Classification criteria for systemic sclerosis subsets

Objective. To evaluate the measurement properties of criteria for systemic sclerosis (SSc) subsets for classification of patients in SSc trials, and to determine if any one criteria set confers measurement advantage over others. Methods. A systematic review of articles describing classification criteria for SSc subsets was performed. Evidence supporting the sensibility (statement of purpose for which the criteria will be used, population, setting, face and content validity, and feasibility), validity, and reliability of the criteria was evaluated. Results. Fourteen sets of criteria for SSc subsets were identified. There is variability in the intended purpose and setting for which criteria sets are to be applied. Although face validity improves with the addition of less commonly encountered subsets or disease manifestations as criteria, the feasibility of implementing such criteria is conversely limited. Content validity for most criteria sets has not been evaluated due to lack of an explicitly stated conceptual framework for SSc. The criteria with 3 or more subsets do not provide incremental predictive validity over the 2-subset criteria. Our ability to compare subset criteria on divergent validity and reliability is limited by a lack of data. Conclusion. The 2-subset criteria of LeRoy, et al have good feasibility, acceptable face validity, and good predictive validity. Further research is needed to compare the content validity, divergent validity, and reliability of these with other subset criteria for use in SSc trials.