Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.

What Do We Really Know about Arthritis Impact on Employment from Prospective Studies? A Systematic Review of the Literature.

Theis,  Kristina A., Wilkie,  Ross, Busija,  Lucy, Elsworth,  Gerald, Osborne,  Richard


Work disability results from many rheumatic conditions and is known to have substantial personal and societal consequences. Much of what we know about arthritis impact on employment comes from condition-specific cross-sectional studies. A comprehensive search strategy was conducted to identify prospective studies demonstrating how arthritis impacts work.


Publications were identified by an online electronic search of 6 databases (MEDLINE, JSTOR, PsychINFO, CINAL, EMBASE, and Social Science Abstracts) conducted in early 2010 together with a hand search of references in all obtained literature. Searches were performed with the following terms: "arthritis (in title or abstract) AND work (in title or abstract)" in combination (independently) with: moderate, mediat*, interaction, buffer, buffering, "path analysis", longitudinal, causal, "risk factor", mitigate, associat*, predictor, and explanatory. Inclusion criteria for review: prospective study design assessing work impact at 2 or more time points in adults (>= 18 years) with an arthritis/rheumatic condition, published in English. Biomedical, scale/instrument development, drug trial, and intervention studies were out of scope. Compliance with the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) consensus statement was examined as a marker for study quality.


Electronic searches returned 2,877 potential publications; after duplicate removal and title/abstract review 90 remained and were retrieved for complete review. An additional 110 records were identified through hand searching and assessed for eligibility. In total, 15 publications met inclusion criteria for abstraction, 10 of which concerned rheumatoid arthritis, 1 each on fibromyalgia and osteoarthritis, 2 on systemic lupus erythematosus, and 1 included multiple conditions. No study included all relevant STROBE items, with the most commonly omitted including discussion of potential biases and clear reporting of participants at each stage. Definitions of work impact varied but were most often related to work cessation; 3 studies examined more subtle impacts (i.e., productivity, sick leave, work transitions). Methodological approaches were diverse including path and survival analyses and logistic and Cox regression. Characteristics examined could be categorized into 4 general groups (demographics, arthritis-specific, general health/disability, and work-specific). Preliminary results indicated that disability/functional limitation and age were the strongest and most consistently identified predictors of work impacts with less consistent results for work-specific, arthritis-specific, and social characteristics.


With the exceptions of age and disability, there is little consistent evidence for predictors of arthritis impact on employment from prospective studies. Robust conclusions are limited by differences in potential predictors examined, work impact definitions, and methods. However, findings from this systematic review indicate reasonable starting places for policy and public health efforts to mitigate arthritis impact on work.

To cite this abstract, please use the following information:
Theis, Kristina A., Wilkie, Ross, Busija, Lucy, Elsworth, Gerald, Osborne, Richard; What Do We Really Know about Arthritis Impact on Employment from Prospective Studies? A Systematic Review of the Literature. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :2079
DOI: 10.1002/art.29844

Abstract Supplement

Meeting Menu