Early Mortality in a Multinational Systemic Sclerosis Inception Cohort

Yanjie Hao, Marie Hudson, Murray Baron, Patricia Carreira, Wendy Stevens, Candice Rabusa, Solene Tatibouet, Loreto Carmona, Beatriz E. Joven, Molla Huq, Susanna Proudman, Mandana Nikpour, the Canadian Scleroderma Research Group, the Australian Scleroderma Interest Group

Research output: Contribution to journalArticleResearchpeer-review

95 Citations (Scopus)


Objective: To determine mortality and causes of death in a multinational inception cohort of subjects with systemic sclerosis (SSc). Methods: We quantified mortality as standardized mortality ratio (SMR), years of life lost, and percentage mortality in the first decade of disease. The inception cohort comprised subjects recruited within 4 years of disease onset. For comparison, we used a prevalent cohort, which included all subjects irrespective of disease duration at recruitment. We determined a single primary cause of death (SSc related or non–SSc related) using a standardized case report form, and we evaluated predictors of mortality using multivariable Cox regression. Results: In the inception cohort of 1,070 subjects, there were 140 deaths (13%) over a median follow-up of 3.0 years (interquartile range 1.0–5.1 years), with a pooled SMR of 4.06 (95% confidence interval [95% CI] 3.39–4.85), up to 22.4 years of life lost in women and up to 26.0 years of life lost in men, and mortality in the diffuse disease subtype of 24.2% at 8 years. In the prevalent cohort of 3,218 subjects, the pooled SMR was lower at 3.39 (95% CI 3.06–3.71). In the inception cohort, 62.1% of the primary causes of death were SSc related. Malignancy, sepsis, cerebrovascular disease, and ischemic heart disease were the most common non–SSc-related causes of death. Predictors of early mortality included male sex, older age at disease onset, diffuse disease subtype, pulmonary arterial hypertension, and renal crisis. Conclusion: Early mortality in SSc is substantial, and prevalent cohorts underestimate mortality in SSc by failing to capture early deaths, particularly in men and those with diffuse disease.

Original languageEnglish
Pages (from-to)1067-1077
Number of pages11
JournalArthritis and Rheumatology
Issue number5
Publication statusPublished - 1 May 2017

Cite this