OBJECTIVE—To assess changes in diabetic lower-extremity amputation rates in a defined relatively static population over an 11-year period following the introduction of a multidisciplinary foot team.
RESEARCH DESIGN AND METHODS—All diabetic patients with foot problems admitted to Ipswich Hospital, a large district general hospital, were identified by twice-weekly surveillance of all relevant in-patient areas and outcomes including amputations recorded.
RESULTS—The incidence of major amputations fell 62%, from 7.4 to 2.8 per 100,000 of the general population. Total amputation rates also decreased (40.3%) but to a lesser extent due to a small increase in minor amputations. Expressed as incidence per 10,000 people with diabetes, total amputations fell 70%, from 53.2 to 16.0, and major amputations fell 82%, from 36.4 to 6.7.
CONCLUSIONS—Significant reductions in total and major amputation rates occurred over the 11-year period following improvements in foot care services including multidisciplinary team work.
Amputation is one of the most feared diabetes complications. The economic burden is immense, and survival is bleak, with a 2-year mortality up to 50% (1–6). There is an urgent need to reduce amputation rates, which vary considerably globally and nationally (7–13). Though often attributed to differences in ethnicity and local practice, such variability could be due to other factors. These include differences in defining minor versus major amputations and in assessing the denominator, i.e., 1) the size of the general population (often ill defined, particularly in cities because of overlap with other institutions) and 2) the diabetic population (sometimes simply estimated from the expected prevalence) (14). Finally, of considerable importance is the accuracy of ascertainment. Nearly all studies have been retrospective, collected from a variety of sources including theater and limb-fitting records, anesthetic databases, and hospital activity coding. In 2004, we demonstrated that compared with prospective data collection, such methods underestimated the incidence by 4.2–90.6% and misclassified 4.5–17.4% of amputations (15). Others have reported similar inaccuracies (16,17). We therefore recommended that future studies should be prospective.
This study of the impact of improvements in foot care on lower-extremity amputations (LEAs) addresses these problems; it is prospective, amputation levels are made clear, and the population is well defined.
RESEARCH DESIGN AND METHODS—
The survey was conducted between 1995 and 2005 in a defined, predominantly Caucasian (95%), mixed rural/urban population served by Ipswich Hospital from which there are few cross-boundary referrals. Over the period, the population grew from 330,462 to 345,890. In contrast, the diabetic population rose dramatically from 6,768 to 11,906. The largest increase followed introduction of the Quality and Outcomes Framework in 2005, whereby practitioners are paid to achieve diabetes targets (18).
An LEA was defined as loss in the transverse anatomical plane of any part of the limb and a major amputation as any above the ankle. When a revision occurred within 3 months, only the later procedure was recorded. Traumatic and tumor-related amputations were excluded.
The data were collected prospectively from 1997 by a diabetes specialist nurse or podiatrist surveying all relevant wards twice weekly to identify and follow patients with diabetic foot problems. We also included retrospective data for 1995 and 1996, based on ICD-9 (International Classification of Diseases, 9th revision) and OPCS4 (Operating and Coding Procedures, 4th revision). Our previous work suggests this may underestimate amputations and, if anything, minimize any improvement (15). In 2001, service cutbacks curtailed prospective collection for that year; the data were obtained retrospectively.
Incidences of amputations were expressed per 100,000 of the general population and per 10,000 people with diabetes. However, we believe the former to be more representative, as rocketing prevalence following improved reporting may bias changes based on the latter.
Incidences of major, minor, and total amputations per 100,000 of the general population are shown in Fig. 1. Over the period, total and major amputations fell, except in 2001, the year the service was reduced, when total amputations rose due to a rise in minor amputations.
Since yearly amputation rates fluctuate, baseline and final rates were calculated by averaging years 1995 to 1997 and 2002 to 2005. Total and major amputations per the 100,000 general population fell 40.3% (10.1 to 6.0) and 61.5% (7.4 to 2.8), respectively, with a small increase in minor amputations (2.7 to 3.3).
The incidence of total, major, and minor amputations per 10,000 people with diabetes fell 70.0% (53.2 to 16.0), 81.6% (41.4 to 6.7), and 21.1% (11.8 to 9.3), respectively. There were no differences in age (mean 70.7 years); proportion with neuropathy, ischemia, or neuro-ischemia; or sex ratio between these periods.
As far as we are aware, this is the longest prospective audit of LEA in people with diabetes. As the data were collected prospectively from 1997 and all cross-boundary referrals accounted for, we are confident that the 62% reduction in major amputations is genuine. Indeed, 2 years of baseline data were retrospective, which was previously shown to underestimate the true rate; therefore, the reduction may have been greater (15).
We favored reporting the change relative to the general population, as incentives to increase diabetes reporting would bias change based on diabetes prevalence. It is of interest that the incidence of diabetic LEAs in the Netherlands was reported to have fallen 34% (55 to 36 per 10,000 people with diabetes), yet actual amputation numbers remained relatively unchanged (9). This was because the diabetic population had increased 50%. For comparison, from 1995 to 2005, our total amputations per 10,000 people with diabetes fell 70%, from 53.2 to 16.0, and major amputations fell 82%, from 36.4 to 6.7.
Large falls in amputations would be expected with modest improvements in care if baseline rates are high. In Denmark, there were dramatic falls from a high baseline of 27.2 major amputations per 100,000 of the general population to 6.9 (13). Our baseline of 7.4 compares well with this and also with major amputation data from other U.K. centers (Leeds, Leicester, Middlesbrough, and Newcastle report 15.4, 5.8, 18.0, and 14.9, respectively) (7). The achieved major amputation rate of 2.8 per 100,000 of the general population is comparable with the lowest published rate of 2.2 from Madrid (12).
These improvements are unlikely to be due to any single factor. There have been improvements in vascular, radiological, and microbiological services and in multidisciplinary working. We also believe that prospective audit, which involved the foot team closely interacting with ward staff, raising the awareness of the at-risk foot, contributed. Furthermore, yearly analysis of performance based on the audit data has driven changes in practice. It is of interest that in the year we were unable to undertake this, amputation rates rose.
In conclusion, we have demonstrated significant improvements in outcomes for one of the most costly and feared complications following improvements in foot care through multidisciplinary team work and continuous prospective audit.
Published ahead of print at http://care.diabetesjournals.org on 12 October 2007. DOI: 10.2337/dc07-1178.
A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C Section 1734 solely to indicate this fact.