We investigated associations between area-level socioeconomic position (SEP), region of residence and the utilisation of primary total knee and/or hip replacement (TKR and THR, respectively) for osteoarthritis.

As part of the Ageing, Chronic Disease and Injury study, adults residing in western Victoria and registered with the Australian Orthopaedic Association National Joint Replacement Registry as having utilised a primary TKR (n=4,179; 56% female) and/or THR (n=3,120; 54% female) 2011-2013 were identified as cases. Residential addresses were matched with the Australian Bureau of Statistics (ABS) 2011 census data, whereby region of residence was defined according to Local Government Areas (LGAs), and area-level SEP (quintiles) were defined using an ABS-derived composite index. The control population (n=591,265; 51% female) was ascertained from ABS data, excluding those individuals identified as cases. Data were analysed using multilevel modelling in a stratified 2-stage cluster design.

In adjusted multilevel models, TKR was greatest for those aged 70-79 years (OR 1.4, 95%CI 1.3-1.5; referent=60-69 years) and in SEP quintile 5 (most advantaged OR 2.1, 95%CI 1.8-2.3; referent=SEP quintile 3); results were similar for THR (70-79 years=OR 1.7, 95%CI 1.5-1.8; SEP quintile 5=OR 2.5, 95%CI 2.2-2.8). The total variance contributed by the variance of LGAs was 2% (SD of random effects ±0.28) and 3% (SD ±0.32), respectively. 

Independent of small between-LGA differences in utilisation, and in contrast to the greater rates of knee and hip osteoarthritis observed in socially disadvantaged populations, we report greater utilisation of TKR or THR in the most advantaged SEP quintiles. In order to address inequities in utilisation, we suggest that any targeted resource allocation may need to focus on people rather than places.