Background/Purpose: Prognosis studies study future outcomes and/or seek to identify predictive or associative factors associated with outcomes. Strong and consistent prognostic factors can be used to individualize management and better outcomes of patients. Many prognostic factors have been identified in SLE but few have been consistent. We hypothesize that this is due to flawed study design. We aim to systematically assess methodological quality of prognosis studies in SLE. 

Methods: A systematic search of prognosis studies in SLE was performed in MEDLINE and EMBASE, from January 1990 to June 2011. Non-English literature, non-original research, non- full length reports and animal studies were excluded. Of 5419 articles subjected to a title and abstract screen, 1039 articles were found. A representative sample of 150 articles was selected using a random number generator and assessed by 2 reviewers. Studies were classified according to design and the clarity of research question was assessed. Each study was assessed by a risk-of-bias tool “QUality In Prognosis Studies” (QUIPS) in 6 domains: study participation, study attrition, measurement of prognostic factors, measurement of outcomes, measurement/adjustment for confounders and appropriateness of statistical analysis. Information on missing data was also collected. 

Results:

Of 150 articles, 15 were pediatric studies, 3 made comparisons of pediatric and adult patients and the remainder were adult studies. The majority were published in rheumatology journals (69%). Cohort design was used in 67% of studies; the remainder used cross-sectional (21%), case-control (5%) and other designs (7%). The research question clearly included study population in 92%, prognostic factor in 54% and outcome in 61% of studies. High risk of bias (QUIPS) was noted in 57% of studies for study participation, 57% for attrition, 20% for prognostic factor, 18% for outcome, 65% for confounders and 36% for statistical analyses. Confounders were named in the methods section in only 12% of studies. Some consideration for confounding was built into the design of 21% of studies. The amount of missing data could not be assessed in 39% of studies. 

Conclusion:

Inadequate articulation of research questions for prognostic factors, poor design addressing confounding, study participation and attrition and inadequately reported information on missing data limited the quality of prognosis studies in SLE. Future prognosis studies should be designed with better consideration to the above factors to improve methodological rigor.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/a-systematic-review-of-quality-of-prognosis-studies-in-systemic-lupus-erythematosus/