Background Whilst strong compression is effective in treating venous leg ulcers, compression systems vary in their effects on ulcer healing and wider outcomes. Four-layer bandage systems (4LB) and two-layer hosiery (2LH) are recognised effective compression systems in practice. The relative effectiveness and cost-effectiveness of two-layer bandage systems (2LB, a commonly used system), compression wraps (CW, a newer option), and short-stretch bandages (SSB) were uncertain. This paper aims to compare the effectiveness on time to venous ulcer healing and cost-effectiveness of 4LB and 2LH (or a choice of these), with 2LB, SSB, and CW. Methods We updated the Ovid MEDLINE search of the Cochrane review on compression systems to identify new – up to 02 February 2026 - relevant randomised controlled trials (RCTs), in addition to the individual patient data from VenUS 6 – a recent large RCT comparing a choice arm of 4LB and 2LH, with 2LB and CW. We identified RCTs evaluating any type of compression systems (bandage or stockings). Study’s risk of bias was assessed using the Cochrane Collaboration tool. We used network metaanalyses (NMA) to estimate relative treatment effects on time to ulcer healing, and used a Markov model to compare the cost-effectiveness of interventions over a lifetime horizon and from a UK NHS perspective. The primary outcome for the network metaanalysis was the hazard ratio (HR) of time to ulcer healing for alternative compression systems. For the cost-effectiveness analysis, total costs, total quality-adjusted life years (QALYs), and Incremental Cost-Effectiveness Ratios (ICERs) were estimated. Results Overall, 21 trials with 2,934 participants were included, of which 5 (24%) and 4 (19%) studies had a moderate and a low risk of bias, respectively. Due to the sparsity of data, fixed-effects models were used. Heterogeneity, transitivity and consistency were explored by comparison of population characteristics in included studies, comparison of fixed- and random-effects models, and assessment of consistency matrix. The HRs of time to ulcer healing compared with 4LB/2LH, when including all studies, were 0.972 (95% CI 0.851 to 1.107) for SSB, 1.044 (95% CI 0.859 to 1.255) for 2LB, 0.927 (95% CI 0.745 to 1.136) for CW. The HRs compared with 4LB/2LH, when including only studies with low and moderate risk of bias, were 0.954 (95% CI 0.823 to 1.099) for SSB, 1.004 (95% CI 0.817 to 1.213) for 2LB, 0.901 (95% CI 0.719 to 1.110) for CW. The HRs compared with 4LB/2LH, when including only studies with low risk of bias, were 1.009 (95% CI 0.856 to 1.187) for SSB, 10.995 (95% CI 0.792 to 1.233) for 2LB, 0.872 (95% CI 0.695 to 1.085) for CW. The cost-effectiveness analysis (when using the NMA results including all studies) showed small differences in total costs and total QALYs across treatments. Compared with 4LB/2LH, the total costs for SSB, 2LB, and CW were higher, of £390 (95% CI: 41 to 985), £345 (95% CI: -102 to 1127), and £348 (95% CI: -133 to 1118), respectively. Total QALYs differences versus 4LB/2LH were -0.0015 (95% CI: -0.0110 to 0.0064) for SSB, 0.0022 (95% CI: -0.0085 to 0.0100) for 2LB, and -0.0044 (95% CI: -0.0207 to 0.0079) for CW. SSB and CW were dominated by 4LB/2LH. The ICER of 2LB versus 4LB/2LH was £159,614/ QALY gain. Interpretation These findings suggest potential small differences in the clinical effectiveness of 4LB/2LH, 2LB, SSB, and CW, but there is uncertainty in the estimates. In the UK, 4LB/2LH emerges as cost-effective.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)