Early mortality outcomes and cost-effectiveness of palliative radiotherapy for bone metastases in the English NHS

Patients with advanced cancer often have symptoms which limit their quality of life. Palliative radiotherapy is widely used to treat many of these symptoms and over 65,000 treatment episodes are delivered each year within the English NHS. Clinical trials have demonstrated that for many patients one single treatment or a short-course works just as well as daily treatments over one to two weeks. In some situations though, longer treatments, delivering higher doses, are needed to improve the disease or symptom control; these can cause more side-effects and inconvenience and limited evidence exists to support this approach. For patients with a limited life expectancy doctors must weigh up the benefits versus the side-effects and inconvenience of treatment; patients who die soon after treatment may not benefit from radiotherapy and longer courses may take up a significant proportion of their remaining life. Such treatments can be considered an avoidable harm and a suboptimal use of healthcare resource. Increasingly, newer radiotherapy technologies, capable of delivering more targeted treatment, are becoming available for the management of metastatic disease such as stereotactic ablative body radiotherapy (SABR). These allow shorter treatments, with higher doses; a potentially beneficial approach. However, the extra benefit of these higher doses has not yet been established in randomised trials and, whilst the treatment burden for patients from SABR is potentially reduced, the cost of treatment is significantly greater and case selection is, therefore, important. It is vital to determine when such treatments are best used, in terms of clinical and cost-effectiveness, not only to guide patient care but also to inform the development of services and future trials in this area.

Pilot work I have undertaken using a local, seven year cohort of radiotherapy treatments, including 14,972 palliative treatment episodes, indicated that 42% of all palliative radiotherapy treatments were delivered to bone metastases. The national Quality Innovation Productivity and Prevention (QIPP) program recommends that 70% of palliative bone treatments should be delivered with a single fraction. However, variation in practice still exists across the English NHS. There is good evidence that single treatments improve symptoms as effectively as longer courses. However, concern about possible increases in the risk of re-treatment and pathological fractures after single treatments may be influencing clinical decision-making, resulting in the variation seen both nationally and internationally. There are currently no guidelines to define when a longer course is appropriate and what is the best treatment; the balance of the costs and treatment burden with the potential benefit, is an under-researched area. Feedback to doctors is also limited and such data could help to inform future decision making, reducing avoidable harm in palliative radiotherapy.

In an under-researched area of oncology, this project aims to formally assess the variation in practice and early mortality following palliative radiotherapy across the English NHS, allowing delivery of clinically useful outcome measures to treating clinicians. Having developed this methodology it will be possible to provide regular feedback. The project will then model 'acceptable' levels of early mortality, taking into account the benefits to patients and the costs of treatment. It will also help to define the potential role of SABR in this setting. As a result it will help to improve the use of palliative radiotherapy, by ensuring the NHS gets the best possible value for money from these treatments and, most importantly, reduce avoidable harm to patients nearing the end-of-life.

30-day mortality (30DM) has been suggested as a clinical indicator for the avoidance of harm in palliative radiotherapy within the NHS. However, no national analysis has been carried out and acceptable outcomes are not known. Alongside this is the increasing use of SABR; a significantly more costly strategy. This project aims to determine acceptable outcomes and optimal treatment strategies both in terms of avoidable harm and cost-effectiveness through four workstreams:

1. Analysis of the variation in fractionation patterns and 30DM following conventional palliative radiotherapy across the English NHS using a 2012-13 cohort of linked data from the National Cancer Data Repository (approx. 120,000 episodes). Factors affecting 30DM will be analysed using multiple logistic regression and clinically useful indicators developed, considering use of both crude and adjusted rates for feedback to centres. This will be carried out in the MRC Medical Bioinformatics centre, University of Leeds.
2. Systematic review of the literature for the outcomes of SABR for uncomplicated bone metastases.
3. Development of a decision analytic model to assess the cost-effectiveness of a range of treatment strategies for uncomplicated bone metastases; best supportive care, single fraction and fractionated conventional radiotherapy and SABR. This is likely to use a Markov model with a cycle time of one week, informed using standard methodologies, alongside the outcomes of the first two workstreams, collaborations and a patient focus group to inform the impact of the treatment process near the end-of-life. A threshold sensitivity analysis for survival will be carried out to determine the point at which each strategy becomes cost-effective over its closest rival. A deterministic sensitivity analysis will assess the impact of other parameters on the model.
4. Application of the model outcomes within a regional budget impact tool to assess the impact of changes on both finance and capacity.

Patients and public:
Patients receiving palliative radiotherapy towards the end of life will benefit from this work which aims to reduce the avoidable harm that may occur due to overly aggressive treatments in this setting. This will help to improve the quality of life of patients receiving palliative radiotherapy. More broadly, and linked to this, the optimisation of healthcare delivery can help to ensure adequate funding is available to other patients receiving any treatment within the NHS. Feedback of the results from the first phase of this work will be delivered within the three years of this fellowship. Charities supporting cancer patients and their relatives (e.g. Macmillan, Marie Curie, diagnosis specific charities) will also value this work which aims to improve the quality of life of cancer patients.

Commissioners and policy makers:
Prior to making policy and commissioning recommendations, it is vital to determine the optimal use of palliative radiotherapy in terms of the balance between health outcomes and the consumption of healthcare resource. Efficient allocation of resources is a high priority. Sub-optimal allocation denies adequately funded healthcare to other patients, particularly in the case of radiotherapy where infrastructure and capacity planning can have long-term repercussions. Policy-makers involved in national radiotherapy capacity and strategy planning will benefit from this work, which will give not only a clear assessment of baseline palliative radiotherapy use across the English NHS but can also guide the development of clinically appropriate, cost-effective outcome targets with the expected costs/savings associated with their delivery. This will help to improve palliative radiotherapy services nationally, improving patient care. This project directly contributes to ongoing endeavours to improve the cost-effectiveness of cancer care. The methodologies used within this project can be used to carry out similar analyses within other healthcare systems globally where the variation in fractionation patterns may vary more widely and healthcare funding differs.

Radiotherapy Centres and Clinical Oncologists:
Radiotherapy Treatment Centres delivering palliative radiotherapy around the country will be able to use the feedback received from workstream one to consider how their use of palliative radiotherapy aligns with their peers. This feedback can contribute to changes in local practice. This can deliver benefits both from a service perspective but also importantly from a patient perspective; reducing the avoidable harm associated with overly aggressive treatment near the end-of-life. Centres aiming to deliver service improvements will value the outcomes of workstream four which can inform capacity and strategy planning.

Individual clinicians will gain feedback of outcomes seen for their disease site within their centre. It may be possible to deliver clinician-specific outcomes in future as more contemporaneous, data becomes available. Such feedback could guide individual's future treatment decisions.

Commercial providers:
Commercial providers of SABR will be interested in this analysis as it will deliver an assessment of the incremental cost-effectiveness of SABR and potentially the cost reductions needed from this to be considered a cost-effective strategy for the treatment of uncomplicated bone metastases within the NHS. This is of value to business planning.


This fellowship will equip me with the skills necessary to develop a future career in Health Services Research. As a clinician with an understanding of the healthcare environment and patient pathways I will develop valuable additional skills which will be used in future, cross-discipline, projects and collaborations. Further opportunities to collaborate at a local, national and international level will also be developed.