- a NHS Centre for Reviews and Dissemination, University of York, York
- b Department of Health Studies, University of York, York Y01 5DD
- Correspondence to: Dr Cullum
- Accepted 14 August 1997
Objective: To estimate the clinical and cost effectiveness of compression systems for treating venous leg ulcers.
Methods: Systematic review of research. Search of 19 electronic databases including Medline, CINAHL, and Embase. Relevant journals and conference proceedings were hand searched and experts were consulted.
Main outcome measures: Rate of healing and proportion of ulcers healed within a time period.
Study selection: Randomised controlled trials, published or unpublished, with no restriction on date or language, that evaluated compression as a treatment for venous leg ulcers.
Results: 24 randomised controlled trials were included in the review. The research evidence was quite weak: many trials had inadequate sample size and generally poor methodology. Compression seems to increase healing rates. Various high compression regimens are more effective than low compression. Few trials have compared the effectiveness of different high compression systems.
Conclusions: Compression systems improve the healing of venous leg ulcers and should be used routinely in uncomplicated venous ulcers. Insufficient reliable evidence exists to indicate which system is the most effective. More good quality randomised controlled trials in association with economic evaluations are needed, to ascertain the most cost effective system for treating venous leg ulcers.
Compression treatment increases the healing of ulcers compared with no compression
High compression is more effective than low compression but should only be used in the absence of significant arterial disease
No clear differences in the effectiveness of different types of compression systems (multilayer and short stretch bandages and Unna's boot) have been shown
Intermittent pneumatic compression appears to be a useful adjunct to bandaging
Rather than advocate one particular system, the increased use of any correctly applied high compression treatment should be promoted
Leg ulceration, usually of venous origin,1 is a common chronic recurring condition usually managed in primary care, and it is expensive to treat.2 3 Compression treatment, in the form of bandaging or hosiery, is regarded as the first line of treatment when venous leg ulceration occurs in the absence of clinically important arterial disease (table 1).
Considerable uncertainty remains, however, about the most effective treatment, and methods vary widely. A systematic review was therefore commissioned by the NHS health technology assessment programme to examine the effectiveness of compression systems for the treatment of venous leg ulcers.
We conducted a systematic review using structured guidelines.4 We undertook an electronic search of 19 specialist databases including Medline, CINAHL, and EMBASE and supplemented our search with hand searching, scrutiny of citations, and contact with relevant manufacturers and original authors (Appendix 1). Details of search terms are available in the Cochrane Library database.
All randomised controlled trials examining the effect of compression on the healing of venous leg ulcers were considered, with no restrictions on publication status, date, or language. Decisions over inclusion of studies were made on the basis of a series of predetermined validity criteria, which, along with data extraction, were checked by a second reviewer. The results of trials making similar comparisons, in which the results were homogeneous, were pooled by using a fixed effects (Peto) method, which weights each study by the inverse of its variance.5
We identified 24 relevant randomised controlled trials, of which six are unpublished (Colgan et al, Gould et al, Kralj et al, London et al, Morrell et al, Taylor et al).6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Details of all these trials will be available in Effective Health Care24 and in the Cochrane Library database.
Compression v no compression (six trials)
Three trials compared the use of compression (provided by Unna's boot) with the use of dressings alone (fig 1). Two of these found a higher proportion of healed ulcers when compression was used. The third, small study showed a non-significant increase in healing with Unna's boot.
Three other studies that compared different forms of compression (short stretch, two layer, and four layer bandages respectively) with treatments using no compression showed that healing improved with compression (fig 1).
Elastic multilayer high compression bandages v inelastic compression (six trials)
Three studies compared elastic, high compression, three layer bandages with low compression. The results of these studies were pooled (test for heterogeneity χ2=0.93, df=2, P>0.7) and showed an overall significant increase in the odds of healing at 3 months with high compression bandages (odds ratio=2.26; 95% confidence interval 1.4 to 3.7) (fig 2).
Three small studies found no difference between multilayer high compression (four layer bandages) and two forms of inelastic compression (Unna's boot and short stretch bandage) (fig 3). Both four layer and short stretch bandages resulted in higher healing rates than a paste bandage plus outer support (44%, 40%, and 23% of cases respectively healed at 3 months).
Multilayer high compression systems v single layer systems (four trials)
Figure 4 shows the results of these trials. Four layer bandages were shown to increase the percentage of ulcers healed at 24 weeks and 12 weeks in two trials when compared with single layer compression bandages (Setopress or Granuflex). Two similar but much smaller studies found no difference in healing between four layer or three layer bandages and self adhesive single layer bandages. Pooling the studies (test for heterogeneity χ2=2.24, df=2, P<0.01) showed that multilayer high compression bandages were associated with a higher rate of complete healing than single layer bandages (fig 4) (odds ratio=2.2; 95% confidence interval 1.3 to 3.5).
Comparisons between different medium and high compression systems (four trials)
The original “Charing Cross” four layer bandage has been compared with a kit that provides all the constituents to make up a four layer bandage11 and with a regimen adapted to achieve similar levels of compression by using materials available on prescription.23 No significant difference in outcome was found in either study, although the second trial was very small. Another small study found no difference in the number of ulcers healed at 12 weeks between four layer bandages (6/10) and a combination of three bandages plus class II compression stockings (7/10) (Colgan et al).
A trial of only 30 patients comparing Unna's boot with moderate compression provided by a single bandage (Coban) found no difference in healing at 12 weeks.9
A trial directly comparing four layer and three layer bandages is in progress at St Thomas's Hospital, London.
Compression hosiery v compression bandaging (two trials)
A trial of 50 patients found that 84% of ulcers treated with a combination of two compression stockings (Thrombo and Sigvaris 503) healed completely at 3 months compared with 52% of those treated with a short stretch bandage (odds ratio 4.9; 95% confidence interval 1.3 to 18.3).14 A small, poor quality trial found no difference between compression stockings and Unna's boot.13
Intermittent pneumatic compression treatment (two trials)
Two small studies evaluated the benefit of adding intermittent pneumatic compression to compression stockings or Unna's boot (fig 5). A pooled estimate (test for heterogeneity χ2=0, df=1) shows that the overall odds of healing increased with intermittent pneumatic compression (odds ratio=10.0; 95% confidence interval 3.0 to 33.8).
The results suggest that the healing of venous ulcers is improved when compression is applied as bandages or hosiery. High compression delivered in three or four layers performs better than systems giving low compression and possibly better than single layer systems. The few small studies that have compared different high compression systems—for example, multilayer and short stretch bandages and Unna's boot—have shown no difference in effectiveness. Intermittent pneumatic compression when added to compression treatment seems to confer a significant benefit.
When clinics have specifically promoted the delivery of four layer high compression treatment their healing rates have improved compared with results for the usual care given by community nurses (Morrell et al, Taylor et al).25 It is not easy, however, to disentangle the effects of the model of care, the associated training, and the type of bandaging used.26
The quality of research in this area is generally poor: trials are often too small, follow up is short, recurrence of ulcers is rarely considered, and sometimes multiple ulcers are incorrectly regarded as independent ulcers (table 2). Several papers do not report the method of bandage application, the experience of staff, other aspects of bandaging, and patients' mobility, which all affect healing. The same system applied by different staff under different circumstances may result in the attainment of widely differing pressures, making interpretation difficult.
It is not clear which of the high compression systems is the most cost effective. Rather than advocate one particular system it would seem more sensible to promote the increased use of any correctly applied high compression treatment in patients with uncomplicated venous disease. Well designed randomised controlled trials incorporating economic evaluations are needed to help determine the optimal strategy.
We are grateful to Anna Semlyen for her collaboration in identifying and summarising economic analyses; to Julie Glanville, who provided information support to the project; and to Dr Fujian Song and Dr Mathew Bradley, who provided technical support. The following members of our advisory panel gave advice throughout the review process and commented on a draft: Dr Mary Bliss, Professor Andrew Boulton, Professor Nick Bosanquet, Dr Richard Bull, Michael Callam, Carol Dealey, Professor Peter Friedman, Brian Gilchrist, Dr Keith Harding, Deborah Hofman, Vanessa Jones, Dr Christina Lindholm, Dr Raj Mani, Andrea Nelson, Dr Steve Thomas, and Dr Ewan Wilkinson. We also acknowledge the useful comments provided by Mr Stephen Blair and Professor Charles McCollum. The authors bear sole responsibility for the content of this paper.
Funding: NHS health technology assessment programme.
Conflicts of interest: None.
ISI Science Citation Index (on BIDS); BIOSIS (on Silver Platter); British Diabetic Association Database; CINAHL (on OVID CD ROM); CISCOM, the database of the Research Council for Complementary Medicine; Cochrane Database of Systematic Reviews (CDSR); Cochrane Wounds Group register of trials; Current Research in Britain (CRIB); Database of Abstracts of Reviews of Effectiveness (DARE); Dissertation Abstracts; DHSS Data (on Knight-Ridder Datastar); EconLit; Embase (on Knight-Ridder Datastar); Index to Scientific and Technical Proceedings (searched on BIDS); Medline (on OVID CD ROM); National Research Register (to locate ongoing research in NHS); NHS Economic Evaluation Database (NHS Centre for Reviews and Dissemination); Royal College of Nursing Database (CD ROM); System for Information on Grey Literature in Europe (SIGLE, on Blaise Line).