PHD POSITION ON DYNAMIC PREDICTION OF BLOODSTREAM INFECTIONS USING EHR DATA

(ref. BAP-2020-623)

The research will be conducted within the Department of Development and Regeneration at KU Leuven, in collaboration with the Management Information & Reporting department of the University Hospitals Leuven. As promoters, prof. Ben Van Calster and prof. Laure Wynants will supervise the research together with Dr. Pieter Stijnen. The promoters are members of EPI-centre, a young group of KU Leuven researchers interested in clinical epidemiology, diagnostic testing and prediction modeling. The project team consists of methodologists/statisticians, data analysts (responsible for hospital quality of care and patient safety reporting), and clinicians from hospital hygiene and intensive care. Further, we have collaborations with Maastricht University (prof. Wynants) and Leiden University Medical Center (prof. Van Calster). This international and interdisciplinary network provides an excellent research environment for the project.

Project

The project focuses on the prediction of central-line associated bloodstream infections (CLABSI) in patients hospitalized at the University Hospitals Leuven. Using electronic health records of recent years, the focus is on making dynamic predictions based on regression and machine learning models for dynamic and competing risks survival analysis. The project has a methodological and an applied component. On the applied level, we aim to develop a robust prediction model, and validate it in new data from Leuven as well as in other Flemish and Dutch hospitals. We plan to combine this with a cost-effectiveness evaluation. Methodologically, there is room to investigate and compare regression and machine learning methods to develop dynamic prediction models in competing risk settings.

The project is funded as a C2 project by KU Leuven, with the following work packages:

• WP1 Development and validation of a dynamic risk model for CLABSI in the University Hospitals Leuven
• WP2 Development and validation of models using flexible machine learning algorithms
• WP3 Evaluation of the cost-effectiveness of CLABSI risk prediction models
• WP4 Transportability (external validation) of the risk model to other hospitals
• WP5 Implementation of the model in the University Hospitals Leuven EHR environment

You will collaborate with a second PhD researcher, and your work will mainly involve WPs 1, 3 and 4. We will collaborate with clinicians and data analysts from the University Hospitals Leuven, as well as with statisticians from Maastricht University and Leiden University Medical Center in the Netherlands.

Profile

• You have a Master’s degree in statistics, or a Master’s degree in another subject in combination with an additional training in statistics.
• You have excellent programming skills (preferably in R or python).
• You have a strong interest in clinical prediction, diagnosis, prognosis,and risk estimation for medical applications. The focus will be on applied methods for risk prediction. You are capable of seeing the bigger picture and reflecting on the clinical context in which the model will be used.
• You are fluent in English, both written and oral.
• You have knowledge of survival analysis, and regression models and their interpretation. You possess skills to critically interpret and write research papers.
• You can work independently as well as in team, and are capable to workout research problems by yourself.
• You are capable to complete a doctoral thesis under appropriate guidance of your supervisors.

Offer

• Full-time employment of 1 year (starting January 2021, can be discussed),which can be extended to a maximum of 4 years with the aim of defending a PhD thesis at the end of this period.
• You will join a diverse and interdisciplinary collaboration.
• You will be able to combine a challenging mix of methodological and clinical research.
• You will be able to collaborate internationally. A stay at Maastricht or Leiden can be discussed.
• You will be able to attend international conferences and write scientific papers.

Interested?

For more information please contact Prof. dr. Ben Van Calster, tel.: +32 16 37 77 88, mail: ben.vancalster@kuleuven.be or Prof. dr. Laure Wynants, tel.: +32 16 32 76 70, mail: laure.wynants@kuleuven.be.

You can apply for this job no later than September 30, 2020 via the online application tool

KU Leuven seeks to foster an environment where all talents can flourish, regardless of gender, age, cultural background, nationality or impairments. If you have any questions relating to accessibility or support, please contact us at diversiteit.HR@kuleuven.be.