1176 - Effect of Knee Osteoarthritis On Outpatient Visits in a Population-Based National Sample

Monday, October 19, 2009: 3:30 PM
112 A (Pennsylvania Convention Center)
E.A. Wright1, J.N. Katz2, M.G. Cisternas3, C.L. Kessler1, A.G. Wagenseller1 and E. Losina4, 1Brigham and Women's Hospital, Boston, MA, 2Brigham and Women's Hospital and Harvard Medical School, Boston, MA, 3MGC Data Services, Carlsbad, CA, 4Brigham and Women's Hospital, BU School of Public Health and Harvard Medical School, Boston, MA
Presentation Number: 1176


The impact of osteoarthritis (OA) on outpatient health care utilization in population-based samples has been rarely studied.  Accurate estimates of resource usage are critical for informed health care policy development.


Using the 2003 Medicare Beneficiary Study (MCBS), we selected a national cohort of persons with primarily knee OA, based on ICD-9 codes (715.x6, 715.x9, 715.x0), and a randomly selected cohort of sex and age-matched OA-free persons.  MCBS is a population-based survey of a stratified random sample of Medicare beneficiaries.  Subjects completed a detailed questionnaire assessing demographic, socioeconomic and other factors.  These data are linked to Medicare claims, which document health services utilization and treated comorbidities. We distinguished two components of outpatient health care utilization: visits to physicians (MD visits) and non-physician providers (non-MD visits, including nurses, physical therapists and laboratory studies).  We built multiple regression models accounting for sampling weights to determine whether knee OA independently affected utilization, controlling for comorbidities (<2 vs. ≥2 conditions), obesity (BMI ≥30), functional limitation (difficulty walking), education (≤ vs. > high school), race and working status (full/part-time or not working).


The MCBS cohort included 12,486 Medicare beneficiaries. Of these, 1,410 (11%) met our definition of OA (OA+) and were matched to 2,820 OA free (OA-) individuals. Mean age in both cohorts was 77 years; 70% were female.  OA+ and OA- differed significantly in obesity (OA+: 29%, OA-: 19%), % with ≥ 2 comorbidities (OA+: 65%, OA-: 45%), and functional limitation (OA+: 38%, OA-: 27%).  The OA+ cohort had significantly more MD visits (15.6) and non-MD visits (11.4) than the OA- cohort (9.8 and 7.5, respectively).  In multivariable  regression models controlling for age, sex, comorbidities, obesity, functional status, race, education and working status, knee OA independently predicted increased MD and non-MD visits.  In particular, the OA+ cohort had on average 4.1 more MD visits (95% CI: 3.5, 4.8) and 2.6 more non-MD visits (95% CI: 2.1, 3.1) than the OA- cohort.  The figure shows the incremental number of visits due to knee OA stratified by education and by number of comorbidities.


This first national, population-based study of health care utilization in OA documents considerable outpatient usage attributable to knee OA, independent of the effects of comorbidity, education and other patient characteristics.  These findings should be considered in policy decisions affecting outpatient resource utilization in the elderly.

Keywords: Knee, economics and osteoarthritis

Disclosure: E. A. Wright, None; J. N. Katz, None; M. G. Cisternas, None; C. L. Kessler, None; A. G. Wagenseller, None; E. Losina, None.