Amputations in diabetes: a changing scene

Lower extremity amputation (LEA) is a life changing and devastating complication of diabetes. For patients with diabetes it occurs 15 times more commonly than in people without diabetes. Diabetes is now present in nearly 50% of all patients undergoing LEA1 in western society. In the UK, between 0.5% and 1.3% of all patients with diabetes have undergone LEA,2,3 but worldwide the incidence is very variable and may be as high as 7%.4 In the majority of cases foot ulcers are a precursor of LEA. It is estimated that 5% of people with diabetes have ever had a foot ulcer2, 4 and around 1.4% of them have a foot ulcer at any one time,4 and that one in six people with diabetes will have a foot ulcer at some time in their life. LEA has a devastating impact on a patient's quality of life. LEA diminishes mobility and can prevent a patient driving, being able to work and socialising, with consequential effects on low mood and functional capacity. Often LEA could have been prevented, thus avoiding these consequences. Diabetes UK has recently highlighted the high rates of amputations due to diabetes in the UK, and that 100 patients from a diabetes population of 2.3 million are undergoing amputation each week.5 A large number of these amputations are potentially avoidable. The hope is that improved foot health care may reduce the rate of amputation, but what is the evidence? Various studies have been reported showing changes in amputation rates. Encouragingly, these generally identify a decrease in amputation rates from population-based studies. Although this may reflect a lack of reporting of negative studies, improvements have been shown across a number of different health care services, although mainly in Europe. Baseline rates of amputations for the 1990s are remarkably consistent at around 500–550 per 100 000 patients with diabetes.1, 6-8 Rates have been reported to decline to 363 per 100 000 of patients with diabetes in the Netherlands,6 around 300 in Germany and Scotland (Krishnan et al8 and local data), and 160/170 in England,1, 8 whilst significant declines in amputations have also been reported in Sweden,9 Denmark,10 Spain,11 and Finland.12 The incidence of diabetes-related amputations can also be reported as the rate per 100 000 of the general population. Some people prefer this, as it avoids difficulties about ascertaining data on the number of patients with diabetes. However, this will underestimate the success of many programmes as the number of patients with diabetes is doubling at least every 10 years. Despite this, some studies have still shown a decrease in amputations using this approach, with decreases from 19.1 to 9.4,9 from 27.2 to 6.9,10 and from 7.4 to 2.88 per 100 000 of the general population. As the number of patients with diabetes increases, many patients will have a short duration of diabetes, which may skew the results of changing amputation rates in a falsely positive direction. However, we have data showing a reduction in major amputation rates even after correcting for diabetes duration, decreasing from 512 to 288 per 100 000 people with diabetes. Many studies show that the main improvements are due to reductions in the rate of major amputation, which include all amputations above the ankle.1, 8, 9 The same studies demonstrate very little change in the rates of minor amputation, which include anything distal to the ankle, and in the majority of cases are toe or ray excisions. Such minor amputations are often judged to be appropriate interventions, as they can prevent subsequent major amputations. Thus, in summary, there appears to be a decline in the rate of amputations for patients with diabetes, which is due to a reduction in numbers of major amputations. Many of these reports come from centres of excellence, and may reflect sporadic and isolated reductions in amputations. Indeed, one study demonstrates that, even in centres of excellence, amputation rates can start to increase again if necessary clinical resources are removed from the diabetes foot-care team (Krishnan et al8 and personal communication). However, the Dutch study is possibly one of the only national data sets to demonstrate widespread improvements in amputations6 which may indicate that the improvements are more than sporadic. In addition, although not established as an epidemiology research project, the Scottish Diabetes Survey has shown a reduction in prevalent amputations across Scotland from 0.8% in 2003 to 0.4% in 2006.2 Although all clinicians are keen to avoid amputation, an amputation is not necessarily the worst outcome for a patient. Making an early decision to amputate can be best for the patient, if it is thought to be an inevitable outcome. It may well be preferable to allowing a patient to continue with a non-healing ulcer. In such a circumstance, amputation can improve the patient's quality of life, improve mobility and decrease dependency on health care professionals in the longer term. As a consequence, interpreting data on amputations needs to done with insight given this proviso. Survival rates for patients with diabetes are increasing.13, 14 As patients with diabetes live longer, and have diabetes for longer, they will be at increased risk of diabetic foot problems. It is thus conceivable that there will be an increase in amputations as patients live longer. However, the hope is that patients' general health will improve as their survival increases, but only time will tell! What are the key elements resulting in a decrease in amputations? Unfortunately, the evidence addressing this issue is weak. Development of multidisciplinary foot clinics (MDFCs) has been associated with improved outcomes for patients with diabetic foot ulcers.9, 15, 16 Many of the studies, described above, demonstrating reductions in amputation rates have attributed this to the existence or development of MDFCs. The key elements of an MDFC appear to be a specialist podiatrist and orthotist, with a dedicated diabetologist, who has rapid access to a vascular radiologist, a vascular surgeon and an orthopaedic surgeon when required. For patients with neuropathic foot ulcers, there is good evidence that total contact casting is beneficial for healing,17 which should reduce amputations. However, a removable cast walker, made irremovable, appears to be as effective at healing ulcers.18 Although purely neuropathic ulcers can lead onto LEA, the neuro-ischaemic foot ulcer is a greater challenge to achieve healing and prevent amputations. Arteriopathy in the patient with diabetes is different from standard peripheral vascular disease. In patients with diabetes, arterial disease tends to occur more in small distal vessels, rather than larger proximal vessels. Diseased distal vessels mean that bypass surgery is frequently not possible because there are no adequate distal arteries to anastamose. Also, angioplasty is more difficult in smaller peripheral vessels as distal vessels are more likely to rupture and distal embolisation is a greater risk. However, small improvements in arterial flow may make a big clinical difference. In a recent study, patients who had undergone revascularisation did end up with fewer major amputations than those who were not revascularised.19 However, this was an observational study, and there may have been bias in patient allocation. Additionally, in the small Danish study, reduced amputations were associated with increased vascular interventions.10 However, the reduction in amputation rates across the Netherlands was associated with no change in rates of vascular interventions.6 Although there are clearly some individual patients who benefit from vascular intervention, the role of surgical vascular intervention in preventing amputations overall in diabetes remains controversial. Of increasing interest is the role of treating modifiable cardiovascular risk factors to prevent amputations. The UK Prospective Diabetes Study demonstrated that tight blood pressure and glycaemic control showed a non-significant trend towards decreased amputation with a relative risk of 0.5120 and 0.6121 respectively. Canavan et al showed an increasing use of statins (5% to 20%), ACE inhibitors (15% to 38%) and aspirin (22% to 33%) during the time period that they demonstrated lowered amputations.1 Locally, we have shown a decrease in mean blood pressure (140/79mmHg to 137/75mmHg), HbA1c (8.1% to 7.5%) and serum cholesterol concentrations (5.0mmol/L to 4.4mmol/L) across the diabetes population at a time when amputations reduced. The MDFC may be the central component of organised diabetes foot care, such that patients with diabetes in the community can gain rapid access to specialised services. Thus, organisation of community foot care and good communication links with the centre may be important for quality foot care and amputation prevention. Amputations are usually preceded by foot ulceration. Diabetic foot screening programmes have been developed22-24 which integrate a number of known risk factors,3, 22 and have been demonstrated to successfully identify patients at risk of foot ulceration in day-to-day clinical settings.23, 24 Lack of adequate health care resources is a problem in most countries and is likely to get worse with the current epidemic of diabetes. Identification of patients at greatest risk of foot problems allows appropriate treatments to be directed towards those at greatest need. Simple education programmes have been shown to prevent amputations in patients with diabetes,25 but the generalisability and sustainability of such programmes remain untested. Orthotic interventions26, 27 have been shown to reduce the rates of foot ulceration and, although it is likely that this will reduce amputation rates, this has not formally been demonstrated. One study of 70 patients with diabetes demonstrated a reduction in major amputations from 33% to 9% with the use of hyperbaric oxygen in patients with neuro-ischaemic foot ulcers.28 Further research is required to identify which treatments are most effective at reducing amputations. Patients with diabetes, who undergo major LEA, have a two-year median mortality of up to 50%.29-32 After one major LEA, patients with diabetes are more likely to undergo a further amputation, and are more likely to suffer cardiac failure, than people without diabetes who have had an LEA.29 Amputation is frequently a surrogate marker of vascular disease. In a recent study, staff who were providing routine foot-ulcer care within an MDFC also focused on optimal management of glycaemic control, blood pressure and dyslipidaemia, and promoting the use of ACE inhibitors. They demonstrated a five-year mortality rate decreasing from 48% to 27%.33 Attention to treating cardiovascular risk factors appears to prolong survival in this group of patients. There are some encouraging data that, in areas where diabetes foot services are well organised and integrated, the rate of amputations may be declining for patients with diabetes. Although there is some evidence of improvement, a lot more can, and needs to be done as many amputations are avoidable. Best practice needs to be shared worldwide. The Scottish Government has invested in a ‘Foot Action Group’ and a two-year Diabetes National Foot Co-ordinator with the aim to improve foot care in Scotland. The Scottish Foot Action Group is working with Foot in Diabetes UK (FDUK) and works with Diabetes UK on various projects. These latter groups are also working with the Joint British Diabetes Societies Inpatient Care Working Group to provide guidance for acute foot problems. In addition, there are various initiatives being undertaken across the rest of Europe and beyond. Hopefully, all this should help to improve outcomes for patients with diabetes who have foot problems in the future.