Welcome to this exclusive interview with Julie Laurent, one of the heads of the Data Scientist department at Carenity, a global CRO with a unique approach to Online Patient-Centered Outcomes Research. Carenity stands at the forefront of connecting with patients worldwide, offering direct access to invaluable insights into patient experiences and preferences in real-world settings.

As a digital CRO, Carenity pioneers in leveraging advanced data science techniques to engage patients and conduct groundbreaking Patient-Centered Outcomes Research. Today, we have the privilege of exploring the technical expertise of Carenity’s Data Science team thanks to the insights of Julie Laurent.

In this interview, we will explore the innovative culture within the data science team, adopting cutting-edge tools and methodologies, envisioning the future of data science at Carenity, and the meticulous approach to project planning and execution. Additionally, we’ll gain valuable perspectives on adapting to business needs, measuring project success through metrics, and their unique utilization of methodologies like Discrete Choice Experiments (DCE) in patient-centered studies.

Join me as we uncover the strategic vision and technical prowess driving Carenity’s data science endeavors.

GT: How do you foster a collaborative and innovative culture within your data science team?

JL: All members of the data science team are involved at some stage in handling and processing the data, from designing the survey to carrying out the statistical analysis. This ensures collaboration between team members, because we all know that each step is a building block towards the successful delivery of a high-quality study.

Innovation is encouraged long before that. We want Carenity to be a major player in the field of Real-World Evidence, and this means leading innovative projects and positioning ourselves as experts in patient-centered outcomes. We always want to provide the best solution to meet our clients’ objectives. This leads us to think collaboratively to propose the best methodology.

GT: Can you describe your approach to adopting new tools or methodologies within your department?

JL: The introduction of new tools and methodologies comes at two key points in the life of the project. Firstly, in order to come up with the best proposal for the future design of a survey, we put our heads together and think, using our existing knowledge and an in-depth analysis of the literature, about how to provide the best answer to the research question we are asked. This leads us to adopt new methodologies within the team, such as all the patient preference panels.

Secondly, when we carry out a project, we are mindful of finding the most appropriate methodology to bring out the best in our data. This means the most appropriate statistical model and the best data visualization tool.

GT: What is your vision for the future of data science in Carenity, and how do you plan to achieve it?

JL: Carenity can rely on high-level in-house skills with different profiles that complement each other well and create great synergies (pharmacists, health engineers, statisticians, etc.). All the conditions are in place for Carenity to grow by giving everyone the place they need to make the most of their abilities. In the near future, Carenity will be focusing on new innovative projects, proposing new methodologies for patient preference studies, and exploiting new data sources (patient-generated health data). This means consolidating our skills and positioning ourselves as experts in data linkage methodologies, and in exploiting and analyzing large volumes of data through clustering and classification methods.

GT: How do you approach project planning and execution to ensure successful delivery?

JL: A clear definition of the different stages of the project and the role of each member of the project team is essential, as it is good communication within the team. Anticipation is also key. Finally, we need to ensure good interaction with the client, to guide him throughout the project and provide him with the recommendations he needs to ensure that the project runs smoothly.

GT: Can you discuss a situation where you had some challenges and had to adapt the study design?

JL: I have two projects in mind. The first is a project where we had to recreate the conditions of a case-control study in order to meet the client’s main objective. This meant creating a new survey design, adapting the sample size and facing a challenging fieldwork.

The second situation was a study of patient preferences readapted to a new scope of emerging countries. We first had to carry out a preliminary study, including interviews with KOLs and feasibility assessments, to ensure that we were going to set the project in the right direction for these markets.

GT: What type of study methodology do you use more often and why?

JL: Patient preference studies are increasingly used, recommended by health authorities, and are therefore key studies for our clients. Among these, we make extensive use of the discrete choice experiment methodology. Why do we do this? Because it is the only patient preference methodology that estimates the trade-offs patients are willing to make when choosing a treatment. DCE is an efficient method to highlight trade-offs made by patients, which is important to support benefit-risk assessments. When people face challenging trade-offs, we learn what is truly important to them.

The results of the DCE give our clients clear indications of the treatment characteristics that are most important to patients, enabling them to better position themselves in the market.

GT: How do you think your use of DCE is unique when conducting studies?

JL: Firstly, we ensure the quality of the first phase of the literature review by mobilizing our medical resources to identify the most appropriate attributes and levels for the DCE. We ensure double

validation of this first stage by conducting qualitative interviews with patients and KOLs. Next, we use the most appropriate software to create an optimal DCE design, which is tested through a pilot study. Finally, we analyze the DCE data by choosing the most appropriate model (conditional logit model, mixed logit model, etc.) and the best data visualization.

—-

In conclusion, collaboration and innovation are key to Carenity’s data science success. Our vision involves leveraging in-house skills for growth and focusing on innovative projects, consolidating expertise in patient preference studies. Effective project planning, clear communication, and client interaction ensure successful study delivery. Our unique use of discrete choice experiment methodology sets us apart, revealing crucial patient trade-offs and providing valuable insights for clients.

By Gilda T.