Improving the Translation of Diagnostic Test Accuracy Research Into Real‑World Clinical Practice Using Routine Health Data and Advanced Meta‑Analysis Methods
Not stated
- Funding
- Self-Funded PhD Students Only
- Application deadline
- Year-round applications
About the project
About the Project Diagnostic tests are central to effective clinical decision‑making, yet the performance reported in research studies often differs markedly from that observed in everyday NHS settings. Variability in disease prevalence, workflow, patient presentation, and clinical thresholds can significantly influence sensitivity, specificity, and predictive values. This PhD project offers an exciting opportunity to develop new methods that bridge the gap between controlled diagnostic research environments and real‑world clinical practice. Under the supervision of Dr Brian Willis, an award‑winning clinician‑scientist whose work in diagnostic test methodology and medical statistics has helped shape international guidance you will explore how routine health data can be harnessed to better understand and predict diagnostic test performance across varied clinical contexts. His background spans medicine, mathematics, biomedical engineering, and advanced biostatistics, providing an exceptional interdisciplinary environment for methodological innovation. Research Aims The project aims to create integrated statistical frameworks that combine traditional diagnostic accuracy evidence with real‑world NHS data. You will examine how clinical factors such as disease prevalence, patient flow, and test‑ordering behaviour modify test accuracy, and will develop methods to recalibrate diagnostic performance estimates for use in everyday practice. The ultimate goal is to provide clinicians with more reliable, context‑specific information to guide diagnostic decision‑making. Methodological Approach You will undertake: Systematic reviews of diagnostic accuracy literature Development of advanced meta‑analytic and regression models Analysis of large routine datasets (e.g., UK primary‑care EHR records) Validation and translation of findings into practical clinical tools This project is ideal for candidates with strong quantitative skills and an interest in diagnostics, biostatistics, epidemiology, or applied health research. Training & Environment You will join a vibrant research community within Applied Health Sciences, with training in advanced statistical methods, evidence synthesis, and the use of large‑scale health databases. Dr Willis’s extensive experience in NHS‑relevant diagnostic research ensures high‑quality mentorship and clear pathways to impactful publications.