Overview

I am a Health Services Researcher working within Primary Care, and particularly Cancer Epidemiology, at the University of Exeter Medical School.  I have expertise in methodology, especially applied statistics, and in the management and analysis of large databases of routine records to answer health research questions.

I am the lead Statistician on the Electronic Risk Assessment for Cancer (ERICA) cluster-randomised controlled trial, exploring whether electronic risk assessment tools (eRATs) embedded into GP clinical software improve cancer diangosis outcomes.  Specifically, through use of the English national cancer registration data, we are investigating whether eRATs improve the proportion of cancers diagnosed at an early stage.

As a methodologist, I have work across various groups within the University of Exeter Medical School.  I am jointly shared between Professor Willie Hamilton in Cancer Diagnostics, and Dr Gary Abel in Primary Care.  I am also affiliated with the Health Statistics Group, and Health Services and Policy Research Group.  Previously I worked in the Epidemiology and Public Health group. 

Before joining the medical school in 2012, I earned a BSc, MSc and PhD in Psychology at the University of Exeter, with my PhD focussing on the longitudinal modelling of health outcomes following traumatic brain injury.

Qualifications

BSc (hons) in Psychology (Exeter, UK)

MSc in Psychological Research Methods (Exeter, UK)

PhD in Psychology (focus: clinical neuropsychology) (Exeter, UK)

Graduate Certificate in Statistics (Sheffield, UK)

Research group links

• Institute of Health Research

Back to top

---

Research

Research interests

I am a methodologist, and seek to translate clinically important research questions into a scientifically rigorous research protocol, and then to carry that out.  As such, I work across a range of fields within health research.  My work has focussed particulary on cancer epidemiology, the causes and consequences of having multiple heatlh conditions (multimorbidity), and patient experience of Primary Care.

I have considerable expertise in the use of routine records, including:

• The Clinical Practice Research Datalink (CPRD), the largest database of electronic primary care records in the world.
• The National Cancer Registration and Analysis Service (NCRAS) cancer registration data, covering over 99% of all cancers diagnosed within England.
• Hospital Episode Statistics
• The GP Patient Survey (GPPS), the largest survey of patient experience of Primary Care in the world, with 2.7 million annual responses.
• The English Longitudinal Study of Ageing (ELSA) database, a national panel survey of adults aged 50 and over, with waves every two years since 2002.

Research projects

The ERICA trial

I am the lead Statistician on the Electronic Risk Assessment for Cancer (ERICA) trial.  ERICA is a pragmatic, cluster-randomised controlled trial (cRCT) that is currently recruiting GP practices to be involved in an investigation of the possible benefits of electronic risk assessment tools (eRATs) for cancer.  The eRATs are embedded into GP clinical software and flag when a patient's recorded symtpoms / test results correspond to a cancer risk of >2% to aid clinical decision making.  The primary outcome measure for the trial, assessed using routine data from the English national cancer registry, is the poroportion of cancers diagnosed at an early stage at the practice-level.

Reliability of a process indicator for quality of cancer care

This project is exploring the possibility of using the percentage of eligible patients with alarm symptoms of cancer who are given an urgent referral for investigation of cancer (a "two week wait" referral) as a process indicator for the quality of cancer care provided at the practice-level.  It will:

1. Explore varaibility in this indicators at the practice-level and clinician level.
2. Investigate whether the indicator can reliably distinguish between high-performing and low-performing practices.
3. Explore who performance on the indicator is affected by a practice's case-mix (e.g. the proportions of different cancers diagnosed).

This project is funded by the Moore Foundation, and is being conducted in collaboration with the University of Houston, Texas, and University College London.

The Community Groups for Post-Stroke Support (CoGS) study

This project, funded by the Stroke Association, is exploring the possible benefits of belonging to a community group for post-stroke support.  In particular, it is exploring the impact of group memebership on loneliness and well-being.

Cancer incidence following a high-normal platelet count

I recently completed the role of lead analyst on a project in cancer diagnostics titled "Are patients with high normal platelet counts at increased risk of cancer? The NORMA study."  Previous work has shown that patients with high platelet counts (thrombocytosis) are at much higher risk of cancer than the 3% criterion indicated in NICE guidelines for referral for investigation for suspected cancer.  The aim of NORMA, therefore, was to investigate the risk of cancer in patients at the high-end of the normal range of a platelet count.  This project involved analysing the anonymised electronic health records of ~300,000 patients, contained in the Clinical Practice Research Datalink with linkage to national cancer registration data, to explore how cancer risk varies by age, gender, and platelet count. We found that men aged 60 and over with a platelet count at the high end of the normal range had cancer incidences exceeding the 3% risk threshold in national guidance, though no female subgroup met the threshold.  Regardless of gender, a high-normal count was associated with up to 5 times the risk for lung cancer, and 4 times the risk of colorectal cancer.  Patients with a high-normal count diagnosed with cancer were 50% more likely to be diagnosed at an advanced stage. NORMA is funded by the Policy Research Unit.

---

I am involved in projects covering a number of other work-streams:

Use of chest x-ray for detecting lung cancer

I am working in collaboration with Stephen Bradley and colleages at the University of Leeds, providing methodological support to several projects exploring the test-characteristics of chest x-ray in the detection of lung cancer.

Use of Cancer antigen 125 for detecting ovarian cancer

I am collaborating with Garth Funstun and colleagues at the University of Cambridge on projects exploring methods for determining CA125 levels, and the use of CA125 in the early detection of ovarian cancer.

Patient Safety

Developing short-form versions of the Patient Reported Experiences and Outcomes of Safety in Primary Care (PREOS-PC) questionnaire.  PREOS-PC is a recently developed patient reported outcome measure (PROM) of experiences safety in Primary Care, consisting of 54 items and 11 scales.  This project is developing a compact version of the instrument, retaining the scales, but using a reduced number of items, as well as a short 6 item screening tool.  This process is being informed by use of both Classical Test Theory and Item Response Theory, and uses data collected for the development of the original measure.

Research networks

I am an associate member of the CanTest Collaborative, an international team of primary care cancer researchers working on early detection and diagnosis of cancer in Primary Care.

Back to top

---

Publications

Back to top

---

Teaching

I was an Associate Teaching Fellow for four years during my PhD in Psychology.  Since joining the Medical School, I am a tutor on the Making Sense of Evidence module for BMBS students, and demonstrate statastics as part of the Fundamental Skills for Medical Scientists module on the Medical Science bachelors degree.

Back to top

---