Portrait - Linda Haroun, clinical data manager

Following a master's degree in public health, Linda Haroun went on to become clinical research associate at AP-HP (Paris public hospital authority). Then two years ago she joined Institut Curie as part of the ESME (UNICANCER) project, to work on databases. "A clinical research associate is the person who guarantees the quality of a project at all stages," she explains. "He/she ensures that regulations are followed and that good clinical practices are in place. This means for example checking that any undesirable effects are declared."

So it was quite natural that she evolved towards a position as clinical data manager, created at Institut Curie in September 2018, at the time the new data team was created. The data manger position mirrors that of clinical research associate, since they are the guarantor for the quality of the data extracted to meet researchers' needs.

I am fortunate to work in a team with a great many complementary skills. I love it and we communicate a lot! I feel that I am being useful and I like to think that my contribution - modest though it may be - will one day benefit research, and therefore patients.

At the heart of data Her remit Linda Haroun is responsible for the Breast database at Institut Curie, which includes data from the Paris and Saint-Cloud sites from 1959 to 2012. Researchers requiring data from after this date may collect the data via data managers. "Part of my work involves entering data to update this base, according to research projects, while ensuring the quality of the data delivered. I am also responsible for creating new clinical databases, in addition to the Breast database." Lind Haroun is there throughout the process of structuring the data to make it accessible to users. After a first step where the applicant checks feasibility, the project goes to CRI Data, a committee that validates and oversees use in research of patient data collected during care, in order to comply with the GDPR. Once the green light is given, the data manager determines the user's needs and the minimum data needed for his/her research. All of this is entered into the REDCap software (Research Electronic Data Capture): the base is created and the data is ready to be used. "Thanks to this tool, we can facilitate secure data collection and storage." The next step is to train and support users. "In 2019 we are offering training for Curie researchers to use the REDCAP system, if they require it. In this way they can enter and collect data securely and avoid using Excel."

Keeping an eye on the data Linda Haroun also works on ConSoRe, a software that collects data in patients' medical records, in an autonomous and intelligent way. But it still has to be checked manually.

We continue to work on the reliability of the system, by validating the data manually through random checks. In addition, machine learning algorithms are being developed to automate this process. We want to use it to supply the other databases automatically, and thus be able to collect as much information as possible in the hospital databases.

The idea is to generate cohorts based on information collected in real time in the computerized patient record. The aim: to save time for both researchers and physicians. "We can respond more quickly to their needs and process more requests."