Background: Globally, healthcare-associated infections (HCAI) continue to occur despite broad IPC measures. Our imperfect understanding of pathogen transmission likely contributes. Consequently, reactionary resource-intensive approaches are required to investigate outbreaks of infection. A higher level of detail is required to understand pathogen spread to support optimising effective and sustainable IPC interventions.
Studies have revealed the high-resolution offered by whole-genome sequencing (WGS) provides detailed characterisation of healthcare outbreaks and uncovers cryptic transmission events. Now translational studies are needed to determine whether WGS can optimise IPC interventions and whether clinical data can support prediction tools to allow effective targeted use of resources.
Methods: The studentship will take place in two parts:
Part A: The student will conduct a prospective clinical study to investigate common HCAI (S. aureus, E.coli, K.pneumoniae, P.aeruginosa). At University Hospital Sussex (UHS) routine clinical isolates will be prospectively sequenced. Real-time genomic relatedness reports will be generated in response to clinically-identified outbreaks and cryptic transmissions determined through genomic investigations. Working with IPC specialists the student will characterise whether WGS data leads to actionable changes in IPC including escalation, de-escalation or further investigation (enhanced environment/patients/staff sampling).
Part B: Using data collected from Part A the student will develop models to predict targeting effective sampling (patients/staff/environment) in outbreaks and develop dynamic graph machine-learning models to predict real-time utility of targeted WGS. The student will learn, and in turn consider, the evolving importance of predictive features using incremental machine-learning. Model evaluation will occur using (i) retrospective data collected during Part A and (ii) prospective data coupled with clinical IPC practices.
Research Plan: In Year 1 the student will conduct a systematic review of prediction tool applications in IPC, lead the clinical study (Part A), and develop modelling skills through supervision and course attendance. In Year 2 the student will analyse the clinical study, complete the retrospective prediction study, develop/instigate prospective predictive modelling evaluation and presentation preliminary work at a conference. In Year 3 the student will complete the prospective prediction study, finalise analyse, and submit publications/thesis.
Inclusivity: Patient groups have been involved in project development. All patients across UHS will be suitable for recruitment. Collaboration with bioartist will support public engagement of results.
Interviews will take place on Friday 18 August.
Continue with Facebook
