Clinical and Economic Impact of a Two-layer Compression System for the Treatment of Venous Leg Ulcers: A Systematic Review

Objective. This study evaluates the clinical and cost effectiveness of a 2-layer compression system (2LBA; 3M Coban Two-Layer Compression System; 3M, St Paul, MN) compared with other 2-layer (2LB) and 4-layer (4LB) compression systems in patients with noninfected venous leg ulcers (VLUs). Methods. The MEDLINE, EMBASE, CINAHL, Cochrane Library, National Health Service (NHS) Economic Evaluation, and EconLit databases were searched from inception up to January 2017. The MEDLINE search was updated on March 31, 2017. Study selection, quality assessment, and data synthesis were undertaken in accordance with recommended standards. Findings were presented narratively. Results. In total, 5 studies (N = 1509 patients) of mixed methodological quality were included. At 6 months, 2LBA achieved better ulcer healing in comparison with 2LBB (odds ratio [OR], 1.57; 95% confidence interval [CI], 1.10–2.24; P = .03) and 4LBA (OR, 1.93, 95% CI, 1.26–2.97; P = .05) in patients with newly diagnosed ulcers only. For a combined population with newly diagnosed and existing VLUs, healing outcomes were OR, 2.87; 95% CI, 1.06–7.77; P = .04, and OR, 16.51; 95% CI, 2.08–131.37; P = .008, for 2LBs and 4LBs, respectively. Results on slippage were inconclusive. Adverse events were infrequent and did not differ significantly between interventions. Lower 6-month NHS costs for the combined population (£2413 vs. £2707 or £2648) and for newly diagnosed patients (£3045 vs. £3842 or £4480) were observed comparing 2LBA with 2LBB or 4LBA. Also, 2LBA was associated with better health-related quality of life (HRQoL) at 6 months. Conclusions. Based on these findings, 2LBA may result in lower treatment costs and better ulcer healing and HRQoL compared with other multicomponent therapies, especially in patients with newly diagnosed VLUs. However, further high-quality research is needed, especially for outcomes such as slippage and bandage wear time.

Venous leg ulcers (VLUs) are associated with significant morbidity and impaired health-related quality of life (HRQoL).^1,2 Prevalence estimates vary by geographical region, with higher rates reported in developed countries.^3-5 Available evidence suggests that 1.0% of the adult population and 3.6% of those older than 65 years develop VLUs mainly as a result of chronic venous insufficiency (CVI).⁵  The prevalence of VLUs in the United Kingdom in 2012 to 2013 was estimated at 277 749; however, methodological differences in epidemiological studies may result in variations within the same region.^6,7 In general, the prevalence of VLUs is higher in women than in men and markedly increases with age.^3,8

Compression therapy is generally accepted as the gold standard therapy for achieving healing in patients with noninfected VLUs.^9-11 The aim of compression treatment is to produce a graduated pressure in the lower limb to improve venous return, which in turn slows down or reverses the mechanisms resulting in CVI. Previous reviews^12-17 assessing the effectiveness of 2-layer bandages (2LBs), 4-layer bandages (4LBs), and other multicomponent compression systems have reported high compression regimens achieve better healing outcomes when compared with low compression therapies. However, there is evidence suggesting treatments resulting in bulky bandages or those requiring frequent bandage changes due to slippage, discomfort, or excessive soiling may result in poor patient compliance, thus longer healing times.¹⁸ These related factors, in addition to the recurring tendency of VLUs, have a negative impact on patients’ HRQoL and overall costs of VLU management.^2,19,20 The 3M Coban Two-Layer Compression System (2LBA; 3M, St Paul, MN) is made up of 2 layers: an inner polyurethane, latex-free layer with a cohesive covering and an outer cohesive, non-woven layer.²¹ Correct application of 2LBA results in an inelastic sleeve, which delivers consistent and adequate pressure to improve venous return.²¹ In addition, reduced bandage slippage within the first week of application has been noted with 2LBA.²¹

The economic burden of managing VLUs is high and costs could increase in the future with an aging population, especially in developing countries. In the United Kingdom, recent estimated annual costs of VLU treatment range between £596.6 and £921.9 million.¹ Without current consensus on the choice of multicomponent compression therapy, the selection of cost-effective options is crucial, especially in light of the resource constraints within the National Health Service (NHS) and the resultant financial impact on clinical procurement departments.²² Therefore, an evaluation of the relative clinical and cost effectiveness of new multicomponent compression treatments for managing VLUs will be helpful to guide the selection of suitable interventions. 

The aim of this systematic review was to evaluate the clinical and cost effectiveness of the 2LBA compared with selected multicomponent compression therapies in the management of patients with noninfected VLUs.

The review was undertaken in line with recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.²³ 

Literature search

Comprehensive database searches were undertaken to identify clinical and cost evidence relating to multicomponent compression systems in adult patients with noninfected VLUs. For the clinical evidence searches, the following electronic databases were searched from inception up to January 4, 2017 to identify prospective, randomized controlled trials (RCTs), cohort studies, and case-control studies evaluating the effectiveness of multicomponent compression systems in patients aged 18 years and older with noninfected VLUs: MEDLINE (via OvidSP), EMBASE (via OvidSP), CINAHL (via EBSCO), Cochrane Library, Science Citation Index Expanded (SCI-EXPANDED) (Web of Science), and the US Food and Drug Administration Manufacturer and User Facility Device Experience (MAUDE). An updated search was conducted in MEDLINE on March 31, 2017 to ensure recently indexed studies were not missed. No language or date restrictions were applied.

In March 2017, the cost-effectiveness searches were conducted using an economics search filter combined with the population and intervention terms used in the clinical-effectiveness searches in the following databases: MEDLINE (via OvidSP), EMBASE (via OvidSP), CINAHL (via EBSCO), NHS Economic Evaluation Database, Science Citation Index Expanded (SCI-EXPANDED) (Web of Science), and EconLit (via OvidSP). No language or date restrictions were applied. 

Details of the search strategies in MEDLINE are presented and available online in Supplementary File 1. Additional searches included searching gray literature, scrutiny of reference lists of included studies, and contacting topic experts. All retrieved records were imported into EndNote (version X8.1; Clarivate Analytics, Philadelphia, PA) for de-duplication and study selection.

Study selection 

Studies were selected based on prespecified criteria as presented in Table 1. Eligible studies were full-text articles available in the English language comparing 2LBA to a 2LB or a 4LB in the management of patients with noninfected VLUs. Outcomes of interest included measures of ulcer healing, bandage wear time or slippage rates, adverse events, HRQoL, resource use, and costs. The choice of comparators and outcomes of interest were discussed and agreed with topic experts.

One researcher reviewed the titles and abstracts of all records. Potentially eligible studies were subsequently obtained in full text and examined in more detail. A second reviewer checked selection of full-text articles; disagreements were resolved by consensus or by referral to a third researcher.

Data extraction and quality assessment

Using a piloted standardized form, data extraction and quality assessment were performed by 1 researcher and checked by a second. Disagreements and inconsistencies were resolved by discussion. Abstracted data included name of first author, publication year, study characteristics, population of interest, compression management, and outcomes of interest. Methodological quality of included RCTs and non-RCTs was assessed using a bespoke appraisal tool with reference to the relevant Critical Appraisal Skills Programme (Oxford, England) tools.^24,25 

Data synthesis and analysis

The primary outcome of the review was ulcer healing, as described by study authors. Other outcomes included time to complete ulcer healing, bandage slippage, adverse events, HRQoL, and costs. Estimates of the relative treatment effects of compression therapies were presented as mean difference (MD) or odds ratio (OR) with 95% confidence interval (CI), where data allowed. Values reported in the studies were used where appropriate to calculate OR, using the Review Manager (RevMan Version 5.3; The Cochrane Collaboration, Copenhagen, Denmark). Additionally, where it was appropriate, descriptive statistics were used to summarize findings.  

Of the 3388 records retrieved, 74 full-text articles were obtained for detailed examination (Figure^26-30). Overall, 5 studies comprising 3 RCTs^27-29 and 2 real-world studies^26,30 were included in the review. A summary of full-text studies with reasons is presented and available online in Supplementary File 2.

Study characteristics

Characteristics of the included studies are summarized in Table 2.^26-30 In total, 3 multicenter RCTs conducted in Belgium,²⁹ Canada,²⁸ Germany,²⁹ Italy,²⁷ the Netherlands,²⁹ United Kingdom,^28,29 and United States²⁸ were eligible for inclusion. One RCT,²⁹ unpublished at the time of developing this manuscript, and 2 real-world studies, were conducted in the United Kingdom.^26,30 

The review included 1509 patients with noninfected VLUs, ranging from 81 patients²⁸ to 675 patients²⁶ per study. Patients were predominantly aged between 62.5 years²⁸ to 76.9 years.²⁶ With the exception of 1 study,²⁸ VLUs were more common in women than men. An ankle-brachial index of > 0.8 was an inclusion criterion applied in all RCT studies.^27-29 However, eligible patients for the 2 real-world studies^26,30 were identified using a read code for VLUs applied to randomly selected case records retrieved from The Health Improvement Network (THIN, 2009 to 2013) database.³¹ This computerized database³¹ holds anonymized information of more than 9 million patients seen in about 500 general practices across the United Kingdom.³¹ A read code refers to a validated identifier for a specific clinical diagnosis used within the NHS in England. One study³⁰ included patients with newly diagnosed ulcers, while the remaining studies^26-29 reported on patients with existing VLUs of different durations. Ulcer duration and measurements reported at baseline in RCTs showed some variations (5.8 months²⁷ to 48.8 months²⁸ and 7 cm²²⁷ to 11.8 cm²,²⁸ respectively).

Treatment switching was described in 3 studies.^26,28,30 Patients in the 8-week RCT reported by Moffatt et al²⁸ were initially randomized to either 2LBA or 4LBA (PROFORE Multi-Layer Compression Bandaging System; Smith & Nephew, Hull, UK) and followed up for 4 weeks. Initial treatment then was switched, and patients were observed for an additional 4 weeks. The authors reported that 12 patients did not crossover at week 4 due to early ulcer healing (n = 9) and withdrawal from the study (n = 3). On the other hand, in 2 studies, the authors stated treatment switching occurred in up to 79% of patients with existing ulcers within 1.2 months of commencing treatment²⁶ and in 3% to 23% of patients with newly diagnosed ulcers within 2 to 3 months of compression therapy.³⁰ Assessment and reporting of relevant outcomes were not presented consistently across studies. For example, ulcer healing was evaluated as a primary outcome^27,29 or secondary outcome^26,28,30 in the different studies. In general, 3 studies contributed data for 2LBA versus other 2LBs (ie, 2LBC [Rosidal K; L&R USA Inc, Milwaukee, WI²⁹] and 2LBB [Urgo KTwo; URGO Medical, Fort Worth, TX^26,30]), while 4 studies reported comparisons between 2LBA and 4LBs (ie, modified Unna Boot [4LBB²⁷] and 4LBA²⁸). Data for estimating cost effectiveness of compression treatment were available from 2 real-world studies.^26,30 

Quality assessment of included studies

The methodological quality of the included studies ranged from high to unclear risk of bias (Table 3). Study populations for RCTs ranged from 81 patients²⁸ to 234 patients²⁹ and from 600 patients³⁰ to 675 patients²⁶ in the real-world studies. Sample size calculations were reported in 2 studies.^28,29 Computer-generated randomization was conducted in all RCTs,^27-29 whereas random selection of case records was undertaken in both real-world studies.^26,30 Generally, recruited patients were comparable in their baseline characteristics across included studies. However, there were more women than men (range, 53%–74%) in 4 studies,^26,27,29,30 whereas in 1 RCT,²⁸ there were more men than women (64% vs. 52%; 2LBA vs. 4LBA, respectively). Study withdrawal rates in the RCTs were generally low. 

Similar dressings were used in all RCTs prior to application of compression therapy. Vanscheidt²⁹ permitted patients to have bandage changes outside of the investigator’s center or at home. The care settings in the studies by Guest et al^26,30 included general practitioner consultations, hospital outpatients, and community nurse visits. The lack of blinding of attending professionals together with limited and unclear reporting across studies made it challenging to assess the impact of the caregiver’s skills and potential confounders (eg, use of antibiotics) on healing rates and other relevant outcomes. In the study by Vanscheidt,²⁹ 35 patients (2LBA group [n = 20/117] and 2LBC group [n = 15/117]) were not included in the per-protocol analysis due to protocol violations or voluntary withdrawals. 

Clinical effectiveness 

Complete ulcer healing. Ulcer healing outcomes were available for 1 month,²⁸ 3 months,²⁷ and 6 months^26,30 of follow-up (Table 4 A^26-30). Data from RCTs demonstrated lack of statistically significant differences in healing rates in the 2LBA group compared with 4LBA group at 1 month (odds ratio [OR], 2.36; 95% confidence interval [CI], 0.55–10.19; P = .25),²⁸ 4LBB at 3 months (OR, 1.39; 95% CI, 0.30–6.57; P = .68),²⁷ or 2LBC at 3 months (OR, 1.08; 95% CI, 0.63–1.84; P = .78).²⁹

For the 2LB compression therapies, statistically significant healing rates were demonstrated when 2LBA was compared with 2LBB at 6 months in a real-world study²⁶ (OR, 1.57; 95% CI, 1.10–2.24; P = .03, for patients with new and existing VLUs who switched treatment after initial allocation; OR, 2.87; 95% CI, 1.06–7.77; P = .04, for patients who did not switch treatment over the entire study duration). Although patients with newly diagnosed VLUs experienced improved healing with 2LBA compared with 2LBB (OR, 1.35; 95% CI, 0.87–2.11; P = 1.18),³⁰ this benefit was not statistically significant. 

Real-world evidence^26,30 showed 2LBA achieved better statistically significant ulcer healing compared with 4LBA at 6 months for patients with existing and new VLUs (OR, 2.70; 95% CI, 1.79–4.08; P = .001, in patients who switched treatment; OR, 16.51; 95% CI, 2.08–131.37; P = .008, in patients who continued with initially allocated compression treatment throughout the study).²⁶ Healing at 6 months also was improved in patients with newly diagnosed ulcers treated with 2LBA compared with those who received 4LBA (OR, 1.93; 95% CI, 1.26–2.97; P = .05).³⁰ 

Time to complete ulcer healing. Evidence for time to ulcer healing at 6 months was mixed. In real-world studies, compared with 2LBs and 4LBs, 2LBA resulted in a statistically shorter mean time to healing for patients with newly diagnosed VLUs (mean difference [MD], months, 0.40; 95% CI, -0.74 to -0.06; P = .02; MD, -0.50; 95% CI, -0.86 to -0.14; P = .007, respectively).³⁰ This trend was absent in study populations with both existing and new ulcers²⁶ (Table 4 B^26-30). An explanation for this observation could be the presence of hard-to-heal ulcers in affected patient populations. Data from RCTs showed patients with existing and new VLUs treated with 2LBA experienced complete ulcer healing slightly earlier than those treated with 2LBC (1.8 months vs. 1.96 months [MD, 0.16; 95% CI, -0.34 to 0.02; P = .08, respectively).²⁹

Bandage slippage and wear time. Bandage slippage was reported as primary outcome³⁰ or secondary outcome²⁹ in 2 RCTs. Moffatt et al²⁸ defined slippage as the change in bandage height (cm), whereas Mosti et al²⁷ did not specify the method of assessment. Moffatt et al²⁸ noted less slippage for 2LBA compared with 4LBA (2.48 cm vs. 4.17 cm; P ≤ .001). Whereas Mosti et al²⁷ reported more bandage slippage for those treated with 2LBA compared with 4LBB (22% vs. 10%; P ≤ .001). The relatively rigid structure of 4LBB may be the likely reason for the observed difference.

In 2 studies, there was inconclusive evidence relating to bandage wear time comparing 2LBA with 4LBA (5.72 days vs. 5.75 days; MD, 0.03 days; P = .72)²⁸ or 2LBC (5.5 ± 1.7 days vs. 4.3 ± 1.9 days; MD, 1.2 days; P = .95),²⁹ respectively.

Adverse events. Reporting of adverse events (AEs) varied across studies. Device-related AEs were few and mainly limited to pain and maceration. Data from 2 RCTs^28,29 showed there were no significant differences in the occurrence of 1 or more AEs between 2LBA and 4LBA²⁸ or 2LBC²⁹ (2LBA vs. 4LBA: 50% vs. 50%; relative risk [RR], 1.0; P = .903; 2LBA vs. 2LBC: 49.6% vs. 51.7%; RR, 0.96; P = .896). Furthermore, 2 RCTs^27,29 demonstrated fewer patients treated with 2LBA experienced pain compared with 2LBC (2.6% [n = 3/117] vs. 11.2% [n = 13/117]; P = .01) (Table 4 C^26-30). 

Cost effectiveness 

Health-related quality of life. The HRQoL was reported as estimates of the Cardiff Wound Impact Scale and HRQoL scores (undefined scale) in RCTs or as quality-adjusted life years (QALYs) in real-world studies. In comparison with 2LBs, 2LBA significantly provided better HRQoL than 2LBB at 6 months (mean gain in QALYs, 0.01; 95% CI, -0.00 to 0.02; P < .12, patients with existing and new ulcers; mean gain in QALYs, 0.02; 95% CI, 0.02–0.02; P < .00001, patients with newly diagnosed ulcers).³⁰ On the other hand, there was no significant difference in HRQoL between 2LBA and 2LBC at 3 months of follow-up.²⁹ 

Evidence from 3 studies^26,28,30 comparing 2LBA with 4LBs showed 2LBA significantly improved patients’ HRQoL at 6 months in real-world studies^26,30; this improvement was not observed in a crossover RCT²⁸ after 1 month of treatment. In the 2 real-world studies,^26,30 patients receiving 2LBA experienced significantly better QALYs compared with those treated with 4LBA (MD, 0.02; 95% CI, 0.02–0.02; P < .00001, patients with existing and new ulcers²⁶; MD, 0.02; 95% CI, 0.01–0.03; P = .003, patients with newly diagnosed ulcers³⁰).

In total, 2 studies^26,30 reported the cost-effectiveness results estimated as the costs and QALYs over a 6-month period based on a retrospective analysis of patient records from the THIN database.³⁰ Patients treated with 2LBA experienced lower 6-month NHS management costs compared with those who received 4LBA and 2LBB.^26,30 In patients with existing and new ulcers, the mean cost for 2LBA was £2413 compared with £2648 for 4LBA and £2707 for 2LBB (9.7% and 12.2% higher than 2LBA, respectively).²⁶ In the patients with newly diagnosed ulcers, the mean cost for 2LBA was £3045 compared with £4480 for 4LBA and £3842 for 2LBB (47.1% and 26.2% higher than 2LBA).³⁰ In addition, 2LBA patients with new or existing VLUs experienced better QALYs (0.374) compared with 2LBB (0.368) and 4LBA (0.353)²⁶; whereas patients with new ulcers on 2LBA experienced better QALYs (0.413) compared with 2LBB (0.404) and 4LBA (0.396), during the 6-month period³⁰ (Supplementary File 3; available online).

A total of 5 studies of mixed methodological quality were included in this systematic review. Overall, 2LBA achieved better ulcer healing rates than other 2LBs or 4LBs, especially in patients with newly diagnosed VLUs. Additionally, 2LBA achieved ulcer healing within a shorter duration (< 2 months) compared with 4LBs (≥ 2 months), although care is required when interpreting the findings due to some variation in treatment protocols and methods of assessment. Further, Moffatt et al²⁸ and Vanscheidt²⁹ allowed bandage wear time of 7 days, but the reasons for early change differed among these studies. Data from these 2 studies showed the wear time for 2LBA was longer than 2LBC (values not reported)²⁹ but similar in comparison to 4LBA (5.72 days vs. 5.75 days, P = .72).²⁸ Less bandage slippage occurred with 2LBA compared with 4LBA (2.48 cm vs. 4.17 cm, P ≤ .001, respectively).²⁸ This finding supports the conclusions of a preclinical study by Collier and Schuren,³² which suggested an enhanced rigidity and improved wear time of 2LBA. However, slippage was less in the 4LBB group compared with 2LBA group (10% vs. 22%, P < .001, respectively).²⁷ 

As a relatively thinner yet rigid dressing, 2LBA may achieve effective high compression and better treatment compliance HRQoL. Fewer bandage changes may also translate into a reduced burden of VLUs for patients, caregivers, and the health care system in general. Generally, tolerance to 2LBA matched that of other compression bandage systems. However, in this study, the paucity of evidence on this outcome should be noted, and more studies are encouraged to establish the relative performance of 2LBA compared with 2LBs and 4LBs. The 2 real-world studies^26,30 based on retrospective analyses of case records may have some limitations relating to the associated data sources.³³ The authors of the 2 real-world studies^26,30 reported that patients were matched on age, gender, ulcer duration, and ulcer size; however, evidence suggests individual factors influencing ulcer healing, such as lack of exercise, nutrition, smoking, injection drug use, or socioeconomic status, may not have been adequately captured or controlled in such studies.^34,35 

Another consideration is the absence of differences in healing rates between 2LBA and 4LBs and 2LBs in RCTs that run for 1 to 3 months, while data from real-world studies^26,30 with 6 months of follow-up showed better healing with 2LBA. This difference could be explained by the length of follow-up in RCTs as well as their narrow eligibility criteria, which often leads to more homogeneity in terms of patient characteristics. Although RCTs are the gold standard for evidence relating to efficacy of interventions, the contribution of real-world studies^26,30 in this review provides more information, such as the history of prior use of the compression bandage systems (therapies) and treatment switching. Differences in skills when applying compression bandages may also explain differences between the findings of the included studies. 

This study was undertaken with agreed recommendations for the conduct of a systematic review.²³ A comprehensive literature search was performed, coupled with a clearly defined inclusion of data from relevant study designs. This review included both RCTs and real-world studies, which the authors consider a strength because recent evidence suggests that about 50% of published reviews on treatment effects include RCTs and non-RCTs.³⁶ However, there are some limitations of this review to consider. It is worth noting that recommended criteria for assessing the methodological quality of reviews indicate literature searching completed within a period of less than 2 years of publication is generally adequate.³⁶ While no limits for language type or publication year were applied for database searching, study inclusion was limited to full-text articles only available in the English language. The review team agreed upon and implemented this approach based on time and resource constraints, though this approach may have resulted in excluding potentially relevant non-English studies.  

Based on the review of the literature, 2LBA achieved better wound healing results over a shorter follow-up period than other multicomponent compression bandages. In addition, findings may suggest 2LBA has better or similar performance properties in terms of wear time and slippage. Overall, 2LBA was associated with 26% to 47% cost reductions compared with other compression therapies, especially in patients with newly diagnosed VLUs. Current evidence, therefore, may suggest 2LBA is a clinically effective and cost-effective compression treatment. Its use in health care settings that manage patients with VLUs may offer cost-effective benefits for treatment costs. Future RCTs should be designed to provide robust data for head-to-head comparison of relevant treatments. Consensus on treatment schedules, procedures, and outcome reporting in trials of VLUs are recommended to support reliable and consistent interpretation of findings of primary studies. 

Note: The authors acknowledge Mrs. Stephanie Bryson for offering administrative support for completing the project.

Authors: Edward A. Goka, PhD, MPH, MRes, BScPh; Edith Poku, MBChB, MSc, FHEA; Praveen Thokala, PhD, MASc; and Anthea Sutton, MA, MCLIP, BA(Hons)

Affiliation: School of Health and Related Research, Faculty of Medicine Dentistry and Health, The University of Sheffield, Sheffield, England

Correspondence: Edward A. Goka, PhD, MPH, MRes, BScPh, The University of Sheffield, School of Health and Related Research, 30 Regent Street, Sheffield, South Yorkshire S1 4DA, United Kingdom; E.A.Goka@sheffield.ac.uk

Disclosure: This research was funded by 3M UK PLC. The funders did not take part in the study design or writing of the manuscript and had no influence on the conclusions reached in this review. All authors declare no conflict of interest.