Projecting the healthcare burden of knee osteoarthritis in New Zealand: results from a computer simulation model

Abstract

Purpose: To forecast the demand for total knee replacement (TKR) surgery and total healthcare expenditure due to knee osteoarthritis (OA) in the New Zealand (NZ) population over the period 2013 to 2043. Methods: We projected the incidence and progression of knee OA, utilization of TKR, and total knee OA-related healthcare expenditure in NZ over 30 years using the Osteoarthritis Policy (OAPol) model. The OAPol model is a computer-based, state-transition, Monte Carlo simulation model that can be used to estimate the projected rates of TKR, health loss, and healthcare spending associated with knee OA.The starting population cohorts in the model were defined by five-year age groups (from 25-29 years up to 75-79 years, and 80 years and over), stratified by ethnicity (Māori or non-Māori), BMI (obese or non-obese), and sex, based on the 2013 NZ population distribution (Census data). Simulated individuals with knee OA pass through up to four regimens in the standard treatment sequence in the model: NSAIDs and physical therapy, corticosteroid injections, TKR, and revision TKR. Data for the costs and utilization of these treatment regimens were obtained from a variety of NZ and international sources. The rate at which patients were offered and accepted TKR was calibrated to TKR provision rates drawn from the New Zealand Joint Registry, by running the model for 5 years and varying the offer and acceptance rates to match actual provision for each cohort.The OAPol model was used to project the proportion of individuals receiving TKR and per-capita OA-related healthcare costs for each five-year period from 2013-2018 to 2038-2043, within the population cohorts stratified by age, sex, obesity, and ethnicity. These rates were then applied to projected population sizes, obtained from Statistics New Zealand population projections and previously-published projections of obesity prevalence rates, to forecast the total burden of OA in each five-year period. As the youngest (25-29 years) cohort would be aged 50-54 at the start of the final (2038-2043) period, we have analysed TKR provision projections for those aged 50 years and over at the start of each period (to ensure an equivalent cohort in each period). Results: The total provision of TKR was projected to increase rapidly from 5400 per year in 2013-2018 to 6800 per year in 2018-2023, and then more slowly to 7300 per year in 2038-2043 (Figure 1). These increases largely reflected projected growth in the size of the older population, with the annual rate of TKR provision in those aged 50 years and over peaking at 410 per 100,000 population in 2018-2023, before falling to 330 per 100,000 in 2038-2042. This trend occurs because the number of TKRs received by those with initial prevalent knee OA (at baseline) rapidly falls away after 2018-2023, as many people in this cohort are by then projected to have either already had a TKR or died.OA-related treatment costs were estimated to be NZ$156 million per year in 2013-2018 (1 NZD ≈ 0.82 USD in 2013), and NZ$8.0 billion (in present-value terms, discounted at 3% per annum) over the 30-year time horizon.After adjusting for differences in the baseline age distribution, Māori had a substantially higher burden of OA, in terms of both TKR utilization and OA healthcare costs (Figure 2). Conclusions: Knee OA inflicts a growing cost to the NZ healthcare system. The projected rapid increase in demand for TKR over the next decade will require additional healthcare resources, to an environment in which the provision of joint replacement within the public healthcare sector is already constrained by capacity limitations.Our estimates of the treatment costs associated with OA, while substantial, represent underestimates of the true cost to the health system and society. Only costs associated with the four treatment regimens (NSAIDs and physical therapy, corticosteroid injections, TKR, and revision TKR) have been modelled, and only for the index knee (the model does not capture contralateral TKRs). Furthermore, indirect and non-healthcare costs (such as lost earnings, informal sector care, and community care) have not been considered. These projections indicate a need for a comprehensive national OA strategy in NZ to manage the demands that an ageing population, with increasing OA prevalence and obesity, will place on the NZ healthcare system.Figure 2Healthcare cost and total joint replacement provision in New Zealand, by age and ethnicity, 2013-20432View Large Image Figure ViewerDownload Hi-res image Download (PPT)