Location: Sustainability Hub, Kirkby Lodge, University of Plymouth, Plymouth, PL4 8AA

9:30am: Registration/coffee

Invited talks

Contributed Talks/Posters (spaces still available see below for details)

Opportunities to find collaborative partnership

Immersive Vision Theater Show

Lunch and refreshments provided

4.30pm close

Registration:  https://www.eventbrite.co.uk/e/impact-labrss-south-west-health-data-science-network-event-tickets-92180314803

Invited Talks:

Rachel Phillips (Imperial College London) An evaluation and application of statistical methods designed to analyse adverse event data in RCTs: An exploration of methods to analyse prespecified secondary harm outcomes and non-specific emerging adverse events. A review of current methods found 73 proposed in the literature, a simulation study is conducted on two of these methods to show adequate power to detect signals for harm.

Harry Parr (The Institute of Cancer Research): Development of Clinical Dynamic Prediction Models to Characterise Prognosis of Localised Prostate Cancer Patients: Using shared-parameter joint models on the CHHiP patients to predict prognosis, providing insights into response to treatment and likelihood of reoccurrence to create bespoke treatment plans.

Sarah Littler (Select Statistics): Developing Medical Decision Tools: Working with an NHS Foundation Trust, analysing historical patient extubation data, we combined factors related to clinical assessments within a statistical model to predict the chance of successful extubation for new intensive care patients. The model was used to identify the most important predictive factors for success and to suggest a cut-off probability beyond which a clinician would be encouraged to extubate. The model provided a tool to help guide clinicians in their decision as to whether to extubate their patient. This published research contributed to improving the objectivity of extubation decisions and, in particular, to helping clinicians with less experience make the right decisions for the patients in their care.

Contributed Talks:

Ben Jones (University of Plymouth): Information borrowing from pilot data in Cluster Randomised Controlled Trials using Bayesian Power Priors: Development of methodology to adaptively borrow information from historical data to improve analysis, and facilitate efficient design of cluster trials.

Lexy Sorrell (University of Plymouth): Using the restricted mean survival time to estimate patient and graft survival following kidney transplant: an application to the national renal database. The difference restricted mean survival time provides the comparison between survival in patients in the time scale compared to the standard risk scale, this may be more intuitive for patients and clinicians. We compare survival between living and deceased donor recipients of renal transplants.

Jade Chynoweth (University of Plymouth): Improving the efficiency of modelling complex physical activity data using novel statistical methods – a literature review: A review of compositional data analysis methods applied to longitudinal data, with a particular focus on accelerometry data.

Lauren Stockley (University of Plymouth): Statistics in Well-Being: A principal component analysis to combine the four subjective wellbeing measures used by the ONS to understand wellbeing in the UK is undertaken, followed by additional analyses to identify variables associated with this combined measure.

Joe Lomax (University of Plymouth): Demographic Representativeness in NHS Trusts: An exploration of demographic representativeness of NHS staff in England compared to the communities they serve, and the impact of this on the perceived performance of the trusts in which they are employed.