Economic modelling of diagnostic/treatment pathways in NICE clinical guidelines

NICE clinical guidelines state which tests and treatments should be offered to patients by the NHS. These services should provide value for money, as well as effective care for patients. But health economists who work on guidelines do not have the time to assess the costs and health benefits of every type of service that could be provided. Instead they select services that are likely to be important ? because the costs or health effects are large - and where guideline groups are unsure of the right choice. But it is possible that this may miss important links between the services. For example, a test will only be useful if followed by the right treatment. And treatment will only work if it is given to the right patients. So the value of tests and treatments may depend on the order or combination in which they are offered to patients. In this study we want to test whether it is possible to assess a system of services recommended in a guideline ?pathway?, rather than looking at single services in isolation. To make things easier, we will start by testing our methods on two existing guidelines which are due to be updated soon. For these guidelines, the pathway of recommended services has already been decided and the evidence about these services already collected together. We will build computer models that will predict the health outcomes for patients and costs to the NHS that are likely to arise if the guidelines are followed. When NICE is deciding whether to update a guideline, it looks to see if any new evidence is available and considers whether there are parts of the guideline that it should look at again. We will collect information that arises during this process, and ask experts and patient representatives to rate the importance of updating suggested topics. We will then use the computer models to estimate the likely value of reconsidering these topics. The results of this analysis will be given to the experts and patient representatives, and they will be asked to consider whether this changes their views. In summary, we want to see if we can model the pathways of care that are recommended in NICE clinical guidelines, use the models to see if NICE should reconsider parts of the guidelines, and see if this adds anything to the current way that NICE reviews its guidelines.

NICE clinical guidelines make recommendations about the diagnosis and care of people with specific diseases and conditions in the NHS in England and Wales. In keeping with the principles of decision-making for NICE advisory bodies, these recommendations should take account of the cost-effectiveness of alternative services, as well as their clinical effectiveness. However, health economists working on clinical guidelines cannot formally evaluate cost-effectiveness for all services under consideration, but instead prioritise selected issues for modelling work. It has been argued that evaluating specific healthcare interventions in isolation may miss important interconnections between services within a broader pathway of care. This project will test whether a more systematic approach to economic modelling can be applied in NICE clinical guidelines. For simplicity, we propose to test this approach initially by modelling two existing guidelines, for which the pathways of care are well-articulated and systematic reviews of evidence are available, but which are shortly due for update. We will build simulation models to estimate the health outcomes (QALYs) and costs associated with the current recommended pathways. We will then collect information about new evidence and suggestions for changes to the pathways that arise during the standard updating process, and people involved in this process will be invited to rate the suggested topics in terms of priority for inclusion in a guideline update. The models will be used to estimate the incremental cost-effectiveness of possible variations to the pathway. A value of information approach will also be used to estimate the potential net benefit from reducing uncertainty over selected model parameters. The topics that are identified as priorities for updating through pathway modelling ? those with a high estimated net benefit and/or high value of information across the relevant population ? will be compared with the priorities obtained from participants in the standard update process. These people will be presented with information about our modelling work and invited reconsider their previously-stated priorities in the light of this information. This will allow us to test: a) whether modelling of NICE guideline pathways is feasible; b) how such models can be used to evaluate the incremental cost-effectiveness of possible variations to the pathway; b) whether this approach can be used to prioritise topics for update; and c) whether the resulting priorities differ from those elicited during the standard update process.