Poster Presentation Clinical Oncology Society of Australia Annual Scientific Meeting 2017

Criterion-based benchmarking (CBB) approach of the appropriate use of radiotherapy in NSW-ACT, Australia (#331)

Gabriel Gabriel 1 2 , Geoff Delaney 1 2 , Michael Barton 1 2
  1. Collaboration for Cancer Outcome, Research and Evaluation (CCORE), Ingham Institute, Liverpool Hospital, LIVERPOOL, NSW, Australia
  2. University of New South Wales, Sydney, NSW, Australia

Introduction:

Planning for radiotherapy services requires information on the proportion of cancer patients who should receive radiotherapy. CCORE has previously estimated optimal rates of radiotherapy utilization (RTU) based on the development of decision trees using evidence-based treatment guidelines and epidemiological data1,2. In Ontario, Mackillop established CBB approach to estimate RTU 3-9.

Objectives:

  • Calculate actual and benchmark RTU 
  • Compare with optimal rates

Methods:

Radiotherapy data were linked to CCR. Calculate road distances between patient residence and nearest RT. Cross-border patients were excluded. Benchmark criteria are:

  1. Patients make no direct payment for radiotherapy,
  2. All RT is provided by site-specialized radiation oncologists (RO) in multi-disciplinary sitting
  3. RO receive salary for their service,
  4. >75% of patients live within 30km
  5. Patients waiting times<=4weeks. 

Results:

Overall, 25.4% of cancer patients received radiotherapy within 1-year of diagnosis in CBB LGAs compared to 22.1% in all LGAs. Actual RTU for NSW, Alberta & USA are comparable but lower than Ontario. CBB RTU rates were > actual RTU but < optimal RTU.

Limitations of the CBB model:

  1. Lack of generalizability:
    • RT payment and RO remuneration arrangements vary in different regions
    • Lack of site-specialized RO in some countries
    • Effect of dispersed populations 
    • Population data on RTU are not available in all jurisdictions
  2. CBB depends upon practice not evidence therefore can be influenced by non-evidence-based practices.

Conclusions:

RTU were 7-16% higher in CBB communities than in all communities but are still 30-65% below optimal RTU. CBB is based on the assumption that there is perfect service delivery in some parts of health service that can be used to benchmark the whole service. This approach may be applicable in well-resourced service delivery model in Ontario, but until feasibility is proved applicable in different regions, CBB approach doesn't seem reproducible and may not be recommended for benchmarking RTU. We recommend evidence-based optimal RTU.