Arthritis & Rheumatism, Volume 63,
November 2011 Abstract Supplement
Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Chicago, Illinois November 4-9, 2011.
Development and Validation of a New Clinical Prediction Rule for 5-Year Survival in Early Scleroderma, a EUSTAR Study.
diaconu-Popa1, Delia, Vonk2, Madelon C., Hesselstrand3, Roger, Carreira4, Patricia E., Valentini5, Gabriele, Beretta6, Lorenzo, Airo7, Paolo
Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands
Department of Rheumatology and Internal Diseases, Medical University of Bialystok, Bialystok, Poland
Paris Descartes University, Rheumatology A department, Cochin Hospital, Paris, France
University of Rome, Medical Clinic and Therapy Department
Rheumatology Unit, University of Aquila, L'Aquila, Italy
Clinica Universitaria Ancona, Ancona, Italy
Charité University Hospital, German Rheumatology Research Center, a Leibniz Institute, Berlin, Germany
University of Florence, Florence, Italy
EBMT, Paris, France
University of Cologne, Cologne, Germany
Department of Rheumatology, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands
Department of Rheumatology, Sint Maartenskliniek, Nijmegen, Netherlands
Radboud University Nijmegen Medical Center, Nijmegen, Netherlands
Lund University & Skåne University Hopsital, Lund, Sweden
Instituto de Investigación Hospital 12 de Octubre (I+12), Madrid, Spain
Second Univ of Napoli, Napoli, Italy
IRCCS Fondazione Policlinico-Mangiagalli-Regina Elena & University of Milan, Milan, Italy
Istanbul University, Istanbul Faculty of Medicine,, Istanbul, Turkey
Rambam Health Care Campus, Haifa, Israel
Systemic sclerosis (SSc) is associated with a significant reduction in life expectancy. The existing clinical prediction model developed in 1999 for 5 year survival in SSc underestimated survival in new patients. Therefore, the objective was to develop and validate a new clinical prognostic model to predict 5-year survival in early systemic sclerosis (SSc).
A multi-centre cohort of SSc patients diagnosed before 2002 was established in EUSTAR centers. Patients were eligible if they were followed for five-years or shorter if they died. The sample (N=955) was split in a development sample (2/3 of deaths) and a validation sample (1/3 of deaths). In the development sample, univariate and multivariate logistic regression and a bootstrap procedure were used to develop a prognostic model to predict 5-year survival after diagnosis of SSc. Model performance (discrimination and calibration) was studied in the validation sample.
The development sample consisted of n=700 patients of whom 59 (8.4%) died and the validation sample consisted of n=255 patients of whom 25 (9.8%) died. After univariate logistic regression, 14 clinical variables (p<0.10) entered the multivariate logistic regression model. Using a backward selection procedure, a combination of 6 variables (p<0.20) remained. The six items were: age >55 at diagnosis (OR 2.16), female gender (OR 1.73), dcSSc (OR 1.93), ECG abnormalities (OR 1.71), low FVC or low DLCO (OR 1.36) and presence of urine protein (OR 2.50). In the validation sample, the 1999 prediction model had an area under the ROC curve of 0.73. The area under the ROC curve of the new model was 0.85 and the predicted probabilities of 5-year survival appropriately ranged from 0.02 to 1.0.
Figure 1. Figure A. Calibrations curves for the prognostic score; B. Receiver operating characteristics curve for the new prognostic model.
A new simple prognostic model using 6 disease factors to predict 5 year survival at diagnosis in SSc was developed. It showed good performance regarding calibration and discrimination in the external validation sample.
To cite this abstract, please use the following information:
diaconu-Popa, Delia, Vonk, Madelon C., Hesselstrand, Roger, Carreira, Patricia E., Valentini, Gabriele, Beretta, Lorenzo, et al; Development and Validation of a New Clinical Prediction Rule for 5-Year Survival in Early Scleroderma, a EUSTAR Study. [abstract]. Arthritis Rheum 2011;63 Suppl 10 :673