UCL Institute of Health Informatics


Phenotyping Methods for Linked EHRs - CALIBER

Primary and Secondary care records are increasingly being linked for use in research. These data, however, are collected as part of routine care or for administrative purposes and a significant amount of work is required to build robust and accurate definitions of clinical concepts that can used to identify cases for further study.

In this course we present the basic theory behind the extraction of phenotype data from combined data resources such as CALIBER.

Learning Objectives

Participants should, by the end of the course: 

  • become familiar with two contemporary primary and secondary data sources (CPRD and HES);
  • learn what types of EHR data are collected and the different ways in which data are recorded;
  • understand how to combine linked EHR data sources to define disease cases. 

Planned Timetable

The following timetable is provisional and it might change closer to the date. 
09:00-09:30Registration and coffeeDr Arturo Gonzalez-Izquierdo
09:30-10:30Phenotyping electronic health records for research - theoryDr Arturo Gonzalez-Izquierdo
10:30-10:45Case study - Phenotyping CancerDr Costantinos Parisinos
11:00-11:45Phenotyping COPDDr Jennifer Quint
11:45-12:15PracticalDr Arturo Gonzalez-Izquierdo, Dr Jennifer Quint, Dr Costantinos Parisinos, Dr Spiros Denaxas, Dr Ghazaleh Fatemifar
13:15-14:15PracticalDr Arturo Gonzalez-Izquierdo, Dr Jennifer Quint, Dr Costantinos Parisinos, Dr Spiros Denaxas, Dr Ghazaleh Fatemifar
14:15-15:15Validating COPD phenotype across linked data sourcesDr Jennifer Quint
15:15-15:30Case study - Phenotyping myocardial infarction (UK Biobank)Dr Ghazaleh Fatemifar
15:45-16:30Alternative methods - next generation phenotypingDr Spiros Denaxas
16:30-16:45Review of the dayDr Arturo Gonzalez-Izquierdo
16:45- 17:00Q&A sessionDr Arturo Gonzalez-Izquierdo, Dr Jennifer Quint, Dr Costantinos Parisinos, Dr Spiros Denaxas, Dr Ghazaleh Fatemifar

Course Team

Dr Arturo Gonzalez-Izquierdo

Arturo Gonzalez
Arturo has a background in Statistics and came to the UK to specialise in the areas of Epidemiology and Public Health (MSc) and Biostatistics (PhD).

In the course of his PhD at Imperial College London and work at UCL over the last nine years, Arturo has gained expertise in studies related to the epidemiology of disease, the identification of patient populations and definition of patient classifications, and the understanding of healthcare utilisation and provision, with particular emphasis on the curation and analysis of large collections of data generated during the delivery of healthcare at national levels.

Dr Jennifer Quint

Jennifer is a Senior Lecturer in Epidemiology at the London School of Hygiene & Tropical Medicine (LSHTM) and an Honorary Consultant in Thoracic Medicine at University College London Hospitals (UCLH) NHS Trust.

Her clinical and research interests focus mainly on COPD and the relationship with cardiovascular disease and exacerbations. This has included validating the identification of COPD and exacerbations of COPD in electronic health records.

She is currently funded on a Medical Research Council Population Health Scientist Fellowship to investigate the causal roles of environmental factors and infections on exacerbations of COPD.

As well as her research she teaches and tutors on the MSc Epidemiology course at LSHTM and as part of her clinical commitment is involved in the COPD service both at UCLH and the local Integrated Community Service. 

Dr Costas Parisinos

Costas Parisinos
Costas is a Wellcome Trust Clinical Research Training Fellow (UCL) and a Specialist Registrar in Gastroenterology & Hepatology (Barts Health NHS Trust).

He is interested in the integration of electronic health records with genomic data, to better understand and prevent gastrointestinal disease and malignancy.

Dr Ghazaleh Fatemifar

Ghazaleh has a background in genetic epidemiology. Her research is focused on using genome-wide association studies to identify instruments for Mendelian randomisation.

She completed her PhD at the University of Bristol before joining the Farr Institute of Health Informatics in October 2015.

Her research interests include genetic epidemiology, Mendelian randomisation and machine learning. 

She was recently awarded an American Heart Association fellowship to work on clustering heart failure patients using unsupervised machine learning algorithms in contemporary electronic health records linked to genetic data.

Dr Spiros Denaxas

Spiros Denaxas
Spiros is Associate Professor in Biomedical Informatics at the Institute of Health Informatics.

His research focuses on Electronic Health Record (EHR) phenotyping methods for translational research. He also manages CALIBER, a linked EHR data warehouse of 2 million adults in the UK.