Towards equity in medicine with big health data, epidemiology, and artificial intelligence

PROJECT OVERVIEW

The Covid-19 pandemic has highlighted inequalities in healthcare systems around the world. Inequalities based on ethnicity are of concern because most of our understanding of many diseases comes from White or Caucasian populations whose risk factors, disease prevalence and incidence can differ from other ethnic groups. Furthermore, those patients with rare diseases and disabilities also endure a lack of equity in the healthcare they receive. In addition to these complexities, imbalances in healthcare technologies such as prediction models can worsen these existing biases. Lack of representation of certain ethnic groups, those with rare diseases or with certain disabilities can create biased artificial intelligence models resulting in a mis-estimation of the health risks leading to groups of patients being under- or over-prioritised for the best care.

We can use real world data such as electronic medical records and genetic data with epidemiological and artificial intelligence methods to obtain more reliable answers regarding equity in healthcare through integrating results from these different approaches. Our group has analysed ethnicity records for over 250 ethnicity groups from >60 million individuals in the UK. This 3-year DPhil project will generate evidence from real world data to describe, analyse and develop prediction models for patients with different ethnic backgrounds, patients with rare conditions and patients with disabilities. The project can involve a variety of descriptive, epidemiological and prediction modelling techniques as well as Mendelian randomisation methods. Patient and public engagement and involvement will also be an important element of this research. There will be flexibility for a student to focus on a specific area of equity specified above based on the expertise within the supervisory team, with preference for cardiometabolic, mental health, cancer, and musculoskeletal diseases within different ethnic groups.

By focussing equity of healthcare, we can provide evidence that can help guide healthcare providers in therapeutic decision making to prioritise the best treatments for the management of diseases in underrepresented groups leading to better quality healthcare in those who need it the most.

TRAINING OPPORTUNITIES

Alongside departmental training opportunities listed below we will ensure hands-on training in real world data analysis and machine learning using medical records and genetic data from the Pharmaco- and Device epidemiology research group and the Planetary Health Informatics group well as in patient and public engagement for research. This interdisciplinary research group contains a variety of students and post-doctoral researchers with expertise in health data science, epidemiology, and machine learning. The student will work on their unique project within an experienced and collaborative supervisory team. The student will also be embedding within our international European Health Data & Evidence Network (EHDEN) and Observational Health Data Sciences and Informatics (OHDSI) networks to ensure additional analytical guidance, training and support. A student would be supported to attend relevant conferences to enrich their studies and financial support will be made available for travel to conferences.

The Centre for Statistics in Medicine is committed to improving the standard of health research through research and training on research and methods development, and home to the UK Equator Centre. This enables and encourages research and education to champion transparent and complete reporting of health research through reporting guidelines and training provision. A core curriculum of lectures will be taken to provide a solid foundation in a broad range of subjects including statistics, epidemiology, and big data analysis. All students will be required to attend a 2-day Statistical and Experimental Design course at NDORMS and the Real World Epidemiology: Oxford Summer School.  Students will also be required to attend regular seminars within the Department and have access to a variety of other courses run by the Medical Sciences Division Skills Training Team and the wider University, such as the UK Equator Publication School and Observational Health Data Sciences Course. Finally, the student(s) will be expected to regularly present data in Departmental seminars, collaborators across the University and internationally.

KEY PUBLICATIONS

1)     Pareek M, Bangash MN, Khunti K et al. Ethnicity and COVID-19: an urgent public health research priority. Lancet; 395(10234) :1421-1422 (2020). doi: 10.1016/S0140-6736(20)30922-3.

2)     Thomasian, N. M., Eickhoff, C. & Adashi, E. Y. Advancing health equity with artificial intelligence. Journal of Public Health Policy 42(4): 602-611 (2021).

https://doi.org/10.1057/s41271-021-00319-5

3)     Khalid, S. et al. A standardized analytics pipeline for reliable and rapid development and validation of prediction models using observational health data. Computer Methods and Programs in Biomedicine 211: 106394 (2021). https://doi.org/10.1016/j.cmpb.2021.106394

4)     Lawlor, D. A., Tilling, K. & Smith, G. D. Triangulation in aetiological epidemiology. International Journal of Epidemiology 45: 1866–1886 (2016).

https://doi.org/10.1093/ije/dyw314

KEYWORDS

Real world data, machine learning, genetics

HOW TO APPLY

It is recommended that, in the first instance, you contact the relevant supervisor(s) and the Graduate Studies Office (graduate.studies@ndorms.ox.ac.uk), who will be able to advise you of the essential requirements.

Interested applicants should have, or expect to obtain, a first or upper second-class BSc degree or equivalent in a relevant subject and will also need to provide evidence of English language competence (where applicable). The application guide and form is found online and the DPhil (aka PhD) will commence in October 2024.

Applications should be made to the following programme using the specified course code:

D.Phil in Clinical Epidemiology and Medical Statistics (course code: RD_NNRA1) 

For further information, please visit http://www.ox.ac.uk/admissions/graduate/applying-to-oxford.