29. Cavanagh PR, Ulbrecht JS. Clinical plantar pressure measurement in diabetes: rationale and methodology. Foot 1994; 4: 123±35. 30. Cavanagh PR, Ulbrecht JS, Caputo GM. Biomechanics of the foot in diabetes mellitus. In Bowker JH, P®efer M (eds), The Diabetic Foot, 6th edn. Philadelphia, PA: WB Saunders, 2000. Foot Biomechanics 59 5 Classi®cation of Ulcers and Its Relevance to Management MATTHEW J. YOUNG Royal In®rmary, Edinburgh, UK The management of diabetic foot ulceration is multidisciplinary in its most effective form, and requires communication between primary and secondary care providers. In addition, the increasing role of research- based practice, audit and clinical effectiveness in the provision of managed health care systems means that accurate and concise ulcer description and classi®cation models are required to improve interdisciplinary collabora- tion and communication and to allow meaningful comparisons between and within centres 1 . The classi®cation of an ulcer should delineate a single type of ulcer with de®nable characteristics which are distinct from other ulcer categories. Examples of potential classi®cation systems are detailed below. They are often related to the risk factors which led to the ulcer and, in at least two cases, they do not use any of the descriptive characteristics of the ulcer to categorize it. As well as being a basis for clinical care, a classi®cation should provide a guide to prognosis and should facilitate audit and research. A good example is the classi®cation of ulcers by their suspected aetiology, such as neuropathic or neuro-ischaemic, or by their perceived severity, for example, super®cial or deep. The classi®cation of an ulcer should be applied once, based on the initial characteristics, and should not alter with the progress of therapy. A description is based upon de®nable character- istics but differs from a classi®cation in that it applies to the ulcer at the exact moment it is seen. It is therefore ephemeral, changing with the progression of the ulcer. It is important to make the distinction between The Foot in Diabetes, 3rd edn. Edited by A. J. M. Boulton, H. Connor and P. R. Cavanagh. & 2000 John Wiley & Sons, Ltd. The Foot in Diabetes. Third Edition. Edited by A.J.M. Boulton, H. Connor, P.R. Cavanagh Copyright  2000 John Wiley & Sons, Inc. ISBNs: 0-471-48974-3 (Hardback); 0-470-84639-9 (Electronic) classi®cation and description of an ulcer. In the future, digital imaging and image transmission may make such systems easier, but at present descriptions are an essential part of working practice. Whilst descriptive terms such as ``uninfected'' or ``infected'' might be used to classify ulcers, most descriptive terms do not lend themselves to a classi®cation with workable numbers of categories and are, therefore, not a basis for auditing the outcome of ulceration or for classifying an ulcer. However, descriptions are very useful in prompting adjustments to ongoing treatment as the nature of the ulcer changes. They are also essential to ensuring that health care professionals can communicate referrals and handover of care in an unambiguous way. Such referrals also need to include patient character- istics other than those of the ulcer. Where such characteristics have been shown to be important in prognosis they are also discussed below. CATEGORIES FOR CLASSIFICATION AND DESCRIPTION OF ULCERATION Location of the Ulcer Ulceration of the lower limb in diabetic patients can occur at any site. However, since the aetiology and treatment of leg ulceration above the ankle is usually different from foot ulceration, this chapter will not discuss this further. It is essential to describe the site of ulceration, as this will often give clues to the cause and often the underlying aetiology for the purpose of guiding therapy. Toe ulceration is often directly shoe-induced; ulceration on the remainder of the foot is often multifactorial. Plantar ulceration is classically neuropathic; marginal ulcers are more commonly associated with ischaemia 2 . In addition, toe ulceration is signi®cantly associated with amputation 3 . Therefore, the location of ulceration can also give a guide to prognosis, although this effect is less signi®cant than the aetiology of the ulcer overall, or Meggitt±Wagner 4,5 grade (see below), irrespective of site 6 . Ulcers which occur in association with signi®cant foot deformity are rarely characterized separately from other ulcers. Deformity forms the basis of a number of foot ulcer risk scoring systems, but once the ulcer has formed, deformity receives little attention as a guide to treatment or prognosis. Only May®eld et al 7 have clearly identi®ed deformity as an additional risk factor for amputation, but this was as part of a pre-ulceration risk strati®cation and not as a direct result of classifying ulcers. Despite this, ulceration and deformity continue to be reported anecdotally in many foot clinics, especially in association with neuro-arthropathy and rocker-bottom foot. 62 The Foot in Diabetes Size and Extent of Ulceration The size of an ulcer, usually de®ned as either two diameters at right angles, or as surface area, is an important descriptive term. Without serial measurements of ulcer size, it is impossible to document change in any meaningful way; therefore, size measurements should be mandatory for all ulcers. However, there is less evidence that ulcer size is a guide to management or prognosis, and none of the widely applied classi®cations of foot ulceration uses ulcer size as a discriminator. Indeed, a recent meta- analysis of wound healing studies showed an absence of effect of ulcer size on prognosis in neuropathic ulceration. 8 The volume of an ulcer is currently almost impossible to assess. However, ulcer depth, either measured or, more commonly, simply described, is an important factor in both descriptive and classi®cation systems. Exposure of bone and tendon is a feature of all classi®cations derived from the Meggitt± Wagner classi®cation. 4,5 The use of sterile blunt probes to fully explore the extent of an ulcer is a useful tool to identify bone and deep tissue involvement in ulcers that do not appear to be extensive upon initial inspection. Probing to bone was shown to identify osteomyelitis with a positive predictive value of 89% in one series 9 . The identi®cation of deep tissue involvement, and in particular deep infection or osteomyelitis, is strongly associated with an increased risk of major amputation 10 ; therefore, probing should be performed in all but the most obviously super®cial ulcers. Aetiology In many classi®cation systems the categorization of foot ulceration is based logically upon the aetiological factors. The management of ulceration has common features, namely pressure relief, debridement and infection control, although these vary depending on the nature of the ulcer. Patients with neuropathy who develop foot ulcers have a signi®cantly better prognosis than patients with vascular insuf®ciency. The simple absence of pulses doubles amputation risk 11 ; ankle pressure indices are lower in patients who have had or will have amputations 12 ; transcutaneous oxygen tensions are associated with delayed healing and amputation if less than 30 mmHg 13 ; and the number of lesions detected on peripheral arteriograms is directly proportional to amputation risk 14 . Therefore, it is clearly important to identify vascular insuf®ciency, so that revascularization can be attempted where appropriate. Even in the absence of these criteria, an ulcer which is not healing despite optimal care should be investigated for vascular insuf®ciency. The coexistence of neuropathy in patients with peripheral vascular disease 15 has led to the use of the term ``neuro-ischaemic foot'' and at least Relevance of Ulcer Classi®cation to Management 63 one classi®cation is based on this distinction 2 . Some patients with peripheral vascular disease do have intact peripheral sensation, which is manifest as rest pain or as pain during ulcer debridement or in the presence of infection. Pain is, itself, an independently poor prognostic indicator in patients with diabetic foot ulceration 11 . However, given the relative paucity of purely ischaemic lesions in diabetic patients and the frequency of coexisting sensory or motor neuropathy, the term ``neuro-ischaemic'' is probably a good one for these patients and will be referred to again later in this chapter. The presence of gangrene is the signi®cant turning point in the Meggitt± Wagner classi®cation system 4,5 , separating the primarily neuropathic from the primarily ischaemic foot. However, the realization that localized gangrene in the toes can occur as a result of infective vasculitis in a foot with normal peripheral pulses highlights the fact that this may be an unduly simplistic approach. The presence of tissue necrosis and gangrene in infected feet should not be taken to imply failure of peripheral circulation without other supporting evidence. Resection of infected tissue necrosis or toe auto-amputation may allow a foot to heal without surgical amputation in an otherwise well-perfused limb. Extensive gangrene, from either peripheral arterial occlusion or infection, is usually a precursor to major amputation, regardless of aetiology. However, it is not clear how much gangrene must be present for it to be de®ned as extensive. Whilst it might appear clinically obvious when a foot needs amputating, the wide disparity in amputation rates between centres suggests that a stricter de®nition might be required. Infection As has been implied above, infection has a signi®cant adverse effect on the diabetic foot with ulceration. Unfortunately, in many cases it is very dif®cult to detect infection in diabetic foot ulcers and to gauge its extent and severity. Few of the descriptive or classi®cation systems that include infection as a parameter give any de®nition as to what constitutes infection. Bacterial colonization of diabetic foot ulcers is the norm in bacteriological surveys and yet it is generally accepted that the classical signs of in¯ammation that typify infective processes elsewhere are signi®cantly reduced in the diabetic foot. For this reason, whilst some regard the presence of bacteria as insigni®cant in the absence of signs of infection, many advocate treating all ulcers as if potentially super®cially infected and use systemic antibiotics in most, if not all patients 16 . However, when osteomyelitis is present, most clinicians, especially surgeons, will advocate surgery 17 , although two recent papers have reported good outcomes with conservative management 18,19 and this approach should be probably used more frequently. 64 The Foot in Diabetes Even extensive infection may be dif®cult to detect. The presence of swelling, heat and pain could indicate a neuro-arthropathic foot (although it is more common to make the converse error). Even if infection is present, there may be little or no supporting systemic features, such as fever or raised white cell count 17,20 . Even the erythrocyte sedimentation rate can be normal. Features such as lymphangitis, frank pus and foul drainage suggest that a foot is severely infected. Osteomyelitis is also dif®cult to detect in the diabetic foot. The typical systemic features of infection may be absent, and radiological and other imaging techniques may be inconclusive or misleading (see Chapters 15 and 17). Therefore, it is important to have a high index of suspicion, to use the probe-to-bone test, and to examine serial radiographs of deep ulcers, which take a long time to heal. If osteomyelitis develops then it is a signi®cant risk factor for amputation, regardless of vascular status 10 . Other factors A number of patient characteristics can be identi®ed from epidemiological surveys as having a signi®cant effect on the outcome of treatment of diabetic foot ulceration. Very few of these are actually independent predictors of amputation but most form part of a multivariate regression. A history of previous foot ulceration, and in particular of previous amputation, is one such independent indicator that there is a high risk of amputation during a subsequent event. In addition there is a need for further evaluation of post-ulcer care if the foot heals 21 . One of the reasons for this is the strong association between patient non-compliance with therapy and amputation in a number of studies. Inability to comply with off-loading strategies and antibiotic therapy, and failure to attend the clinic, may all compromise the foot. In addition, late presentation to clinic with an ulcer carries a high risk of subsequent amputation, although this may be as much due to primary care delays as patient delays 22 . Irrespective of these factors it is more common for men to have foot ulcers and to have amputations compared to women. The elderly, especially if they live in institutionalized care or have a low walking tolerance, and patients with longer duration of diabetes, are at greater risk of major amputation 23 . Although one study did not identify end-stage renal disease as a factor that in¯uenced healing 24 , in most studies, amputation risk is generally higher in patients with other major diabetes complications, particularly renal impairment and visual impairment 7,12,20,21,23,24 . Type 2 patients on insulin, higher glycated haemoglobin, and random glucose levels are also associated with a greater risk of amputation or re- ulceration in some studies, and may again re¯ect a lower degree of patient compliance with therapy 21 . Relevance of Ulcer Classi®cation to Management 65 THE MYTH OF THE NON-HEALING ULCER? Many reports have tried to categorize ulcers as ``healing'' and ``non- healing''. It is important to be able to identify those patients in whom treatment is failing and for whom a new approach should be used. This is particularly true with the advent of very expensive advanced wound- healing technologies, such as growth factors or skin replacements, which are targeted at the chronic non-healing primarily neuropathic foot ulcer. If no objective measure of ulcer healing is used, there is no possibility that such patients will be detected and, once again, the need for measurement and standardized descriptions of ulcers cannot be stressed too highly. Based on a review of all of the studies included in the discussion above, it is clear that the primary reasons for failure of the diabetic foot ulcer to heal are inadequate or inappropriate pressure relief, inadequate debridement and infection control, failure to recognize or treat vascular insuf®ciency and patient non-compliance. An ulcer can truly be described as non-healing only when all of these factors have been addressed, including angiography and reconstruction where necessary, or by the implementation of non- weightbearing regions, using inpatient bed-rest or a non-removable cast. Such ulcers will be rare. This is discussed further in a review by Cavanagh et al 25 . CURRENT CLASSIFICATION SYSTEMS The most widely used and validated foot ulcer classi®cation system is the Meggitt±Wagner classi®cation 4,5 , which divides foot ulcers into ®ve categories. Grade 1 ulcers are super®cial ulcers limited to the dermis. Grade 2 ulcers are transdermal with exposed tendon or bone, and without osteomyelitis or abscess. Grade 3 ulcers are deep ulcers with osteomyelitis or abscess formation. Grade 4 is assigned to feet with localized gangrene con®ned to the toes or forefoot. Grade 5 applies to feet with extensive gangrene. A signi®cant problem with the Meggitt±Wagner classi®cation is that it does not differentiate between those Grade 1±3 ulcers which are associated with arterial insuf®ciency and might be expected to heal less well, or those Grade 1 and 4 ulcers which are signi®cantly infected and which might also be expected to have a poorer prognosis. Despite this, the Meggitt±Wagner classi®cation has been shown to give an accurate guide to risk of amputation in a number of studies and remains the standard by which other classi®cations have to be judged 6,26 . In an effort to improve upon the Meggitt±Wagner classi®cation, Harkless et al 27 proposed an expansion of the grading system to allow for ischaemia in the early grades 27 . Each of the original Meggitt±Wagner Grades 1±3 are subdivided into A (without ischaemia) or B (with signi®cant ischaemia). 66 The Foot in Diabetes Although the prognosis of the various foot lesions is postulated, there does not appear to be any validation of this system or the newer Texas system 28 which has superseded it. CLASSIFICATIONS BASED ON FOOT ULCER DESCRIPTION CATEGORIES The limitations of the Meggitt±Wagner classi®cation were demonstrated by Reiber et al 29 , who tried to classify their patients retrospectively using a number of different systems and found that between one-®fth and a half of their patients could not be categorized satisfactorily. To accommodate this, a number of descriptive systems have been devised, most notably from the Nottingham group 30,31 . However, as they state in their most recent version 31 , these are descriptions rather than classi®cations. Their proposed system has three main categoriesÐthe person, the foot, and the lesionÐtogether with 14 variables. To classify an ulcer on such a basis would lead to at least 2Â 10 14 categories, even if they were only dichotomous variables, and indeed, many are multifactorial. Therefore, most systems based on descriptions concentrate on the ulcer and aetiological factors alone. The Gibbons classi®cation includes ulcer depth and infection but ignores aetiology and, in particular, vascular impair- ment 32 . The most validated of this type of system is the classi®cation proposed by Lavery et al 28 . This classi®cation excludes factors other than those in¯uencing the wound, since the authors felt such parameters were dif®cult to measure or categorize, despite the fact that some of those factors are known to in¯uence outcome at least as much as the parameters they chose to include 33 . Indeed, the authors have addressed the outcome problem elsewhere 34 . The three main categories are related to the relative depth of the ulcer. Grade 1 is a super®cial ulcer not involving capsule or bone. Grade 2 is an ulcer which extends to tendon or joint capsule. Grade 3 is a lesion which extends into joint or bone. Each of these grades is then subdivided into one of four stages: (a) uninfected and not ischaemic; (b) infected but not ischaemic; (c) ischaemic but not infected; (d) ischaemic and infected. Thus, an ulcer could be placed into one of 12 categories. Two years later the same group reviewed their classi®cation in practice and demonstrated that amputation risk was clearly and independently linked to both increasing depth and grade of ulcer. No uninfected and non-ischaemic patients had an amputation in the follow-up period, whereas patients with both infection and ischaemia were 90 times more likely to have a midfoot or higher amputation than patients with lower- graded lesions, despite following clearly de®ned treatment protocols which are described in this paper 33 . Relevance of Ulcer Classi®cation to Management 67 CLASSIFICATIONS BASED ON FOOT ULCER RISK CATEGORIES The third main type of foot ulcer classi®cation system relies on the underlying foot ulcer risk categories of peripheral neuropathy, peripheral vascular disease and deformity. These have been used in various combinations in a number of classi®cations which are further reviewed by Harkless et al 27 . Ultimately these are screening tools for education and pre-ulcer intervention. Patients with ulcers are grouped together in the ®nal category as the highest risk of amputation in population surveys, but this method of classifying foot lesions gives little information as to how to approach an individual ulcer or about the variable prognoses between ulcers. A minimalist approach to foot ulcer classi®cation was proposed by Edmonds and Foster 2 in the previous edition of this book. Foot ulcers were divided into neuropathic and neuro-ischaemic on the basis of clinical tests, mono®laments and Doppler ultrasound, understanding the limitations of this test in the diabetic foot 35 . This has the advantage of simplicity and also identi®es patients with vasculopathy, which is the principal adverse prognostic indicator for amputation of the diabetic foot and which may require revascularization. This classi®cation approach provides a very simple means for rapid comparison of outcomes across clinics, but it may be limited if used for more detailed prognostication and treatment planning. CAN THE CURRENT CLASSIFICATION SYSTEMS BE IMPROVED? With such a variety of classi®cation systems available, it is clear that no one system offers an ideal compromise between comprehensive applicability and simplicity. The reviewers of classi®cation systems usually want each system to include their own particular facet. For example, the Texas system was reviewed by Levin 36 , who noted that site of ulceration was missing, despite the fact that this has been shown to be an uncertain predictor of outcome 36 . A good classi®cation system would seem to require some allowance for patient factors and inclusion of a deformity index, particularly in relation to ulceration in association with Charcot feet. At present, most of the current classi®cations force the user to become totally foot-centred at the expense of the patient as a whole. Whilst this is not likely to create problems in multidisciplinary practice, it is a possible cause of fragmented care when the foot clinic is separate from diabetology and other support. Addressing the social as well as the diabetes related issues of patients is likely to improve foot ulcer outcomes 29 . 68 The Foot in Diabetes At present, the de®nitions of neuropathy, ischaemia, infection, deep ulceration, etc. are still open to interpretation. Clear and explicit standards for these parameters in the context of foot ulceration would aid the evolution of classi®cations and improve their prognostic reliability. Such a classi®cation might then form the basis of an integrated care pathway for foot ulceration for each patient. THE VALUE OF CLASSIFICATION SYSTEMS IN CLINICAL PRACTICE At the beginning of this chapter, a distinction was made between descriptions and classi®cations of ulcers. At times the boundaries are blurred, but descriptions in general are more detailed and apply to individuals, while classi®cations are pigeonholes which facilitate research and audit in groups of patients. Individuals within the same classi®cation grade will have other characteristics, principally the presence or absence of other diabetes complications, diabetic control, social factors and treatment compliance levels, which may in¯uence their treatment and outcome. In general, however, increasing severity of ulceration has been clearly shown in most systems to in¯uence prognosis and amputation rate. It is a major step from that premise to a decision to amputate on the basis of a poor classi®cation grading. Despite various multi- and univariate analyses of potential risk factors, no classi®cation yet devised can aid such decision making and all decisions have to be made on an individual basis 23 . Treatment regimens are in many ways the same for all ulcers and should not normally be in¯uenced by classi®cation grade alone. Some principles of managementÐfor example, pressure relief (including pressure from shoes) and debridementÐapply to all ulcers. The value of scoring and grading systems in planning treatment is that they prompt the clinician to search for the depth of the ulcer, to consider whether infection is present, and to seek evidence of vascular insuf®ciency. Thus, the care of the patient is improved simply because all the major relevant factors in the healing of the ulcer are considered during classi®cation 1 . For this reason alone it should be the standard practice for all clinicians treating diabetic foot ulcers to adopt a classi®cation system, either their own or one chosen from those outlined above. Unfortunately, none of the present foot ulcer classi®cations discussed above has been validated prospectively outside of their originating centre. A multicentre prospective study of diabetic foot ulceration, using one or more classi®cations or examining a number of potential candidate criteria for inclusion in a ®nal classi®cation, would be of immense help in answering the question of whether or not the classi®cation of diabetic foot Relevance of Ulcer Classi®cation to Management 69 [...]... sensation, foot pulses and training in self-care of The Foot in Diabetes, 3rd edn Edited by A J M Boulton, H Connor and P R Cavanagh & 2000 John Wiley & Sons, Ltd 96 The Foot in Diabetes the foot2 It is to be hoped that recently introduced changes in the organization of primary care in the UK will also produce more speci®c requirements ST VINCENT DECLARATION TARGET FOR REDUCING AMPUTATION The St Vincent... English Dictionary, 3rd edn Oxford: Oxford University Press, 19 73 13 Burden ML Barriers to integration of foot care services Diabet Foot 1999; 2: 27 32 The Foot in Diabetes Third Edition Edited by A.J.M Boulton, H Connor, P.R Cavanagh Copyright  2000 John Wiley & Sons, Inc ISBNs: 0-4 7 1-4 897 4 -3 (Hardback); 0-4 7 0-8 4 63 9-9 (Electronic) 6 Providing a Diabetes Foot Care Service (b) Establishing a Podiatry Service... identifying who does what and when There is little communication and liaison between the different professional disciplines in the community (nursing, podiatry and tissue viability), each developing their own working methods The community disciplines do not involve the hospital foot care team in the planning and delivery of diabetic foot care and the hospital does not involve the community Working practices... in amputations in the UK These include: 1 2 3 Screening for the ``at-risk foot' ' in primary care Special review and extra education for those at risk Prompt referral to a multidisciplinary diabetes foot care team should ulceration or infection occur RECOGNITION OF THE ``AT-RISK'' FOOT IN PRIMARY CARE The detection of risk factors for ulceration and amputation requires regular visual inspection of the. .. R, McInnes A The ``at risk'' foot: the role of the primary care team in achieving St Vincent targets for amputation Diabet Med 1998; 15(suppl 3) : S61±4 15 Currie L, Harvey G Care pathways development and implementation Nursing Standard 1998; 12 (30 ): 35 ±8 The Foot in Diabetes Third Edition Edited by A.J.M Boulton, H Connor, P.R Cavanagh Copyright  2000 John Wiley & Sons, Inc ISBNs: 0-4 7 1-4 897 4 -3 (Hardback);... services may have the organizational machinery to ensure integration of the different elements However, in the ever-changing National Health Service it is easy to overlook important sections: vigilance and team communication are needed to prevent this Recent examples in the UK include the threat to orthotic services in different parts of the country In the UK the various agencies involved in the provision... elements of research, audit and training in addition to a willingness to be innovative Ideally, the specialist clinician will also work in community clinics to ensure good liaison with the community-based podiatry staff Communication is an essential role, informing members of the primary care team of developments and ensuring integration of services Training The quality of training of podiatrists varies greatly... follow-up there was a signi®cant reduction in major amputations in the index group compared with the control group ( p50.01) There were fewer minor amputations and new ulcers in the index group, but the differences were not statistically signi®cant, possibly because any patients who developed new ulcers during the study were automatically transferred to the foot protection clinic The foot protection clinic... up-to-date information about foot care management Practice nurses are especially important, as they have a central role in diabetes foot education and screening in many general practices The arrangements for providing the 100 The Foot in Diabetes education and training will vary from one country to another, and even within countries depending on local circumstances In parts of Germany, reimbursement... disciplines who regularly meet to plan and provide a service The team is often dominated by hospital-based specialists, and it is easy to forget the many others who are involved in diabetic foot care outside the hospital and who must also be considered as part of the team If, as Knight has pointed out9, the interrelationships between members of a hospital team are often complex, the 76 The Foot in Diabetes . developing their own working methods. The community disciplines do not involve the hospital foot care team in the planning and delivery of diabetic foot care and the hospital does not involve the. this. Recent examples in the UK include the threat to orthotic services in different parts of the country. In the UK the various agencies involved in the provision of diabetic foot care services. & Sons, Inc. ISBNs: 0-4 7 1-4 897 4 -3 (Hardback); 0-4 7 0-8 4 63 9-9 (Electronic) of optimum mobility''. This accords with the St Vincent Declaration target of reducing by one-half the rate