Hrev_master Veins and Lymphatics 2013; volume 2:e15 [Veins and Lymphatics 2013; 2:e15] [page 49] An international survey on the interpretation of pigmentation using the C class of the Clinical, Etiological, Anatomical, Pathophysiological Classification Christopher R. Lattimer, Evi Kalodiki, Mustapha Azzam, George Geroulakos Josef Pflug Vascular Unit, Ealing Hospital & Imperial College, London, UK Abstract Skin changes over the gaiter area like pigmen- tation, lipodermatosclerosis and eczema are a clinical sign of advanced chronic venous disorder. This is documented as C4 in the Clinical, Etiological, Anatomical, Pathophysio logical (CEAP) classification. The hypothesis was that there is great variability whether skin changes are recorded as early or advanced disease. The aim was to evaluate different patterns of skin changes by delegates at 3 international venous conferences. Seven high-definition, A4-sized, color photographs were taken of untreated legs with skin changes from patients attending a pub- lic hospital venous clinic. They all had venous dis- ease confirmed on duplex with deep or superficial vein reflux >0.5 s. The photographs were dis- played and a questionnaire was provided. Delegates familiar with CEAP were asked to choose from 3 C class options for each photo- graph. The responses were summarized by group- ing them into mild (C0-3) and severe (C4-6). A total of 117 delegates completed the questionnaire from 30 countries. A percentage of 60 had prac- ticed phlebology >10 years. The percentages of responders scoring mild (C0-3) and severe disease (C4-6) were: mild/severe=3/96 (photo 1), 65/33 (photo 2), 31/67 (photo 3), 56/34 (photo 4), 74/21 (photo 5), 89/10 (photo 6) and 37/59 (photo 7). The median percentage measure of agreement was 36.8 [95% confidence interval (CI): 24.8- 48.9]. The range was 23.2 (95% CI: 10.5-36.0) to 94.8 (95% CI: 90.7-98.9), P<0.001/image, Fisher exact test). This indicates a significant difference of opinion between the appearances of mild and severe clinical disease. Clinical decisions using the C class as a sign of advanced disease may be unreliable if used alone for recording severity, grouping patients or rationing treatment. Introduction The Clinical, Etiological, Anatomical, Pathophysiological (CEAP) classification for chronic venous disorders (CVD) was set up fol- lowing an international ad hoc committee.1,2 It was revised in 2004 when it was stated that revision of CEAP is an ongoing process and that recommendations for change in the CEAP standard be supported by solid research.3 In 2007, a Joint Statement of the American Venous Forum and the Society of Interventional Radiology reported that this revision should be included as a baseline patient characteristic prior to endovenous treatments.4 This would facilitate comparison between the results of different studies and improve the overall quality of research on venous disease. The C component is the most widely used part of the CEAP classification and it is based solely on clinical appearance. Skin changes secondary to CVD are classified as C4 with C4a representing pigmentation or eczema and C4b representing lipodermatosclerosis (LDS) or atrophie blanche. In 2004, a questionnaire study without photographs was sent to 3681 phlebologists around the world. They conclud- ed from 206 responders that future work would be required on discriminating between C1 and C2 varicose vein sizes and which C (C1-3) to assign for corona phlebectatica.5,6 Discrepan - cies on the C placement of varicose veins of differing sizes were also apparent in an inter- observer reproducibility study between 3 clini- cians on 54 limbs.7 In that study there was dis- agreement between C0 and C2 or C1 and C2 in 13 legs (24%) but only in 6 legs (11%) between C2 and C4 or C3 and C4. The above studies5-7 suggest that further clarification should focus on the definitions of C0-3 rather than the later stages. Furthermore, it has been commented in an international consensus from experts that patients with more advanced venous dis- ease were most reliably classified.8 The hypothesis in this study is that pigmen- tation is also difficult to classify and that any discrepancy could have greater clinical signifi- cance if it spans several points across the 8- point C scale. This is because wide variations in the interpretation of skin changes could inappropriately place a leg into a mild (C0-3) or severe (C4-6) category. The aim was to investi- gate the definition of C4 by asking dedicated phlebologists to use their judgment and expe- rience to classify 7 color photographs of skin changes using the C of CEAP. Materials and Methods Study design This was an international observational study involving experienced phlebologists familiar with the CEAP classification. The majority of delegates were vascular surgeons/angiologists with an interest in phle- bology. There were no responders whose main practice was dermatology. They were asked to complete a short questionnaire in order to grade 7 photographs on the C of CEAP. These photographs were displayed on a table, placed at strategic locations, throughout the duration of 3 international venous conferences: the Royal Society of Medicine venous forum (RSM- VF), the European Venous Forum (EVF) and the venous section of the World Congress of the International Union of Angiology (IUA). Participants were either self-selected when they took an interest in the display table or when known specialists, prominent in the venous world, were invited specifically to com- plete the questionnaire. Patients All 7 photographs were from patients with Correspondence: Christopher R. Lattimer, Josef Pflug Vascular Laboratory, 7th Floor, Ealing Hospital, Uxbridge Road, Southall, Middlesex, UB1 3HW, United Kingdom. Tel. +44.7960.502253 - Fax: +44.20.72627681. E-mail: c.lattimer09@imperial.ac.uk Key words: CEAP classification, chronic venous disorder, hyperpigmentation, skin changes, vari- cose veins. Acknowledgements: we are grateful to the dele- gates, presenters and faculty at the RSM venous forum, the EVF and the IUA world congress who classified the photographs for this study. We also acknowledge Joseph Eliahoo, Statistical Consultant, Statistical Advisory Service, Imperial College, SW7 2AZ, for his help in the analysis of the data. Contributions: CL, conception and design, article drafting; CL, EK, MA, collection of data; CL, JE, statistical analysis; CL, EK, MA, GG, analysis and interpretation of data, manuscript final approval; EK, MA, GG, critical revision; GG, overall respon- sibility. Conflict of interests: the authors declare no potential conflict of interests. Conference presentation: oral presentation at the XXII Meeting of the Mediterranean League of Angiology and Vascular Surgery (MLAVS), Civitavecchia, Italy, October 11-13, 2012. Received for publication: 16 November 2012. Revision received: 21 February 2013. Accepted for publication: 10 June 2013. This work is licensed under a Creative Commons Attribution 3.0 License (by-nc 3.0). ©Copyright C.R. Lattimer et al., 2013 Licensee PAGEPress, Italy Veins and Lymphatics 2013; 2:e15 doi:10.4081/vl.2013.e15 No n- co mm er cia l assign for corona phlebectatica. No n- co mm er cia l assign for corona phlebectatica. cies on the C placement of varicose veins of No n- co mm er cia l cies on the C placement of varicose veins ofdiffering sizes were also apparent in an inter- No n- co mm er cia l differing sizes were also apparent in an inter- observer reproducibility study between 3 clini- No n- co mm er cia l observer reproducibility study between 3 clini- cians on 54 limbs. No n- co mm er cia l cians on 54 limbs.7 No n- co mm er cia l 7 In that study there was dis- No n- co mm er cia l In that study there was dis- agreement between C No n- co mm er cia l agreement between C0 No n- co mm er cia l 0 and C No n- co mm er cia l and C 13 legs (24%) but only in 6 legs (11%) between No n- co mm er cia l 13 legs (24%) but only in 6 legs (11%) between and C No n- co mm er cia l and C No n- co mm er cia l graph. The responses were summarized by group- No n- co mm er cia l graph. The responses were summarized by group- ). A total No n- co mm er cia l ). A total of 117 delegates completed the questionnaire No n- co mm er cia l of 117 delegates completed the questionnaire from 30 countries. A percentage of 60 had prac- No n- co mm er cia l from 30 countries. A percentage of 60 had prac- ticed phlebology >10 years. The percentages of No n- co mm er cia l ticed phlebology >10 years. The percentages of ) and severe disease No n- co mm er cia l ) and severe disease ) were: mild/severe=3/96 (photo 1), 65/33No n- co mm er cia l ) were: mild/severe=3/96 (photo 1), 65/33 (photo 2), 31/67 (photo 3), 56/34 (photo 4), 74/21No n- co mm er cia l (photo 2), 31/67 (photo 3), 56/34 (photo 4), 74/21 4 No n- co mm er cia l 4 or C No n- co mm er cia l or C3 No n- co mm er cia l 3 and C No n- co mm er cia l and C suggest that further clarification should focus No n- co mm er cia l suggest that further clarification should focus on the definitions of C No n- co mm er cia l on the definitions of C stages. Furthermore, it has been commented No n- co mm er cia l stages. Furthermore, it has been commented in an international consensus from experts No n- co mm er cia l in an international consensus from experts that patients with more advanced venous dis- No n- co mm er cia l that patients with more advanced venous dis- us e ed from 206 responders that future work would us e ed from 206 responders that future work would be required on discriminating between C us e be required on discriminating between C1 us e 1 and us e andvaricose vein sizes and which C (C us e varicose vein sizes and which C (C1-3 us e 1-3) to us e ) to 5,6 us e 5,6 Discrepan -us e Discrepan - cies on the C placement of varicose veins ofus e cies on the C placement of varicose veins of on ly study without photographs was sent to 3681 on ly study without photographs was sent to 3681 phlebologists around the world. They conclud- on ly phlebologists around the world. They conclud- on ly Contributions: CL, conception and design, article on ly Contributions: CL, conception and design, article drafting; CL, EK, MA, collection of data; CL, JE, on ly drafting; CL, EK, MA, collection of data; CL, JE, statistical analysis; CL, EK, MA, GG, analysis and on lystatistical analysis; CL, EK, MA, GG, analysis andinterpretation of data, manuscript final approval; on lyinterpretation of data, manuscript final approval; EK, MA, GG, critical revision; GG, overall respon- on ly EK, MA, GG, critical revision; GG, overall respon- sibility.on ly sibility. Conflict of interests: the authors declare no on ly Conflict of interests: the authors declare no Article [page 50] [Veins and Lymphatics 2013; 2:e15] leg symptoms who were attending the varicose vein clinic at a single public hospital. Patients were referred in the later stages of their dis- ease because uncomplicated varicose veins do not fill the referral criteria under the current public health rationing system. Five patients had duplex evidence of reflux in the saphenous trunks greater than 0.5 s9 without evidence of deep venous reflux. Two photographs were taken from the same patient (leg elevated and dependant) who had a previous calf vein thrombosis with significant tibial vein reflux. None of the patients had a healed venous ulcer or had received endovenous intervention for their condition. Photographs The photographs were high-resolution, taken at close range and printed in color on high-quality photographic A4 print-paper (Figures 1-7). Dissimilar backgrounds were chosen to avoid direct comparisons between the pictures. Each photograph was cropped to highlight the gaiter and ankle areas. The first photograph of lipodermatosclerosis was used as a quality control to determine the standard of the participants at classifying skin changes and to familiarize them with the task. The remaining 6 were selected to highlight contro- versial areas in the C-class. Although they were considered to be representative of venous disease, it is for the reader to decide on the prevalence of these skin changes in their day- to-day practice. All 7 photographs were dis- played simultaneously on a table which enabled each participant to pick them up and view each from different angles in order to make their judgment. Conferring was not for- bidden and occurred occasionally. A copy of the clinical classification of the revision of CEAP summary was also placed on the table for infor- mation purposes to remind each person of its precise wording as shown below.3 Clinical classification (C class) According to CEAP classification, C class is divided as follows: - C0: no visible or palpable signs of venous disease; - C1: telangiectasies or reticular veins; - C2: varicose veins; - C3: edema; - C4a: pigmentation or eczema; - C4b: lipodermatosclerosis or atrophie blanche; - C5: healed venous ulcer; - C6: active venous ulcer. Questionnaire design The questionnaire occupied half a page of A4 paper from which participants were asked to check 9 boxes and state their country of practice (Figure 8). This restriction on 10 questions/replies was purposeful to prevent lapses of concentration or questionnaire fatigue thereby ensuring data quality with maximal completion.10 Participants were given 3 C-class options for each photograph from which to pick their answer. They also had the option of stating an alternative C-class option or checking the don’t know box for whatever reason, including poor quality of photographs. The question on whether a clinical history would influence their judgement was included because it is uncertain whether clinicians should rely on clinical appearance alone or use supplementary information in deciding the C of CEAP. This may be relevant because pigmen- tation could be the result of treatment. Discoloration could also be caused by exten- sive telangectasiae following a deep venous thrombosis (DVT). The clinical history in both of these situations may encourage an enthusi- astic C class score of C4a rather than C0 or C1. Data analysis Data were transferred manually from the questionnaires onto spreadsheets at the end of the study and then imported into the IBM® SPSS® statistics software version 19 (IBM Corp., Armonk, NY, USA) for statistical analy- sis. The results on the C-class determination from the 7 pictures were reported in a similar way to how the C of CEAP is used to stratify patients in clinical trials: percentages in each C-class and/or stratification into mild and severe venous disease. Similarly, the results were reported in two ways. Firstly, specifically, as the percentage of responders choosing each C class (frequency distribution). Secondly, generally, as the percentage choosing mild (C0- 3) or severe (C4-6) venous disease (binary out- come). The percentage agreement between mild versus severe disease was determined using the risk difference value of the Fisher exact test. Full agreement, where all the raters scored either mild or severe, would be repre- sented as 100%, whereas equivalence would be represented as 0% agreement. Results Participant characteristics A total of 117 delegates completed the ques- tionnaire out of the 120 that were returned. Three were excluded because the answers to the 7 picture questions were incomplete. It was interesting that 2 responders ticked multiple boxes for each question in line with the recom- mendations of the advanced CEAP. In this case the single highest descriptor was used for the clinical classification.3 Delegates of 30 differ- ent nationalities completed this questionnaire, the top 5 being: UK (17), Italy (16), USA (10), Figure 1. The control image depicting lipo- dermatosclerosis. Survey result: C4 (16%), C4a (39%), C4b (42%). Figure 2. Corona phlebectatica paraplan- taris with an ankle flare. Survey result: C1 (16%), C2 (48%), C4a (28%). Figure 3. Pigmentation over extensive vari- cose veins. Survey result: C1 (0%), C2 (29%), C4a (66%). No n- co mm er cia l summary was also placed on the table for infor- No n- co mm er cia l summary was also placed on the table for infor- mation purposes to remind each person of its No n- co mm er cia l mation purposes to remind each person of its Clinical classification (C class)No n- co mm er cia l Clinical classification (C class) According to CEAP classification, C class isNo n- co mm er cia l According to CEAP classification, C class is Corp., Armonk, NY, USA) for statistical analy- No n- co mm er cia l Corp., Armonk, NY, USA) for statistical analy- sis. The results on the C-class determination No n- co mm er cia l sis. The results on the C-class determinationfrom the 7 pictures were reported in a similar No n- co mm er cia l from the 7 pictures were reported in a similar way to how the C of CEAP is used to stratify No n- co mm er cia l way to how the C of CEAP is used to stratify patients in clinical trials: percentages in each No n- co mm er cia l patients in clinical trials: percentages in each C-class and/or stratification into mild and No n- co mm er cia l C-class and/or stratification into mild and severe venous disease. Similarly, the results No n- co mm er cia l severe venous disease. Similarly, the results were reported in two ways. Firstly, specifically, No n- co mm er cia l were reported in two ways. Firstly, specifically, as the percentage of responders choosing each No n- co mm er cia l as the percentage of responders choosing each C class (frequency distribution). Secondly, No n- co mm er cia l C class (frequency distribution). Secondly, generally, as the percentage choosing mild (C No n- co mm er cia l generally, as the percentage choosing mild (C 3 No n- co mm er cia l 3) or severe (C No n- co mm er cia l ) or severe (C come). The percentage agreement between No n- co mm er cia l come). The percentage agreement between mild No n- co mm er cia l mild using the risk difference value of the Fisher No n- co mm er cia l using the risk difference value of the Fisher us e questionnaires onto spreadsheets at the end of us e questionnaires onto spreadsheets at the end ofthe study and then imported into the IBM us e the study and then imported into the IBM® us e ®statistics software version 19 (IBM us e statistics software version 19 (IBM Corp., Armonk, NY, USA) for statistical analy-us e Corp., Armonk, NY, USA) for statistical analy- sis. The results on the C-class determinationus e sis. The results on the C-class determinationus e o nlyon lyFigure 1. The control image depicting lipo- on lyFigure 1. The control image depicting lipo-dermatosclerosis. Survey result: C on lydermatosclerosis. Survey result: C on ly C on ly C4a on ly 4a (39%), Con ly (39%), C4bon ly 4bon ly on ly Article [Veins and Lymphatics 2013; 2:e15] [page 51] Czech Republic (8) and France (7). This distri- bution reflected the location of the confer- ences: London (RSM-VF: 20/117 responders, 17%), Florence (EVF: 63/117 responders, 54%) and Prague (IUA: 34/117 responders, 29%). The experience of the delegate was deter- mined by their number of years in phlebology practice which were: less than 2 (6%), between 2 and 10 (31%), between 10 and 30 (43%), over 30 (17%) and failure to answer (3%). Of the 27 original members of the ad hoc committee on the revision of the CEAP classi- fication,3 12 (44%) were recognized and invit- ed in person into the study. Nine members completed the questionnaire and 3 were unable to take part for whatever reason. A fur- ther index of experience was provided by the answers to the quality control picture 1 depict- ing lipodermatosclerosis. A total of 96% of del- egates recognized this correctly as C4/C4a/C4b, with 3% as edema (C3) and 1% checking don’t know for whatever reason. C class response stratification The percentage of participants checking each of the three given choices is displayed in the legends underneath each picture for con- venience (Figures 1-7). Pictures 4, 5 and 7 caused the greatest amount of uncertainty with the percentage of participants checking the don’t know box at 10%, 5% and 4%, respectively. The full spectrum is illustrated in Table 1. Mild and severe response stratification The percentages of participants scoring mild (C0-3) and severe disease (C4-6) for each photograph from 1 to 7 were: mild/severe=3/96 (photo 1), 65/33 (photo 2), 31/67 (photo 3), 56/34 (photo 4), 74/21 (photo 5), 89/10 (photo 6) and 37/59 (photo 7), respectively. Apart from the control picture 1 which was classed as severe according to 96% of participants, there was a significant lack of agreement between mild and severe clinical disease for the remaining legs. The percentage agreement (risk difference) of mild versus severe disease is displayed in the last column of Table 1. There was clinical uncertainty in classifying picture 2 (Figure 2) with 32% of participants choosing advanced disease. However, when the same leg was elevated in picture 6 (Figure 6), this was reduced to 10%. Importance of a clinical history In response to the influence a clinical histo- ry had to judgment of C class (Figure 8) the participants chose YES (67%), NO (26%), don’t know (3%) with 4% leaving this question unanswered. The fact that two-thirds of partic- ipants stated that they would use a history was surprising because the C of CEAP was designed to be judged solely from clinical appearance. Discussion The CEAP classification remains the gold standard classification of CVD. This was con- firmed in a recent review article at which they conducted a Medline analysis retrieving 266 publications using CEAP.11 This review also discussed the limitations of CEAP firstly stat- ing that CEAP was not a severity classification and then pointed out the controversial areas as C0-3. The current study has demonstrated that the controversial areas can extend into the higher categories of CEAP. The reality is that C of CEAP is frequently used to group patients into categories and is also used to discriminate patients with mild and severe disease. The C4- 6 group has been discussed as an individual disease.12 The C stratification has been used in epidemiological studies,13 longitudinal stud- ies14 and as a comparator against symptoms and signs,15 quality of life questionnaires16 and hemodynamic assessments.17 Many clinical papers stratify patients’ legs into mild/severe or uncomplicated/complicated based on this division between C0-3 and C4-6.18-24 This stratifi- cation is also used for rationing treatment in most public hospitals and in cost calcula- tions.25,26 The CEAP and venous clinical severity score (VCSS) are different tools and do not measure the same items equally. The existence of simi- lar items with different definitions revised or otherwise should be clear in the mind of the assessor to avoid substitution error. For exam- ple, C2 uses a definition of >3 mm for a vari- cose vein whereas the VCSS uses a cut-off point of 4 mm.27 Furthermore, pigmentation defined by the VCSS is more strict than the CEAP because focal pigmentation over varicose veins does not qualify, and a focal low intensi- ty (tan) is not considered by the VCSS as indicative of significant skin pigmentation.27 Eczema is C4a but not a VCSS attribute unless it is synonymous with inflammation. The current research demonstrates that there are substantial discrepancies in the clin- ical classification of CVD using the C of CEAP and the distinction between mild and severe venous disease is also unclear. Each photo- graph is commented upon below in order to focus on the controversial areas. Picture 1: This is the control photograph which was correctly identified as C4/C4a/C4b by 96% of participants. A plaque of LDS is seen in the gaiter region with deeply situated varicose veins above this area. Although LDS is con- firmed by palpation, this was not possible using photographs, a fact probably realized by 16% of participants who decided on choosing C4 alone. Nevertheless, the highest percentage score was C4b (42%) indicative of LDS. Pictures 2 and 6: This is the same leg Figure 4. Mild pigmentation with eczema at the gaiter region. Survey result: C0 (32%), C2 (21%), C4a (34%). Figure 5. Mild retro-malleolar pigmenta- tion over a normal vein. Survey result: C0 (50%), C2 (16%), C4a (21%). Figure 6. The same as shown in Figure 2 but the leg is now elevated. Survey result: C1 (86%), C4a (8%), C5 (2%). No n- co mm er cia l (photo 1), 65/33 (photo 2), 31/67 (photo 3), No n- co mm er cia l (photo 1), 65/33 (photo 2), 31/67 (photo 3), 56/34 (photo 4), 74/21 (photo 5), 89/10 (photo No n- co mm er cia l 56/34 (photo 4), 74/21 (photo 5), 89/10 (photo 6) and 37/59 (photo 7), respectively. Apart from No n- co mm er cia l 6) and 37/59 (photo 7), respectively. Apart from the control picture 1 which was classed as No n- co mm er cia l the control picture 1 which was classed as severe according to 96% of participants, there No n- co mm er cia l severe according to 96% of participants, there was a significant lack of agreement between No n- co mm er cia l was a significant lack of agreement between mild and severe clinical disease for theNo n- co mm er cia l mild and severe clinical disease for the remaining legs. The percentage agreementNo n- co mm er cia l remaining legs. The percentage agreement most public hospitals and in cost calcula- No n- co mm er cia l most public hospitals and in cost calcula-The CEAP and venous clinical severity score No n- co mm er cia l The CEAP and venous clinical severity score (VCSS) are different tools and do not measure No n- co mm er cia l (VCSS) are different tools and do not measure the same items equally. The existence of simi- No n- co mm er cia l the same items equally. The existence of simi- lar items with different definitions revised or No n- co mm er cia l lar items with different definitions revised or otherwise should be clear in the mind of the No n- co mm er cia l otherwise should be clear in the mind of the assessor to avoid substitution error. For exam- No n- co mm er cia l assessor to avoid substitution error. For exam- ple, C No n- co mm er cia l ple, C2 No n- co mm er cia l 2 uses a definition of >3 mm for a vari- No n- co mm er cia l uses a definition of >3 mm for a vari- cose vein whereas the VCSS uses a cut-off No n- co mm er cia l cose vein whereas the VCSS uses a cut-off point of 4 mm. No n- co mm er cia l point of 4 mm. defined by the VCSS is more strict than the No n- co mm er cia l defined by the VCSS is more strict than the CEAP because No n- co mm er cia l CEAP because us e papers stratify patients’ legs into mild/severe us e papers stratify patients’ legs into mild/severe or uncomplicated/complicated based on this us e or uncomplicated/complicated based on thisThis stratifi- us e This stratifi- cation is also used for rationing treatment inus e cation is also used for rationing treatment in most public hospitals and in cost calcula-us e most public hospitals and in cost calcula- on ly Many clinical on ly Many clinical on lyFigure 4. Mild pigmentation with eczema on lyFigure 4. Mild pigmentation with eczemaat the gaiter region. Survey result: C on lyat the gaiter region. Survey result: C on ly (32%), C on ly (32%), C2on ly 2on ly on ly Article [page 52] [Veins and Lymphatics 2013; 2:e15] dependent (Figure 2) and elevated (Figure 6) in a patient who has deep venous reflux in the calf veins following a DVT. The dependent ankle is discolored with a venous flare/corona phlebectatica and small varicose veins but on elevation pigmentation is not present which confirms the C1-2 status of this leg. The dependent leg was reported as C4a by 28% of participants but this reduced to 8% after eleva- tion. Elevation was used here to discriminate apparent pigmentation from venous conges- tion against true pigmentation from extravasa- tion. Although corona phlebectatica is current- ly C1 there are recommendations by many phle- bologists to consider it as C3.6 The lack of a marker scale prohibits the sizing of varicose veins. This may be necessary for establishing if they are >3 mm in diameter, in which case they would belong to C2. However, skin thick- ness and depth of vein should also be taken into consideration. For example it would be dif- ficult to compare vein size in Figure 1 with those in Figure 2. Picture 3: Obvious mid-calf pigmentation over extensive varicose veins was reported by only 66% of participants using C4a in compari- son to the 29% who reported C2. This may be because pigmentation in CEAP is defined as brownish darkening of the skin and occurs in the ankle region but may extend to leg and foot.3 Should pigmentation arising de novo over a mid-calf varicose vein without having extended from the ankle be classed as C4a? These factors may explain the reluctance of choosing C4a as an option. Pictures 4 and 5: These legs demonstrate mild (Figure 4) and very mild (Figure 5) degrees of pigmentation with eczema. This is reflected in the percentage of participants choosing C0 at 32% and 50%, respectively. Once again, it has been left for the observer to decide on what constitutes brownish darken- ing and how much eczema is significant to qualify as C4a. Both patients also had a normal looking vein present beneath the medial malle- olus which may have prompted the choice of C2 at 21% and 16%, respectively. Picture 7: Many patients have different baseline skin colors which may cause addition- al problems in defining increased pigmenta- tion rather than basing a judgment on its pres- ence or absence. This has been highlighted in this picture of gaiter telangectasiae where 56% of participants classified the accompanying hyperpigmentation as C4a. In cases of doubt there are two additional ways to evaluate pigmentation. The first is to make a comparison with the normal other side because this will indicate the natural color of the skin. The second is to elevate the leg to confirm a real pigmentation that does not dis- appear. A published limitation of CEAP is that the patient’s venous history is not taken into account.28 The C class according to the current definition is about clinical appearance, not medical history. It is clear from the current study that 67% of participants would use a clin- ical history in making their judgment on C. This supports the rationale of this study because the C of CEAP is nothing more than clinical appearance and was not intended to co-ordinate historical features. However, in everyday practice the judgment of a C class cannot be devoid of a clinical history. Pigmentation in the gaiter region may have other etiological factors like post-inflammatory hyperpigmentation occurring after trauma or a skin infection. Similarly, eczema may be caused by an allergic reaction or an insect bite rather than venous insufficiency. A C2 classifi- cation assessed by a doctor in the morning may become a C3 in the evening. These may only become apparent from the clinical history. Furthermore, the presence of minor reflux within a small caliber vein may not be enough to cause pigmentation. It would be interesting if a group of dermatologists were asked to com- plete the survey since they have a focused interest in pigmentation disorders. The pictures represent legs seen in common practice, which are difficult to classify because of lack of agreement using the C class. Improvements in C stratification could be Figure 7. Telangectasiae and reticular veins with infra-malleolar pigmentation. Survey result: C1 (21%), C2 (15%), C4a (56%). Figure 8. The questionnaire which dele- gates were asked to complete. Picture 1 is Figure 1 and likewise for the remaining 6 illustrations. Table 1. Percentage of C classes chosen by 117 participants for each picture. The 3 given choices are highlighted in italics. Risk differ- ence represents agreement, from no agreement (0%) to full agreement (100%). Picture C0 C1 C2 C3 C4 C4a C4b C5 Don’t know Mild/severe* P value° Risk difference % (CI: 95%) 1 - - - 2 16 39 42 - 1 3/113 <0.0005 94.8 (90.7-98.9) 2 - 16 48 1 - 28 2 2 3 76/38 <0.0005 33.3 (21.1-45.6) 3 - 0 29 1 - 66 1 - 3 36/78 <0.0005 36.8 (24.8-48.9) 4 32 - 21 3 - 34 - - 10 65/40 0.0009 23.8 (10.7-36.9) 5 50 4 16 4 - 21 - - 5 87/24 <0.0005 56.8 (45.9-67.6) 6 - 86 1 1 - 8 - 2 1 104/12 <0.0005 79.3 (71.5-87.2) 7 - 21 15 1 - 56 3 - 4 43/69 0.0008 23.2 (10.5-36.0) *Ratio in absolute numbers of C0-3/C4-5; °Fisher exact test. CI, confidence interval. No n- co mm er cia l marker scale prohibits the sizing of varicose No n- co mm er cia l marker scale prohibits the sizing of varicoseveins. This may be necessary for establishing No n- co mm er cia l veins. This may be necessary for establishing if they are >3 mm in diameter, in which case No n- co mm er cia l if they are >3 mm in diameter, in which case they would belong to C No n- co mm er cia l they would belong to C2 No n- co mm er cia l 2. However, skin thick- No n- co mm er cia l . However, skin thick- ness and depth of vein should also be taken No n- co mm er cia l ness and depth of vein should also be taken into consideration. For example it would be dif- No n- co mm er cia l into consideration. For example it would be dif- ficult to compare vein size in Figure 1 with No n- co mm er cia l ficult to compare vein size in Figure 1 with those in Figure 2. No n- co mm er cia l those in Figure 2. Picture 3: No n- co mm er cia l Picture 3: Obvious mid-calf pigmentation No n- co mm er cia l Obvious mid-calf pigmentation over extensive varicose veins was reported by No n- co mm er cia l over extensive varicose veins was reported by only 66% of participants using C No n- co mm er cia l only 66% of participants using C son to the 29% who reported C No n- co mm er cia l son to the 29% who reported C because pigmentation in CEAP is defined as No n- co mm er cia l because pigmentation in CEAP is defined as brownish darkening of the skin No n- co mm er cia l brownish darkening of the skin No n- co mm er cia l Figure 7. Telangectasiae and reticular veins No n- co mm er cia l Figure 7. Telangectasiae and reticular veins with infra-malleolar pigmentation. Survey No n- co mm er cia l with infra-malleolar pigmentation. Survey No n- co mm er cia l (56%). No n- co mm er cia l (56%). us e The lack of aus e The lack of a marker scale prohibits the sizing of varicoseus e marker scale prohibits the sizing of varicoseus e o nly 36/78 <0.0005 36.8 (24.8-48.9) on ly 36/78 <0.0005 36.8 (24.8-48.9) on ly on ly10 65/40 0.0009 23.8 (10.7-36.9) on ly10 65/40 0.0009 23.8 (10.7-36.9)87/24 <0.0005 56.8 (45.9-67.6) on ly87/24 <0.0005 56.8 (45.9-67.6) on ly on ly 1 104/12 <0.0005 79.3 (71.5-87.2)on ly 1 104/12 <0.0005 79.3 (71.5-87.2)on ly on ly 43/69 0.0008 23.2 (10.5-36.0)on ly 43/69 0.0008 23.2 (10.5-36.0) Article [Veins and Lymphatics 2013; 2:e15] [page 53] made by defining the controversial areas as this work has demonstrated. A consensus statement from a panel of experts using pub- lished photographs may improve the reliability and agreement of CEAP. Limitations This is an observational study where patients with an indeterminate C class were selected deliberately because they would invoke disagreement between different partic- ipants. However, the objective of this study was to highlight controversial areas because recog- nition of a limitation is an essential step prior to an improvement. Although all the patients were photographed within the same week the true prevalence of their leg appearances and the extent to which these patients are repre- sentative of a diseased venous cohort should be determined by the readers’ individual clini- cal practice. However, patients attend the clin- ic because of their varicose veins rather than hyperpigmentation per se. A further limitation is that photographs are not patients. Differences in lighting, back- ground and angles are known to have profound effects on the interpretation of varicose veins. The quality of the photographs appears poor from a professional viewpoint, and this is a factor which may have caused difficulties in participant’s choice. However, an A4 photo- graph at high resolution is much better than its on screen image. Each participant had the option to check the don’t know box, for whatev- er reason, but this rarely happened. Care was taken to ensure that each photograph accu- rately represented the clinical features of each patient. This may have advantages over ques- tionnaire studies which use descriptors with- out photographs5 and disadvantages in com- parison to studies where patients are exam- ined in a clinical setting.7 However, the use of photographs outside a clinical setting may be beneficial because it standardizes the avail- able information from which judgments are made. Clinicians are therefore less likely to have their judgments on the C of CEAP influ- enced by the patients’ medical records, symp- toms or duplex findings. Conclusions Clinical trials using the C class as a means of stratifying legs into mild and severe clinical disease should be interpreted with caution because of the difficulties in weighting the importance of pigmentation based solely on appearance. This information is of value in clinical situations where the C of CEAP may be used to ration treatment and in research situ- ations were it is often used as a benchmark or comparator for hemodynamic and quality-of- life validations. The results of this work have also indicated that the C of CEAP may be improved by using the same rater throughout clinical studies, unifying the CEAP definitions with those of the VCSS and by using leg eleva- tion to discriminate between telangectasiae and pigmentation. This work also confirms that the C class should not be used as a sever- ity classification. References 1. Porter JM, Moneta GL. Reporting stan- dards in venous disease: an update. International Consensus Committee on Chronic Venous Disease. J Vasc Surg 1995;21:635-45. 2. Beebe HG, Bergan JJ, Bergqvist D, et al. Classification and grading of chronic venous disease in the lower limbs. A con- sensus statement. Eur J Vasc Endovasc Surg 1996;12:487-91; discussion 91-2. 3. Eklof 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. 4. Kundu S, Lurie F, Millward SF, et al. Recommended reporting standards for endovenous ablation for the treatment of venous insufficiency: joint statement of The American Venous Forum and The Society of Interventional Radiology. J Vasc Surg 2007;46:582-9. 5. Antignani PL, Cornu-Thenard A, Allegra C, et al. Results of a questionnaire regarding improvement of ‘C’ in the CEAP classifica- tion. Eur J Vasc Endovasc Surg 2004;28: 177-81. 6. Uhl JF, Cornu-Thenard A, Carpentier PH, et al. Clinical and hemodynamic signifi- cance of corona phlebectatica in chronic venous disorders. J Vasc Surg 2005;42: 1163-8. 7. Uhl J-F, Cornu-Thenard A, Carpentier PH, et al. Reproducibility of the “C” classes of the CEAP classification. J Phlebol 2001;1: 39-43. 8. Allegra C, Antignani PL, Bergan JJ, et al. The “C” of CEAP: suggested definitions and refinements: an International Union of Phlebology conference of experts. J Vasc Surg 2003;37:129-31. 9. van Bemmelen PS, Bedford G, Beach K, Strandness DE. Quantitative segmental evaluation of venous valvular reflux with duplex ultrasound scanning. J Vasc Surg 1989;10:425-31. 10. Rathod S, LaBruna A. Questionnaire length and fatigue. ESOMAR Panel Research Conference, Budapest, 2005. Available from: http://www.esomar. org/web/research_papers/web-Panel_ 1092_Questionnaire-length-and-fatigue. php Accessed: 12 July 2012. 11. Rabe E, Pannier F. Clinical, aetiological, anatomical and pathological classification (CEAP): gold standard and limits. Phlebology 2012;27:114-8. 12. Bradbury AW. Epidemiology and aetiology of C4-6 disease. Phlebology 2010;25:2-8. 13. Maurins U, Hoffmann BH, Losch C, et al. Distribution and prevalence of reflux in the superficial and deep venous system in the general population—results from the Bonn Vein Study, Germany. J Vasc Surg 2008;48:680-7. 14. Stucker M, Reich S, Robak-Pawelczyk B, et al. Changes in venous refilling time from childhood to adulthood in subjects with apparently normal veins. J Vasc Surg 2005;41:296-302. 15. Chiesa R, Marone EM, Limoni C, et al. Chronic venous disorders: correlation between visible signs, symptoms, and presence of functional disease. J Vasc Surg 2007;46:322-30. 16. Shepherd AC, Gohel MS, Lim CS, Davies AH. A study to compare disease-specific quality of life with clinical anatomical and hemodynamic assessments in patients with varicose veins. J Vasc Surg 2011;53: 374-82. 17. Lattimer CR, Azzam M, Kalodiki E, et al. Saphenous pulsation on duplex may be a marker of severe chronic superficial venous insufficiency. J Vasc Surg 2012;56:1338-43. 18. O’Hare JL, Parkin D, Vandenbroeck CP, Earnshaw JJ. Mid term results of ultra- sound guided foam sclerotherapy for com- plicated and uncomplicated varicose veins. Eur J Vasc Endovasc Surg 2008;36:109-13. 19. Rasmussen LH, Lawaetz M, Bjoern L, et al. Randomized clinical trial comparing endovenous laser ablation, radiofrequency ablation, foam sclerotherapy and surgical stripping for great saphenous varicose veins. Br J Surg 2011;98:1079-87. 20. Darvall KA, Sam RC, Bate GR, et al. Changes in health-related quality of life after ultrasound-guided foam sclerothera- py for great and small saphenous varicose veins. J Vasc Surg 2010;51:913-20. 21. Neglen P, Egger JF, Olivier J, Raju S. Hemodynamic and clinical impact of ultra- sound-derived venous reflux parameters. J Vasc Surg 2004;40:303-10. 22. Yamaki T, Nozaki M, Fujiwara O, Yoshida E. Comparative evaluation of duplex- derived parameters in patients with chron- ic venous insufficiency: correlation with clinical manifestations. J Am Coll Surg 2002;195:822-30. 23. Lattimer CR, Kalodiki E, Azzam M, Geroulakos G. Reflux time estimation on air-plethysmography may stratify patients No n- co mm er cia l tionnaire studies which use descriptors with- No n- co mm er cia l tionnaire studies which use descriptors with- and disadvantages in com- No n- co mm er cia l and disadvantages in com- parison to studies where patients are exam- No n- co mm er cia l parison to studies where patients are exam- However, the use of No n- co mm er cia l However, the use of photographs outside a clinical setting may be No n- co mm er cia l photographs outside a clinical setting may be beneficial because it standardizes the avail- No n- co mm er cia l beneficial because it standardizes the avail- able information from which judgments areNo n- co mm er cia l able information from which judgments are made. Clinicians are therefore less likely toNo n- co mm er cia l made. Clinicians are therefore less likely to 4. Kundu S, Lurie F, Millward SF, et al. No n- co mm er cia l 4. Kundu S, Lurie F, Millward SF, et al.Recommended reporting standards for No n- co mm er cia l Recommended reporting standards for endovenous ablation for the treatment of No n- co mm er cia l endovenous ablation for the treatment of venous insufficiency: joint statement of No n- co mm er cia l venous insufficiency: joint statement of The American Venous Forum and The No n- co mm er cia l The American Venous Forum and The Society of Interventional Radiology. J Vasc No n- co mm er cia l Society of Interventional Radiology. J Vasc Surg 2007;46:582-9. No n- co mm er cia l Surg 2007;46:582-9. 5. Antignani PL, Cornu-Thenard A, Allegra C, No n- co mm er cia l 5. Antignani PL, Cornu-Thenard A, Allegra C, et al. Results of a questionnaire regarding No n- co mm er cia l et al. Results of a questionnaire regarding improvement of ‘C’ in the CEAP classifica- No n- co mm er cia l improvement of ‘C’ in the CEAP classifica- tion. Eur J Vasc Endovasc Surg 2004;28: No n- co mm er cia l tion. Eur J Vasc Endovasc Surg 2004;28: us e 3. Eklof B, Rutherford RB, Bergan JJ, et al. us e 3. Eklof B, Rutherford RB, Bergan JJ, et al.Revision of the CEAP classification for us e Revision of the CEAP classification forchronic venous disorders: consensus us e chronic venous disorders: consensus statement. J Vasc Surg 2004;40:1248-52.us e statement. J Vasc Surg 2004;40:1248-52. 4. Kundu S, Lurie F, Millward SF, et al.us e 4. Kundu S, Lurie F, Millward SF, et al. on ly 15. Chiesa R, Marone EM, Limoni C, et al. on ly 15. Chiesa R, Marone EM, Limoni C, et al. Chronic venous disorders: correlation on lyChronic venous disorders: correlationbetween visible signs, symptoms, and on lybetween visible signs, symptoms, andpresence of functional disease. J Vasc on lypresence of functional disease. J Vasc Surg 2007;46:322-30.on ly Surg 2007;46:322-30. 16. Shepherd AC, Gohel MS, Lim CS, Davieson ly 16. Shepherd AC, Gohel MS, Lim CS, Davies Article [page 54] [Veins and Lymphatics 2013; 2:e15] with early superficial venous insufficien- cy. Phlebology 2013;28:101-8. 24. Navarro TP, Delis KT, Ribeiro AP. Clinical and hemodynamic significance of the greater saphenous vein diameter in chron- ic venous insufficiency. Arch Surg 2002; 137:1233-7. 25. Lattimer CR, Kalodiki E, Azzam M, Geroulakos G. The Aberdeen Varicose Vein Questionnaire may be the preferred method of rationing patients for varicose vein surgery. Angiology 2013. [Epub ahead of print]. 26. Ratcliffe J, Brazier JE, Campbell WB, et al. Cost-effectiveness analysis of surgery ver- sus conservative treatment for uncompli- cated varicose veins in a randomized clin- ical trial. Br J Surg 2006;93:182-6. 27. Rutherford RB, Padberg FT Jr, Comerota AJ, et al. Venous severity scoring: An adjunct to venous outcome assessment. J Vasc Surg 2000;31:1307-12. 28. Cornu-Thenard A, Uhl JF, Carpentier PH. Do we need a better classification than CEAP? Acta Chir Belg 2004;104:276-82. No n- co mm er cia l u se on ly