THE EPIDEMIOLOGY OF REVISION HIP ARTHROPLASTY IN AUSTRALIA

Abstract

Introduction There are no reports on the epidemiology of revision hip arthroplasty in Australia. The aim of this study was to characterise the epidemiology of revision hip arthroplasty in relation to primary hip replacements in Australia.

Methods This study covered the seven year period 1993/1994 to 1999/2000. Data on all primary hip replacements and revisions done were obtained from the Australian Institute of Health and Welfare using the ICD-9 (81.53) and ICD-10 (Block No. 1492) cartegorisation. The data was stratified by age (five year age groups), sex, year and state or territory. Log linear modelling was used to examine the rate of revision procedures out of the total number of procedures (primary and revision). The effects of gender, age and year were examined in a series of hierarchical log-rate models (Poisson Loglinear Regression).

Results For the period of the study there were 18,027 revision cases and 122,595 cases of primary hip replacement, representing a revision rate of 14%. The rate of increase of primary hip replacements was significantly higher than the rate of increase of revisions (t= −12.1, p< 0.0005). The number of primary hip replacements performed nationally increased by 810 (95% confidence intervals 658,964) a year. The number of revisions increased by only 62 (95% confidence intervals 21,104) a year. The proportion of revisions decreased by 0.3% per year as determined by regression analysis. The hierarchical log-rate models indicate significant interactions between age and gender and beween age and year.

Conclusions The number of primary and revision hip replacements has been increasing with time. The rate of increase of revision hip replacements has been lower. The proportion of hip replacements that are revisions has been gradually dropping, probably due to a greater increase in the number of primary hip replacements.