Compression improves healing of venous leg ulcers compared with no compression, with differences between different compression systems
- Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria, Australia
- Correspondence to
: Dr Carolina D Weller
Department of Epidemiology and Preventive Medicine, Monash University, Level 6 Alfred Centre, 99 Commercial Road, Prahran, Melbourne, VIC 3004, Australia;
Implications for practice and research
Multicomponent compression systems are more effective than single-component systems.
Multicomponent systems containing elastic bandage are more effective than inelastic material.
High compression stocking system healing needs further research.
Compression is used to treat venous leg ulcers (VLU) that are secondary to venous insufficiency; it assists by reducing venous hypertension, enhancing venous return and reducing peripheral oedema. Healing potential is decreased if compression is not applied. Many randomised controlled trials (RCTs) included in the present study have reported on healing outcomes of different types of compression systems. A large number of compression options are now available, and the choice is further complicated by different application techniques of single and multiple components. Laboratory studies have reported differences in compression sub-bandage pressure (SBP) although laboratory performance of bandage may not reflect clinical application.
Eight RCTs were identified with comparisons between compression systems. These included compression when compared with: primary dressing alone, non-compressive bandages or usual care that did not routinely include compression. Six categories of compression bandage system were compared, including: single-component, two-component, three-component and compression systems comprising four components (4LB) including elastic, and adjustable compression boots compared with compression bandages, and compression stockings or tubular devices. The review included 48 RCTs with a total of 4321 participants from the UK, the USA, Canada, Europe and Australia. Forty per cent of included trials had sample sizes of 50 or fewer, and 67% recruited 100 patients or fewer. Of the included trials, 31 RCTs did not report statistical power or sample size estimation. Bandaging application technique was seldom reported.
Evidence suggests that VLUs heal more quickly with compression than without and that multicomponent systems achieve better healing outcomes. Comparison between four layer bandage (4LB) and multicomponent systems that include inelastic short stretch bandage (SSB) indicated the estimated probability of healing with 4LB was about 1.3 times that of the SSB (Cox regression analysis incorporated between-group differences). No significant differences were found between 4LB and paste bandages for complete healing. When compared with SSB, high-compression stockings were associated with better healing and better outcomes for pain and discomfort. No significant between-group differences were found for complete healing at 4 months and at 18 months for single-layer stockings and paste bandages. There was limited evidence on ulcer recurrence or compression adherence.
At present, it is difficult to classify and compare different compression systems as suggested by an international consensus statement1 because information on SBP is generally not available from clinical trial reports. The observed treatment effect of included trials may have been further influenced by imbalance of treatment groups at baseline, particularly with independent prognostic factors (ulcer size and duration). Some recent trials2 ,3 have addressed the imbalance of treatment groups at baseline by stratified randomisation and using analyses that adjusted for covariates. Many trials included in this review did not measure bandage skill; however, this would be useful if reported in future trials as a key variable in determining treatment effect. Methodological quality of future RCTs can be further improved by ensuring appropriate methods of randomisation, generating balanced groups at baseline, allocation concealment, using intention to treat analyses, ensuring blinded outcome assessment and designing studies with sufficient power to detect true treatment effect.4 ,5 To date, no published RCT has systematically examined reasons for compression treatment non-adherence from the patient's perspective. Low or non-adherence affects all compression therapies and, if numbers of efficacious treatments continue to be marketed, then future trials should report relevant outcomes such as healing and recurrence as well as possible reasons for non-adherence from the patient's perspective.