Centre for Electronic Health Records research And Methodology (CEHRAM)
What we do
The team includes epidemiologists and statisticians who work on a range of clinical and methodological topics in collaboration with colleagues in London, UK and around Europe.
Using different data sources and methods we work on topics such as:
- Does antidepressant treatment in pregnancy increase the risk of congenital heart defects?
- What is the incidence and prevalence of type 2 diabetes in the UK and how is it pharmacologically managed?
- Is it possible to predict postnatal depression using pregnant women's electronic health records?
- How many people are prescribed antibiotics in a year, and how many antibiotics are they prescribed?
- Does diabetes mellitus increase the risk of caesarean section?
- Does smoking increase the risk of schizophrenia?
- Common infections in patients prescribed systemic glucocorticoids.
- How women use primary care services after childbirth.
- New methods to deal with missing data in electronic health records.
- Innovative study designs and analytical methods to answer clinical/epidemiological questions with routinely collected data
- These methods include machine and deep learning tools, statistical methods, sampling procedures,
Members
Brendan Hallam
Holly Smith is a PhD student, and her research interests include using electronic health records for research, women’s health after childbirth and the mental health of parents. She is funded by the NIHR School for Primary Care Research.
Kingshuk Pal
Takahiro Itaya
Christina Avgerinou
Danielle Nimmons
Laura Horsfall
Doug McKechnie
Kristian Svendsen
Siti Watiqah Samsuddin is a PhD student currently interested in researching the psychotropic polypharmacy use in individuals with depression using UK and Brunei electronic health records. She holds a Masters in Pharmacy degree (MPharm, UK) and prior to doing her PhD research full-time, she had worked as a responsible pharmacist with a Mental Health Trust in England.
Karan Mehta is a PhD student researching the development of all-cause mortality risk prediction algorithms using classical and machine learning modelling techniques and an electronic healthcare primary care dataset.
Rini Veeravalli
Rachael Hunter
Cini Bhanu (THINK Group Coordinator)
Elizabeth O’Nions (THINK Group Coordinator) Liz is a post-doctoral researcher based at the Research Department of Clinical, Educational and Health Psychology. She working on the AUDIT-50 project funded by the Dunhill Medical Trust. She is investigating underdiagnosis of autism in the UK, plus physical and mental health in UK autistic people.
Caroline Clarke (Methodology Group Coordinator) Caroline is a Senior Research Fellow in Health Economics based in Priment CTU at PCPH who works with the HEART (Health Economics Analysis and Research methods Team) on a number of clinical trials and other studies and projects across UCL and elsewhere.
JC Bazo-Alvarez (CEHRAM Co-Director) is a Research Fellow based at the UCL Department of Primary Care & Population Health. His methodological research is mainly focused on interrupted time series analysis and missing data handling, both of them for individual-level data (e.g., electronic health records). His applied research is focused on mental health and its connection with long-term physical health, with special emphasis on health inequalities.
Irene Petersen (CEHRAM Director) is a Professor of Epidemiology and Health Informatics at UCL. r almost two decades, she has focused largely on the use of electronic health records for aetiological and epidemiological research. Since she moved to UCL in 2003, she built a vibrant and productive research environment around analysis of primary care databases. Irene has led and supported several projects funded by MRC, NIHR and various charities and have co-authored around 200 papers based on electronic health records and population registries.
Resources
Some recent publications
- Laugesen, K., Sørensen, H. T., Jørgensen, J. O. L., & Petersen, I. (2022). In utero exposure to glucocorticoids and risk of anxiety and depression in childhood or adolescence. Psychoneuroendocrinology, 141, 105766. doi:10.1016/j.psyneuen.2022.105766
- Bhanu, C., Petersen, I., Orlu, M., Davis, D., & Walters, K. (2022). Incidence of postural hypotension recorded in UK general practice: an electronic health records study. British Journal of General Practice, BJGP.2022.0111. doi:10.3399/bjgp.2022.0111
- Smith HC, Saxena S, Petersen I (2022). Maternal Postnatal Depression and Completion of Infant Immunizations: A UK Cohort Study of 196,329 Mother-Infant Pairs, 2006–2015. The Journal of Clinical Psychiatry. 2022;83(00):20m13575.
- Pal, K., Sharma, M., Mukadam, N. M., & Petersen, I. (2022). Initiation of antidepressant medication in people with type 2 diabetes living in the UK – a retrospective cohort study. Pharmacoepidemiology & Drug Safety. doi:10.1002/pds.5484
- Hallam, B., Petersen, I., Cooper, C., Avgerinou, C., & Walters, K. (2022). Time Trends in Incidence of Reported Memory Concerns and Cognitive Decline: A Cohort Study in UK Primary Care. Clinical Epidemiology, 2022 (14), 395-408. doi:10.2147/clep.s350396
- Bazo-Alvarez, J. C., Pal, K., Pham, T. M., Nazareth, I., Petersen, I., & Sharma, M. (2021). Cardiovascular outcomes of type 2 diabetic patients treated with DPP‑4 inhibitors versus sulphonylureas as add-on to metformin in clinical practice. Scientific Reports, 11, 23826. doi:10.1038/s41598-021-02670-9
- Bazo-Alvarez, J. C., Morris, T. P., Carpenter, J. R., & Petersen, I. (2021). Current Practices in Missing Data Handling for Interrupted Time Series Studies Performed on Individual-Level Data: A Scoping Review in Health Research. Clinical Epidemiology, 13, 603-613. doi:10.2147/CLEP.S314020
- Smith HC, Saxena S, Petersen I. Postnatal checks and primary care consultations in the year following childbirth: an observational cohort study of 309 573 women in the UK, 2006–2016 BMJ Open 2020;10:e036835. doi: 10.1136/bmjopen-2020-036835
- Bazo-Alvarez, J. C., Morris, T. P., Pham, T. M., Carpenter, J. R., & Petersen, I. (2020). Handling Missing Values in Interrupted Time Series Analysis of Longitudinal Individual-Level Data. Clinical Epidemiology, Volume 12, 1045-1057. doi:10.2147/clep.s266428
See also a full list of publications.
Courses
- Introduction to primary care data: This course includes an overview of primary care datasets characteristics, basics on programming and data management, codelists creation, incidence estimation, cohort construction and sampling. Practical and examples are provided with Stata codes. The course is usually given once a year for PhD students, post-docs and other colleagues starting in primary care data analysis.
- Introduction to Stata + R: This course focuses on teaching the basics on Stata and R coding for people with previous knowledge on statistics. No previous knowledge on coding is required. Main topics are data management, data visualisation, basic statistical tests, and regression models. Each topic is developed in Stata and R simultaneously.
- Introduction to interrupted time series with electronic health records: In this course, students explore the basis of interrupted time series (ITS) as a study design, how ITS design can be applied to both population and individual level data (e.g., electronic health records), and the way it can also be used to evaluate interventions applied at the population or individual level. The course includes details on ITS analysis with Stata and R.
Recorded presentations
May 2017: Communicating risks and evidence to patients in a clear and balanced way - presented by Alex Freeman from the Winton Centre for Risk and Evidence Communication (slides)
Links
- THINK Research Group
- Primary Care Methodology Research Group
- THINK GitHub
- Others
Joint de group
- Register in a course
- PhD opportunities
- Contact us
Broaden research areas (PhD/ Post-Doc)
If you are interested in study a PhD within CEHRAM, these are the research areas we are delighted to support:
- Methods for missing data handling in electronic health records
- Complex Interrupted Time Series designs applied to individual-level data
- Machine Learning tools applied to advanced cluster analysis
- Strengths and limitations of synthetic data
- Machine and Deep Learning to improve health events predictions with electronic health records
- Validation of codelists and algorithms for health conditions and study of long-term drug treatment prescriptions
For informal questions about PhD/Post-Doc opportunities, please contact Prof Irene Petersen (irene.petersen@ucl.ac.uk) or Dr JC Bazo-Alvarez (juan.alvarez.16@ucl.ac.uk)