Premium
Derivation and Validation of a Prediction Rule for Two‐Year Mortality in Early Diffuse Cutaneous Systemic Sclerosis
Author(s) -
Domsic Robyn T.,
Nihtyanova Svetlana I.,
Wisniewski Stephen R.,
Fine Michael J.,
Lucas Mary,
Kwoh C. Kent,
Denton Christopher P.,
Medsger Thomas A.
Publication year - 2014
Publication title -
arthritis and rheumatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 4.106
H-Index - 314
eISSN - 2326-5205
pISSN - 2326-5191
DOI - 10.1002/art.38381
Subject(s) - cohort , medicine , logistic regression , cutoff , stepwise regression , proportional hazards model , receiver operating characteristic , statistics , mathematics , physics , quantum mechanics
Objective Systemic sclerosis (SSc) is associated with a reduction in life expectancy, but there are no validated prognostic models for determining short‐term mortality. The objective of this study was to derive and validate a prediction rule for 2‐year mortality in patients with early diffuse cutaneous SSc (dcSSc). Methods We studied a prospectively enrolled cohort of 387 US Caucasian patients with early dcSSc (<2 years from the appearance of the first symptom), randomly divided into a derivation cohort (n = 260) and a validation cohort (n = 127). Predefined baseline predictor variables were analyzed in a stepwise multivariable logistic regression model in order to identify factors independently associated with 2‐year all‐cause mortality using a cutoff of P < 0.05. We rounded the beta values to the nearest integer and summed the points assigned to each variable in order to stratify patients into low‐risk, moderate‐risk, and high‐risk groups. We then applied this rule to an external validation cohort of 110 Caucasian patients with early dcSSc from a single UK center and compared stratum‐specific mortality using chi‐square statistics. Results Four independent predictor variables (with assigned integer values) comprised the model: age at first visit (points allotted: –1, 0, or 1), skin thickness progression rate (points allotted: 0 or 1), gastrointestinal tract severity (points allotted: 0, 1, or 2), and anemia (points allotted: 0 or 2). The prediction model performed well, with no significant differences between the derivation cohort and the US or UK validation cohorts in the low‐risk and moderate‐risk groups. Conclusion We derived a 4‐variable prediction rule that can be used to stratify patients with early dcSSc into groups by risk of 2‐year mortality, and we validated that prediction rule in US and UK cohorts.