Diabetes-related foot ulceration is a serious complication of diabetes, affecting the quality of life of both the person with diabetes and their carer. It is also known to impose a heavy economic burden on the healthcare system; the relative cost of care for those with foot ulceration being 5.4 times that for people with diabetes but without foot ulceration (Ramsey et al, 1999). In Australia, 20% of people with diabetes are considered to be at high risk of foot ulceration (International Diabetes Institute, 2000), with this figure more than doubling for those who have had diabetes for more than 20 years (Tapp et al, 2003). The number of diabetes-related amputations in Australia is approximately 3400 per year (2004–2005) and admissions for diabetic foot disease are longer than for any other diabetic complication (Australian Institute of Health and Welfare, 2008).
A multidisciplinary foot clinic was established at the Royal Prince Alfred Hospital in Sydney, Australia, more than 20 years ago, with the aim of providing care for people with diabetes who suffer from foot ulceration, infection or acute Charcot arthropathy. Having started with a simple model comprising a podiatrist and an endocrinologist, the clinic is now staffed by healthcare professionals from a variety of disciplines, including regular on-site visits from an orthopaedic surgeon, a vascular surgeon, a pedorthist (trained in design and fabrication of footwear and foot orthoses), as well as dietitians and social workers. Once their ulcers have healed, people are discharged and referred to a public or private podiatry clinic in their local area for routine foot care.
While the proportion of people whose ulcers heal, and the rate at which this is achieved, is good, there appears to be a sub-population treated at the foot clinic who have very little time free of foot ulceration. It has been proposed that minimising ulcer recurrence, both in terms of incidence and time taken to next ulcer, provides an alternative measure of overall treatment success to traditional end points, such as ulcer healing rate (Pound et al, 2005). Hence, the aim of this study was to determine the rate of ulcer recurrence among people attending the Royal Prince Alfred Hospital multidisciplinary diabetic foot ulcer clinic, and to examine the possible predictors of re-ulceration.
Method
Data were retrieved for all patients with a diabetic foot ulcer referred over a 3-year period (n=227), with the exception of 21 individuals who attended only a single visit or had unknown outcomes. Relevant information was prospectively entered into a purpose-built electronic database. This included various clinical parameters, past history of ulceration, amputation and the number of previous ulcers. The stage and grade of ulcers were determined according to the Texan Grading System (Armstrong, 1998). Neuropathy was defined by vibratory perception threshold (measured using a biothesiometer) and inability to feel the 10g monofilament. Ischaemia was defined by the absence of pedal pulses, or an ankle–brachial pressure index of less than 0.8. Ulcers were graded C or D, in accordance with the Texan Grading System. Data were grouped according to whether the subject had ever become ulcer free during the 3-year study period, and whether the subject had recurrent ulceration. An ulcer was considered healed if the lesion became completely epithelialised and free from exudate for more than 2 weeks.
According to the standard foot clinic protocol, ulcers were managed, as appropriate, with traditional wound care methods, systemic antibiotics, regular sharp debridement, vascular assessment and intervention, and pressure off-loading. During treatment, and at the time of discharge, participants received foot-care education. This included how to select suitable footwear, the need for podiatry care, daily foot inspection and avoidance of common causes of foot injury. Booklets containing this information were also given to participants.
Individuals identified as having significant foot deformity, or abnormal foot shape, were referred for medical grade footwear (MGF) and/or foot orthoses. In this study, an MGF referral was used as a surrogate measure of clinically significant foot deformity. People with Charcot arthropathy or partial foot amputation were automatically included in this category. Information on MGF referral was available for 196 of those included in the study group. Referrals were made directly to the supplier or, for those who were financially disadvantaged, an application for government assistance for the provision of shoes and orthoses was made.
As an indicator of adherence to treatment, clinic appointment attendance was recorded.
Statistical analysis
Data were analysed using NCSS 2004 software (NCSS, Kaysville, UT, USA). Continuous data were checked for normality and presented as mean ± standard deviation, or median and interquartile range (IQR). Data not normally distributed were transformed for analysis. ANOVA was used to compare means. Categorical data were represented as percentages. The Chi-square test was used to compare groups. A Kaplan–Meier curve was constructed to determine the time to re-ulceration. Cox regression was used to determine the independent predictors for re-ulceration. Predictors examined included age, gender, duration of diabetes, mean HbA1c level, neuropathy status, ischaemic status, referral for MGF, number of missed clinic appointments, extent of ulceration using the Texan Grading System and past history of foot ulceration. The risk of re-ulceration was calculated using Cox regression and expressed as a hazard ratio with a 95% confidence interval (CI). Data were further stratified by these predictors and a Kaplan–Meier curve was constructed, and a log-rank test carried out, to compare the groups. Significance was accepted when P<0.05.
Results
Clinical characteristics of the participants are shown in Table 1. There was a preponderance of males in the study group (68%) and the degree of neuropathy was severe, with most unable to detect vibration at the maximum biothesiometer reading of 50 volts.
Of the 227 people referred to the foot clinic with ulceration during the 3-year study period, there was a total of 559 ulcers. The median follow-up time for the total cohort was 26.4 months (IQR 16.4–33.7), and the minimum follow-up time was 6.0 months.
Two hundred and five people (90%) healed and became ulcer free for a period of time during the study period (Figure 1). About half of those who healed (n=104) re-ulcerated, the rest remained ulcer free for the remainder of the study period. More participants in the recurrent-ulcer group required an amputation than those who remained ulcer free (11% versus 1%, P=0.006).
During the follow-up period, 71% of participants were ulcer free at 6 months, 59% at 12 months and 46% at 24 months (Figure 2). For those who healed and re-ulcerated, the median time to re-ulceration was 3.9 months (IQR 0.7–10.8).
Multivariate analysis showed that there were three significant predictors for re-ulceration: (i) past history of ulceration, (ii) ischaemia and (iii) foot deformity. These predictors caused 1.9 (95% CI 1.2–4.4), 2.2 (95% CI 1.1–4.4) and 2.1 (95% CI 1.2–3.7) fold increases, respectively, in the risk of ulcer recurrence, accounting for 4.6%, 2.3% and 4.6%, respectively, of the total variance.
When grouped by ischaemic status and analysed by Kaplan–Meier survival curve, the difference in ulcer-free period approached statistical significance (log-rank test=3.6; P=0.06; Figure 3). Seventy percent of people who experienced ulcer recurrence were referred for MGF, while fewer (51%) in the no recurrent ulcer group were deemed to require MGF.
Participants who experienced recurrent ulceration missed more foot clinic appointments than those who remained ulcer free, with a median of four versus one missed appointment, respectively. Those without ulcer recurrence were also more likely to have attended all their appointments (Table 1).
Twenty-two people were never ulcer free during the study period. The clinical characteristics of these people were similar to those of people whose ulcer healed at some stage (Table 1), except that the prevalence of peripheral arterial disease was greater (P<0.0001). Wounds in this group were more severe in terms of their grading for depth, though the difference was not statistically significant. Of the participants that never became ulcer free, nearly half died (n=10, 45%) during the study period, compared to 3% (n=7) among those who became ulcer free at some stage during the study (Chi-squared=50.7; P<0.0001).
Discussion
The Royal Prince Alfred Hospital multidisciplinary foot clinic is effective in healing ulcers in people with diabetes, but the 51% recurrence rate is disappointing. High rates of ulcer recurrence have been reported elsewhere (Apelqvist et al, 1993; Busch and Chantelau, 2003; Pound et al, 2005; Ghanassia et al, 2008) and previous studies have sought to identify the causes. Connor and Mahdi (2004) found that the presence of neuroarthropathy, poor glycaemic control, sex (preponderance of men), failure to attend clinic appointments, living alone and being non-compliant with footwear and foot care were associated with ulcer recurrence. Severity of neuropathy, poor glycaemic control, increased alcohol intake and waiting longer than 24 hours to report a foot problem have also been implicated in ulcer recurrence (Mantey et al, 1999; Jude et al, 2001). Interestingly, ischaemia was not found to be significant risk factor in these studies.
Peripheral vascular disease was shown to be a significant risk factor for recurrent foot ulceration by Peters et al (2007), who also assessed the role of ulcer location. In their study, the site of ulceration was important, with plantar hallux and plantar metatarsal ulceration recurring far more often than ulcers elsewhere. This finding did not appear to be fully explained by the higher amputation rate for ulcers on the lesser toes.
The most notable result of the current study is that factors usually associated with foot ulceration, such as severe neuropathy and peripheral arterial disease, could not adequately explain recurrence. This finding does not negate the importance of these two factors. All participants experienced severe neuropathy, but the biothesiometer measurement of neuropathy used was not sufficiently discriminating to differentiate between people with severe loss of sensation and those with even more profound sensory loss. Likewise, one of the definitions of ischaemia used was an ankle–brachial pressure index of <0.8. Using this threshold yields a categorical outcome, but does not grade the degree of tissue perfusion and therefore the extent of ischaemia. Participants with severe ischaemia may have failed to heal, and hence not been included in the analysis of ulcer recurrence.
Selection bias could also have affected the results. Individuals with more severe ischaemia may have been referred directly for surgery rather than to the foot clinic.
Unfortunately, when considered collectively, these results suggest that the clinical variables that can be determined objectively and with relative ease, by physical examination or through patient histories, do not allow confident prediction of long-term outcomes for people presenting to a foot clinic of this nature. Previous foot ulceration predisposes people to re-ulceration, but these findings indicate that the predictive value of previous ulceration is modest.
In the authors’ clinical experience, foot deformity and biomechanical problems predispose people to re-ulceration. Clinical guidelines cite the importance of accommodating foot deformity with appropriate footwear (International Working Group on the Diabetic Foot, 2007) and some studies have shown that re-ulceration is reduced when MGF is worn as part of a comprehensive foot-care programme (Edmonds et al, 1986; Chantelau et al, 1990; Faglia et al, 2001; Busch and Chantelau, 2003). However, due to its pleomorphic nature, deformity is remarkably difficult to quantitate and, therefore, evidence of its importance in relation to ulceration remains in the domain of “clinical experience”. Thus, in this study, significant foot deformity was defined by whether a clinical judgement had been made that MGF was necessary for a participant. These results support the notion that deformities are indeed a cause of recurrent ulceration. However, further studies, preferably with more quantitative measures, would be of great interest in this area. It is important to stress that in the context of this study, recommendation of MGF was used as a surrogate measure of significant foot deformity and no attempt has been made to correlate the actual use and appropriateness of the prescribed footwear with ultimate outcomes.
Anecdotally, the authors have observed that many people attending the foot clinic have difficulty adhering to treatment. This is, of course, the very nature of diabetic foot disease, where the absence of pain prevents the normal behavioural response to injury and illness. The data from this study showed that attendance of scheduled clinic appointments was worse in those who experienced ulcer recurrence. Although clinic attendance may not be entirely indicative of overall concordance, modifying behaviours and improving strategies to help people with diabetic ulcers adhere to treatment may be as valuable as better clinical management in preventing re-ulceration. Judging from the small percentage of total variance explained by the factors examined here, it is axiomatic that more systematic studies are needed to explore other possible factors.
Conclusion
Studying diabetic foot ulcer recurrence has allowed us to identify the characteristics of those who experience recurrent foot ulceration over time and examine possible predictors. This study found that previous foot ulceration, ischaemia and foot deformity were associated with an increased risk of recurrent foot ulceration, but these risk factors did not explain a high degree of variance overall. Other factors, in other populations, related to re-ulceration remain to be determined by future studies. The authors concur with Pound et al (2005) that if ulcer-free survival is to be used as a measure of the effectiveness of a foot clinic, then patient population characteristics must be taken into consideration.
