E P I DE M I O L O G Y A N D HE A L T H S C I E N C E R ES E A RC H
BJD
British Journal of Dermatology
Prognostic factors associated with healing of venous leg ulcers: a multicentre, prospective, cohort study G. Chaby,1 P. Senet,2 O. Ganry,3 A. Caudron,1 D. Thuillier,1 C. Debure,4 S. Meaume,5 F. Truchetet,6 P. Combemale,7 F. Skowron,8 P. Joly9 and C. Lok;1 for the Angio-Dermatology Group of the French Society of Dermatology 1
Department Department 4 Department 5 Department 6 Department 7 Department 8 Department 9 Department 2
of of of of of of of of
Dermatology, H^opital Sud, and 3Department of Epidemiology and Public Health, H^opital Nord, University of Amiens, 80054 Amiens, France Dermatology, H^opital Tenon, Assistance Publique–H^opitaux de Paris (AP–HP), University of Paris X, 75020 Paris, France Vascular Rehabilitation, H^opital Corentin-Celton, AP–HP, 92133 Issy-les-Moulineaux CEDEX, France Dermatology and Gerontology, H^opital Rothschild, AP–HP, University of Paris VI, 75571 Paris, France Dermatology, H^opital Bon-Secours, 57038 Metz, France Dermatology, Centre Leon-Berard, 69373 Lyon CEDEX 08, France Dermatology, Centre Hospitalier de Valence, 26953 Valence, France Dermatology, H^opital Charles-Nicolle, University of Rouen, 76038 Rouen, France
Summary Correspondence Guillaume Chaby. E-mail: [email protected]
Accepted for publication 31 July 2013
Funding sources The study was funded by a grant from the French Society of Dermatology in 2003 and was accorded a research grant from Servier Medical, France, in 2003.
Conflicts of interest P.S. has received consulting fees from Servier and Sanofi, and fees from Urgo and Pierre Fabre as an investigator in clinical studies on venous ulcer dressings; G.C. has received fees from Urgo Medical as an investigator; F.T. has received fees from Urgo as an investigator and expert; P.C. has received fees from Urgo and Coloplast as an expert; C.L. has received consulting fees from Actelion, Servier, Coloplast, Urgo, Pierre Fabre and M€olnlycke Healthcare as an investigator in clinical studies on venous ulcer dressings. DOI 10.1111/bjd.12570
Background Some prognostic markers of venous leg ulcer (VLU) healing have been evaluated, mostly in retrospective studies. Objectives To identify which clinical characteristics, among those known as possible prognostic factors of VLU healing, and which VLU-associated sociodemographic and psychological factors, are associated with complete healing at week 24 (W24). Methods A prospective, multicentre, cohort study was conducted in 22 French dermatology departments between September 2003 and December 2007. The end point was comparison between healed and nonhealed VLUs at W24, for patient clinical and biological characteristics; psychological, cognitive and social assessments; affected leg inclusion characteristics; venous insufficiency treatment and percentage of initial wound area reduction during follow-up. Results In total, 104 VLUs in 104 patients were included; 94 were analysed. The mean VLU area and duration were 368 555 cm2 and 248 457 months, respectively. At W24, 41/94 VLUs were healed. Univariate analysis significantly associated complete healing with superficial venous surgery (P = 0001), adherence to compression therapy at W4 (P = 003) and W24 (P = 001), ankle-joint ankylosis (P = 001) and mean percentage of VLU area reduction at W4 (P = 004). Multivariate analysis retained superficial venous surgery during follow-up [odds ratio (OR) 84, 95% confidence interval (CI) 19–482] and percentage reduction of the VLU area at W4 (OR 16, 95% CI 10–214) as being independently associated with healing. Conclusions These results indicate that complete healing of long-standing, large VLUs is independently associated with ablation of the incompetent superficial vein and percentage of wound area reduction after the first 4 weeks of treatment.
What’s already known about this topic?
• •
1106
Despite advances in venous leg ulcer (VLU) management, the time to ulcer healing generally remains long. Some prognostic markers of healing have been evaluated, mostly in retrospective studies.
British Journal of Dermatology (2013) 169, pp1106–1113
© 2013 British Association of Dermatologists
Prognostic factors of venous leg ulcer healing, G. Chaby et al. 1107
What does this study add?
•
In this prospective, multicentre, cohort study, complete healing of long-standing, large VLUs at 6 months was independently associated with ablation of the incompetent superficial vein and percentage of wound area reduction after the first 4 weeks of treatment.
Venous leg ulcers (VLUs) represent a major complication of chronic venous insufficiency and a common source of morbidity, especially in those aged over 60 years. The prevalence of leg ulcers with a predominant venous cause has been estimated at 1–15% of the adult population.1 Despite advances in the management of these patients, the time to ulcer healing generally remains long: 30–70% of VLUs heal only within 6 months, and 10–20% require therapy lasting > 1 year.2,3 Several factors have been associated with patients with VLUs, compared with general populations – poorer socioeconomic status, lower educational level and social isolation and/or depression – but their impact on healing has not been investigated.4,5 In current practice, several factors are commonly thought to be associated with failure of VLUs to heal, but have not yet been documented. These include poor adherence to compression therapy, conditions that may decrease compliance with wound care management (e.g. impaired mobility), number of comorbidities, calf muscle-pump dysfunctions (e.g. ankle-joint ankylosis) and factors limiting tissue remodelling, such as protein deficiency. Some prognostic markers of healing have been evaluated, mostly in retrospective studies. Margolis et al.2 published a simple statistical prognostic model based on wound area (≥ 10 cm2) and duration (≥ 12 months), validated on a patient database, whereas Kantor and Margolis6 showed that complete ulcer healing at week 24 (W24) was associated not with initial ulcer size, but with the percentage of area reduction after 4 weeks of treatment. Therefore, our study was undertaken to identify prospectively which clinical VLU characteristics known to be possible prognostic markers, and which VLU-associated sociodemographic and psychological factors, are associated with complete healing at W24.
Patients and methods Design overview This prospective, multicentre, cohort study, conducted in 22 French centres from September 2003 to December 2007, included 24 weeks of follow-up and was performed in accordance with the Declaration of Helsinki. The Institutional Review Board of Picardie, Centre Hospitalier Universitaire d’Amiens, and regulatory authorities approved its protocol (NCT 01673412). All subjects received detailed information about the study and gave their written consent to participate. © 2013 British Association of Dermatologists
Setting and participants The study population consisted of ambulatory or hospitalized patients managed at or referred to each centre for VLU management. Patients were included with the following criteria: (i) > 50 years old; (ii) more than one VLU lasting ≥ 1 month and with wound area > 1 cm2; (iii) clinical findings consistent with established venous disease (skin hyperpigmentation, varicose veins, lipodermatosclerosis) confirmed by venous Doppler duplex ultrasonography performed during the preceding 6 months; (iv) absence of significant arterial insufficiency assessed by clinical findings (intermittent claudication or resting pain, necrotic or distal foot wound) and ankle brachial index (ABI) ≥ 08; and (v) ability to give informed consent and to be followed for 24 weeks. Exclusion criteria were (i) ongoing systemic diseases known to be associated with pyoderma gangrenosum or necrotizing vasculitis; (ii) corticosteroid, or cytotoxic or immunosuppressant drug use during the preceding 3 months; (iii) hypertensive leg ulcer; (iv) foot ulcer; and (v) ABI < 08. When several VLUs were present, the largest was chosen as the reference VLU (rVLU). Study procedures For all subjects satisfying the entry criteria, the following baseline information was recorded: (i) sociodemographic characteristics; (ii) medical and surgical history; (iii) concomitant illness(es); (iv) current medications; (v) mobility and autonomy status; (vi) psychological profile; (vii) serum albumin and haemoglobin concentrations; (viii) wound aetiology, Doppler duplex ultrasonography and ABI results; (ix) CEAP (clinical– etiology–anatomy–pathophysiology) classification; (x) wound features, including rVLU duration, location, size and related pain; and (xi) characteristics of the affected lower leg. All patients were prescribed standard therapy, including daily application of high-compression elastic bandages (30– 40 mmHg at the ankle), and dressings depending on the wound stage.7 At inclusion, patients received detailed information about compression therapy. The frequency of dressing changes depended on the wound stage and was left to the investigator’s judgement. Patients with isolated superficial reflux and/or perforating veins were referred during the study to a vascular surgeon for an opinion on surgical treatment of venous insufficiency. The same investigator saw patients monthly until W24, regardless of whether or not they had undergone surgery. British Journal of Dermatology (2013) 169, pp1106–1113
1108 Prognostic factors of venous leg ulcer healing, G. Chaby et al.
Outcomes and measurements The study judgement criterion was the comparison between healed and nonhealed VLUs at W24, with respect to the following characteristics: (i) general patient characteristics: age, sex, body mass index (BMI), presence of anaemia (haemoglobin < 12 g dL 1 in women, < 13 g dL 1 in men) and hypoalbuminaemia (albumin < 35 g L 1), level of autonomy and mobility, social status (educational level, marital status, occupation, health insurance coverage, income, living conditions); (ii) presence of the following comorbidities: hypertension, diabetes, cardiac insufficiency, rheumatological disease(s) involving the affected leg, history of orthopaedic surgery (hip or knee), prior deep-vein thrombosis, renal insufficiency, previous vein surgery; (iii) psychological profile; (iv) affected lower-leg characteristics: presence of white atrophy, lipodermatosclerosis or ankle-joint ankylosis; (v) rVLU characteristics: initial area, duration, recurrent nature, granulation tissue > 50%, evolution of ulcer-related pain during dressing change; (vi) venous Doppler duplex ultrasonography findings: presence of obstruction and/or reflux in the superficial- and/ or deep-venous system; (vii) rVLU management during follow-up: compression adherence, modalities of compression use at W4 and W24, superficial vein surgery; and (viii) percentage wound area reduction at W4 – ulcer areas were calculated by measuring tracings of planimetry images, according to the same standardized protocols. The total ulcer area of the affected leg was estimated by adding the area of each ulcer. Areas were determined at inclusion and every 4 weeks until W24. Change of ulcer-related pain was assessed with a pain visual analogue scale (range 0– 100). Autonomy and mobility were assessed with Katz’s basic Activities of Daily Living (ADL) score, which measures basic activities for a 1-month period prior to the interview; a score < 3 (out of 6) indicates major difficulties in performing normal self-care activities. Patients were evaluated for compression adherence and acceptance at W4 and W24. The patient was considered to be compression adherent when (i) compression bandages were applied < 1 h after getting up; (ii) they were kept in place for 80% of the time during the day and for > 5 days per week; (iii) they covered the dressing at each visit; and (iv) bandages were in good condition. Compression acceptance was evaluated at W4 and W24 on a four-point Likert scale: 1, easily; 2, with minor difficulties; 3, with difficulties; 4, unbearably. The patient was considered to have understood the compression benefit when they responded ‘no’ to the statements ‘the main interest of compression is to maintain the dressing’ and ‘the compression is also effective if it is worn only a few hours a day’; and ‘yes’ to ‘compression therapy is an essential treatment of venous ulcers’ and ‘compression therapy decreases venous stasis of the lower leg’. Psychological and cognitive assessments were performed at inclusion. Mini-Mental State Examination (MMSE; range 0– 30) and Raven’s Coloured Progressive Matrices (RCPM; range 0–36) were used to evaluate cognitive capacity and inductive British Journal of Dermatology (2013) 169, pp1106–1113
capacity, respectively. Depression and self-esteem were evaluated with the Beck Depression Inventory (BDI) and the Coopersmith Self-Esteem Inventory (CSEI), respectively. The BDI is a 21-question self-reporting inventory; each question has a set of at least four possible responses, measuring the severity of depression: 0–9, minimal depression; 10–18, mild depression; 19–29, moderate depression; 30–63, severe depression. The CSEI is a self-assessment score of self-esteem with four domains and 50 items: general (26 items), social (eight items), familial (eight items) and professional scales (eight items); a score ≤ 33/50 indicates very low selfesteem. Patients’ monthly incomes were self-reported, and classified using the median income in France as the reference (> or ≤ €1524 per month), with the poverty threshold set at 50% of the median.8 Statistical analyses Continuous variables (age, BMI, and VLU area and duration) are expressed as the mean SD (when normally distributed) or as medians; categorical variables (all others) are expressed as number (%). Possible independent predictors of healing were first identified by univariate analysis using Student’s t-test for continuous variables, or the v2 or Fisher exact test for qualitative variables. P < 005 defined significance. All statistical tests were two sided. Single variable logistic regression analyses assessed the magnitude of riskfactor effects, giving estimated odds ratios (ORs) with 95% confidence intervals (CIs). Multivariate models, including variables with P < 010 in univariate analyses, estimated the independence of prognosis-associated parameters. Reaching 80% power required inclusion of 140 patients, with a 65% expected cure rate. Database management and all statistical analyses were performed with SPSS software (IBM, Armonk, NY, U.S.A.).
Results Baseline characteristics Of the 104 patients with VLU, 10 did not complete the entire study and were excluded from the statistical analyses: one was lost-to-follow-up after cellulitis; five withdrew their consent because they considered the mandatory consultations and psychological tests too much for them to commit to, and four gave no specific reason (Fig. 1). The mean age at inclusion was 735 99 years (median 76) and most patients were female (71%). The mean number of VLUs on the affected leg was 16 10 (median 20), with no between-group difference; 51% of the patients had only one VLU at inclusion. The mean rVLU area and duration at inclusion were 368 555 cm2 and 248 457 months, respectively. Notably, 63% of rVLUs were located on the medial malleolus. The patient and VLU baseline characteristics are summarized in Table 1. © 2013 British Association of Dermatologists
Prognostic factors of venous leg ulcer healing, G. Chaby et al. 1109
104 VLUs in 104 patients Visit 1 History and physical examination Affected leg characteristics Wound area measurement Pain visual analogue scale Activities of Daily Living score Questionnaires: Socioeconomic level BDI CSEI MMSE RCPM Biology and Doppler duplex Detailed information on compression therapy Week 4 Ulcer area calculated on planimetry tracing Adherence to compression Compression acceptance Weeks 8, 12, 16, 20 Ulcer area calculated on planimetry tracing Week 24 Ulcer area calculated on planimetry tracing Adherence to compression Compression acceptance
Fig 1. Flow diagram of patients with venous leg ulcers (VLUs) throughout the study. BDI, Beck Depression Inventory; CSEI, Coopersmith Self-Esteem Inventory; MMSE, Mini-Mental State Examination; RCPM, Raven’s Coloured Progressive Matrices.
Follow-up and overall outcome Factors affecting complete healing at W24 were analysed for the 94 patients still being followed at the end of the study. Nine patients withdrew their participation and one was hospitalized. At W24, 41 rVLUs were completely healed and 53 were nonhealed. Follow-up compression therapy and surgical treatment findings are summarized in Table 1. Factors identified by univariate analysis as predicting venous leg ulcer healing or nonhealing Concerning baseline characteristics and comorbidities, anklejoint ankylosis was the only between-group difference found, being significantly more frequent in patients with nonhealed VLUs at 6 months (60% vs. 37%; P = 003). Comparing the healed and nonhealed VLU groups, respectively, complete rVLU healing was significantly associated with superficial venous system surgery during follow-up (27% vs. 4%, P = 0001), compression therapy adherence at W4 (87% vs. 76%, P = 003) and W24 (88% vs. 59%, P = 001), and mean percentage rVLU area reduction at W4 (48% vs. 27%, P = 004), with the difference becoming significant at the 30% threshold (P = 003). In addition, complete rVLU healing at W24 was associated with complete healing of all VLUs on the affected leg, for all 21 healed patients with multiple © 2013 British Association of Dermatologists
94 Patients completed follow-up 10 Patients excluded (lost to follow-up before week 4) 9 Patient decisions 1 Lost to follow-up
41 VLUs healed
53 VLUs not healed
lesions. Among the 49 patients eligible for venous surgery, 13 underwent surgery and 36 were treated with compression alone. Among the latter, 31% of VLUs were healed at W24 and 69% were not (P = 0007). No significant between-group differences were found for VLU area (P = 023), duration (P = 007) or the presence of deep-venous insufficiency (P = 079). Compression was applied less frequently by patients themselves or a home helper than by nurses or physicians for the healed vs. nonhealed group, but the difference reached significance only at W4 (29% vs. 49%, P < 001). Compression acceptance was considered easy by most healed and nonhealed patients with VLU (respectively, 56% vs. 58% at W4 and 72% vs. 57% at W24), with no significant between-group difference. For the healed and nonhealed groups, correct understanding of the benefit of compression therapy was reported, respectively, by 20% vs. 22% of the patients at W4 (P = 071), and by 25% vs. 29% at W24 (P = 067). Sociodemographic data were similar for the healed and nonhealed groups: a majority of patients had median or lower socioeconomic status, i.e. monthly income < €1524, and nearly half of the patients in both groups lived alone (Table 2). Meanwhile, significantly more patients with nonhealed VLUs received wound care at home, rather than office-based care, compared with those with healed VLUs (P = 001). MMSE, BDI, CSEI and RCPM scores were comparable for the two groups (Table 3). British Journal of Dermatology (2013) 169, pp1106–1113
1110 Prognostic factors of venous leg ulcer healing, G. Chaby et al. Table 1 Baseline and follow-up characteristics of patients according to venous leg ulcer (VLU) status Characteristic General Age (years), mean (SD) Women, n (%) Body mass index (kg m 2), mean (SD) Albuminaemia (albumin < 35 g L 1), n (%) Anaemia, n (%) ADL score (out of 6), mean (range) ADL score < 3, n Comorbidity Cardiovascular risk factors (HT, diabetes), n (%) Cardiac insufficiency, n (%) Rheumatological disease of affected leg, n (%) Orthopaedic surgery (hip or knee), n (%) Prior deep-vein thrombosis, n (%) Renal insufficiency, n (%) Prior venous surgery, n (%) VLUs Area (cm2), mean (SD) Duration (months), mean (SD) Recurrent, n (%) Granulation tissue > 50%, n (%) VAS score, mean (range) Venous duplex ultrasonography Isolated SVI, n (%) DVI, isolated or not, n (%) Popliteal vein reflux, n (%) Leg characteristic Lipodermatosclerosis, n (%) White atrophy, n (%) Ankle-joint ankylosis, n (%) Follow-up VLU area reduction at week 4 (%) ≤ 20%, n ≤ 30%, n ≤ 40%, n Adherence to compression Week 4 (%) Week 24 (%) Superficial vein surgery, n (%)
Healed (n = 41)
Nonhealed (n = 53)
OR (95% CI)
P-value
739 (101) 38 (72) 302 (61) 34 (64) 22 (42) 56 (2–6) 1
103 105 088 207 123 – –
(077–134) (039–283) (066–127) (083–521) (049–310)
089 091 062 010 057 084 087
(73) (12) (27) (44) (59) (15) (37)
33 (62) 7 (13) 17 (32) 23 (43) 29 (55) 4 (8) 25 (47)
060 110 129 098 086 048 155
(023–160) (028–389) (048–349) (045–242) (035–211) (010–328) (061–382)
032 081 054 094 079 046 043
289 (522) 161 (232) 32 (78) 20 (49) 46 (0–90)
427 (577) 315 (567) 45 (85) 16 (30) 48 (0–100)
214 288 158 045
(122–437) (135–623) (049–513) (018–115) –
023 007 040 028 071
22 (54) 19 (46) 14 (34)
25 (47) 28 (53) 21 (40)
077 (031–189) 130 (053–319) 127 (050–323)
053 056 054
32 (78) 22 (54) 15 (37)
47 (89) 33 (62) 32 (60)
220 (069–787) 142 (057–355) 264 (105–670)
034 045 003
48 11 12 12
27 22 27 31
289 194 251 341
(127–584) (074–514) (098–653) (132–893)
004 014 003 0005
87 88 11 (27)
76 59 2 (4)
045 (020–099) 022 (010–046) 011 (002–057)
003 001 0001
734 (97) 29 (71) 304 (73) 19 (46) 15 (37) 55 (3–6) 0 30 5 11 18 24 6 15
ADL, Activities of Daily Living; CI, confidence interval; DVI, deep-vein insufficiency; HT, hypertension; OR, odds ratio; SVI, superficial vein insufficiency; VAS, visual analogue scale for pain.
Independent prognostic factors of healed and nonhealed venous leg ulcers Our multivariate analyses retained two prognostic factors as being significantly associated with complete VLU healing at W24: surgery for superficial venous insufficiency during follow-up (OR 84, 95% CI 19–482) and percentage rVLU area reduction at W4 (OR 16, 95% CI 10–214).
Discussion The main objective of this prospective study was to identify clinical, social, psychological and cognitive factors associated with VLU healing after 24 weeks. Multivariate analyses British Journal of Dermatology (2013) 169, pp1106–1113
retained surgical treatment of superficial venous incompetence during follow-up and the healing rate at W4, assessed as the percentage VLU area change, as significant and independent factors predicting complete VLU healing at W24. Notably, complete rVLU healing was associated with complete healing of all lesions on the affected leg of all patients with multiple lesions at inclusion. The mean age, sex ratio and BMI were the same as those previously described as being associated with VLU.1,2,4 Herein, the high frequencies of patients without high-school diplomas and with monthly incomes below the French median suggest, as shown previously in other European countries, that VLU might also be associated with low social status in France.4,9 © 2013 British Association of Dermatologists
Prognostic factors of venous leg ulcer healing, G. Chaby et al. 1111 Table 2 Socioeconomic characteristics of patients according venous leg ulcer status Socioeconomic characteristic
Healed (N = 41), n (%)
Nonhealed (N = 53), n (%)
Home owner Rented home Free universal healthcarea Long-term illnessb Additional private health insurance Urban Rural Living alone Home wound carec Income (€ per month) < 762 762–1524 1524–2286 > 2286 Education level No diploma Less than high school High school and above Marital status Single Married/partnership Widowed Employment Working Not working On disability benefit Retired
25 (61) 14 (34) 2 (5)
26 (49) 23 (43) 2 (4)
28 (68) 37 (90)
42 (79) 48 (91)
32 (78) 9 (22) 21 (51) 28 (68)
41 12 34 48
12 (29) 20 (49) 9 (22) 0
12 (23) 24 (45) 13 (25) 4 (8)
17 (41) 22 (54) 2 (5)
24 (45) 26 (49) 3 (6)
5 (12) 19 (46) 17 (41)
4 (8) 29 (55) 20 (38)
1 (2) 2 (5) 2 (5) 36 (88)
2 (4) 4 (8) 4 (8) 43 (81)
(77) (23) (64) (91)
a Universal healthcare coverage (Couverture Maladie Universelle; CMU) is a free French social welfare programme for people on low incomes; the CMU offers 100% healthcare coverage that is added to the standard National Health Insurance. bLong-term illness (affection de longue duree) includes long-term or major disease for which all costs are fully reimbursed by the National Health Insurance. cBetween-group comparison is significant (P < 005).
Clinical leg characteristics indicated that our population had severe venous disease. Indeed, lipodermatosclerosis and white atrophy are signs of severe chronic venous disorders, and were observed in 84% and 59% of the study patients, respectively.10 VLUs had been present for a mean of 248 months, with a mean area of 368 cm2, and were recurrent for 82% of the patients, thereby meeting the literature definition of hardto-heal ulcers.2,11 Indeed, using the statistical prognostic model of Margolis et al.,2 based on ulcer size and duration, 77% (72/94) of the VLUs in our study had < 50% chance of healing at W24. The study setting in tertiary care centres might partially explain the high rate of hard-to-heal ulcers, further emphasizing that long-lasting and relapsing VLUs are resistant to standard care, and are more frequently referred to specialized facilities than ‘usual’ VLUs, which heal within a © 2013 British Association of Dermatologists
few months.2,12,13 Thus, our results might not be applicable to patients with milder disease treated in primary-care settings. A limitation of our study is that relatively fewer VLUs than expected were evaluated at the end of the study. Indeed, numerous patients declined participation given the length of the study. This limitation might have created a selection bias, leading to a lack of power to establish some differences, including VLU area or duration, which are usually associated with late healing. Logistic regression analysis yielded ORs > 1, i.e. close to those previously reported, for VLU area and duration, whereas Student’s t-test, which is more sensitive to the study power, showed no significant between-group differences for those parameters. Moreover, multivariate analysis did not retain these two variables as being independently associated with healing. According to well-validated psychometric instruments (BDI, CSEI), our results showed that almost half of the patients were depressed and that one-third had low self-esteem. These psychological problems might be explained by the chronic, relapsing course of the venous disease investigated in our study, but we are unable to provide any evidence supporting a relationship between these psychological factors and complete VLU healing at W24. Conflicting results have been published concerning the association of a history of deep-vein thrombosis or popliteal reflux with late healing.13–17 Some of those studies were criticized because of the systematic lack of vascular imaging or their retrospective design. The VLUs of our patients with residual deep-vein obstruction and deep-valve incompetence were not less significantly healed at W24. Our univariate analysis identified ankle-joint ankylosis, which increases venous hypertension, as prospectively predictive of failure to heal at W24, but it was not retained in the multivariate analysis, possibly because it is merely a marker of the severe disease that characterized our patients. Ankle-joint ankylosis had not been previously evaluated as a prognosis factor, making any comparisons difficult. Even though it is accepted that good adherence to compression therapy is associated with faster healing, few studies have examined compression compliance as a factor predictive of VLU healing.1,18 According to our univariate analyses, poor compression adherence was significantly associated with nonhealed VLUs. Because compression compliance was assessed by a questionnaire, we cannot exclude that it might have been overevaluated, with some responses corresponding to expected, socially desirable behaviour. Despite written information and oral instructions emphasizing the importance of compression therapy, lower adherence to compression therapy in the nonhealed group during follow-up might be explained by a loss of motivation reflecting impaired healing or lower compression efficacy in this group, whose participants or entourage more frequently applied the bandages, rather than nurses or physicians. Our multivariate analysis retained superficial vein surgery during follow-up as being associated with complete VLU healBritish Journal of Dermatology (2013) 169, pp1106–1113
1112 Prognostic factors of venous leg ulcer healing, G. Chaby et al.
Psychological test
Healed
Nonhealed
Beck Depression Inventory 0–11: no depression 12–19: light depression 20–27: moderate depression 28–63: severe depression Coopersmith Self-Esteem Inventory Score (out of 50), mean SD Score ≤ 33, low self-esteem Raven’s Coloured Progressive Matrices Mean SD (median) Mini-Mental State Examination Mean SD (range)
N = 40
N = 53
22 (55) 13 (32) 4 (10) 1 (2) N = 39 355 78 13 (33) N = 33
31 (58) 13 (25) 6 (11) 3 (6) N = 53 354 78 15 (28) N = 47
245 84 (27) N = 41 261 43 (16–30)
208 91 (21) N = 53 266 36 (16–30)
Table 3 Psychological characteristics of patients according to venous leg ulcer status
Values are n (%) unless stated otherwise. All between-group comparisons were nonsignificant (P ≥ 005).
ing at W24. In a recent, nonrandomized, prospective study, endovenous ablation of incompetent superficial and perforator veins was associated with a shorter time to healing and a higher complete healing rate, compared with medical treatment alone.19 However, this point is still being debated, because surgical treatment of superficial vein reflux was shown to lower the VLU recurrence rate but not to increase the complete healing rate in the prospective, randomized ESCHAR study.20 Finally, according to our multivariate analysis, the percentage VLU area change at W4 was predictive of complete healing at W24. The healing rate at W4 was shown to be a reliable criterion to predict VLU healing in a retrospective study on 306 VLUs by Cardinal et al.,21 and in a prospective study on 104 patients by Kantor and Margolis.6 In our study, a mean area reduction ≤ 30% at W4 was predictive of incomplete healing at W24, thereby confirming that the percentage wound area change at W4 can be used as a surrogate marker of healing in clinical trials to predict longer-term outcomes of healing technologies.22,23 Our results suggest that, for this cohort of long-standing, large VLUs, the prognosis of complete healing at W24 was independently associated with ablation of the incompetent superficial vein, which can rapidly lower venous hypertension, rather than with inherent VLU characteristics, e.g. wound area and duration, or with patient characteristics that are not easily modifiable, e.g. general mobility, protein deficiency, and psychological and socioeconomic factors. The healing rate at W4 under optimal treatment might predict complete healing at W24 and enable consideration of alternative therapies for VLUs not responding to this regimen.
Acknowledgments The authors thank the investigators from the Angio-Dermatology Group of the French Society of Dermatology: Henri Adamski, Jean-Marie Bonnetblanc, Jacqueline Chevrant-Breton, British Journal of Dermatology (2013) 169, pp1106–1113
Anne Dompmartin, Gerard Lorette, Herve Maillard, Philippe Modiano, Geraldine Perceau, Marie Aleth Richard-Lallemand, Barbara Roth, Pascal Toussaint and Marie-Francßoise VendeaudBusquet.
References 1 van Gent WB, Wilschut ED, Wittens C. Management of venous ulcer disease. BMJ 2010; 341:c6045. 2 Margolis DJ, Allen-Taylor L, Hoffstad O, Berlin JA. The accuracy of venous leg ulcer prognostic models in a wound care system. Wound Repair Regen 2004; 12:163–8. 3 O’Meara S, Tierney J, Cullum N et al. Four layer bandage compared with short stretch bandage for venous leg ulcers: systemic review and meta-analysis of randomized controlled trials with data from individual patients. BMJ 2009; 338:b1344. 4 Moffatt CJ, Franks PJ, Doherty DC et al. Sociodemographic factors in chronic leg ulceration. Br J Dermatol 2006; 155:307–12. 5 Scott TE, LaMorte WW, Gorin DR, Menzoian JO. Risk factor for chronic venous insufficiency: a dual case–control study. J Vasc Surg 1995; 22:622–8. 6 Kantor J, Margolis DJ. A multicentre study of percentage change in venous leg ulcer area as a prognostic index of healing at 24 weeks. Br J Dermatol 2000; 142:960–4. 7 Chaby G, Senet P, Vaneau M et al. Dressings for acute and chronic wounds: a systematic review. Arch Dermatol 2007; 143:1297–304. 8 Report by the National Observatory on Poverty and Social Exclusion. Available at: www.onpes.gouv.fr (last accessed 3 October 2013). 9 Herberger K, Rustenbach SJ, Grams L et al. Quality-of-care for leg ulcers in the metropolitan area of Hamburg – a community-based study. J Eur Acad Dermatol Venereol 2011; 26:495–502. 10 Ekl€ of B, Rutherford RB, Bergan JJ et al. Revision of the CEAP classification for chronic venous disorders: consensus statement. J Vasc Surg 2004; 40:1248–52. 11 Chaby G, Viseux V, Ramelet AA et al. Refractory venous leg ulcers: a study of risk factors. Dermatol Surg 2006; 32:512–19. 12 Margolis DJ, Berlin JA, Strom BL. Risk factors associated with the failure of a venous leg ulcer to heal. Arch Dermatol 1999; 135:920–6. 13 Franks P, Moffatt CJ, Connoly M et al. Factors associated with healing leg ulceration with high compression. Age Ageing 1995; 24:407–10. © 2013 British Association of Dermatologists
Prognostic factors of venous leg ulcer healing, G. Chaby et al. 1113 14 Skene AI, Smith JM, Dore CJ et al. Venous leg ulcers: a prognostic index to predict time to healing. BMJ 1992; 305:b1119–21. 15 Brittenden J, Bradbury AW, Allan PL et al. Popliteal vein reflux reduces the healing of chronic venous ulcer. Br J Surg 1998; 85:60–2. 16 Barwell JR, Ghauri ASK, Taylor M et al. Risk factors for healing and recurrence of chronic venous ulcers. Phlebology 2000; 15:49–52. 17 Moffatt CJ, Doherty DC, Smithdale R, Franks PJ. Clinical predictors of leg ulcer healing. Br J Dermatol 2010; 162:51–8. 18 O’Meara S, Cullum N, Nelson EA, Dumville JC. Compression for venous leg ulcers. Cochrane Database Syst Rev 2012; 11:CD000265. 19 Harlander-Locke M, Lawrence PF, Alktaifi A et al. The impact of ablation of incompetent superficial and perforator veins on ulcer healing rates. J Vasc Surg 2012; 55:458–64.
© 2013 British Association of Dermatologists
20 Barwell JR, Davies CE, Deacon J et al. Comparison of surgery and compression with compression alone in chronic venous ulceration (ESCHAR study): randomized controlled trial. Lancet 2004; 363:1854–9. 21 Cardinal M, Eisenbud DE, Phillips T, Harding K. Early healing rates and wound area measurements are reliable predictors of later complete wound closure. Wound Repair Regen 2008; 16:19–22. 22 Gelfand JM, Hoffstad O, Margolis DJ. Surrogate endpoints for the treatment of venous leg ulcers. J Invest Dermatol 2002; 119:1420–5. 23 Kurd SK, Hoffstad OJ, Bilker WB, Margolis DJ. Evaluation of the use of prognostic information for the care of individuals with venous leg ulcers or diabetic neuropathic foot ulcers. Wound Repair Regen 2009; 17:318–25.
British Journal of Dermatology (2013) 169, pp1106–1113
BJD
British Journal of Dermatology
Prognostic factors associated with healing of venous leg ulcers: a multicentre, prospective, cohort study G. Chaby,1 P. Senet,2 O. Ganry,3 A. Caudron,1 D. Thuillier,1 C. Debure,4 S. Meaume,5 F. Truchetet,6 P. Combemale,7 F. Skowron,8 P. Joly9 and C. Lok;1 for the Angio-Dermatology Group of the French Society of Dermatology 1
Department Department 4 Department 5 Department 6 Department 7 Department 8 Department 9 Department 2
of of of of of of of of
Dermatology, H^opital Sud, and 3Department of Epidemiology and Public Health, H^opital Nord, University of Amiens, 80054 Amiens, France Dermatology, H^opital Tenon, Assistance Publique–H^opitaux de Paris (AP–HP), University of Paris X, 75020 Paris, France Vascular Rehabilitation, H^opital Corentin-Celton, AP–HP, 92133 Issy-les-Moulineaux CEDEX, France Dermatology and Gerontology, H^opital Rothschild, AP–HP, University of Paris VI, 75571 Paris, France Dermatology, H^opital Bon-Secours, 57038 Metz, France Dermatology, Centre Leon-Berard, 69373 Lyon CEDEX 08, France Dermatology, Centre Hospitalier de Valence, 26953 Valence, France Dermatology, H^opital Charles-Nicolle, University of Rouen, 76038 Rouen, France
Summary Correspondence Guillaume Chaby. E-mail: [email protected]
Accepted for publication 31 July 2013
Funding sources The study was funded by a grant from the French Society of Dermatology in 2003 and was accorded a research grant from Servier Medical, France, in 2003.
Conflicts of interest P.S. has received consulting fees from Servier and Sanofi, and fees from Urgo and Pierre Fabre as an investigator in clinical studies on venous ulcer dressings; G.C. has received fees from Urgo Medical as an investigator; F.T. has received fees from Urgo as an investigator and expert; P.C. has received fees from Urgo and Coloplast as an expert; C.L. has received consulting fees from Actelion, Servier, Coloplast, Urgo, Pierre Fabre and M€olnlycke Healthcare as an investigator in clinical studies on venous ulcer dressings. DOI 10.1111/bjd.12570
Background Some prognostic markers of venous leg ulcer (VLU) healing have been evaluated, mostly in retrospective studies. Objectives To identify which clinical characteristics, among those known as possible prognostic factors of VLU healing, and which VLU-associated sociodemographic and psychological factors, are associated with complete healing at week 24 (W24). Methods A prospective, multicentre, cohort study was conducted in 22 French dermatology departments between September 2003 and December 2007. The end point was comparison between healed and nonhealed VLUs at W24, for patient clinical and biological characteristics; psychological, cognitive and social assessments; affected leg inclusion characteristics; venous insufficiency treatment and percentage of initial wound area reduction during follow-up. Results In total, 104 VLUs in 104 patients were included; 94 were analysed. The mean VLU area and duration were 368 555 cm2 and 248 457 months, respectively. At W24, 41/94 VLUs were healed. Univariate analysis significantly associated complete healing with superficial venous surgery (P = 0001), adherence to compression therapy at W4 (P = 003) and W24 (P = 001), ankle-joint ankylosis (P = 001) and mean percentage of VLU area reduction at W4 (P = 004). Multivariate analysis retained superficial venous surgery during follow-up [odds ratio (OR) 84, 95% confidence interval (CI) 19–482] and percentage reduction of the VLU area at W4 (OR 16, 95% CI 10–214) as being independently associated with healing. Conclusions These results indicate that complete healing of long-standing, large VLUs is independently associated with ablation of the incompetent superficial vein and percentage of wound area reduction after the first 4 weeks of treatment.
What’s already known about this topic?
• •
1106
Despite advances in venous leg ulcer (VLU) management, the time to ulcer healing generally remains long. Some prognostic markers of healing have been evaluated, mostly in retrospective studies.
British Journal of Dermatology (2013) 169, pp1106–1113
© 2013 British Association of Dermatologists
Prognostic factors of venous leg ulcer healing, G. Chaby et al. 1107
What does this study add?
•
In this prospective, multicentre, cohort study, complete healing of long-standing, large VLUs at 6 months was independently associated with ablation of the incompetent superficial vein and percentage of wound area reduction after the first 4 weeks of treatment.
Venous leg ulcers (VLUs) represent a major complication of chronic venous insufficiency and a common source of morbidity, especially in those aged over 60 years. The prevalence of leg ulcers with a predominant venous cause has been estimated at 1–15% of the adult population.1 Despite advances in the management of these patients, the time to ulcer healing generally remains long: 30–70% of VLUs heal only within 6 months, and 10–20% require therapy lasting > 1 year.2,3 Several factors have been associated with patients with VLUs, compared with general populations – poorer socioeconomic status, lower educational level and social isolation and/or depression – but their impact on healing has not been investigated.4,5 In current practice, several factors are commonly thought to be associated with failure of VLUs to heal, but have not yet been documented. These include poor adherence to compression therapy, conditions that may decrease compliance with wound care management (e.g. impaired mobility), number of comorbidities, calf muscle-pump dysfunctions (e.g. ankle-joint ankylosis) and factors limiting tissue remodelling, such as protein deficiency. Some prognostic markers of healing have been evaluated, mostly in retrospective studies. Margolis et al.2 published a simple statistical prognostic model based on wound area (≥ 10 cm2) and duration (≥ 12 months), validated on a patient database, whereas Kantor and Margolis6 showed that complete ulcer healing at week 24 (W24) was associated not with initial ulcer size, but with the percentage of area reduction after 4 weeks of treatment. Therefore, our study was undertaken to identify prospectively which clinical VLU characteristics known to be possible prognostic markers, and which VLU-associated sociodemographic and psychological factors, are associated with complete healing at W24.
Patients and methods Design overview This prospective, multicentre, cohort study, conducted in 22 French centres from September 2003 to December 2007, included 24 weeks of follow-up and was performed in accordance with the Declaration of Helsinki. The Institutional Review Board of Picardie, Centre Hospitalier Universitaire d’Amiens, and regulatory authorities approved its protocol (NCT 01673412). All subjects received detailed information about the study and gave their written consent to participate. © 2013 British Association of Dermatologists
Setting and participants The study population consisted of ambulatory or hospitalized patients managed at or referred to each centre for VLU management. Patients were included with the following criteria: (i) > 50 years old; (ii) more than one VLU lasting ≥ 1 month and with wound area > 1 cm2; (iii) clinical findings consistent with established venous disease (skin hyperpigmentation, varicose veins, lipodermatosclerosis) confirmed by venous Doppler duplex ultrasonography performed during the preceding 6 months; (iv) absence of significant arterial insufficiency assessed by clinical findings (intermittent claudication or resting pain, necrotic or distal foot wound) and ankle brachial index (ABI) ≥ 08; and (v) ability to give informed consent and to be followed for 24 weeks. Exclusion criteria were (i) ongoing systemic diseases known to be associated with pyoderma gangrenosum or necrotizing vasculitis; (ii) corticosteroid, or cytotoxic or immunosuppressant drug use during the preceding 3 months; (iii) hypertensive leg ulcer; (iv) foot ulcer; and (v) ABI < 08. When several VLUs were present, the largest was chosen as the reference VLU (rVLU). Study procedures For all subjects satisfying the entry criteria, the following baseline information was recorded: (i) sociodemographic characteristics; (ii) medical and surgical history; (iii) concomitant illness(es); (iv) current medications; (v) mobility and autonomy status; (vi) psychological profile; (vii) serum albumin and haemoglobin concentrations; (viii) wound aetiology, Doppler duplex ultrasonography and ABI results; (ix) CEAP (clinical– etiology–anatomy–pathophysiology) classification; (x) wound features, including rVLU duration, location, size and related pain; and (xi) characteristics of the affected lower leg. All patients were prescribed standard therapy, including daily application of high-compression elastic bandages (30– 40 mmHg at the ankle), and dressings depending on the wound stage.7 At inclusion, patients received detailed information about compression therapy. The frequency of dressing changes depended on the wound stage and was left to the investigator’s judgement. Patients with isolated superficial reflux and/or perforating veins were referred during the study to a vascular surgeon for an opinion on surgical treatment of venous insufficiency. The same investigator saw patients monthly until W24, regardless of whether or not they had undergone surgery. British Journal of Dermatology (2013) 169, pp1106–1113
1108 Prognostic factors of venous leg ulcer healing, G. Chaby et al.
Outcomes and measurements The study judgement criterion was the comparison between healed and nonhealed VLUs at W24, with respect to the following characteristics: (i) general patient characteristics: age, sex, body mass index (BMI), presence of anaemia (haemoglobin < 12 g dL 1 in women, < 13 g dL 1 in men) and hypoalbuminaemia (albumin < 35 g L 1), level of autonomy and mobility, social status (educational level, marital status, occupation, health insurance coverage, income, living conditions); (ii) presence of the following comorbidities: hypertension, diabetes, cardiac insufficiency, rheumatological disease(s) involving the affected leg, history of orthopaedic surgery (hip or knee), prior deep-vein thrombosis, renal insufficiency, previous vein surgery; (iii) psychological profile; (iv) affected lower-leg characteristics: presence of white atrophy, lipodermatosclerosis or ankle-joint ankylosis; (v) rVLU characteristics: initial area, duration, recurrent nature, granulation tissue > 50%, evolution of ulcer-related pain during dressing change; (vi) venous Doppler duplex ultrasonography findings: presence of obstruction and/or reflux in the superficial- and/ or deep-venous system; (vii) rVLU management during follow-up: compression adherence, modalities of compression use at W4 and W24, superficial vein surgery; and (viii) percentage wound area reduction at W4 – ulcer areas were calculated by measuring tracings of planimetry images, according to the same standardized protocols. The total ulcer area of the affected leg was estimated by adding the area of each ulcer. Areas were determined at inclusion and every 4 weeks until W24. Change of ulcer-related pain was assessed with a pain visual analogue scale (range 0– 100). Autonomy and mobility were assessed with Katz’s basic Activities of Daily Living (ADL) score, which measures basic activities for a 1-month period prior to the interview; a score < 3 (out of 6) indicates major difficulties in performing normal self-care activities. Patients were evaluated for compression adherence and acceptance at W4 and W24. The patient was considered to be compression adherent when (i) compression bandages were applied < 1 h after getting up; (ii) they were kept in place for 80% of the time during the day and for > 5 days per week; (iii) they covered the dressing at each visit; and (iv) bandages were in good condition. Compression acceptance was evaluated at W4 and W24 on a four-point Likert scale: 1, easily; 2, with minor difficulties; 3, with difficulties; 4, unbearably. The patient was considered to have understood the compression benefit when they responded ‘no’ to the statements ‘the main interest of compression is to maintain the dressing’ and ‘the compression is also effective if it is worn only a few hours a day’; and ‘yes’ to ‘compression therapy is an essential treatment of venous ulcers’ and ‘compression therapy decreases venous stasis of the lower leg’. Psychological and cognitive assessments were performed at inclusion. Mini-Mental State Examination (MMSE; range 0– 30) and Raven’s Coloured Progressive Matrices (RCPM; range 0–36) were used to evaluate cognitive capacity and inductive British Journal of Dermatology (2013) 169, pp1106–1113
capacity, respectively. Depression and self-esteem were evaluated with the Beck Depression Inventory (BDI) and the Coopersmith Self-Esteem Inventory (CSEI), respectively. The BDI is a 21-question self-reporting inventory; each question has a set of at least four possible responses, measuring the severity of depression: 0–9, minimal depression; 10–18, mild depression; 19–29, moderate depression; 30–63, severe depression. The CSEI is a self-assessment score of self-esteem with four domains and 50 items: general (26 items), social (eight items), familial (eight items) and professional scales (eight items); a score ≤ 33/50 indicates very low selfesteem. Patients’ monthly incomes were self-reported, and classified using the median income in France as the reference (> or ≤ €1524 per month), with the poverty threshold set at 50% of the median.8 Statistical analyses Continuous variables (age, BMI, and VLU area and duration) are expressed as the mean SD (when normally distributed) or as medians; categorical variables (all others) are expressed as number (%). Possible independent predictors of healing were first identified by univariate analysis using Student’s t-test for continuous variables, or the v2 or Fisher exact test for qualitative variables. P < 005 defined significance. All statistical tests were two sided. Single variable logistic regression analyses assessed the magnitude of riskfactor effects, giving estimated odds ratios (ORs) with 95% confidence intervals (CIs). Multivariate models, including variables with P < 010 in univariate analyses, estimated the independence of prognosis-associated parameters. Reaching 80% power required inclusion of 140 patients, with a 65% expected cure rate. Database management and all statistical analyses were performed with SPSS software (IBM, Armonk, NY, U.S.A.).
Results Baseline characteristics Of the 104 patients with VLU, 10 did not complete the entire study and were excluded from the statistical analyses: one was lost-to-follow-up after cellulitis; five withdrew their consent because they considered the mandatory consultations and psychological tests too much for them to commit to, and four gave no specific reason (Fig. 1). The mean age at inclusion was 735 99 years (median 76) and most patients were female (71%). The mean number of VLUs on the affected leg was 16 10 (median 20), with no between-group difference; 51% of the patients had only one VLU at inclusion. The mean rVLU area and duration at inclusion were 368 555 cm2 and 248 457 months, respectively. Notably, 63% of rVLUs were located on the medial malleolus. The patient and VLU baseline characteristics are summarized in Table 1. © 2013 British Association of Dermatologists
Prognostic factors of venous leg ulcer healing, G. Chaby et al. 1109
104 VLUs in 104 patients Visit 1 History and physical examination Affected leg characteristics Wound area measurement Pain visual analogue scale Activities of Daily Living score Questionnaires: Socioeconomic level BDI CSEI MMSE RCPM Biology and Doppler duplex Detailed information on compression therapy Week 4 Ulcer area calculated on planimetry tracing Adherence to compression Compression acceptance Weeks 8, 12, 16, 20 Ulcer area calculated on planimetry tracing Week 24 Ulcer area calculated on planimetry tracing Adherence to compression Compression acceptance
Fig 1. Flow diagram of patients with venous leg ulcers (VLUs) throughout the study. BDI, Beck Depression Inventory; CSEI, Coopersmith Self-Esteem Inventory; MMSE, Mini-Mental State Examination; RCPM, Raven’s Coloured Progressive Matrices.
Follow-up and overall outcome Factors affecting complete healing at W24 were analysed for the 94 patients still being followed at the end of the study. Nine patients withdrew their participation and one was hospitalized. At W24, 41 rVLUs were completely healed and 53 were nonhealed. Follow-up compression therapy and surgical treatment findings are summarized in Table 1. Factors identified by univariate analysis as predicting venous leg ulcer healing or nonhealing Concerning baseline characteristics and comorbidities, anklejoint ankylosis was the only between-group difference found, being significantly more frequent in patients with nonhealed VLUs at 6 months (60% vs. 37%; P = 003). Comparing the healed and nonhealed VLU groups, respectively, complete rVLU healing was significantly associated with superficial venous system surgery during follow-up (27% vs. 4%, P = 0001), compression therapy adherence at W4 (87% vs. 76%, P = 003) and W24 (88% vs. 59%, P = 001), and mean percentage rVLU area reduction at W4 (48% vs. 27%, P = 004), with the difference becoming significant at the 30% threshold (P = 003). In addition, complete rVLU healing at W24 was associated with complete healing of all VLUs on the affected leg, for all 21 healed patients with multiple © 2013 British Association of Dermatologists
94 Patients completed follow-up 10 Patients excluded (lost to follow-up before week 4) 9 Patient decisions 1 Lost to follow-up
41 VLUs healed
53 VLUs not healed
lesions. Among the 49 patients eligible for venous surgery, 13 underwent surgery and 36 were treated with compression alone. Among the latter, 31% of VLUs were healed at W24 and 69% were not (P = 0007). No significant between-group differences were found for VLU area (P = 023), duration (P = 007) or the presence of deep-venous insufficiency (P = 079). Compression was applied less frequently by patients themselves or a home helper than by nurses or physicians for the healed vs. nonhealed group, but the difference reached significance only at W4 (29% vs. 49%, P < 001). Compression acceptance was considered easy by most healed and nonhealed patients with VLU (respectively, 56% vs. 58% at W4 and 72% vs. 57% at W24), with no significant between-group difference. For the healed and nonhealed groups, correct understanding of the benefit of compression therapy was reported, respectively, by 20% vs. 22% of the patients at W4 (P = 071), and by 25% vs. 29% at W24 (P = 067). Sociodemographic data were similar for the healed and nonhealed groups: a majority of patients had median or lower socioeconomic status, i.e. monthly income < €1524, and nearly half of the patients in both groups lived alone (Table 2). Meanwhile, significantly more patients with nonhealed VLUs received wound care at home, rather than office-based care, compared with those with healed VLUs (P = 001). MMSE, BDI, CSEI and RCPM scores were comparable for the two groups (Table 3). British Journal of Dermatology (2013) 169, pp1106–1113
1110 Prognostic factors of venous leg ulcer healing, G. Chaby et al. Table 1 Baseline and follow-up characteristics of patients according to venous leg ulcer (VLU) status Characteristic General Age (years), mean (SD) Women, n (%) Body mass index (kg m 2), mean (SD) Albuminaemia (albumin < 35 g L 1), n (%) Anaemia, n (%) ADL score (out of 6), mean (range) ADL score < 3, n Comorbidity Cardiovascular risk factors (HT, diabetes), n (%) Cardiac insufficiency, n (%) Rheumatological disease of affected leg, n (%) Orthopaedic surgery (hip or knee), n (%) Prior deep-vein thrombosis, n (%) Renal insufficiency, n (%) Prior venous surgery, n (%) VLUs Area (cm2), mean (SD) Duration (months), mean (SD) Recurrent, n (%) Granulation tissue > 50%, n (%) VAS score, mean (range) Venous duplex ultrasonography Isolated SVI, n (%) DVI, isolated or not, n (%) Popliteal vein reflux, n (%) Leg characteristic Lipodermatosclerosis, n (%) White atrophy, n (%) Ankle-joint ankylosis, n (%) Follow-up VLU area reduction at week 4 (%) ≤ 20%, n ≤ 30%, n ≤ 40%, n Adherence to compression Week 4 (%) Week 24 (%) Superficial vein surgery, n (%)
Healed (n = 41)
Nonhealed (n = 53)
OR (95% CI)
P-value
739 (101) 38 (72) 302 (61) 34 (64) 22 (42) 56 (2–6) 1
103 105 088 207 123 – –
(077–134) (039–283) (066–127) (083–521) (049–310)
089 091 062 010 057 084 087
(73) (12) (27) (44) (59) (15) (37)
33 (62) 7 (13) 17 (32) 23 (43) 29 (55) 4 (8) 25 (47)
060 110 129 098 086 048 155
(023–160) (028–389) (048–349) (045–242) (035–211) (010–328) (061–382)
032 081 054 094 079 046 043
289 (522) 161 (232) 32 (78) 20 (49) 46 (0–90)
427 (577) 315 (567) 45 (85) 16 (30) 48 (0–100)
214 288 158 045
(122–437) (135–623) (049–513) (018–115) –
023 007 040 028 071
22 (54) 19 (46) 14 (34)
25 (47) 28 (53) 21 (40)
077 (031–189) 130 (053–319) 127 (050–323)
053 056 054
32 (78) 22 (54) 15 (37)
47 (89) 33 (62) 32 (60)
220 (069–787) 142 (057–355) 264 (105–670)
034 045 003
48 11 12 12
27 22 27 31
289 194 251 341
(127–584) (074–514) (098–653) (132–893)
004 014 003 0005
87 88 11 (27)
76 59 2 (4)
045 (020–099) 022 (010–046) 011 (002–057)
003 001 0001
734 (97) 29 (71) 304 (73) 19 (46) 15 (37) 55 (3–6) 0 30 5 11 18 24 6 15
ADL, Activities of Daily Living; CI, confidence interval; DVI, deep-vein insufficiency; HT, hypertension; OR, odds ratio; SVI, superficial vein insufficiency; VAS, visual analogue scale for pain.
Independent prognostic factors of healed and nonhealed venous leg ulcers Our multivariate analyses retained two prognostic factors as being significantly associated with complete VLU healing at W24: surgery for superficial venous insufficiency during follow-up (OR 84, 95% CI 19–482) and percentage rVLU area reduction at W4 (OR 16, 95% CI 10–214).
Discussion The main objective of this prospective study was to identify clinical, social, psychological and cognitive factors associated with VLU healing after 24 weeks. Multivariate analyses British Journal of Dermatology (2013) 169, pp1106–1113
retained surgical treatment of superficial venous incompetence during follow-up and the healing rate at W4, assessed as the percentage VLU area change, as significant and independent factors predicting complete VLU healing at W24. Notably, complete rVLU healing was associated with complete healing of all lesions on the affected leg of all patients with multiple lesions at inclusion. The mean age, sex ratio and BMI were the same as those previously described as being associated with VLU.1,2,4 Herein, the high frequencies of patients without high-school diplomas and with monthly incomes below the French median suggest, as shown previously in other European countries, that VLU might also be associated with low social status in France.4,9 © 2013 British Association of Dermatologists
Prognostic factors of venous leg ulcer healing, G. Chaby et al. 1111 Table 2 Socioeconomic characteristics of patients according venous leg ulcer status Socioeconomic characteristic
Healed (N = 41), n (%)
Nonhealed (N = 53), n (%)
Home owner Rented home Free universal healthcarea Long-term illnessb Additional private health insurance Urban Rural Living alone Home wound carec Income (€ per month) < 762 762–1524 1524–2286 > 2286 Education level No diploma Less than high school High school and above Marital status Single Married/partnership Widowed Employment Working Not working On disability benefit Retired
25 (61) 14 (34) 2 (5)
26 (49) 23 (43) 2 (4)
28 (68) 37 (90)
42 (79) 48 (91)
32 (78) 9 (22) 21 (51) 28 (68)
41 12 34 48
12 (29) 20 (49) 9 (22) 0
12 (23) 24 (45) 13 (25) 4 (8)
17 (41) 22 (54) 2 (5)
24 (45) 26 (49) 3 (6)
5 (12) 19 (46) 17 (41)
4 (8) 29 (55) 20 (38)
1 (2) 2 (5) 2 (5) 36 (88)
2 (4) 4 (8) 4 (8) 43 (81)
(77) (23) (64) (91)
a Universal healthcare coverage (Couverture Maladie Universelle; CMU) is a free French social welfare programme for people on low incomes; the CMU offers 100% healthcare coverage that is added to the standard National Health Insurance. bLong-term illness (affection de longue duree) includes long-term or major disease for which all costs are fully reimbursed by the National Health Insurance. cBetween-group comparison is significant (P < 005).
Clinical leg characteristics indicated that our population had severe venous disease. Indeed, lipodermatosclerosis and white atrophy are signs of severe chronic venous disorders, and were observed in 84% and 59% of the study patients, respectively.10 VLUs had been present for a mean of 248 months, with a mean area of 368 cm2, and were recurrent for 82% of the patients, thereby meeting the literature definition of hardto-heal ulcers.2,11 Indeed, using the statistical prognostic model of Margolis et al.,2 based on ulcer size and duration, 77% (72/94) of the VLUs in our study had < 50% chance of healing at W24. The study setting in tertiary care centres might partially explain the high rate of hard-to-heal ulcers, further emphasizing that long-lasting and relapsing VLUs are resistant to standard care, and are more frequently referred to specialized facilities than ‘usual’ VLUs, which heal within a © 2013 British Association of Dermatologists
few months.2,12,13 Thus, our results might not be applicable to patients with milder disease treated in primary-care settings. A limitation of our study is that relatively fewer VLUs than expected were evaluated at the end of the study. Indeed, numerous patients declined participation given the length of the study. This limitation might have created a selection bias, leading to a lack of power to establish some differences, including VLU area or duration, which are usually associated with late healing. Logistic regression analysis yielded ORs > 1, i.e. close to those previously reported, for VLU area and duration, whereas Student’s t-test, which is more sensitive to the study power, showed no significant between-group differences for those parameters. Moreover, multivariate analysis did not retain these two variables as being independently associated with healing. According to well-validated psychometric instruments (BDI, CSEI), our results showed that almost half of the patients were depressed and that one-third had low self-esteem. These psychological problems might be explained by the chronic, relapsing course of the venous disease investigated in our study, but we are unable to provide any evidence supporting a relationship between these psychological factors and complete VLU healing at W24. Conflicting results have been published concerning the association of a history of deep-vein thrombosis or popliteal reflux with late healing.13–17 Some of those studies were criticized because of the systematic lack of vascular imaging or their retrospective design. The VLUs of our patients with residual deep-vein obstruction and deep-valve incompetence were not less significantly healed at W24. Our univariate analysis identified ankle-joint ankylosis, which increases venous hypertension, as prospectively predictive of failure to heal at W24, but it was not retained in the multivariate analysis, possibly because it is merely a marker of the severe disease that characterized our patients. Ankle-joint ankylosis had not been previously evaluated as a prognosis factor, making any comparisons difficult. Even though it is accepted that good adherence to compression therapy is associated with faster healing, few studies have examined compression compliance as a factor predictive of VLU healing.1,18 According to our univariate analyses, poor compression adherence was significantly associated with nonhealed VLUs. Because compression compliance was assessed by a questionnaire, we cannot exclude that it might have been overevaluated, with some responses corresponding to expected, socially desirable behaviour. Despite written information and oral instructions emphasizing the importance of compression therapy, lower adherence to compression therapy in the nonhealed group during follow-up might be explained by a loss of motivation reflecting impaired healing or lower compression efficacy in this group, whose participants or entourage more frequently applied the bandages, rather than nurses or physicians. Our multivariate analysis retained superficial vein surgery during follow-up as being associated with complete VLU healBritish Journal of Dermatology (2013) 169, pp1106–1113
1112 Prognostic factors of venous leg ulcer healing, G. Chaby et al.
Psychological test
Healed
Nonhealed
Beck Depression Inventory 0–11: no depression 12–19: light depression 20–27: moderate depression 28–63: severe depression Coopersmith Self-Esteem Inventory Score (out of 50), mean SD Score ≤ 33, low self-esteem Raven’s Coloured Progressive Matrices Mean SD (median) Mini-Mental State Examination Mean SD (range)
N = 40
N = 53
22 (55) 13 (32) 4 (10) 1 (2) N = 39 355 78 13 (33) N = 33
31 (58) 13 (25) 6 (11) 3 (6) N = 53 354 78 15 (28) N = 47
245 84 (27) N = 41 261 43 (16–30)
208 91 (21) N = 53 266 36 (16–30)
Table 3 Psychological characteristics of patients according to venous leg ulcer status
Values are n (%) unless stated otherwise. All between-group comparisons were nonsignificant (P ≥ 005).
ing at W24. In a recent, nonrandomized, prospective study, endovenous ablation of incompetent superficial and perforator veins was associated with a shorter time to healing and a higher complete healing rate, compared with medical treatment alone.19 However, this point is still being debated, because surgical treatment of superficial vein reflux was shown to lower the VLU recurrence rate but not to increase the complete healing rate in the prospective, randomized ESCHAR study.20 Finally, according to our multivariate analysis, the percentage VLU area change at W4 was predictive of complete healing at W24. The healing rate at W4 was shown to be a reliable criterion to predict VLU healing in a retrospective study on 306 VLUs by Cardinal et al.,21 and in a prospective study on 104 patients by Kantor and Margolis.6 In our study, a mean area reduction ≤ 30% at W4 was predictive of incomplete healing at W24, thereby confirming that the percentage wound area change at W4 can be used as a surrogate marker of healing in clinical trials to predict longer-term outcomes of healing technologies.22,23 Our results suggest that, for this cohort of long-standing, large VLUs, the prognosis of complete healing at W24 was independently associated with ablation of the incompetent superficial vein, which can rapidly lower venous hypertension, rather than with inherent VLU characteristics, e.g. wound area and duration, or with patient characteristics that are not easily modifiable, e.g. general mobility, protein deficiency, and psychological and socioeconomic factors. The healing rate at W4 under optimal treatment might predict complete healing at W24 and enable consideration of alternative therapies for VLUs not responding to this regimen.
Acknowledgments The authors thank the investigators from the Angio-Dermatology Group of the French Society of Dermatology: Henri Adamski, Jean-Marie Bonnetblanc, Jacqueline Chevrant-Breton, British Journal of Dermatology (2013) 169, pp1106–1113
Anne Dompmartin, Gerard Lorette, Herve Maillard, Philippe Modiano, Geraldine Perceau, Marie Aleth Richard-Lallemand, Barbara Roth, Pascal Toussaint and Marie-Francßoise VendeaudBusquet.
References 1 van Gent WB, Wilschut ED, Wittens C. Management of venous ulcer disease. BMJ 2010; 341:c6045. 2 Margolis DJ, Allen-Taylor L, Hoffstad O, Berlin JA. The accuracy of venous leg ulcer prognostic models in a wound care system. Wound Repair Regen 2004; 12:163–8. 3 O’Meara S, Tierney J, Cullum N et al. Four layer bandage compared with short stretch bandage for venous leg ulcers: systemic review and meta-analysis of randomized controlled trials with data from individual patients. BMJ 2009; 338:b1344. 4 Moffatt CJ, Franks PJ, Doherty DC et al. Sociodemographic factors in chronic leg ulceration. Br J Dermatol 2006; 155:307–12. 5 Scott TE, LaMorte WW, Gorin DR, Menzoian JO. Risk factor for chronic venous insufficiency: a dual case–control study. J Vasc Surg 1995; 22:622–8. 6 Kantor J, Margolis DJ. A multicentre study of percentage change in venous leg ulcer area as a prognostic index of healing at 24 weeks. Br J Dermatol 2000; 142:960–4. 7 Chaby G, Senet P, Vaneau M et al. Dressings for acute and chronic wounds: a systematic review. Arch Dermatol 2007; 143:1297–304. 8 Report by the National Observatory on Poverty and Social Exclusion. Available at: www.onpes.gouv.fr (last accessed 3 October 2013). 9 Herberger K, Rustenbach SJ, Grams L et al. Quality-of-care for leg ulcers in the metropolitan area of Hamburg – a community-based study. J Eur Acad Dermatol Venereol 2011; 26:495–502. 10 Ekl€ of B, Rutherford RB, Bergan JJ et al. Revision of the CEAP classification for chronic venous disorders: consensus statement. J Vasc Surg 2004; 40:1248–52. 11 Chaby G, Viseux V, Ramelet AA et al. Refractory venous leg ulcers: a study of risk factors. Dermatol Surg 2006; 32:512–19. 12 Margolis DJ, Berlin JA, Strom BL. Risk factors associated with the failure of a venous leg ulcer to heal. Arch Dermatol 1999; 135:920–6. 13 Franks P, Moffatt CJ, Connoly M et al. Factors associated with healing leg ulceration with high compression. Age Ageing 1995; 24:407–10. © 2013 British Association of Dermatologists
Prognostic factors of venous leg ulcer healing, G. Chaby et al. 1113 14 Skene AI, Smith JM, Dore CJ et al. Venous leg ulcers: a prognostic index to predict time to healing. BMJ 1992; 305:b1119–21. 15 Brittenden J, Bradbury AW, Allan PL et al. Popliteal vein reflux reduces the healing of chronic venous ulcer. Br J Surg 1998; 85:60–2. 16 Barwell JR, Ghauri ASK, Taylor M et al. Risk factors for healing and recurrence of chronic venous ulcers. Phlebology 2000; 15:49–52. 17 Moffatt CJ, Doherty DC, Smithdale R, Franks PJ. Clinical predictors of leg ulcer healing. Br J Dermatol 2010; 162:51–8. 18 O’Meara S, Cullum N, Nelson EA, Dumville JC. Compression for venous leg ulcers. Cochrane Database Syst Rev 2012; 11:CD000265. 19 Harlander-Locke M, Lawrence PF, Alktaifi A et al. The impact of ablation of incompetent superficial and perforator veins on ulcer healing rates. J Vasc Surg 2012; 55:458–64.
© 2013 British Association of Dermatologists
20 Barwell JR, Davies CE, Deacon J et al. Comparison of surgery and compression with compression alone in chronic venous ulceration (ESCHAR study): randomized controlled trial. Lancet 2004; 363:1854–9. 21 Cardinal M, Eisenbud DE, Phillips T, Harding K. Early healing rates and wound area measurements are reliable predictors of later complete wound closure. Wound Repair Regen 2008; 16:19–22. 22 Gelfand JM, Hoffstad O, Margolis DJ. Surrogate endpoints for the treatment of venous leg ulcers. J Invest Dermatol 2002; 119:1420–5. 23 Kurd SK, Hoffstad OJ, Bilker WB, Margolis DJ. Evaluation of the use of prognostic information for the care of individuals with venous leg ulcers or diabetic neuropathic foot ulcers. Wound Repair Regen 2009; 17:318–25.
British Journal of Dermatology (2013) 169, pp1106–1113
