Heel‐Free Compression Bandaging for Venous Leg Ulcers: A Non‐Inferiority Trial of Interface Pressure

ABSTRACT

To determine whether heel‐free compression bandaging is non‐inferior to heel‐wrapped bandaging for maintaining interface pressure in adults with venous leg ulcers. Randomised, controlled, parallel‐group, non‐inferiority trial. Adults with open venous leg ulcers (CEAP C6) treated at a French university hospital were randomised to heel‐free or heel‐wrapped (control) bandaging for 4–6 weeks. The primary outcome was interface pressure 48 h after the final bandage application, analysed in the per‐protocol population. Secondary outcomes were ankle oedema, change in interface pressure and patient preference. Of 102 randomised patients, 85 were included in the per‐protocol analysis. At V3, mean interface pressure was 30.1 mmHg in the heel‐free group and 28.4 mmHg in the control group (difference, 1.7 mmHg; 95% CI, −1.3 to 4.6), meeting the non‐inferiority criterion. After adjustment for baseline circumference and sex, ankle circumference was 1.7 cm lower in the heel‐free group (95% CI, −3.35 to 0.04). Among 48 patients who had experienced both techniques, 23 (47.9%) preferred heel‐free bandaging. Heel‐free bandaging was non‐inferior to heel‐wrapped bandaging for interface pressure at 48 h. The technique may be considered when compatible with ulcer location and patient preference; effects on healing and adherence require further study.

Trial Registration: NCT04046120.

Key Points

• Medical compression bandages have to be applied using a technique that maintains the target pressure and ensures maximum patient comfort.

• This trial provides the first evidence that leaving the heel free is not inferior to wrapping the heel regarding the interface pressure 48 h after bandage application.

• This result supports allowing nurses to decide the heel wrapping technique according to patient's preference, in line with Evidence Based Nursing practice.

1. Introduction

Venous leg ulcers are a common complication of chronic venous insufficiency and affect approximately 4% of adults aged 65 years or older [1]. Compression is a cornerstone of treatment, and multicomponent or short‐stretch bandages are recommended [1, 2]. Recommended interface pressures vary across guidelines: the French Haute Autorité de Santé recommends 30–40 mmHg at application, whereas German consensus guidance recommends 40–60 mmHg [3, 4]. Pressures below 30 mmHg may be insufficient, whereas excessive pressure can cause discomfort or tissue injury [5, 6, 7, 8]. Because applied pressure depends partly on bandage tension and technique, practitioner training is essential [9].

1.1. Background

Whether the heel should be wrapped remains uncertain. Some guidelines recommend covering it, although supporting comparative evidence is limited [10, 11]. In a French observational study of 261 patients, nurses left the heel uncovered in 48% of applications, most often at the patient's request [12]. Heel wrapping may also restrict footwear and reduce treatment acceptability [12, 13]. Comparative pressure data are therefore needed to determine whether heel‐free bandaging is a reasonable alternative.

1.2. Aim and Objectives

The primary objective was to determine whether heel‐free compression bandaging was non‐inferior to heel‐wrapped bandaging for interface pressure, using a prespecified margin of −5 mmHg for the between‐group difference (heel‐free minus control).

Secondary objectives were to compare ankle oedema, change in interface pressure from V2 to V3, and patient preference between techniques.

2. Methods

2.1. Design

This randomised, controlled, parallel‐group, unblinded non‐inferiority trial compared heel‐free and heel‐wrapped compression bandaging in patients with open CEAP C6 venous ulcers [14].

2.2. Study Setting and Sampling

The trial was conducted from October 2019 through June 2022 at a vascular ulcer referral centre in a French university hospital. The sample‐size calculation used interface‐pressure data from 35 healthy volunteers treated with the UrgoK2 composite bandage (Urgo Medical, Paris, France) [15].

Because no previous study had directly compared heel‐free and heel‐wrapped techniques, clinicians selected a 5‐mmHg non‐inferiority margin by consensus within the 30‐ to 40‐mmHg range recommended by the French Haute Autorité de Santé [3, 16]. Assuming equal mean pressures and an SD of 6.9 mmHg in both groups, a one‐sided alpha of 0.025, and 90% power, 84 patients were required. The target was increased by 20% for attrition, giving a planned sample of 102 patients (51 per group).

2.3. Inclusion and Non‐Inclusion Criteria

Patients seen between 11 October 2019, and 12 April 2022, for unilateral or bilateral CEAP C6 venous ulcers were screened. Inclusion criteria were age 18 years or older, ankle‐brachial index greater than 0.8, ulcer duration of at least 4 weeks, ambulatory status and coverage by the French statutory health insurance system. Non‐inclusion criteria were legal guardianship, major cognitive impairment incompatible with protocol adherence, pregnancy and breastfeeding.

Eligible patients received oral and written study information and provided written informed consent before enrolment and randomisation.

2.4. Randomisation

Patients were randomised 1:1 in permuted blocks to heel‐free bandaging or heel‐wrapped bandaging (control). A clinical research nurse obtained the assignment through the secure Clinsight randomisation system.

2.5. Study Interventions

The protocol included three visits: baseline (V1), 4–6 weeks after enrolment (V2) and 48 h after V2 (V3) (Figure S1).

At V1, the study nurse confirmed eligibility and randomised the patient. After wound care, the nurse measured ankle circumference using the figure‐of‐eight method [17] (Figure S2) and applied the allocated technique. With the ankle at 90°, the bandage was applied spirally from the base of the toes to 2 cm below the popliteal crease [5, 10, 18]. Bandage type followed the medical prescription. For bilateral ulcers, the assigned technique was used on both legs, but only left‐leg pressure was analysed. The bandage was reapplied if initial interface pressure was outside the 30‐ to 40‐mmHg target. The nurse provided a study logbook, informed the home care nurse of the assigned technique, arranged delivery of the PicoPress device for V2, and recorded baseline clinical and bandaging data.

Between V1 and V2, home care nurses provided wound care every 2 or 3 days as prescribed and reapplied the allocated bandaging technique. Dates and bandage types were recorded in the study logbook.

At V2, 4–6 weeks after enrolment and 48 h before V3, the home care nurse positioned the pressure sensor, applied the allocated bandage, measured interface pressure according to protocol and recorded the result.

At V3, the study nurse connected the PicoPress device to the sensor placed at V2 and measured interface pressure before removing the bandage. The nurse then measured ankle circumference with the figure‐of‐eight method, and the patient completed the preference questionnaire.

2.6. Intervention Fidelity

One of three nurses experienced in compression bandaging applied the initial bandage and marked the sensor position at V1. Each home care nurse received a tutorial covering the assigned technique, sensor placement and pressure measurement.

Home care nurses documented protocol deviations in the study logbook. The study nurse delivered the PicoPress device in person and reviewed its use and the protocol procedures with each home care nurse.

2.7. Outcomes and Definitions

The primary outcome was interface pressure (mmHg) at the B1 point at V3, measured with a PicoPress system (Microlab Elettronica, Villatora, Italy). Measurements were standardised across visits as follows.

At V1, the research nurse positioned the sensor at the B1 point before bandage application. This point, defined as the transition from the Achilles tendon to the calf muscles approximately 10–15 cm proximal to the medial malleolus, is the recommended site for interface‐pressure measurement [19]. The site was marked with a skin pencil for subsequent measurements (Figure S2).

At V2, the home care nurse placed the sensor at the marked B1 point and applied the bandage. At V3, the patient rested supine for 2 min with the arms alongside the body and the heel slightly raised to avoid calf contact with the examination table; the nurse then recorded left‐leg interface pressure [18].

Secondary outcomes were ankle circumference at V3, change in interface pressure from V2 to V3, and patient preference. Ankle circumference was measured with the validated figure‐of‐eight method using four landmarks: (1) the midpoint between the tibialis anterior tendon and lateral malleolus, (2) tuberosity of the navicular, (3) the base of the 5th metatarsal and (4) the distal tip of the lateral malleolus [17] (Figure S3). At V2, interface pressure was recorded after a 2‐min rest. A two‐item questionnaire assessed previous experience with each technique, technique preference and who should choose the technique.

2.8. Statistical Analysis

Continuous variables are reported as mean ± SD and categorical variables as n (%). The intention‐to‐treat (ITT) population included all randomised patients, regardless of protocol adherence or treatment received. The per‐protocol (PP) population included eligible patients with an available primary outcome and no major protocol deviation. Major deviations comprised death or hospitalisation during follow‐up, failure to attend V3, use of the non‐allocated technique at V2, or use of the non‐allocated technique more than four times during follow‐up.

The primary analysis used the PP population. Interface pressure at V3 was analysed by linear regression with bandaging technique as the independent variable. The treatment effect was defined as the mean difference between heel‐free and control groups. Non‐inferiority was established if the lower bound of the two‐sided 95% CI exceeded −5 mmHg. Because no placebo‐controlled or direct comparative data were available to preserve a known fraction of effect, the 5‐mmHg margin was selected by clinical consensus among physicians and nurses [16].

Figure‐of‐eight measurements were compared by linear regression, with between‐group differences and 95% CIs. Change in interface pressure from V2 to V3 was analysed with a mixed‐effects model including a random intercept and fixed effects for technique, time point and their interaction.

Three sensitivity analyses were performed for the primary outcome: sex‐adjusted linear regression in the PP population; a complete‐case analysis in the ITT population; and a worst‐case/best‐case analysis in the ITT population, imputing missing pressure as the control‐group mean plus 2 SDs for control patients and the heel‐free‐group mean minus 2 SDs for heel‐free patients.

2.9. Ethical Considerations

The Comité de Protection des Personnes Ouest II–Angers approved the trial (reference 2019/48; favourable opinion dated July 10, 2019). The study was conducted in accordance with the Declaration of Helsinki and subsequent amendments.

Patients received oral and written information and provided written informed consent. The trial was prospectively registered at ClinicalTrials.gov (NCT04046120).

3. Results

3.1. Participants and Procedures

All 102 enrolled patients were randomised (51 per group) and included in the ITT population. Seventeen patients (nine in the heel‐free group and eight in the control group) were excluded from the PP population, leaving 85 patients for the primary analysis (Figure 1).

FIGURE 1.

Study flow diagram.

The PP population comprised 42 patients in the heel‐free group and 43 in the control group. Mean age was similar between groups, whereas the proportion of women was higher in the heel‐free group (Table 1).

TABLE 1.

Baseline patient characteristics and bandaging measures.

	Heel‐free (N = 42)	Control (N = 43)
Age (years), mean ± SD	77.2 ± 13.5	76.1 ± 13.4
Sex, n (%)
Female	26 (61.9%)	15 (34.9%)
Figure‐of‐eight circumference, cm, mean ± SD	61.6 ± 4.8	64.4 ± 4.6
Interface pressure, mmHg mean ± SD	37.8 ± 5.0	38.2 ± 5.1
Number of bandaging attempts a , n (%)
1	30 (71.4%)	37 (86.0%)
2	12 (28.6%)	5 (11.6%)
3	0	1 (2.3%)

Abbreviation: SD, standard deviation.

^a Number of attempts required to achieve the recommended interface‐pressure range of 30–40 mmHg.

3.2. Interface Pressure at V3

In the PP population, mean interface pressure at V3 was 30.1 mmHg (95% CI, 28.0 to 32.2) in the heel‐free group and 28.4 mmHg (95% CI, 26.4 to 30.5) in the control group, yielding an estimated between‐group difference of 1.7 mmHg (95% CI, −1.3 to 4.6) (Table 2). Because the lower confidence limit (−1.3 mmHg) exceeded −5 mmHg, the non‐inferiority criterion was met.

TABLE 2.

Interface pressure at V3 and sensitivity analyses.

	Heel‐free (N = 42)	Control (N = 43)	Difference [95% CI] Heel‐free − Control
Interface pressure, mmHg Mean [95% CI]
Primary analysis #1	30.1 [28.0, 32.2]	28.4 [26.4, 30.5]	1.7 [−1.3, 4.6]
Sex‐adjusted analysis #2	30.1 [28.0, 32.3]	28.4 [26.2, 30.5]	1.8 [−1.3, 4.8]
Complete‐case analysis #3	30.1 [28.0, 32.2]	28.2 [26.1, 30.2]	1.9 [−1.0, 4.8]
Worst‐case/best‐case analysis #4	27.7 [25.5, 29.9]	29.4 [27.2, 31.6]	−1.7 [−4.8, 1.5]

Note: #1, per‐protocol population; #2, per‐protocol population adjusted for sex; #3, complete‐case intention‐to‐treat population; #4, intention‐to‐treat population with worst‐case/best‐case imputation.

Abbreviation: CI, confidence interval.

After adjustment for sex in the PP population (n = 85), the between‐group difference was 1.76 mmHg (95% CI, −1.30 to 4.83). In the complete‐case ITT analysis (n = 89), the difference was 1.92 mmHg (95% CI, −1.00 to 4.84). With worst‐case/best‐case imputation in the full ITT population (n = 102), the difference was −1.68 mmHg (95% CI, −4.83 to 1.47). All lower confidence limits exceeded −5 mmHg (Figure 2).

FIGURE 2.

Difference in interface pressure between the heel‐free and control groups at V3.

3.3. Secondary Outcomes

3.3.1. Ankle Oedema

At V3, unadjusted mean figure‐of‐eight circumference was 58.9 cm (95% CI, 57.4 to 60.4) in the heel‐free group and 62.4 cm (95% CI, 60.9 to 63.9) in the control group, a difference of −3.5 cm (95% CI, −5.57 to −1.40). After adjustment for baseline circumference and sex, estimated mean circumference was 60.0 cm (95% CI, 58.8 to 61.2) and 61.7 cm (95% CI, 60.5 to 62.9), respectively, a difference of −1.7 cm (95% CI, −3.35 to 0.04) (Figure 3).

FIGURE 3.

Estimated mean figure‐of‐eight circumference at V1 and V3 by bandaging technique and sex.

3.3.2. Interface Pressure at V2

At V2, 48 h before V3, mean interface pressure was 39.7 mmHg (95% CI, 37.9 to 41.6) in the heel‐free group and 36.4 mmHg (95% CI, 34.6 to 38.2) in the control group, a difference of 3.23 mmHg (95% CI, 0.6 to 6.0).

3.3.3. Change in Interface Pressure From V2 to V3

Over the final 48 h, mean interface pressure changed by −9.7 mmHg (95% CI, −11.5 to −7.9) in the heel‐free group and −8.0 mmHg (95% CI, −9.8 to −6.2) in the control group. The between‐group difference in change was −1.7 mmHg (95% CI, −4.3 to 0.8).

In the analysis adjusted for sex, the estimated changes were −9.7 mmHg (95% CI, −11.8 to −7.5) and −8.0 mmHg (95% CI, −10.1 to −5.9), respectively; the between‐group difference was −1.7 mmHg (95% CI, −4.7 to 1.3) (Figure 4).

FIGURE 4.

Estimated mean values of the interface pressure (mmHg) at V1, V2 and V3, by bandaging technique and by sex.

3.3.4. Patient Preferences

Thirty‐five patients had experienced only one technique. Among the 48 who had experienced both, 23 (47.9%) preferred heel‐free bandaging, 14 (29.2%) preferred heel‐wrapped bandaging and 11 (22.9%) had no preference (Table 3).

TABLE 3.

Patient preferences by randomised group.

	Heel‐free (N = 42)	Control (N = 43)	Overall (N = 85)
Experience and preferred technique, n/N (%)
I have only experienced one of the two techniques	5/40 (12.5%)	30/43 (69.8%)	35/83 (42.2%)
Among patients who had experienced both techniques a
I prefer the heel‐free technique	20/35 (57.1%)	3/13 (23.1%)	23/48 (47.9%)
I prefer the heel‐wrapped technique	6/35 (17.1%)	8/13 (61.5%)	14/48 (29.2%)
No preference	9/35 (25.7%)	2/13 (15.4%)	11/48 (22.9%)
Missing	2	0	2
Who should choose the technique?
The nurse	23 (57.5%)	29 (67.4%)	52 (62.7%)
The patient and the nurse jointly	16 (40.0%)	14 (32.6%)	30 (36.1%)
The patient	1 (2.5%)	0	1 (1.2%)
Missing	0	2	2

^a Among patients who had experienced both techniques (heel‐free group, n = 35; control group, n = 13).

4. Discussion

The primary finding was that heel‐free bandaging was non‐inferior to heel‐wrapped bandaging for interface pressure 48 h after application. This result was consistent across the PP analysis and three sensitivity analyses. The finding pertains to interface pressure and should not be interpreted as equivalence in ulcer healing or other clinical outcomes.

Interface pressure decreased over 48 h in both groups. At V3, mean values were 30.1 mmHg with heel‐free bandaging and 28.4 mmHg with heel‐wrapped bandaging, within or near the 30‐ to 40‐mmHg range recommended in French guidance. Higher pressures may accelerate healing in some settings but may also increase discomfort and tissue injury [6, 9]. The present results indicate that leaving the heel uncovered did not produce a clinically relevant additional loss of interface pressure.

Among patients who had experienced both techniques, 47.9% preferred heel‐free bandaging, 29.2% preferred heel‐wrapped bandaging, and 22.9% had no preference. Because the questionnaire did not assess reasons for preference or adherence, these data do not establish that heel‐free bandaging is more comfortable or improves adherence. They nevertheless support incorporating informed patient preference when either technique is clinically appropriate.

Ulcer location may modify the clinical implications of these findings. Retromalleolar and ankle ulcers have been associated with slower healing and may require targeted local compression, including direct pressure over the wound bed [20, 21]. Because ulcer location was not included in the present analysis, the findings should not be generalised to these anatomical sites without further evidence.

Bandages after V1 were applied by multiple home care nurses, which reflects routine practise but may have increased variability in application. Standardised training and access to pressure‐measurement devices may help clinicians achieve and maintain target pressure across care settings.

4.1. Strengths and Limitations

Strengths include the randomised non‐inferiority design, prespecified power, standardised B1 measurements and concordant PP and sensitivity analyses. Limitations include the single‐centre setting, lack of blinding, variation in community‐nurse experience and a non‐inferiority margin based on clinical consensus rather than direct comparative efficacy data [16]. The trial was not designed to evaluate ulcer healing, treatment adherence or comparative clinical safety. These limitations constrain conclusions about therapeutic equivalence.

4.2. Implications for Nursing Practise

Heel‐free bandaging may be considered when both techniques are clinically appropriate, particularly when heel wrapping interferes with footwear or patient acceptability. The choice should account for ulcer location, limb anatomy, professional judgement and patient preference. The present findings support comparable interface pressure, not equivalent healing or safety.

5. Conclusion

Heel‐free compression bandaging was non‐inferior to heel‐wrapped bandaging for interface pressure 48 h after application. It may therefore offer a flexible alternative when clinically appropriate and preferred by the patient. Trials assessing ulcer healing, adverse events, adherence, quality of life and outcomes by ulcer location are needed before broader clinical recommendations are made.

Funding

This study was supported by a grant from the French Ministry of Health (PHRIP 2018_18‐0550 2019_ID‐RCB : A00785‐52).

The pressure measurements were obtained using the PicoPress (Microlab Elettronica, Villatora, Italy) graciously loaned by the laboratories Lohmann‐Rauscher (Germany) and Urgo (France).

Ethics Statement

This trial was conducted in accordance with the Ethic Committee (reference 2019/48), the French Agency for drug safety and with Good Clinical Research Practises.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Location and type of work planned for each study visit.

Figure S2: B1 measuring point.

Figure S3: Measure of the Figure of 8.

Mauvilain A., Durant C., Voisine A., Perrot B., and Cartron E., “Heel‐Free Compression Bandaging for Venous Leg Ulcers: A Non‐Inferiority Trial of Interface Pressure,” International Wound Journal 23, no. 10 (2026): e71052, 10.1111/iwj.71052.

Data Availability Statement

The data is hosted by the CHU de Nantes (the Nantes University Hospital). They are available on reasoned request from the research unit at Nantes University Hospital: recherche-paramedicale@chu-nantes.fr.