C. S. Crowson

OBJECTIVE: We report the results of a population based analysis of all health services used and charges incurred over a one-year period among a community based cohort of persons with a diagnosis of arthritis [including both osteoarthritis (OA) and rheumatoid arthritis (RA)] compared to a similar cohort of individuals from the same community who have never had a diagnosis of arthritis (NA), to examine the attributable costs of this chronic condition. METHODS: The unique resources of the Rochester Epidemiology Project were used to assemble the arthritis prevalence cohorts and the population based control cohort. The Olmsted County Health Care Utilization and Expenditures Database was used to collect information on health services utilization and charges. RESULTS: The average direct medical charges for the RA, OA, and NA cohorts were $3,802.05, $2,654.51, and $1,387.83, respectively (age and sex adjusted, p < 0.0001 for both the RA vs NA and OA vs NA comparisons). The median charges for these 3 groups were $1,050.00, $663.55, and $232.04 for the RA, OA, and NA groups, respectively (age and sex adjusted p < 0.0001 for both the RA vs NA and OA vs NA comparisons). These analyses indicated that, compared to the NA cohort both the OA and the RA prevalence cohorts incurred statistically significantly more charges, not only for the musculoskeletal disease care, but also for the care of numerous other conditions including respiratory, cardiovascular, gastrointestinal, neurological, and psychiatric conditions; and for general medical care. Individuals with arthritis (both OA and RA) also incurred statistically significantly more charges for diagnostic and therapeutic procedures, in-hospital care, imaging studies, physician services, equipment, and laboratory studies. Use of prescription medications was statistically significantly more common in the RA and OA groups compared to NA (96.3, 96, and 83%, respectively; age and sex adjusted p = 0.006 for the OA vs NA comparison and p = 0.015 for RA vs NA). CONCLUSION: These results emphasize the importance of considering all health services utilization (rather than only disease specific use) when estimating the economic effect of a chronic illness such as arthritis.

OBJECTIVE: Compared to rheumatoid arthritis (RA), osteoarthritis (OA) is considered much more benign and much less costly. We sought to describe the economic effects of RA and OA, in terms of the indirect and nonmedical expenditures, compared to nonarthritic controls. METHODS: Using our unique population based data resources, we developed a model for estimating and comparing disease specific costs among 2 randomly selected, community based samples of 200 patients each with RA and OA and a control group of 200 individuals from the same community who do not have arthritis. Data were collected using a pretested postal survey. Age and sex adjusted comparisons were conducted across the 3 groups, and predictors of cost and utilization were identified using logistic regression modeling. RESULTS: There were 123, 116, and 94 respondents among the RA, OA, and nonarthritis groups, respectively. The average age and the female-to-male ratios were higher in the OA and RA groups compared to the nonarthritis group. Patients with RA and OA required 3 times more days of care for their conditions compared to nonarthritics (p < 0.0001) and incurred significantly more expenditures for home or child care (p = 0.01) and other services (p = 0.001) (i.e., medical equipment, assistive devices, or home remodeling) compared to nonarthritics. In addition, patients with RA were significantly more likely to have lost their job or to have retired early due to their illness (p = 0.001); were the most likely to have reduced their work hours or stopped working entirely due to their illness (p = 0.003); and were 3 times more likely to have had a reduction in household family income than either individuals with OA or those without arthritis (p = 0.0001). Fifteen percent of respondents with RA were unable to get a job because of their illness, while 3% of respondents with OA and only 1% of nonarthritic respondents reported this experience (p = 0.001). Logistic regression analysis revealed that functional status and pain score, as well as the presence of either RA or OA, were significant predictors of cost and health services utilization. CONCLUSION: Disease specific indirect and nonmedical costs for OA are substantial and approach those for RA. This has important societal implications, given the high prevalence of OA.