of even correctly targetedantimicrobial therapy. Appropriate and aggressive attentionneeds to be given to the wound bed and the surroundingtissue. This includes removal of necrotic or poorly vascularizedtissue from the surface and edges of the wound, which maybe done in a clinic environment by sharp debridement, but ifthe nature of the wound dictates, may require a more invasiveprocedure in the controlled environment of the operatingtheatre. A major factor in the failure of chronic wounds to healis the continued, unchallenged trauma to the wound bed as aresult of continued weight bearing in an insensate neuropathicfoot. Effective redistribution of this pathological pressure is a criti-cal adjunct to successful healing This may be as intensive as atotal contact plaster cast to redistribute the force but can beas simple as instigating a non-weight bearing policy on theaffected foot. A lesion occurring on a limb with clinical evidenceof impaired blood flow mandates consideration of revasculariza-tion, via either bypass or endovascular intervention. This is particu-larly important with infection in the presence of arterial disease, aspenetration of antibiotic into infected tissue may be poorer.33 It isworthwhile highlighting the sinister combination of ischaemia andinfection. In an Italian study looking at 564 patients presentingwith critical limb ischaemia, two-thirds of them also had clinicalevidence of infection. Despite 95% of the patients undergoing arevascularization procedure, at follow-up 6 years later, 50% ofthe patients were dead.34 The fact that coronary artery diseasewas the leading cause of death is also testament to the fact thatdiabetic foot disease has a predilection to occur in patientsalready overburdened with disabling co-morbidities, in particularcardiovascular pathologies.35

Successful delivery of these key interventions can only bedone with the involvement of a multidisciplinary team, whichrequires the expertise of a whole range of professionals rangingfrom diabetologist, orthopaedic surgeon and specialist podiatristright through to orthotist, diabetes specialist nurse and plastertechnician. The concept of a team approach is core to recentdocumentation produced by a Diabetes UK specialist foot careservices working group,36 in which effective integration of theinput of different healthcare professionals is highlighted asbeing essential for effective management of disease of thefoot in diabetes.

The lack of robust validated protocols to guide treatment ofinfection in the diabetic foot can engender a feeling of thera-peutic nihilism, but to take such a stance risks unjustly dismissingbeneficial interventions and condemning a fragile patient groupto undeserved suffering and debility. We welcome the recentinitiative supported by the UK National Institute for Health andClinical Excellence (NICE) to develop clinical practice guidelineson the inpatient management of diabetic foot problems37 andlook forward to the results of properly constructed trials toinform treatment of this difficult clinical problem.

Transparency declarationsThis article is part of a Supplement sponsored by the BSAC.

References1 Gonzalez ELM, Johansson S, Wallander M-A et al. Trends in theprevalence and incidence of diabetes in the UK: 1996–2005.J Epidemiol Comm Health 2009; 63: 332–6.

2 Yorkshire and Humber Public Health Observatory. Browsing PBSDiabetes Population Prevalence Model Phase 3 – Collection. http://www.yhpho.org.uk/resource/view.aspx?RID=9906 (17 May 2010, datelast accessed).

3 Kumar S, Ashe HA, Parnell LN et al. The prevalence of foot ulcerationand its correlates in type 2 diabetic patients: a population-based study.Diabet Med 1994; 11: 480–4.

4 Lawrence SM, Wraight PR, Campbell DA et al. Assessmentand management of inpatients with acute diabetes-relatedfoot complications: room for improvement. Intern Med J 2004; 34:229–33.

5 Currie CJ, Morgan CL, Peters JR. The epidemiology and cost of inpatientcare for peripheral vascular disease, infection, neuropathy, and ulcerationin diabetes. Diabetes Care 1998; 21: 42–8.

6 Pecoraro RE, Reiber GE, Burgess EM. Pathways to diabetic limbamputation. Basis for prevention. Diabetes Care 1990; 13: 513–21.

7 Edmonds ME. The diabetic foot: pathophysiology and treatment. ClinEndocrinol Metab 1986; 15: 889–916.

8 Wagner FW Jr. The dysvascular foot: a system for diagnosis andtreatment. Foot Ankle 1981; 2: 64–122.

9 Lavery LA, Armstrong DG, Harkless LB. Classification of diabetic footwounds. J Foot Ankle Surg 1996; 35: 528–31.

10 Lipsky BA, Berendt AR, Deery HG et al. Diagnosis and treatment ofdiabetic foot infections. Clin Infect Dis 2004; 39: 885–910.

11 Lipsky BA, Polis AB, Lantz KC et al. The value of a wound score fordiabetic foot infections in predicting treatment outcome: a prospectiveanalysis from the SIDESTEP trial. Wound Repair Regen 2009; 17: 671–7.

12 Citron DM, Goldstein EJ, Merriam CV et al. Bacteriology ofmoderate-to-severe diabetic foot infections and in vitro activity ofantimicrobial agents. J Clin Microbiol 2007; 45: 2819–28.

13 Sotto A, Lina G, Richard JL et al. Virulence potential of Staphylococcusaureus strains isolated from diabetic foot ulcers: a new paradigm.Diabetes Care 2008; 31: 2318–24.

14 Lipsky BA. New developments in diagnosing and treatingdiabetic foot infections. Diabetes Metab Res Rev 2008; 24 Suppl 1:S66–71.

15 Yates C, May K, Hale T et al. Wound chronicity, inpatient care, andchronic kidney disease predispose to MRSA infection in diabetic footulcers. Diabetes Care 2009; 32: 1907–9.

16 Richard JL, Sotto A, Jourdan N et al. Risk factors and healing impact ofmultidrug-resistant bacteria in diabetic foot ulcers. Diabetes Metab 2008;34: 363–9.

17 Game FL, Boswell T, Soar C et al. Outcome in diabetic foot ulcers withand without methicillin resistant Staphylococcus aureus (MRSA). DiabetMed 2003; 20 Suppl 2: 30.

18 Jeffcoate WJ, Lipsky BA, Berendt AR et al. Unresolved issues in themanagement of ulcers of the foot in diabetes. Diabet Med 2008; 25:1380–9.

19 Leese G, Nathwani D, Young M et al. Use of antibiotics in people withdiabetic foot disease: a consensus statement. Diabetic Foot J 2008; 12:62–74.

20 Lipsky BA. Osteomyelitis of the foot in diabetic patients. Clin Infect Dis1997; 25: 1318–26.

Powlson and Coll

iii8

by guest on August 12, 2013

http://jac.oxfordjournals.org/D

ownloaded from

21 Lavery LA, Armstrong DG, Wunderlich RP et al. Risk factors forfoot infections in individuals with diabetes. Diabetes Care 2006; 29:1288–93.

22 Butalia S, Palda VA, Sargeant RJ et al. Does this patient with diabeteshave osteomyelitis of the lower extremity? JAMA 2008; 299: 806–13.

23 Teh J, Berendt T, Lipsky BA. Rational Imaging. Investigating suspectedbone infection in the diabetic foot. BMJ 2009; 339: b4690.

24 Berendt AR, Peters EJ, Bakker K et al. Diabetic foot osteomyelitis: aprogress report on diagnosis and a systematic review of treatment.Diabetes Metab Res Rev 2008; 24 Suppl 1: S145–61.

25 Game F. Management of osteomyelitis of the foot in diabetesmellitus. Nat Rev Endocrinol 2010; 6: 43–7.

26 Tan JS, Friedman NM, Hazelton-Miller C et al. Can aggressivetreatment of diabetic foot infections reduce the need for above-ankleamputation? Clin Infect Dis 1996; 23: 286–91.

27 Henke PK, Blackburn SA, Wainess RW et al. Osteomyelitis of the footand toe in adults is a surgical disease: conservative managementworsens lower extremity salvage. Ann Surg 2005; 241: 885–92;discussion 92–4.

28 Ha Van G, Siney H, Danan JP et al. Treatment of osteomyelitis in thediabetic foot. Contribution of conservative surgery. Diabetes Care 1996;19: 1257–60.

29 Aragon-Sanchez FJ, Cabrera-Galvan JJ, Quintana-Marrero Y et al.Outcomes of surgical treatment of diabetic foot osteomyelitis: a series of185 patients with histopathological confirmation of bone involvement.Diabetologia 2008; 51: 1962–70.

30 Jeffcoate WJ, Lipsky BA. Controversies in diagnosing and managingosteomyelitis of the foot in diabetes. Clin Infect Dis 2004; 39 (Suppl 2):S115–22.

31 Game FL, Jeffcoate WJ. Primarily non-surgical management ofosteomyelitis of the foot in diabetes. Diabetologia 2008; 51: 962–7.

32 Valabhji J, Oliver N, Samarasinghe D et al. Conservative managementof diabetic forefoot ulceration complicated by underlying osteomyelitis:the benefits of magnetic resonance imaging. Diabet Med 2009; 26:1127–34.

33 Raymakers JT, Houben AJ, van der Heyden JJ et al. The effect ofdiabetes and severe ischaemia on the penetration of ceftazidime intotissues of the limb. Diabet Med 2001; 18: 229–34.

34 Faglia E, Clerici G, Clerissi J et al. Long-term prognosis of diabeticpatients with critical limb ischemia: a population-based cohort study.Diabetes Care 2009; 32: 822–7.

35 Prompers L, Huijberts M, Apelqvist J et al. High prevalence ofischaemia, infection and serious comorbidity in patients with diabeticfoot disease in Europe. Baseline results from the Eurodiale study.Diabetologia 2007; 50: 18–25.

36 Putting feet first; Commissioning specialist services for the managementand prevention of diabetic foot disease in hospitals, June 2009. http://www.diabetes.org.uk/Documents/Reports/Putting_Feet_First_010709.pdf (30March 2010, date last accessed).

37 Diabetic foot problems - inpatient management: draft scope forconsultation, Feb 2010. http://www.nice.org.uk/guidance/index.jsp?action=folder&o=47369 (30 March 2010, date last accessed).

Diabetic foot infections

iii9

JAC

by guest on August 12, 2013

http://jac.oxfordjournals.org/D

ownloaded from