Actual versus Forecast Burden of Primary Hip and Knee Replacement Surgery in Australia: Analysis of Data from the Australian Orthopaedic Association National Joint Replacement Registry.

National projections of future joint replacement use can help us understand the changing burden of severe osteoarthritis. This study aimed to compare actual utilisation rates for primary total hip replacement (THR) and total knee replacement (TKR) to previously forecast estimates for Australia. Data from the Australian Orthopaedic Association National Joint Replacement Registry and Australian Bureau of Statistics were used to calculate 'actual' THR and TKR utilisation rates for the years 2014-2019, by sex and age group. 'Forecast' utilisation rates for 2014-2019 were derived from an earlier study that modelled two alternate scenarios for THR and TKR in Australia: Scenario 1 assumed a constant rate of surgery; Scenario 2 assumed continued growth in surgery rates. Actual utilisation rates were compared descriptively to forecast rates for females and males (overall and by age group). Rate ratios were calculated to indicate the association between actual and forecast THR and TKR rates, with a rate ratio of 1.00 reflecting perfect alignment. Over the study period, 191,996 THRs (53% in females) and 312,203 TKRs (55% in females) were performed. For both sexes, actual rates lay clearly between the Scenario 1 and 2 forecast estimates. In 2019, actual THR utilisation rates were 179 per 100,000 females (Scenario 1: 156; Scenario 2: 200) and 158 per 100,000 males (Scenario 1: 139; Scenario 2: 191). Actual TKR utilisation rates in 2019 were 289 per 100,000 females (Scenario 1: 275; Scenario 2: 387) and 249 per 100,000 males (Scenario 1: 216; Scenario 2: 312). Age-specific rate ratios were close to 1.00 for all age groups, indicating good alignment between forecast and actual joint replacement rates. These validation analyses showed that linear plus exponential growth forecasting scenarios provided an efficient approximation of actual joint replacement utilisation. This indicates our modelling techniques can be used to judiciously predict future surgery demand, including for age groups with high surgery rates.