The Health Economics Unit collaborated with Janssen, The Association of the British Pharmaceutical Industry (ABPI) and DATA-CAN: The Health Data Research UK Hub for Cancer to understand not only if cancer treatments work to save patients’ lives, but also the quality of life they leave them with.
Endpoints – or markers in the data to say that something has been successful or not – allow more informed decisions to be made around patient care. An obvious example of an endpoint in cancer treatment is whether patients die or not. However, with survival rates rising, the question we looked at was how to measure the impact of a specific intervention, such as a drug or type of surgery.
Defining and analysing other, more complicated endpoints, leads to more helpful outcomes. If someone needs to decide whether to use a certain drug or have a type of surgery, we wanted the data to be able to show whether it could also help them have a good quality of life.
Methodology / approach
The project looked to understand more about the ongoing quality of life of those who have survived from lung cancer and multiple myeloma.
Our team carried out a feasibility study, engaging with stakeholders including leading clinicians on lung cancer and multiple myeloma, data teams and NICE. We also conducted an in-depth literature search to identify and validate new endpoints that could be used for future studies.
The feedback gathered from clinicians and other stakeholders revealed a need to identify new ‘surrogate’ endpoints that could be put into the data to show how patients are recovering.
Our experienced team brought new ideas to the project, such as using machine learning to look at the data and letting it tell us what the endpoints could be. For example, the team might look at a patient who is deteriorating. By analysing the data, they can look for markers – like increased visits to hospital – to indicate if this could have been predicted. This intelligence could then be applied in future for other patients.
Our analysts apply their expertise and experience in data analysis and machine learning to create meaningful insights which can be used to inform future treatments, adding value to the original proposal. Instead of just answering the question we were given, we added our own thoughts and suggested other ways that it could be done.
The principles of our work here can be applied to other cancers and other diseases.