Objective We sought to develop the first Damage Index (DI) in systemic sclerosis (SSc).
Methods The conceptual definition of ‘damage’ in SSc was determined through consensus by a working group of the Scleroderma Clinical Trials Consortium (SCTC). Systematic literature review and consultation with patient partners and non-rheumatologist experts produced a list of potential items for inclusion in the DI. These steps were used to reduce the items: (1) Expert members of the SCTC (n=331) were invited to rate the appropriateness of each item for inclusion, using a web-based survey. Items with >60% consensus were retained; (2) Using a prospectively acquired Australian cohort data set of 1568 patients, the univariable relationships between the remaining items and the endpoints of mortality and morbidity (Physical Component Summary score of the Short Form 36) were analysed, and items with p<0.10 were retained; (3) using multivariable regression analysis, coefficients were used to determine a weighted score for each item. The DI was externally validated in a Canadian cohort.
Results Ninety-three (28.1%) complete survey responses were analysed; 58 of 83 items were retained. The univariable relationships with death and/or morbidity endpoints were statistically significant for 22 items, with one additional item forced into the multivariable model by experts due to clinical importance, to create a 23-item weighted SCTC DI (SCTC-DI). The SCTC-DI was predictive of morbidity and mortality in the external cohort.
Conclusions Through the combined use of consensus and data-driven methods, a 23-item SCTC-DI was developed and retrospectively validated.
- systemic sclerosis
- organ damage
- outcome measure
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
What is already known about this subject?
Organ damage is part of the natural history of systemic sclerosis (SSc).
Organ damage occurs early in the disease course of SSc.
Based on a systematic review of the literature, no tool currently exists to measure organ damage in SSc.
What does this study add?
The first ever composite disease Damage Index in SSc has been developed.
The methodology used to develop the index is somewhat unique, with the combination of both consensus methods and statistical analysis of patient data.
How might this impact on clinical practice or future developments?
The Damage Index can potentially be used in observational studies and clinical trials as an endpoint, and for the enrichment of clinical trials.
Systemic sclerosis (SSc; ‘scleroderma’) is a multiorgan disease wherein the pathogenic processes of inflammation, vasculopathy and fibrosis lead to multiple disease manifestations.1 Unlike other rheumatological diseases that have a relapsing-remitting course, the course of SSc is often one of progressive damage to multiple organs including the skin, joints, heart, lungs, gastrointestinal (GI) tract and kidneys. Using a preliminary ad hoc set of criteria to define ‘organ damage’ in SSc, we found that irreversible organ damage accrues very early in the disease course, with 40% of patients having damage in one or more organ systems within 2 years of disease onset.2
Despite the fundamental importance of organ damage in SSc, there has been no clinical tool to identify and quantify damage in SSc, with current tools assessing disease severity, activity or combined treatment responses.3 The lack of an existing tool, together with the observation of damage accrual early in the disease course of SSc, led to an initiative to develop the multiorgan Scleroderma Clinical Trials Consortium Damage Index (SCTC-DI). Potential applications of this instrument include use as an outcome measure in interventional drug trials and in observational studies, and as a selection tool to enrich clinical trials for patients at risk of developing organ damage who are most likely to benefit from therapy.
The development of the SCTC-DI was an international collaboration using a combined approach of consensus and data-driven methods applied to data from the prospective Australian Scleroderma Cohort Study (ASCS). External validation of the SCTC-DI was performed in the prospective Canadian Scleroderma Research Group (CSRG) registry.
The SCTC-DI working group
The SCTC is an international consortium representing researchers and clinicians who have particular interest and expertise in care and research of SSc. The goal of the SCTC is to conduct, sponsor or facilitate clinical research in SSc, in particular projects aimed at improving and developing outcome measures for clinical trials and observational studies.
In 2013, under the auspices of the SCTC, a working group comprised 22 international experts in SSc, representing Australasia, the Americas and Europe, was assembled to develop a disease DI in SSc. The working group was led by a steering committee comprised 6 members (MN, MB, MH, WS, SMP, NF) overseeing the day-to-day development of the DI.
Non-rheumatology expert advisors
A panel of seven non-rheumatologist experts with an interest in SSc was assembled to give recommendations regarding what constitutes significant damage in each organ system. This panel comprised two cardiologists, two respiratory physicians, two gastroenterologists and one nephrologist. Input from this panel was sought for: (1) item generation for a survey of SSc experts and (2) ensuring face and content validity of the final instrument.
Three experienced SSc patient partners (from Australia, Canada and the UK) provided insight from the patients’ perspective. Input from these patient partners was sought at several stages in the process of instrument development: (1) conceptual definition of organ damage; (2) item generation for a survey of SSc experts; (3) ensuring face and content validity of the final instrument.
Development of the SCTC-DI
Conceptual definition of organ damage
The working group members participated in a two-round survey to achieve consensus regarding the conceptual definition of organ damage in SSc. In the second round, responses from the first round were provided, with the survey focusing on resolution of contentious areas. The definition was then presented to each working group member to ensure further comments and feedback were exhausted. All working group members had to approve the final definition.
A systematic review of existing measures of disease status in SSc was performed to generate items for potential inclusion in the DI.3 The working group members, expert advisors and patient partners, made further revisions to item definitions and provided further items for inclusion in a survey.
Item reduction using consensus survey
Three hundred and thirty-one expert members of the SCTC, European Scleroderma Trials and Research Group, CSRG and the Australian Scleroderma Interest Group (ASIG) were invited by email to take part in an online survey (Research Electronic Data Capture (REDCap), Vanderbilt University, Nashville Tennessee; survey included as an online supplementary file) to rate the appropriateness (on a scale of 1–7 with 1=very inappropriate, 2=inappropriate, 3=somewhat inappropriate, 4=neither inappropriate or appropriate, 5=somewhat appropriate, 6=appropriate, 7=very appropriate) in terms of both importance and feasibility of recording in a day-to-day setting, of including each item in an SSc DI. The survey items were categorised according to organ systems and for each organ system, a link was provided within the survey to a repository of 64 articles from the systematic literature review.4–66 An a priori cut-off of 60% votes for ‘appropriate’ (score of 5 or above) was used to reduce the items. Where applicable, responders were also asked to select an option regarding a time period for which the item needed to be present in order to constitute damage (present ever, 6 months, 12 months or 2 years).
Statistical analysis to achieve further item reduction and item weighting
Data from 1568 patients in the ASCS fulfilling the European League Against Rheumatism (EULAR)/American College of Rheumatology (ACR) criteria for SSc4 were used to retrospectively assess the univariable relationship of each damage item with the endpoint of mortality using time-dependent Cox hazards regression models. ASCS is an ongoing longitudinal study of patients with SSc recruited from nine sites in Australia since 2007, in whom standardised clinical (including the health-related quality of life (HRQoL) questionnaire Short Form 36 (SF-36))67 and laboratory data are collected at each annual visit and recorded in a customised database.
As there is no specific measure of morbidity in SSc per se, the Physical Component Summary (PCS) score of the SF-36 questionnaire was used as a surrogate for morbidity. SF-36 data were available for 1877 visits and panel logistic regression was used to determine the univariable relationship between each damage item and the SF-36 PCS (dichotomised for analysis based on median value in the cohort) measured at each visit for each patient. This method allows for the expected correlation between repeated measures over time within an individual. As not all items in the ASCS were defined exactly according to the definitions used in the DI, the closest possible equivalent of each item was used in item reduction and weighting. Items not demonstrating a univariable relationship (p<0.10) with the mortality or morbidity endpoints were removed. In addition, items that were highly clinically collinear with other items or lacked feasibility were removed as per steering committee consensus.
In order to create a weighted score, the multivariable coefficients were then calculated for the remaining items against each of the mortality and morbidity end points. The following rules were used to combine the correlation coefficients from each of the mortality and morbidity analyses in order to create a weighted score for each item: (1) if both coefficients were positive, an average was taken; (2) if one coefficient was negative and one positive, we took the positive only; (3) if both coefficients were negative, indicating protection from damage, the item was discarded. For scoring items, the following methods were used: (1) all the coefficients were divided by the smallest coefficient and multiplied by 0.25. The multiplier of 0.25 was chosen to ensure that if the smallest coefficient was very small (<0.10), the scores did not become too large; (2) the result was rounded to the nearest integer.
The performance of the SCTC-DI in the derivation cohort (internal validation) was determined using receiver operating characteristic (ROC) curves. Kaplan-Meier survival estimates against mortality and morbidity endpoints with patients divided into risk groups based on their SCTC-DI scores were generated. Risk groups were determined by looking for clear demarcation of the cumulative frequency percentage of total scores within the derivation cohort. Both of these analyses were used to demonstrate criterion validity. Frequency of damage items in a subset of patients with early disease at each year of follow-up was determined to demonstrate construct validity and responsiveness to change over time.
External validation was conducted using the CSRG patient registry. The CSRG is an ongoing longitudinal study of patients with SSc fulfilling the EULAR/ACR criteria4 recruited from 14 sites in Canada and 1 site in Mexico since 2004, using a standardised data collection protocol that is comparable to that of ASCS. Once again, ROC analyses were performed. Kaplan-Meier survival estimates were performed using the SCTC-DI risk scores generated in the derivation cohort. Damage accrual was assessed using the mean SCTC-DI score at each year of follow-up in each organ domain and in the overall cohort.
Development of the SCTC-DI
Conceptual definition of organ damage
A total of 20 (of 22) responses from the working group members were received (for a summary of results see online supplementary tables S1 and S2). The resulting definition of damage in SSc was as follows: ‘Damage in SSc is the permanent and irreversible loss of anatomical structure or physiological function, caused by SSc and not secondary to its treatment or comorbidities. Damage should be differentiated from disease activity, which is potentially reversible, and disease severity which encompasses both activity and damage. The DI will not include patient-reported outcomes. The DI should be reflective of morbidity and/or mortality. The duration that each abnormality must be present before constituting damage may depend on the particular item.’
It must be noted that damage items caused by treatment of SSc were originally included in the definition, but later removed (see below).
A total of 83 items for potential inclusion in the DI were generated through the systemic review of the literature3 and consultation with expert advisors and patient partners (all items included in the survey can be seen in online supplementary file 2). The variables come from six different organ system groupings, as follows: musculoskeletal and skin, vascular, GI, respiratory, cardiovascular and renal.
Item reduction using consensus survey
A total of 93 of 331 (28.1%) international SSc experts provided complete responses. Of the 83 items presented, 58 items had greater than 60% consensus with 36 items removed (online supplementary table S3). Of these 58 items, nine items (bony erosions seen on radiographic imaging, acro-osteolysis on radiographic imaging, sicca signs confirmed on objective testing, surgical limb amputation, insertion of permanent pacemaker or implantable cardioverter defibrillator, histologically proven Barrett’s oesophagus, rectal prolapse, lung malignancy with SSc interstitial lung disease (ILD), left ventricular failure on clinical assessment and confirmed on transthoracic echocardiogram [TTE]) were not collected in the ASCS database and could not be analysed. Overall, the working group was comfortable with the removal of these items given the inclusion of similar alternatives, and in most cases, more clinically appropriate items.
Removal of ‘damage due to treatment’ items
A preliminary conceptual definition of damage included ‘damage caused by the treatment of SSc’. Inclusion of these items was deemed unnecessary as most drug treatment clinical trials would be collecting extensive data on drug-related adverse outcomes. Furthermore, ascertaining attribution of damage to treatment was deemed unreliable.
Removal of 6-minute walk distance
The 6 min walk distance is used in some countries for various indications, including screening and monitoring of cardiopulmonary disease in SSc, and access to drugs that treat right heart catheterisation-proven pulmonary arterial hypertension. However, because of the lack of widespread use in clinical practice and accessibility across all centres, the working group members and patient partners decided to remove this item.
Inclusion of renal items
The items in the renal domain, such as ‘scleroderma renal crisis’ (SRC) and ‘need for renal replacement therapy’ (dialysis), occur rarely in SSc, and therefore, their relationship with mortality did not reach statistical significance in our derivation data set. Two of the three renal items (SRC and SRC requiring renal replacement therapy) did, however, have a significant univariable relationship with morbidity. Given its clinical importance, the working group decided to also include the third renal item (SRC and persistent renal impairment) despite the lack of a statistically significant relationship with either morbidity or mortality in the derivation data set.
Items removed due to clinical collinearity
The following 12 items were removed as their definition was very similar to that of other item(s) retained in the DI that were either more strongly associated with the mortality and morbidity endpoints, or were more feasible to record reliably in a clinical care setting: joint contracture in any large and/or small joint; joint contracture of the fingers with finger to palm distance >1 cm; calcinosis on physical examination; proximal muscle atrophy; digital pitting scar; modified Rodnan Skin Score >20, left ventricular systolic dysfunction on TTE alone, right ventricular dilation and/or enlargement on TTE, right ventricular failure on clinical assessment and confirmed on TTE, ILD of any extent on imaging, symptoms of gastroparesis and malabsorption syndrome (online supplementary tables S4 and S5). Persistent erectile dysfunction was also removed given its lack of applicability to all patients.
Items removed due to insufficient observations
The following three items were removed as they had insufficient observations to allow for meaningful statistical analysis: SRC with proteinuria, need for intravenous hyperalimentation and need for enteral feeding. Please refer to figure 1, which summarises the item generation and reduction steps.
Statistical analysis to achieve further item reduction and item weighting
The univariable relationship with mortality and morbidity (dichotomised using the cohort median SF-36 PCS=37.2) was significant (p<0.10) in 17 and 22 items, respectively (online supplementary tables S4 and S5). These items were then weighted and scored using multivariable regression (table 1). Four items were removed at this stage as they did not reach statistical significance in both the mortality and morbidity analyses (late-stage nailfold capillaroscopic changes, moderate to severe left ventricular diastolic dysfunction on TTE, any conduction defect on ECG or Holter monitor and faecal incontinence).
Scleroderma Clinical Trials Consortium Damage Index
The 23-item multiorgan weighted SCTC-DI (SCTC-DI) is presented in table 2. The scores are additive with a maximum possible total score of 55. In the final product, items addressing the same issue, for example SRC, are grouped together and scores are added if further clinical markers of severity are present. Members of the working group reviewed the final SCTC-DI for face and content validity and potential feasibility of application of each item.
Validation of the SCTC-DI
Characteristics of the ASCS and CSRG cohorts
The characteristics of the ASCS (derivation data set) and CSRG (validation data set) cohorts and the frequency of items together with percentage missing within each cohort are presented in online supplementary tables 6-9.
Performance of the SCTC-DI in the derivation and external cohorts
The mean (SD), median, minimum and maximum DI scores within the ASCS derivation data set were 6.65 (4.8), 6, 0 and 33, respectively. Within the external cohort of CSRG patients, the mean (SD), median, minimum and maximum DI score were 6.9 (4.7), 6, 0, 33, respectively.
ROC analysis revealed a good discriminatory capability of the SCTC-DI with ROC area under the curve of 0.77 (95% CI 0.73 to 0.82) and 0.73 (95% CI 0.69 to 0.77) for mortality in the ASCS and CSRG cohorts, respectively (figure 2A and B). For morbidity (SF-36 PCS), ROC area under the curve was 0.69 (95% CI 0.65 to 0.72) and 0.72 (95% CI 0.69 to 0.74) in the ASCS and CSRG cohorts, respectively (figure 2C and D).
Damage accrual over time within the derivation cohort was demonstrated in a subset of patients recruited within 2 years of disease onset, where the percentage of items present at each year of follow-up shows a steady increase (figure 3A). Steady accrual of the mean DI score was also demonstrated in the CSRG cohort at each year of follow-up (figure 3B). When total scores were categorised into risk groups (low damage score <5, moderate damage score 6–12, high damage score ≥13), Kaplan-Meier survival estimates showed a statistically significant graded relationship of damage score and time to death or first reduction (ie, worsening) of SF-36 PCS below the median score of 37.2 (log rank p<0.001) in the derivation cohort. Similar results were seen when the SCTC-DI was assessed in the CSRG validation cohort (figure 4A and B).
Through use of consensus methods with data-driven item reduction and weighting, an international working group of the SCTC has produced the first instrument developed specifically for the purpose of quantifying organ damage in SSc. The 23-item SCTC-DI has been designed to capture damage that is reflective of both morbidity and mortality in each of the six major organ systems affected. Although it was considered that separate damage indices each reflective of mortality or morbidity be created, the working group felt that the creative approach of combining the two analyses as a single index would be more practical in the research and clinical care setting. Relative to other damage indices such as the Vasculitis Damage Index (VDI)68 and the Systemic Lupus International Collaborative Clinics (SLICC)/ACR DI in Systemic Lupus Erythematosus,69 with 65 and 39 items, respectively, the SCTC-DI of 23 items is shorter and would be easy to use in clinical trials. While the total possible damage score is 55, the maximum score in the derivation data set was 33, indicating that the SCTC-DI is not limited by a ceiling effect and that damage scores exceeding 33 are likely incompatible with life. In the derivation data set, the SCTC-DI demonstrated good discriminatory capability for mortality risk and also morbidity.
We acknowledge that we did not follow the exact consensus methodology prescribed by EULAR to generate comparable indices, and this may be considered a limitation. However, the international collaborative approach of the working group, the survey of a large number of SSc experts outside of the working group, contributions of experts in non-rheumatology specialties and patient partners, together with robust statistical methods used to reduce, weight and score the items make the development of the SCTC-DI rigorous and unique. While the SCTC survey response rate was 28%, we received 93 complete survey responses from international experts to work with. The development of the SCTC-DI also benefited from a recent mortality study that we undertook using both ASIG and CSRG data where we traced lost to follow-up and confirmed all deaths.2
Existing multiorgan outcome measures in SSc include the European Scleroderma Study Group (EScSG) activity index and the Combined Response Index in Systemic Sclerosis (CRISS).70–72 As the name suggests, the EScSG activity index measures disease activity, which inherently implies that it measures potentially reversible components of SSc and is not cumulative in nature. The CRISS on the other hand is a set of outcome measures used to measure the effect of treatments in SSc. While both are valuable clinical trial tools, they do not capture the concept of cumulative, irreversible organ damage which invariably occurs in the natural history of SSc. Experience with both the VDI and the SLICC/ACR DI as endpoints in clinical trials also demonstrates the potential applications of a DI in SSc.
The value of the patient perspective in the development of disease indices has become increasingly recognised.73 Although the conceptual definition of damage in SSc stipulated that patient-reported outcomes not be included in the DI, the patient perspective was still taken into account through inclusion of patient partners who were consulted at all steps of instrument development. Furthermore, the use of the HRQoL measure SF-36 as an endpoint for weighting the scores allowed for a patient-reported outcome to be included in instrument development.
Our conceptual definition of ‘damage’ in SSc has inherent limitations. However, this could be said of all complex multiorgan connective tissue diseases where assignment of direct attribution to the disease and irreversibility is based on physician judgement. Some items in the DI are required to be present for a minimum of 6 months to ensure they are ‘irreversible’. The future availability of truly disease modifying drugs may mean that some items in the DI may indeed be made reversible. In such a situation, an updated revision of the DI will be needed. The use of an existing prospectively acquired clinical database for statistical analyses meant that a handful of items were not present or did not have a definition identical to that presented in the survey of experts. However, where this was the case, a surrogate definition, which was as close as possible to the original item, was used. Despite this limitation, both internal and external validity were demonstrated against the endpoints of mortality and morbidity, as was responsiveness to change over time. To complete validation of the SCTC-DI, future studies need to evaluate reliability/reproducibility and feasibility of application of the SCTC-DI.
The use of retrospective data in a prevalent cohort underscores the need for further validation of the SCTC-DI in a prospective data set where all items in the DI, together with those that did not make the final index but were deemed appropriate by experts surveyed, are collected. In addition, the early mortality74 and early accrual of damage in SSc has provided a rationale for the creation of an international inception cohort (International Systemic Sclerosis Inception Cohort) of patients with SSc in order to prospectively validate disease outcome measures, including the SCTC-DI.
We thank Kerry McKenna for assistance with the survey of experts. In addition to members of the SCTC Damage Index Working Group, we thank all members of the SCTC, EUSTAR, CSRG and ASIG who took part in our survey. We thank the following non-rheumatology expert advisors for their input: DP (cardiologist), St Vincent’s Hospital, Melbourne, Australia; David Celermajer (cardiologist), University of Sydney, Australia; Professor Robyn Langham (nephrologist), Monash University, Melbourne, Australia; Eli Gabbay (respiratory physician), University of Western Australia, Australia; Trevor Williams (respiratory physician), Alfred Health, Australia; Michael Kamm (gastroenterologist), University of Melbourne, Australia and Peter De Cruz (gastroenterologist), Austin Health, Melbourne, Australia. We thank the following patient partners for their input: Kim Fligestone and Maureen Sauve. We also thank our third patient partner who wishes to remain anonymous.
MB and MN contributed equally.
Handling editor Prof Josef S Smolen
Collaborators Scleroderma Clinical Trials Consortium Damage Index (SCTC-DI) Working Group members: Mandana Nikpour (Co-chair), The University of Melbourne at St. Vincent’s Hospital Melbourne, Australia; Murray Baron (Co-Chair), Lady Davis Institute for Medical Research and Jewish General Hospital Montreal, Canada; Nava Ferdowsi, The University of Melbourne at St. Vincent’s Hospital Melbourne, Australia; Tracy Frech, University of Utah, USA; Marie Hudson, Lady Davis Institute for Medical Research and Jewish General Hospital Montreal, Canada; Susanna Proudman, The University of Adelaide at Royal Adelaide Hospital, Australia; Wendy Stevens, St Vincent’s Hospital, Melbourne, Australia; Tien Tay, The University of Melbourne, Australia; Souyma Chatterjee, Cleveland Clinic, Ohio, USA; Lorinda Chung, Stanford University, USA; Jessica K Gordon, Hospital for Special Surgery New York, USA; Anna Haemel, University of California San Francisco, USA; Sindhu R Johnson, University of Toronto, Canada; Dinesh Khanna, University of Michigan, USA; Thomas A Medsger, University of Pittsburgh, USA; Peter Merkel, University of Pennsylvania, USA; John Pauling, Royal National Hospital for Rheumatic Diseases, Bath, UK; Janet E Pope, University of Ontario, Canada; Tatiana Rodriguez-Reyna, National Institute of Medical Sciences, Mexico; Lesley Saketkoo, Louisiana State University, USA; James R Seibold, Scleroderma Research Consultants, USA; Ankoor Shah, Duke University, USA; Virginia Steen, Georgetown University, USA; Gemma Strickland, Barwon Health, Australia. Australian Scleroderma Interest Group (ASIG) members: Catherine Hill, Adelaide, South Australia; Gene-Siew Ngian, Melbourne, Victoria; Mandana Nikpour, Melbourne, Victoria; Susanna Proudman, Adelaide, South Australia; Maureen Rischmueller, Adelaide, South Australia; Janet Roddy, Perth, Western Australia; Joanne Sahhar, Melbourne, Victoria; Wendy Stevens, Melbourne, Victoria; Gemma Strickland, Geelong, Victoria; Jenny Walker, Adelaide, South Australia; Peter Youssef, Sydney, New South Wales. Canadian Scleroderma Research Group (CSRG) members: J. Pope, London, Ontario; M. Baron, Montreal, Quebec; J. Markland, Saskatoon, Saskatchewan (deceased); D. Robinson, Winnipeg, Manitoba; N. Jones, Edmonton, Alberta; N. Khalidi, Hamilton, Ontario; P. Docherty, Moncton, New Brunswick; E. Kaminska, Calgary, Alberta; A. Masetto, Sherbrooke, Quebec; E. Sutton, Halifax, Nova Scotia; J-P. Mathieu, Montreal, Quebec; M. Hudson, Montreal, Quebec; S. Ligier, Montreal, Quebec; T. Grodzicky, Montreal, Quebec; S. LeClercq, Calgary, Alberta; C. Thorne, Newmarket, Ontario; G. Gyger, Montreal, Quebec; D. Smith, Ottawa, Ontario; P.R. Fortin, Quebec, Quebec; M. Larché, Hamilton, Ontario; M. Abu-Hakima, Calgary; TS Rodriguez-Reyna, Mexico City, Mexico; AR Cabral, Mexico City, Mexico; M. Fritzler, Mitogen Advanced Diagnostics Laboratory, Cumming School of Medicine, Calgary, Alberta.
Contributors NF: study design, data collection, data analysis, interpretation of results, preparation of manuscript; MoH (biostatistician): data analysis, interpretation of results, preparation of manuscript; WS: study design, data collection, interpretation of results, preparation of manuscript; MaH: study design, data collection, data analysis, interpretation of results, preparation of manuscript; TT: study design, data collection, interpretation of results, preparation of manuscript; JLB: study design, data analysis, interpretation of results, preparation of manuscript; SM: study design, data analysis, interpretation of results, preparation of manuscript; CR: data collection, interpretation of results, preparation of manuscript; VS: study design, data analysis, interpretation of results, preparation of manuscript; DP: study design, interpretation of results, preparation of manuscript; SMP: study design, data collection, interpretation of results, preparation of manuscript; MB: study design, data collection, data analysis, interpretation of results, preparation of manuscript; MN: study design, data collection, data analysis, interpretation of results, preparation of manuscript.
Funding The development of the SCTC-DI was supported by SCTC Working Group grants. The Australian Scleroderma Cohort Study is supported by Scleroderma Australia, Arthritis Australia, Actelion Australia, Bayer, CSL Biotherapies, GlaxoSmithKline Australia and Pfizer. A/Prof Nikpour holds an NHMRC Fellowship (APP1126370). The Canadian Scleroderma Research Group (CSRG) is funded by the Canadian Institutes of Health Research (CIHR) (grant #FRN 83518), the Scleroderma Society of Canada and its provincial Chapters, Scleroderma Society of Ontario, Scleroderma Society of Saskatchewan, Sclérodermie Québec, Cure Scleroderma Foundation, INOVA Diagnostics (San Diego, CA), Fooke Laboratorien (Neuss, Germany), Euroimmun (Lubeck, Germany), Mikrogen (Neuried, Germany), Fonds de la rechercheen santé du Québec (FRSQ), the Canadian Arthritis Network (CAN) and the Lady Davis Institute of Medical Research of the Jewish General Hospital, Montreal, QC. The CSRG has also received educational grants from Pfizer and Actelion pharmaceuticals.
Competing interests None declared.
Patient consent for publication Not required
Ethics approval Ethics approval for the survey of international experts was obtained from the Human Research Ethics Committee of St. Vincent’s Hospital Melbourne. Ethics approval for the ASCS was obtained from the Human Research Ethics Committees of each of the participating centres, led by St. Vincent’s Hospital Melbourne. Ethics approval for the CSRG registry was obtained from the Human Research Ethics Committees of each of the participating centres, led by the Jewish General Hospital, Montreal.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement All data that can be made available have been included in online supplementary files.