Abstract
Background With tremendously improved survival, many adults live with childhood-onset (onset <18 years) systemic autoimmune rheumatic diseases (ChildSARDs) which are systemic lupus erythematosus (SLE), Sjogren’s syndrome, idiopathic inflammatory myositis (IIM), systemic sclerosis, and systemic vasculitides. Little is known about ChildSARD outcomes in adulthood. We systematically reviewed prognosis studies performed on ChildSARD adults to summarize the scope of outcomes studied, and evaluate for methodological issues.
Methods With an academic librarian, we developed search strategies in MEDLINE and Embase (January 1990 – May 2017) to search for full-length English articles. Our strategies were iteratively fine-tuned and finalized after peer-review by librarians. We supplemented the search with hand-searching of the references of review articles and included articles. If a study included both pediatric and adult subjects, we included the study only if the median or mean age at outcome ascertainment was ≥18 years. Information about outcomes and study designs was recorded. All studies were graded independently by two reviewers using the Quality in Prognosis Studies risk-of-bias tool.
Results Of 14 100 articles, 26 publications were included in this review. Of these, only 10 (36%) focused on adults; while the rest studied mixed adult and paediatric populations. Two studies (7%) were longitudinal (repeat measures on ≥3 occations) while the remainder assessed outcomes on a single occasion. About 80% of studies were published within 2010–2017. The most commonly studied diseases were SLE (61%), IIM (18%), and systemic sclerosis (11%). The most commonly reported primary outcomes were organ damage (29%), cardiovascular outcomes (14%), and mortality (14%). The mean ages at outcome assessment were 19.5–46.8 years for adult-only studies and 19.3–35 years for mixed studies. Moderate to high risk-of-bias was found in all studies for study participation, 90% for study attrition, 61% for prognostic factor measurement, 36% for outcome measurement, 89% for confounding, and 54% for statistical analysis.
Conclusion There is need for more information about adulthood outcomes of ChildSARDs. Longitudinal data was especially lacking. We recommend that future studies on ChildSARD outcomes be undertaken in the framework of a longitudinal cohort. Adult outcomes should be separately reported from pediatric outcomes in a mixed cohort. Study populations should be clearly defined to allow for accurate MESH coding so as to facilitate easy searching for such information and for knowledge dissemination. Careful attention should be paid during study design to reduce bias in choice of study populations, especially in accounting for attrition and confounding.
