Download files
Abstract
Background: Adequate radiotherapy services provision entails systematic planning due to their high capital costs and the requirement for specialised staff. A treatment attendance (called a fraction) is a fundamental unit of productivity in a radiotherapy department. There is variation in radiotherapy fractionation practices, however, there is no evidence-based benchmark for appropriate activity. A radiotherapy utilisation model was previously constructed and estimated that 52.3% of cancer patients should receive external beam radiotherapy at least once during their illness. The next challenge is to translate an overall radiotherapy utilisation rate into a more practical estimate of radiotherapy demand.
Aim: To construct an evidence-based model to estimate the optimal number of fractions for the first course of radiotherapy, building on the radiotherapy utilisation model.
Methods: Evidence-based treatment guidelines, meta-analyses and randomised controlled trials were reviewed for fraction number recommendations for each indication of radiotherapy for notifiable cancers with an incidence of ≥ 1%. The previously published radiotherapy utilisation tree was adapted so that the most appropriate evidence-based fraction number was added to each branch. Epidemiological data previously used were updated. For each cancer type, the optimal fraction number was then calculated using the TreeAge software, taking into account the frequency of specific clinical conditions where radiotherapy is indicated and the recommended fraction number for each condition. One-way sensitivity analyses were performed to assess the impact of uncertainties on the model.
Results: For each cancer type, the optimal number of fractions for the first course of radiotherapy ranged from 0 to 26.1 per cancer patient, and 0 to 30.8 per course. Head and neck, brain and anal cancers had the highest number of fractions per course. Overall, the optimal fraction number was 9 per cancer patient and 18 per course. Sensitivity analysis showed that this ranged from 8.6 to 9.6 per cancer patient, and 17.2 to 19.2 per course.
Conclusion: These results represent the first evidence-based benchmark for radiotherapy services delivery, and allow comparisons with actual practices. The model can be used to predict workload to aid in radiotherapy services planning, and adapted to future changes in cancer incidence, stage distribution and fractionation recommendations.