Modelling the Natural History of Rare Diseases Using Disparate Data Sources: Application to Patients with Duchenne Muscular Dystrophy.

Background Health technology assessment (HTA) compares the effect and value of two or more treatments for a disease.

Often, the disease in question will have a number of key phases that a patient progresses through, with different treatments/resource requirements\quality of life at each stage. These stages can be represented by a multi-state model.

To compare treatments in progressive diseases, a baseline model (termed the natural history) is estimated to represent the underlying transitions made by the average patient with standard/current treatment, to which different treatments can be compared.

This project will address the deficiencies of current methodology when modelling the trajectory of rare progressive diseases, such as Duchenne Muscular Dystrophy (DMD), as data sources will scarcely be available in either sufficient quantity or quality and current methods do not suitably account for this.

Aims The aims of this project are to: Identify methodology for combining evidence sources to construct and populate multi-state models of disease histories.

Develop the methods to ensure robustness to a variety of data sources of differing quality, as is often the case in rare disease research. Evaluate the methods by a simulation study and application to DMD data.

Evaluate the impact of the methods on economic models via a sensitivity analysis and provide recommendations for future HTA of rare diseases, and DMD in particular.

Methods I will achieve these aims with the following methods: I will first perform a literature review of the current methods for combining a variety of data types for use in multi-state models.

I will extend a review of the appropriate sources for populating natural history models to consider cases when these sources are unavailable, or are obtained from different populations.

Where these methods do not suitably account for heterogeneity between data sources, I will extend them accordingly to ensure their suitability in the context of rare diseases. I will create freely available software packages implementing new methodology for use by fellow researchers.

I will apply the methods to simulated data representing different real-world possibilities of evidence availability, and to a variety of DMD data, evaluating sensitivity and robustness.

I will use the developed natural history modelling to fit health economic models, comparing HTA outcomes and conclusions from the different methodologies. I will make recommendations on future study design and analysis of rare diseases.

Impact The methods developed will be applicable to future evidence synthesis of rare diseases, which will greatly assist HTA for such diseases.

They will be disseminated to researchers via publications, conference presentations and a workshop at the end of the fellowship.

The work will assist methodological and statistical developments, and software packages will enable implementation of the methods in HTA decision-making, which will translate into a benefit in health care for patients that suffer from numerous rare diseases.

The work will also be regularly disseminated to Duchenne UK Project HERCULES, a multi-stakeholder collaboration for DMD, and to broader settings such as the European Rare Diseases Conference, who will see the substantial benefits of the project from their perspectives.