The wealth of biological, clinical and epidemiological data now available, coupled with the resolution afforded by current high throughput technologies and the longitudinal follow-up of individuals, has led to a shift towards ever-finer patient/participant stratification to inform clinical practice. Building on the Unit’s expertise in clustering, dimension reduction, longitudinal and risk prediction modelling, causality and addressing the challenges of analysing and integrating complex and varied data types, members of the PREM Theme will develop and apply statistical methods to (i) uncover biologically and clinically meaningful subpopulations of individuals within heterogeneous disease and clinical populations; and (ii) more accurately and differentially diagnose, monitor and predict disease course, risk and response to treatment.
PREM will focus on the methodological issues arising from the development of precision medicine. These include the need to stratify complex disease phenotypes, to perform integrative clustering based on multiple data types, to tailor risk prediction, to estimate treatment effects and to analyse and make subject-level inference from complex longitudinal data. Our work is directed towards better understanding of mechanisms of disease, informing of patient care and management and patient selection into trials, and guiding therapeutic development and application. Our aspiration is to realise the potential of precision/stratified medicine and ensure that ‘the right patient gets the right treatment at the right time’.
Precision/stratified medicine is currently identified by the MRC (and other organisations) to be of strategic importance. There has been a large investment in this area in the UK (e.g. through the MRC Stratified Medicine Initiative), Europe (e.g. through the Innovative Medicines Initiative 2) and worldwide (e.g. The Precision Medicine Initiative in the USA). Common disease areas such as dementia, oncology, inflammatory and autoimmune disease and infectious disease, which individually have a large societal impact/burden, and rarer diseases, which collectively affect a significant proportion of the UK population, have been targeted. In these areas there is significant unmet clinical need and the necessity for patient stratification is recognised. The MRC Biostatistics Unit is committed to developing statistical methodology under this strategic research agenda and is actively engaged in forming, fostering and sustaining partnerships with researchers/stakeholders working in precision medicine. We have major involvement in a number of ongoing precision medicine projects in anti-microbial resistance, cardiovascular disease, cystic fibrosis, delirium, dementia, hepatology, and rheumatology that motivates much of our methodological work. These projects and others will form a test-bed in which our newly developed and generic statistical methodology can be applied after tailoring to the specific substantive problem.
(Previous name of theme: Methods for the Analysis of Complex Observational and Longitudinal Data – COLD)