Present and new techniques and devices in the treatment of DFU: a critical review of evidence
Finn Gottrup, Department of Dermatology, D42, Copenhagen Wound Healing Center, Bispebjerg University Hospital, Copenhagen NV, Denmark.
Management of foot ulcer in individuals with diabetes remains a major therapeutic challenge throughout the world. We performed a critical review of evidence of present and new techniques and devices in the treatment of diabetic foot ulcer. The golden standard for optimal evidence in the Cochrane system is level I – randomized controlled trials, and meta-analyses of several randomized controlled trials. Available evidence on different types of wound debridement; use of antimicrobials; use of dressings in wounds; topical negative pressure, hyperbaric oxygen treatment; electrical, electromagnetic, laser, shockwave, and ultrasound therapies; growth and cell biology factors; cell products and tissue engineering; bioengineered skin and skin grafts; and adjuvant therapies were evaluated. The results of this review show that there is limited evidence on the highest level to justify a change in routine clinical practice. There is a paucity of high-quality evidence, because the studies are often based on inadequate sample size, short follow-up, nonrandom allocation to treatment arms, nonblinded assessment of outcomes, poor description of control, and concurrent intervention. The heterogeneity of the population (of both people and ulcers), with multiple factors contributing to both ulcer onset and failure to heal, makes the trial design difficult in this field. Another fundamental reason for the lack of evidence is the general use of the outcome measure ‘complete healing’. In conclusion, when the results of this updated review are taken together with those of the earlier reports, they provide limited evidence to justify a change in routine clinical practice. For this reason, there is an urgent need to increase the quality of clinical studies. A re-evaluation of which type of research is acceptable for producing evidence in the wound area may be important in the future. Copyright © 2012 John Wiley & Sons, Ltd.
Management of foot ulcer in individuals with diabetes remains a major therapeutic challenge throughout the world. There is an urgent need for a review of strategies and treatments for this patient group to reduce the burden of care in an efficient and cost-effective way. The question is which type of intervention, technology, and dressings are the best from those available. Before implementation in common practice, the effectiveness of different treatments, procedures, or products should be rigorously assessed and determined. Consequently, there has been an increasing focus on the available evidence for the effectiveness and quality of the type of intervention, technology, and especially dressing materials used in the wound healing area [1, 2]. The accepted way to assess effectiveness and quality in health care is the evidence-based practice (EBP), which focuses on the use of the current best evidence in the decision-making and the care of the individual patient. The present golden standard for optimal evidence in the Cochrane system is level I – randomized controlled trials (RCTs), and meta-analyses of several RCTs.
However, there are many difficulties and controversies using RCTs in wound healing research . An upcoming report from the International Working Group of the Diabetic Foot, which is an update of an earlier systematic review , has shown that evidence presents considerable challenges in the wound area .
Evidence in wound care/management
Although the term ‘evidence’ is quite informal, EBP aims to apply the best available evidence to support clinical decision-making with practitioners reviewing information from rigorous data instead of relying on single observations or customs.
EBP means integrating individual clinical expertise with the best available clinical evidence from systematic research. Different types of evidence are available, and their relative importance for changing clinical practice has been organized into a hierarchy such that a well-constructed meta-analysis of several well-conducted randomized clinical trials is considered by many to be the most robust type of evidence on which to base changes in clinical practice .
The question for wound care practitioners is which type of intervention, which type of technology, and which type of dressing materials are the best from the perspective of a single patient or group of patients, with the primary focus on healing and the absence of complications.
This article focuses on the present evidence of selected types of treatments, techniques, and devices from the perspective of diabetic foot ulcer (DFU) and emerging therapies. Present review literature has been examined, and future perspective of evidence strategies is discussed
The present review is based on published Cochrane reviews and recent systematic reviews of treatment of DFUs [4, 5] and the systematic overview by the patient outcome group (POG) with regard to outcome in RCTs on wound management in hard to heal ulcers .
What is the available evidence?
Debridement of wounds is considered essential as a part of wound bed preparation to improve healing by production of granulation tissue. There are substantial numbers of techniques suggested in the literature ; however, very few comparative studies on this topic are available.
The scientific evidence to confirm the benefit of sharp debridement in the literature is not strong. For the debridement of surgical wounds, there is insufficient valid evidence based on research to recommend any on this particular method . A Cochrane review  of DFU concluded that surgical debridement shows no significant benefit over standard treatment of pressure relief and dressings.
The evidence of the benefit of sharp debridement in DFUs is based on a single study comprising a subgroup analysis of cases in an RCT of another intervention . Healing at 12 weeks more likely follows a more vigorous debridement.
Some studies of relatively poor quality have been performed, in which an effective debridement of wounds by maggots was reported [11, 12]. In another article, no improvement of healing and amputations was found . A Cochrane review  on DFU has concluded that there is insufficient evidence with regard to the effect of larvae therapy.
A comparative study using the hydrotherapy Versajet® was identified . Wound debridement time was shorter with hydrotherapy, whereas no benefit was observed in healing at 12 weeks.
The available data for the different debridement procedures demonstrate that these therapies are not effective . This confirms the findings of a recent Cochrane review . It is essential that better studies are undertaken, and until the benefit of these interventions is confirmed, their cost should be considered before they are recommended for routine clinical practice.
There is little evidence on the highest level that sharp or maggot debridement has any effect. In a Cochrane review , it was suggested that a hydrogel dressing may increase healing rate of DFU compared with gauze. However, this finding has not been confirmed in recent studies.
Use of antimicrobials
There is an increasing interest regarding the use of antimicrobials especially in the management of DFUs because of the high infection rate and the ongoing debate on biofilms especially in the perspective of the increased awareness of methicillin-resistant staphylococcus aureus (MRSA) and extended spectrum beta-lactamaser (ESBL) and the relative ineffectiveness of systemic antibiotic treatment in nonhealing ulcers.
In terms of healing by 24 weeks, a recent RCT of good quality reported no differences among the following three products: Aquacel® (hydrofibre), a surface antiseptic (Inadine®), and a nonadherent product (N-A®) . Similarly, another large, but nonblinded, RCT reported no difference between a silver impregnated carboxymethylcellulose hydrofibre dressing (Aquacel Ag®) and an alginate in the incidence of complete healing over 8 weeks, healing velocity, change in ulcer area, and time to healing . In a recent systemic review , it was concluded that in the absence of supportive evidence, the additional expense of the newer hydrofibre product was not justified. A single nonblinded RCT on the use of superoxidized solution (Dermacyn®)  focused on the infected surgical wounds of the foot. Although the results suggested an improvement with regard to healing using superoxidized solution or povidone iodine, the study was found to be of poor methodological quality. Despite its widespread use in clinical practice, it was possible to identify only one study in which the effect of honey was assessed in a controlled way. This study was a small, nonblinded study of poor design and reported no difference in the use of povidone iodine .
The use of surface antimicrobials (tobramycin beads) in the wound after a forefoot amputation was studied in a nonrandomized cohort study [5, 19]. The positive effects described could, however, have been a result of confounding influences.
There is little evidence on the highest level for the use of antimicrobials in wound care. The POG concluded that healing may not be the most proper outcome for the evaluation of an antimicrobial strategy in the treatment of wounds .
Use of dressings in wounds
There are a substantial number of various types of dressings available for treating wounds. However, they have frequently not been able to met expectations with regard to evidence-based medicine.
In a review from 2008 , it was suggested that hydrogel also may hasten healing, but no other studies on hydrogel or other dressings were identified to support the statement. No other specific differences between uses of dressings in high-quality trials have been published. Game et al.  concluded that there is little evidence that any one dressing or wound application has a place in the management of chronic foot ulcers in diabetes.
There is little evidence on the highest level, except for hydrogel dressing increasing the healing rate of DFU compared with gauze.
Topical negative pressure
Negative pressure is one of the most interesting emerging therapies that have been studied in more than 1000 publications frequently in acute and postoperative wounds. Topical negative pressure (TNP) has been introduced during recent years into the treatment of nonhealing chronic wounds.
Topical negative pressure
A significant benefit both with regard to healing rate and healing time has been reported [20-22] in individuals with diabetes and postamputation wounds and cavity ulcers. Recent studies have been focused on reduced time to granulate , reduced time to closure, and increased incidence of healing . Despite that a sufficient number of patients were included in this study, the NICE Guidelines Development Group in UK (NICE 2011) concluded that the available data were insufficient to demonstrate that the use of negative-pressure wound therapy (NPWT) was cost-effective and advised that it should be used only in exceptional circumstances. A similar conclusion was presented in a Cochrane review , and both reviews asked for better quality research and further evidence. The analysis from the POG concluded that a correct primary endpoint may not be healing, but percentage granulation, and that all outcome assessments should be undertaken by observers blinded to the randomization group.
Vacuum compression therapy has been suggested as a significant benefit on postoperative wounds . Recent studies have reported a significant reduction in wound area , a significantly higher incidence of complete healing , and a significant reduction in time to healing . These studies, however, were generally of poor quality.
There is insufficient evidence to demonstrate cost-effectiveness of TNP, and it should only be used in special occasions.
Hyperbaric oxygen treatment
Oxygen is vital for healing of wounds, and hyperbaric oxygen treatment (HBOT) has been debated for decades as a treatment to achieve healing in complex wounds. Data have suggested that systemic, but not topical, hyperbaric oxygen may reduce the rate of major amputation in people who have chronic foot ulcers complicating diabetes [5, 29]. More recently, small studies have shown a significant difference between groups in terms of the number of major amputations  and healing . These studies were retrospective and nonblinded, and as a consequence, there was a limitation with regard to valid conclusions. Recently, a high-quality double-blind RCT demonstrated a significantly (p = 0.03) improved outcome in the intervention group , which were more likely to heal within 12 months: 25/48 (52%) versus 12/42 (27%). Although the limitations of this study have been listed , its potential implications are far reaching, and the study needs to be repeated with full health economic analysis and with an attempt being made to define the population most likely to benefit. A recent systematic review by the NICE Guidelines Development Group in UK concluded that the available data were insufficient to demonstrate that the intervention was cost-effective (NICE 2011).
There is increasing evidence on the highest level for an effect of HBOT in DFU patients. However, the cost-effectiveness of the treatment is still discussed, and there is limited information with regard to potential responders.
Electrical, electromagnetic, lasers, shockwaves, and ultrasound
Several studies related to the effect of electrical, electromagnetic, laser, and ultrasound therapies have been published. There are interesting data from animal/laboratory research regarding these technologies' potential value in wound healing. However, there have been substantial challenges to prove the value in nonhealing wounds. Only few RCTs comparing groups are found, and no convincing evidence has been demonstrated . Recent studies in electrical stimulation have shown a reduction in ulcer area at 45 days but not at 60 days  and a significant reduction in ulcer area . These studies were, however, small and methodologically of poor quality. Recently, two studies on shockwave therapy have shown no difference in healing, although the time to healing in the small number of patients that were healed was reported to be significantly shorter in the intervention group . In another low-grade study, shockwave treatment showed improved healing compared with hyperbaric oxygen . None of these studies give convincing evidence for this type of treatment to be introduced in clinical practice.
There is insufficient evidence on the highest level to demonstrate that electrical, electromagnetic, laser, shockwave, and ultrasound therapies have any effect on DFU.
Growth and cell biology factors
There have been a substantial number of studies in the perspective of wound bed preparation to stimulate granulation and matrix formulation in nonhealing wounds to achieve healing.
There are a number of studies on growth factors but only a few major RCTs. The initial RCT of becaplermin in noninfected neuropathic ulcers  indicated a significant effect, and this was confirmed in the later definitive phase III study . However, a subsequent equally large RCT had negative results, but the results of this study have not been published. A study using basic fibroblast growth factor has reported a significant difference in reduction of wound area but no difference in healing rate . The difference was, however, not significant when intention to treat (ITT) analysis was used . Promising date has been published for the effect of epidermal growth factor (EGF), and plasmid containing the gene for phVEGF165 with regard to ulcer area and prevalence of granulation tissue [41-43], but more data are needed. One large RCT on a collagen/oxidized regenerated cellulose dressing product (Promogran®), which is believed to modify the balance between matrix metalloproteinases and their tissue inhibitors, failed to confirm an effect on healing . There have been a small number of high-quality trials on the effect of GCSF even though they have been directed primarily at the eradication of infection more than wound healing. Amputation was recorded as an endpoint, and an apparent trend towards increased limb salvage rate was noted in a separate meta-analysis . Studies on protease modulating dressings  and the dressing added to an autologous platelet supernatant  have shown benefits in healing, but the data are difficult to interpret . Studies on talactoferrin, a recombinant human form of the breast milk protein, lactoferrin, acellular dermal regenerative tissue matrix, and Chrysalin, a ligand for thrombin binding sites, were poor or showed no differences between the groups [48-52]. However, most of these studies with various success have been performed in plantar neuropathic foot ulcers in which offloading altered biomechanics is the cornerstone for treatment success.
There is insufficient evidence on the highest level to demonstrate that growth and cell biology factors have any effect on the DFU.
Cell products and tissue engineering
There is a great interest in platelet-derived products, stem cells, and bioengineered skin products because they have shown initial promising results in selected preclinical studies especially with regard to patients with DFU. Studies on the use of platelet-derived products and stem cells have been limited by methodological problems and uncontrolled and observational studies, and no firm conclusion has been drawn. Recent studies have shown a significant improvement in healing rate [53-55]. Further studies will reveal the potential of this treatment.
There is insufficient evidence on the highest level to demonstrate that cell products and tissue engineering have any effect on the DFU.
Bioengineered skin and skin grafts
Several articles have been published in this area. Both dermal fibroblast cultures and fibroblast/keratinocyte co-cultures were associated with improved healing rate of clean neuropathic ulcers. The major concerns in these studies were the variable rates of healing in the placebo groups and the lack of comparison with the outcome of using split skin grafts. In a recent RCT, fibroblast/keratinocyte co-culture (Apligraf®) was compared with polyamide and saline-moistened gauze . An improvement in healing rate was found, but methodological problems made these findings questionable. Recent studies focusing on the effect of a keratinocyte delivery system and split skin grafting have shown better healing rate, but the studies were small and of poor methodological quality [57, 58].
There is insufficient evidence on the highest level to demonstrate that bioengineered skin and skin grafts have any effect on the DFU.
Herbal preparations in China have been studied in patients with necrotic/gangrenous ulcers without any effect on ulcer granulation and incidence of amputation . In Iran, herbal extract (ANGIPARS™) has been studied in DFU patients both administrated oral and locally  and intravenously . Significant reductions in surface area were reported within two intervention groups but not for the control group. The quality of the trial methodology was, however, not good .
There is insufficient evidence on the highest level to demonstrate that herbal extracts have any effect on the DFU
Evidence in the wound area has always been a challenge [62, 63], and the results of this review also show that there are limited evidence on the highest level to justify a change in routine clinical practice. There may be different reasons for this. Earlier systematic reviews have indicated that there are substantial deficiencies in the quality of clinical research (www.cochrane.org, www.nice.org.uk). Also, wound management has a paucity of high-quality evidence, because the studies are often based on inadequate sample size, short follow-up, nonrandom allocation to treatment arms, nonblinded assessment of outcomes, poor description of control, and concurrent intervention. The heterogeneity of the study population (of both people and ulcers), with multiple factors contributing to both ulcer onset and failure to heal, makes the trial design very difficult. Because an intervention is likely to be directed at only one of these factors, it is possible that an agent can have an effect, but this may be hidden. Also, analysis of evidence presents considerable difficulties in this field. The complexity of the clinical condition and its frequently slow response to intervention pose particular problems in trial design . Thus, a therapy or dressing may have a significant stimulatory effect for only a definite period during the course of the lifespan of the ulcer being treated. Antibiotics may have a clear benefit in the management of any ulcer that is infected but may not influence the overall rate of healing if this is critically affected by other factors, such as poor peripheral circulation. Nevertheless, these problems do not mean that studies of effectiveness should not be undertaken; they just need to be designed with great care.
Another fundamental reason for the lack of evidence at the highest level is the validity of the outcome measure used . To date, the obvious outcome measure in evaluating interventions in wound healing has been complete healing. Healing, time to healing, avoidance of amputation, and survival may be the most clinically relevant measures, but these may not necessarily reflect the effect of a chosen intervention and therefore may not be the only appropriate outcome in wound healing studies. Other endpoints such as clinical, intermediate, and surrogate outcomes (e.g. infection rate, bacterial contamination, wound pain, resource utilization, and economic cost) also need to be considered. Because ulcers also typically take many weeks to heal, it is likely that the factor most responsible for delayed healing may be different at different stages of the process, and this means that an agent may be effective at one stage, but not necessarily all the time. This leads to uncertainty concerning the choice of endpoints. Clinically relevant endpoints (such as ulcer healing or amputation) may mean more in practice but may be only partially dependent on any effect of the chosen intervention. On the other hand, surrogate endpoints (such as change in wound bed appearance or in ulcer area) may be more closely related to the effect of the product being tested but have little relevance to clinical outcome.
The insufficient sample size and outcome measurements may result to a type II error or ‘false negative result’, leading to rejection of treatments or products with a positive effect. From the above-mentioned reasons, this is a fundamental problem in the wound area. In clinical practice experience, case reports and cohort studies may show an effect of a treatment or a product, but almost all evaluations that are based on level IA evidence in the Cochrane system and that use as an outcome measure ‘complete healing’ have resulted in the statement ‘insufficient evidence’. This statement can be interpreted that a given treatment/strategy is of no value for the individual patient and that, as a consequence, it does not receive reimbursement and will not be used. However, it has to be remembered that these studies showing ‘insufficient evidence’ with regard to the specific outcome chosen are telling nothing about the usefulness or value of a specific treatment.
These difficulties in the Cochrane collaboration and similar survey systems resulted in low-evidence evaluation for treatment of DFU (and other types of wounds). It is concluded that there is lacking evidence on the highest level, and it is nearly almost concluded: ‘There is insufficient evidence of the effects of …’ and ‘more research is needed to evaluate the effects of range of …’ Clinically, however, the relevance of such statements may be questionable in all cases. For a clinician working with DFU patients, it is surprising to realize from a Cochrane review on debridement of DFU that hydrogel dressing is the only treatment, which has shown evidence on the highest level . Surgical and larval debridement has shown insufficient evidence, and the conclusion is that more research is needed. Do these results mean that we should use hydrogel dressing and not surgical/larval debridement for DFU until the evidence is available? Even for the two topics for which a significant benefit (TNP and HBOT) was shown, the NICE Guideline Development Group and the Cochrane Collaboration concluded that the available data were insufficient to demonstrate any cost-effectiveness, and these treatments should only be used in exceptional circumstances. The message from these authorities is that clinical benefit is not sufficient unless it is supported by solid health economic data. These conclusions raise a major concern for wound treatment in the clinical world, because the lack of modern treatments and technologies in wound management will result in a dramatic decrease of the quality of life and an increased suffering of the patient.
How can this be improved in the future? How can we get the necessary evidence, and how can we minimize the risk of errors of decisions? This is not a simple issue, but the focus should be on the main factors involved: sample size, type of outcome measures, and a re-evaluation of which type of research is acceptable for evidence. Also, there are some minimum data requirements for demonstrating comparability of procedures and devices, which have to be filled (Table 1). Before these topics have been sufficiently discussed, accepted, and implemented, there will be evidence problem in the wound area. In the POG document , statements of how to improve evaluation of treatment strategies regarding outcome, to meet the requirements for evidence-based information in wound management, have been described. Similar documents providing recommendations for other parts of the process to achieve evidence should in the future be produced to establish evidence usable in the daily clinical life. Finally, it may be stated that, in general, RCTs are the correct method to establish evidence. However, in a few situations, the extended definition of Sackett et al.  may be more relevant in the wound area. Evidence-based medicine is not restricted to randomized trials and meta-analysis but involves exploration of all types of best external evidence with which to answer our clinical question. Prospective cohort studies may be particularly helpful as background information on the natural progression towards healing, especially when cost and resource use are the major outcomes of interest.
Table 1. Data requirements for demonstration of comparability of procedures and devicesa
|Similar conditions of use|
|Similar specifications and properties (e.g. tensile strength, viscosity, surface characteristics)|
|Must be of similar design|
|Similar deployment methods (if relevant)|
|Similar principles of operation|
|See also the harmonized norms for wound dressings and medical devices (further information is available on the European Wound Management Association (EWMA) website)|
|Use same materials in contact with the same human tissues or body fluids|
|Used for the same clinical condition or purpose|
|Used at the same body site|
|Used in similar population (including age, anatomy, physiology)|
|Similar relevant critical performance in terms of the expected clinical effect for its specific intended use|
When the results of this updated review are taken together with those of the earlier reports, they provide limited evidence to justify a change in routine clinical practice. There are a substantial number of emerging technologies of potential value in the treatment of complex wounds, but there is an urgent need to increase the quality of clinical studies. RCTs should be optimized by using sufficient sample size (to avoid a type II error), and correct outcome measures should be chosen. A re-evaluation of which type of research is acceptable for producing evidence in the wound area may be important.
Conflict of interest