Big Data

For Better Hearts

BigData@Heart

Interview with Alicia Uijl

Dr Alicia Uijl

Postdoc at the University Medical Center Utrecht 

Utrecht, Netherlands

Alicia is a key member of case study 4 “Epidemiology of Heart Failure in Europe” working group in the IMI BigData@Heart (BD@H) project.

 

How did you become interested in epidemiology?

I first became interested in epidemiology during my bachelor’s degree in biomedical sciences. I found out that all courses in epidemiology came very naturally and every other student always wanted to have me in their group. My Master research project was an epidemiological study in diabetes and obesity at Leiden University. Thanks to my supervisor and colleagues this was a great experience, including participation in conferences and writing scientific abstracts - privileges of experienced PhD students. I was determined this was what I wanted to do and I got a PhD position in Utrecht University with prof. dr. Folkert Asselbergs and prof. dr. Arno Hoes as supervisors. First year of my PhD I spent at University College London getting deeper into big data science and its application in epidemiology and particularly cardiovascular diseases.

When did you get involved in BD@H?

It was in 2017, I was invited to the first BD@H annual meeting in Utrecht. In 2018 we submitted an application for projects for what is now known as case study 4 “Epidemiology of Heart Failure in Europe” and received extra PhD funding to complete this project. In our project we compare different indicators of the HF patients across four European countries: Sweden, the Netherlands, Spain and the UK.

Why did you choose these four countries?

Within the BigData@Heart we collaborate with medical universities located in these countries and they helped us to organize access to the real-world patient’s data bases.

How long does it take to get an access to the health registry?

It is a long process and differs from country to country, but on average it takes several months from starting with a research proposal and receiving the data.

On which stage of the project are you now?

We have already analyzed the data from Sweden, Spain and the UK, and now we are working with the Dutch data.  

Why is it important to compare data registries from different countries?

Health care systems vary by country and our main goal is to understand if there any difference that related to the health care system and not to the patients. Important to mention that the only countries with relevant level of health care may be compared.

Tell us about your recently published paper (June 2020) in simple words

Electronical health records (EHRs) frequently lack some specific information that researchers may need. As an example - EF (Left ventricular ejection fraction) – is an important marker, that is used in heart failure (HF) for diagnosis and treatment selection, is often missing in EHRs, thereby preventing analyses of specific EF subphenotypes and limiting EHRs use in HF research.

In this project we proposed an algorithm which can predict EF type based on available characteristics, such as patient demographics, clinical characteristics and implemented treatments. We used 22 variables to correctly predict HF subphenotype with an accuracy of 70%. Previous algorithms have been developed using ICD-9 codes (International Classification of Diseases) and therefore are inapplicable for EF predictions in trials data sets, registries and EHRs in many European countries, as it is more common to use ICD-10. 

We are now trying to implement our model in a hospitalized cohort to show that it is valid in different populations.

What are your future plans?

I have recently finished my PhD and now continuing my research for the BigData@Heart project at Utrecht University as a postdoc. I would like to thank to my team in achieving this wonderful milestone. Together with UMCU (Prof. dr. Folkert Asselbergs) we have applied for a 2-year grant to work in Australia. There situation with COVID-19 gave rise to telemedicine development and we want to research how the distant monitoring affects the HF patients. Furthermore, we want to create a platform of international research collaboration that would combine data from different countries in the world that would help people to understand the EHR data better.

From your point of view, why is the BigData@Heart project important?

There are many questions on relevance of using real world data, not everyone agrees that the quality of this data is sufficient for usage in research. 

The overall goal of our project is to show that big data is much more than you imagine, it is infinite source of information, that can be used in various approaches. All the small projects, including mine, are contributing towards this. We are not working personally with the patients in clinics, but our work is dedicated to improve patient healthcare.