Comparison of four methods for estimating actual radiotherapy utilisation using the 45 and Up Study cohort in New South Wales, Australia

Aim: To compare four methods for estimating actual radiotherapy utilisation (A-RUR) reported in the literature. Materials and methods: Participants in the 45 and Up Study in New South Wales (NSW) Australia completed a baseline questionnaire during 2006-2009 and consented to record linkage with administrative health datasets. Incident primary cancers (2006-2010) were identified through linkage with the NSW Cancer Registry. Radiotherapy receipt was identified through linkage with the Medicare Benefits Schedule and/or NSW Admitted Patient Data Collection (2006-2014). The four methods for estimating A-RUR were: 1" crude proportion; 2" crude proportion for patients followed for a defined period; 3" life table without censoring of deaths; 4" life table with censoring of deaths. Results: There were 9817 participants with a diagnosis of cancer between recruitment and end of 2010, median follow-up 5.4 years. Crude A-RUR for the cancer cohort was 30.2%, below the "optimal" 48%. The 5 yr A-RUR was 29.7%, 29.8% and 33.4% using methods 2-4 respectively. Conclusions: A-RUR estimates differed depending on the method used and all were below optimal. The method for estimating A-RUR for future studies should depend on the availability of the data as well as the intended audience for the results.