Background:
Evidence-based radiotherapy utilisation (RTU) models are used to estimate demand for radiotherapy and for planning radiotherapy services. These models can be further expanded to estimate the benefit of radiotherapy at the population level for individual cancer sites.
Aim:
To quantify the benefits of definitive and adjuvant radiotherapy in terms of locoregional control (LC) and overall survival (OS) in the population of gastro-intestinal (GI) cancers based on the indications of radiotherapy recommended in evidence-based treatment guidelines.
Methods:
Previously developed RTU models for oesophagus, stomach, colon, rectum, gall bladder and pancreatic cancers were merged and extended to incorporate an estimate of benefit of radiotherapy alone (RT) and in conjunction with concurrent chemotherapy (CRT). A literature review (1990-2013) was conducted to identify the 5-year LC and OS benefit proportions for radiotherapy indications in the individual cancer site models and summed to estimate the population-based gains for both outcomes. Palliative benefits were not considered.
Results:
The model predicted that the application of guideline-recommended radiotherapy in GI cancer population has overall gains in LC and OS of 3% and 1% respectively compared with a population where no patients receive radiotherapy. An additional 2% LC and 1% OS advantage was estimated with CRT to bring the total benefit to 5% and 2% respectively. This could be interpreted as an incremental benefit in 5-year LC of 1000 patients and OS of 400 patients in every 20,000 patients (average yearly number of new cases of selected GI cancers) in the Australia population.
Conclusions:
The radiotherapy survival benefit estimation applied to the cancer patient population can be utilised by the health service planners to evaluate the effect of a treatment modality, especially, the shortfall in survival expected with underutilisation of therapy for a particular population of patients and is adaptable to other populations with known epidemiological parameters.