Randomised controlled trial of the use of three dressing preparations in the management of chronic ulceration of the foot in diabetes

Article history

The research reported in this issue of the journal was commissioned by the HTA programme as project number 01/74/03. The contractual start date was in June 2003. The draft report began editorial review in October 2007 and was accepted for publication in January 2009. As the funder, by devising a commissioning brief, the HTA programme specified the research question and study design. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the referees for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

None

Copyright statement

© 2009 Queen’s Printer and Controller of HMSO. This monograph may be freely reproduced for the purposes of private research and study and may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NETSCC, Health Technology Assessment, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2009 Queen’s Printer and Controller of HMSO

Introduction

Ulceration of the foot of people with diabetes (diabetic foot ulcers) is common, and widely acknowledged to be a source of major distress and morbidity in a predominantly elderly population, as well as an enormous drain on health-care resources. 1–4 Not only does diabetes make the foot more liable to ulceration, but it also impairs the process of healing, and diabetic foot ulcers readily develop into chronic wounds. There are approximately 24,000 admissions for diabetic foot ulcers each year in the UK,5 and approximately 15% of all ulcers in the UK result in some form of amputation. 6 Diabetic foot ulcers also have a significant negative impact on health-related quality of life (HRQoL). 1,4

While the pathobiology of chronic wounds remains poorly understood, there is no logical framework to underpin many strategies of care. 7 The choice of dressings, in particular, is largely empirical and based more on professional experience and preference than on evidence of proven efficacy. The principal reason for this is the lack of available evidence, which is itself partly the result of the difficulty in conducting controlled trials in this field.

Evidence base for effectiveness of management strategies

The paucity of the evidence base for the treatment of diabetic foot ulcers has been highlighted in several recent reviews. 6–11 O’Meara et al. 12 could find no good evidence to substantiate the use of any of the preparations in widespread use, and this finding has been confirmed in a recent systematic review undertaken by the International Working Group on the Diabetic Foot of the International Diabetes Federation. 13 The effectiveness of some of the more recently introduced therapeutic agents (including growth factor preparations and bioengineered human skin products) has been suggested in some (but not all) industry-funded trials and remains to be confirmed. Even if effective, they are expensive in terms of both product costs and professional time, and in the absence of robust evidence of cost-effectiveness, they have not been widely adopted in the UK. The more recent introduction of a number of silver-impregnated dressings has been undertaken without evidence of effectiveness in this population. A recent trial of one such product suggested that the product tested was no more effective than conventional therapy. 14

It is very necessary, therefore, to establish whether any difference can be demonstrated between the efficacy and cost-effectiveness of products which are currently in widespread use, including those which are well-established and of low material cost as well as those which are newer and more expensive. If any product is shown to have greater effectiveness and is relatively cost-effective, then this evidence should be used to underpin routine clinical practice in the UK. If no difference in effectiveness can be demonstrated, then clinical choice should be based primarily on issues of patient acceptability and on cost. Moreover, if no difference in effectiveness can be shown, the data will provide an invaluable benchmark in the later evaluation of newer technologies.

The aim of this study was, therefore, to compare the effectiveness and cost-effectiveness of three dressing products which are widely used in routine management in the UK: comparing two traditional preparations, a non-adherent, knitted, viscose filament gauze product [N-A^® (Johnson & Johnson Medical, Berkshire, UK)] and an iodine-impregnated dressing [Inadine^® (Johnson & Johnson Medical, Berkshire, UK)], with a newer product of higher unit cost [Aquacel^® (ConvaTec Ltd, Middlesex, UK)]. N-A is thought to be metabolically inert, and is designed simply to be a non-adherent dressing which is easily changed, with minimal discomfor t and trauma to the regenerating wound bed. Inadine is a knitted viscose fabric impregnated with a polyethylene glycol base containing 10% povidone-iodine, equivalent to 1.0% available iodine. The potent antimicrobial, povidone-iodine, is released when in contact with wound fluid. Aquacel is marketed as a textile fibre which is bonded into the form of a fleece. The dressing is designed to ‘absorb and interact with wound exudate to form a soft, hydrophilic, gas-permeable gel that traps bacteria and conforms to the contours of the wound while providing a micro-environment that is believed to facilitate healing’. 15 In one small, short-term, randomised trial Aquacel has previously been shown to be more effective in the management of deeper diabetic foot ulcers than saline-moistened gauze,16 but saline-moistened gauze is rarely used in clinical practice in Europe.

Issues surrounding the choice of outcome measures

The principal aim of dressing products is to promote healing, and hence the primary measure of effectiveness should be ulcer healing. Moreover, newly healed ulcers often break down within the first few weeks and so the chosen definition of healing should take this into account. Secondary measures of effectiveness comprise those that are ulcer-related, process-related and patient-related. Ulcer-related outcomes include time to healing, adverse events, incidence of recurrence and improvement in the appearance of the wound bed, incidence of secondary infection of the index ulcer and incidence of both minor and major amputation. Process-related outcomes include those relating to frequency of dressing changes.

Patient-related outcomes include mortality, pain, serious adverse events (SAEs) and quality of life (QoL). Some assessment of the profound implications of diabetic foot ulcers on mood and QoL has been produced in recent years,17 but even though the need for a robust condition-specific QoL assessment tool has been highlighted, none has yet been fully published and validated for diabetic foot ulcers. The (Rand) short form 36 (SF–36)18 has been shown to discriminate between those with and without ulcers, but not between those whose ulcers are either active or healed.The Euroqol–5D (EQ–5D)19 has been shown to discriminate between patients with active and former ulcers, despite its simple structure.One factor likely to contribute significantly to the frustration and anxiety of having an ulcer is dependence on the frequent attention of health-care professionals. There are few condition-specific tools in this area, but work on the Cardiff Wound Impact Schedule (CWIS) has demonstrated poor QoL responses from patients with active ulceration,20–22 reflecting the qualitative work of Brod. 23

Cost-effectiveness

Assessment of cost is complex because of its dependence on material unit cost, the frequency of dressing changes and the time of professional staff. 11,17 In the case of traditional, less expensive dressings, the relative contribution made by professional time is potentially much greater, especially if healing is delayed. Similarly, the possible need for more frequent changes of traditional dressing products can outweigh the relatively low material costs because of the professional time involved. 11 On the other hand, it would be wrong to assume that all dressing changes are actually performed by professional staff in routine clinical practice. Unpublished data from Nottingham University Hospitals indicate that 55% of dressings are undertaken in the community by non-professional staff.

In assessing the cost-effectiveness of treatment strategies, and in particular dressings, cognizance has therefore to be taken not only of the unit cost of the dressings but also of the number and frequency of dressings used and the time of professionals and others involved in the process. Furthermore, the implications of non-healing have to be encompassed in the assessment of relative cost-effectiveness. It has been estimated that up to 15% of patients with diabetic foot ulcers require an amputation6 – with direct costs ranging from USD 20,000–60,00011 – which emphasises the need to maximise effective treatment.

The health economic evaluation in this study was planned primarily from the perspective of the UK NHS, but with some consideration given to travel costs incurred by patients. The full impact that the treatment and care of diabetic foot ulcers has on the family, friends and carers of the patients was not considered. The costs associated with dressings were assessed for the duration of active participation in the trial, and were not extended beyond healing or withdrawal.

Objectives

The overall objective of this study was therefore to determine the comparative effectiveness and cost-effectiveness of three dressing products in common clinical use for patients with diabetic foot ulcers in the UK, as well as the feasibility and consequences of less frequent dependence on dressings by health-care professionals. This study had five specific objectives:

1. To test whether a modern dressing product is more clinically effective than traditional dressings in the treatment of diabetes-related foot ulcers. The dressings compared were: a simple non-adherent preparation (N-A), a widely used modern antiseptic preparation (Inadine), and a new hydrocolloid preparation of higher unit cost (Aquacel). All three dressings are widely used in clinical practice in the UK.

2. To investigate changes in the condition of each ulcer during the study period associated with each dressing, and the incidence of recurrence after healing.

3. To determine the relative cost-effectiveness of the three dressings by:

   1. identifying and assessing the cost components associated with the treatment of diabetic foot ulcers

   2. assessing the relative effectiveness of the three dressings, based on findings from the randomised controlled trial

   3. estimating the relative cost-effectiveness of the three dressings

   4. determining the extent to which the cost-effectiveness is affected by changes in costs and effects.

4. To assess patients’ HRQoL, physical and social functioning, and pain associated with each of the dressings.

5. To investigate the contribution made by patient and carer in terms of involvement with self-care.

Design

This was a multicentre, prospective, observer-blinded, parallel group, randomised controlled trial, with three arms. Patients with ulcers were randomised to treatment with N-A, Inadine or Aquacel.

The study was undertaken in accordance with the Declaration of Helsinki and followed the guidelines published by the Medical Research Council. The conduct of the study was supervised by a Trial Steering Committee with an independent chairman, and issues of recruitment, randomisation, retention and adverse events were scrutinised by an independent Data Monitoring and Ethics Committee.

Randomisation

Randomisation was stratified both by centre and by size, using a block size of nine. Randomisation was stratified across the whole population by ulcer area into three groups: 25–100 mm², 101–250 mm² and 251–2500 mm². Randomisation lists were created using spss (SPSS Inc., Version 14), using blinded dressing codes. The lists were held at Cardiff University and each recruiting centre telephoned a designated number during working hours; they were required to identify the centre and size of wound only. Records of the allocation details were kept at Cardiff University for data verification and checking at monitoring visits.

Setting

Patients were recruited from those attending, or newly referred to, established expert multidisciplinary clinics for the management of diabetic foot ulcers in Blackburn, Cardiff and Newport (University Hospital of Wales, Llandough and Royal Gwent Hospitals), Hull, Ipswich, Nottingham and Derby (City and University Hospitals, Nottingham and Derbyshire Royal Infirmary), London (Kings College Hospital), Leeds (Leeds General Infirmary and St James’ Hospital), Swansea (Singleton and Morriston Hospitals) and Bristol (Southmead and Frenchay Hospitals) – each of which receive in excess of 100 new referrals each year. These centres reflect both NHS Trusts and University Teaching Hospitals across the UK.

Target population

Patients over age 18 with either type 1 or type 2 diabetes with a chronic (present for at least 6 weeks) full-thickness foot ulcer (on or below the malleoli) not penetrating to tendon, periosteum or bone, and with a cross-sectional area between 25 and 2500 mm² were invited to participate. If there was more than one ulcer on the foot, the largest ulcer that conformed to the inclusion criteria was selected as the index ulcer.

Inclusion criteria

• Type 1 or 2 diabetes.

• 18 years of age or more.

• A foot ulcer which had been present for at least 6 weeks and had a cross-sectional area of between 25 and 2500 mm².

• Able and willing to give informed consent.

• Reasonably accessible by car to the hospital base.

• Under routine review by the multidisciplinary clinic.

Exclusion criteria

• Those with a known allergy to any of the trial preparations (including iodine).

• Any ulcer on either foot extending to tendon, periosteum or bone.

• Infection of bone.

• Soft tissue infection requiring treatment with systemic antibiotics.

• An ulcer on a limb being considered for revascularisation.

• Those chosen for management with a non-removable cast without a dressing window.

• Gangrene on the affected foot.

• Eschar which was not removable by clinical debridement.

• Those with evidence of a sinus or deep track.

• Those in whom the hallux had been amputated on the affected side (preventing the measurement of toe pressure).

• Those with an ankle:brachial pressure index (ABPI) of less than 0.7 or toe systolic pressure less than 30 mmHg.

• Ulceration judged to be caused primarily by disease other than diabetes.

• Patients with any other serious disease likely to compromise the outcome of the trial.

• Patients with critical renal disease (creatinine greater than 300 μmol/l), and those receiving immunosuppressants, systemic corticosteroid therapy (other than by inhalation) or any other preparation which could, in the opinion of the supervising clinician, have interfered with wound healing.

• Those living at such a distance (generally further than 10 miles) from the clinic as would have made frequent assessment visits inappropriately expensive and/or impractical.

• Those who withheld consent.

Baseline assessment

Those who satisfied the inclusion and exclusion criteria and gave written informed consent to participate were assessed by a research nurse (the term research nurse is used to apply to any health-care professional involved in the conduct of the study, including research podiatrists) and their basic demographic and medical details noted.

The foot was examined and the following additional information recorded:

• toe pressure (systolic pressure in the hallux)

• ABPI

• peripheral sensation using a 10-g Semmes–Weinstein monofilament at four specified sites on the sole, as well as vibration perception threshold.

Following debridement in the clinic, details of the ulcer were recorded, including:

• history (cause, duration)

• pain at or close to the ulcer (10-cm visual analogue scale)

• cross-sectional area using a sterile marked acetate sheet

• the appearance of the surface of the wound: percentage granulation, percentage slough, percentage necrosis

• a digital image was made.

Clinical care

Patients remained under the supervision of the staff at their respective multidisciplinary clinic throughout the study. The frequency of clinic visits was determined by clinical need and was not affected by the trial. Ulcer management was in line with current guidelines for good practice, including appropriate and regular use of debridement and with a removable fibreglass or polyester boot being recommended for off-loading. In the absence of any significant deterioration or adverse event, clinic staff made no decision concerning dressings. Dressings were removed prior to examination by investigators who were not involved in the conduct of the trial and who were blind to the randomisation group.

Withdrawal

Participants were withdrawn from the study at their request, in the event of a significant adverse event (including deterioration in the condition of the ulcer), other serious illness (such that it was either not appropriate or not possible for them to remain in the study) and protocol violation. Protocol violation was deemed to have occurred if two or more consecutive non-trial dressings had been applied (Figure 1).

FIGURE 1.

Study design. If an ulcer appeared healed at any point up to and including 24 weeks, its status was checked after a further 2 and 4 weeks. It was only labelled as healed if it did not recur in that time. Healing was then verified by a blinded assessor and the patient was asked to complete the questionnaires. If the ulcer recurred within 4 weeks of apparent healing and at any point up to 24 weeks, the patient was re-entered into the study, using the allocated dressing. If the occurrence happened at or after 24 weeks, the ulcer was recorded as unhealed.

End points

Economic evaluation

A bottom-up approach to costing was employed to construct a profile of costs associated with the treatment and care of diabetic foot ulcers. Only the costs that depended on and varied according to the dressings used were included in the cost profile; those that were fixed and unrelated to dressing choice (such as equipment) were excluded.

Sample size

As healing was the primary objective, this was the basis for the calculation of sample size. Calculation of sample size was difficult because of the paucity of data on the healing rate of different types of ulcer, and although data are available for neuropathic ulcers on the plantar surface, they are inconsistent. Thus, Katz et al. 27 reported 61–89% healing of plantar neuropathic ulcers within 12 weeks, while an earlier meta-analysis of the control arm of published trials of similar (but not all identical) ulcers reported only 24% healing with accepted good clinical practice by 12 weeks, and 31% at 20 weeks. 28 Moreover, neuropathic ulcers with good vascular supply form a minority of ulcers cared for in the UK and, despite the lack of much published information, it is accepted that they heal more quickly than other types. The experience at the City Hospital, Nottingham, was that of all 449 individuals referred in the 4 years between January 2000 and December 2003, only 55% of index ulcers healed without amputation within 6 months of referral. ²⁹ It is on these bases that we calculated that in order to demonstrate a 20% difference in healing between groups, with 80% power, and with alpha = 0.05, and allowing for 25% dropout, 300 recruits were required. This was based on equal distribution of the sample to the three arms of the study. The N-A group was treated as the reference arm of the study, with an anticipated healing rate of 30%. The size was powered to indicate a 20% increase in healing for those in the Inadine group (50% healed at 24 weeks), and a 25% increase for those receiving Aquacel (55% healed at 24 weeks).

Data management

All files were checked by hand, with outstanding data questions addressed with each individual site. All data were entered into spss version 14 by research staff at the Department of Wound Healing, School of Medicine, Cardiff University and random checks were completed on an entry basis. All variables were checked for valid entries, i.e. within the expected range for that variable. Fifty per cent of files were double checked for errors by a different research assistant: the error rate was less than 1% over all entries for all variables over 165 files.

Deviations from the planned protocol

EQ–5D was excluded to reduce patient burden in relation to questionnaires. It was anticipated that SF–36 scores could be converted into SF–6D scores in order to assess impact on HRQoL and utility scores for derivation of quality-adjusted life-years (QALYs), if considered appropriate.

Recruitment, retention and primary outcome

A total of 317 patients were recruited to the trial, with relatively equal allocation of different dressings to each of the nine centres (Table 1).

Allocation to the different dressings was also relatively equal when analysed by cross-sectional area at baseline (Table 2). As there was no statistical difference between the groups in terms of distribution by ulcer size at baseline, the two larger groups were combined for the purposes of analysis such that there were two final groups of roughly similar size (Table 3). There remained no difference between groups.

The index ulcers of 135 participants (42.6%) healed within the 24-week intervention phase (Table 4). Eighty-eight participants were withdrawn (27.8%), which is more than originally anticipated. Two hundred and twenty-nine participants completed the full study, however, and were evaluable (meeting the 80% power target) (Table 4). There were 19.4% withdrawals for Inadine, compared with 29.1% for Aquacel and 34.9% for N-A, and this difference was statistically significant different (Table 5). The flow of patients through the study is outlined in Figure 2.

FIGURE 2.

Consort diagram for HTA dressing trial for diabetic foot ulceration.

Demographics of participants

The distribution of baseline demographics between the groups was very similar by intervention (Table 6). The proportion of male to female participants was higher than expected, with a 3.2:1 ratio in the study overall. It should be noted that one subject, being managed with N-A, underwent gender realignment during the course of the study and is not listed in the table as either male or female. The majority of the participants presented with type 2 diabetes mellitus, in a ratio of 3.7:1, but with an equal distribution across dressing groups. Mean age was 60 years, and there were no statistical differences in age by group. The mean duration of known diabetes was 16 years, with no differences between groups. Approximately 8% were on diet alone, while one third were on oral hypoglycaemic agents (OHAs), 38% on insulin and 21% on a combination of OHAs and insulin. One third of participants had never smoked, while 17% were current smokers. Sixteen per cent had had a previous cerebrovascular complication, while 39% had cardiovascular complications, 57% had known retinopathy and 21% had nephropathy.

At least one additional significant medical problem was reported by 255 (80%) participants at baseline, which was either unrelated or partially related to diabetes, with the most frequently reported being hypertension (n = 159). Two additional significant medical problems were reported by 175 participants, while three were reported by 112 (Table 7).

The 10 most frequently reported additional medical problems at baseline were:

• hypertension 159 participants

• hyperlipidaemia 69 participants

• asthma 29 participants

• neuropathy 29 participants

• arthritis 28 participants

• angina 24 participants

• depression 13 participants

• anaemia 11 participants

• obesity 9 participants

• hypothyroidism 9 participants.

Figure 3 outlines a summary of the specific details of the reasons for patient withdrawal from each arm of the study. The most frequent reason for withdrawal in each group was ‘adverse events’, followed by ‘protocol violation’. A more detailed account of the reasons for withdrawal, by dressing type, can be found in Appendix 4.

FIGURE 3.

Reasons for patient withdrawal.

A more detailed section on patient withdrawals is presented in Withdrawals.

Total medications prescribed

The number and type of medications taken was recorded for each participant, and changes to medication were noted throughout the trial. Table 8 outlines the details of the total number of medications taken by participants during the 24 weeks of the intervention phase. There are no differences in the median number of medications across the groups, with all groups taking a median of 8–8.5 different types of medication at some stage during the study. The median number of changes of medication was also similar for all groups. A detailed list of these medications and the conditions for which they were prescribed or taken can be found in Appendices 1 and 2.

Ulcer characteristics at baseline

The ulcer-specific details by intervention group are presented in Tables 9 and 10. Sixty-four per cent of participants had had a previous foot ulcer, and 19.9% had undergone a previous amputation, the majority of which were single toe or ray. There was equivalent presentation of ulcers on the right and left limbs, with the majority of ulcers being on either the toe or forefoot. Approximately half of the ulcers were small as per the definition in the protocol (25–100 mm²), with an even distribution across dressing groups. The majority of participants had palpable dorsalis pedis and post-tibial pulses. Seventy-seven per cent of participants had loss of sensation under the first metatarsal head using the 10-g monofilament, while 70% had loss of sensation under the fifth metatarsal head, and 74% and 62% on the plantar aspect of the hallux and heel respectively.

The appearances of the wound bed are outlined in Table 11. There were no significant differences in the clinical condition of the ulcers by intervention. The majority of ulcers were not odorous, with 45% having 100% granulation tissue and no slough. The majority of ulcers had light or moderate exudate, with some associated callus, and approximately 30% reported maceration of the surrounding skin. Just over 20% of the participants reported pain in the region of the ulcer.

Methods of off-loading used at visit 1

The methods of off-loading used at visit 1 were variable. Only 132 (42%) of the 317 patients were provided with a removable fibreglass or polyester casting device, which was defined as the preferred method in the protocol. A further 97 (31%) had a proprietary removable device. There was no difference between dressing groups (Table 12). There was no apparent difference in the use of off-loading devices for those whose index ulcers were on the plantar aspect of the foot or were not (Table 13). For those patients (n =  11) who were identified as having a plantar ulcer and no off-loading at visit 1, five were reported as either awaiting bespoke footwear (n = 3) or considering a casted device (n =  2).

The use of off-loading devices by different centres is shown in Table 14. Two of the centres (centres 7 and 9) did not issue a casted device for any of their participants and one centre (centre 8) only issued one – indicating possible centre differences either in attitudes to off-loading or in the availability of resources.

Primary outcome – incidence of healing

The sample numbers recruited to this study were based on the following assumptions of estimated healing and non-healing at end point by intervention. These assumptions were that by visit 13 (24 weeks) 30% of ulcers managed with N-A would be healed, compared with 50% managed with Inadine and 55% managed with Aquacel. Intention to treat (ITT) analysis was carried out using the last value carried forward method, with strict adherence to the protocol such that only those who attended for a healing verification visit and reported as still healed at 28 days have been coded as ‘healed’ for the outcome classification.

Secondary outcomes – ulcer-related outcomes

Secondary outcomes – patient-related outcomes

Secondary outcomes – process-related outcomes

Health economic analysis

Data not presented

Cost-effectiveness analysis

The only statistically significant difference between the three dressing types was in relation to their acquisition cost and the number of dressings used. The additional cost of Aquacel was not offset by a reduced frequency of dressing changes. There was no significant difference observed in the frequency of dressing changes between groups nor in the proportion of dressing changes undertaken by professionals, with nearly 70% of dressing changes undertaken by non-professionals. It might be the case that non-professionals changed Aquacel dressings more frequently than professionals would have done, but in the trial the overall mean material cost of using Aquacel per patient was significantly higher: approximately £44 compared with £15 for N-A. If these findings were generalisable across the UK, where the incidence of new ulcers is estimated at 40,000, it is possible to derive a potential increased annual cost that is attributable to using a product such as Aquacel in preference to one such as N-A, which would exceed £1 million in any 6-month period.

Further analysis of the effectiveness of the different dressing products was hampered by the difference in withdrawal rates between the groups, with N-A being withdrawn significantly more often. The reason for this increased rate of withdrawal is not clear because there was no difference in incidence of adverse events and SAEs between groups. It is possible, but unproven, that participants using N-A had their product withdrawn because either they, or their professional advisors, felt that this product was unsuitable for their chronic wound. Analysis of the reasons for withdrawal documented for each of the three dressings reveals that 17 of the withdrawals from the N-A group were attributable to protocol violation or withdrawal of consent, which was more than for both Aquacel (12) and Inadine (7). These two reasons could account for the difference in rate of withdrawal that was observed. There was no difference in the incidence of adverse events between the three groups, and in the incidence of infective episodes, in particular.

This difference in rate of withdrawal has implications for the health economic analyses that could be undertaken, because the analyses assume that the increased rate of withdrawal relates to the properties of the dressing rather than to – as may be the case – the beliefs of either participants or professionals concerning its properties. The implications of withdrawals (and non-healing) would add significantly to the overall costs associated with treating diabetic foot ulcers, if the withdrawals (and non-healing) contributed to an increased risk of amputation. There is, however, no evidence that this was the case. Further work is needed to determine the extent of the relationship between non-healing of ulcers and risk of amputation in order to fully assess the cost implications.

Two final caveats must be added to the assessment of these findings. The first is that the comparison of the secondary outcomes involved a very large number of statistical analyses, and caution must be attached to the significance of any differences found. The study was powered only for the primary end point, and other apparent differences must be regarded as being simply suggestive. The other caution relates to the part played by dressings in overall strategies regarding the management of chronic wounds. Despite the attention paid to choice of dressings in clinical practice, it is likely that the contribution that it makes to healing is relatively limited.

Participating centres

Progress of the study

Start date – June 2003

Appointment of research staff – October 2003

First participants recruited – January 2004

New centres recruited – October 2004

Final participant recruited – June 2006

End of data collection – March 2007

Data entry and analysis – March–August 2007

Presentation of draft final report – September 2007

Receipt of reviewers’ comments – September 2008

Presentation of revised final report – December 2008

Contribution of authors and collaborators

Trial Steering Committee

Professor Geoff Gill, Royal Liverpool Hospitals, Chairman; Dr Phil Weston, Royal Liverpool Hospitals; Mr Collin Thompson, Consumer Representative; Professor Tim Dornan, Salford Royal Hospitals NHS Trust; Professor William Jeffcoate, Chief Investigator.

Data Monitoring and Ethics Committee

Dr Peter Watkins, Royal College of Physicians, Chairman; Professor John Fuller, University College, London; Dr Martin Shipley, University College, London; Dr Jane Lewis, Podiatry Department, Cardiff and Vale NHS Trust.

Disclaimers

The views expressed in this publication are those of the authors and not necessarily those of the HTA programme or the Department of Health.

Numbers refer to individual patient code.