Cutting Through the Noise: a Q&A with Director of Data Science at Kinetic
An Interview with Bharathi Ramachandran, Director of Data Science for Kinetic™ at Syneos Health®
Working with big data to solve healthcare problems is a vital part of the work being done at Syneos Health. As part of that effort, it is key to the success of Kinetic, where a team of PhD and Masters’ level data scientists and analysts from around the globe disseminate vast amounts of healthcare data and analyze findings to help life sciences companies reach healthcare providers (HCP) and improve patient outcomes across the asset lifecycle.
Leading this important data science team is Bharathi Ramachandran, Director of Data Science. With a wealth of knowledge in the healthcare data science space, Bharathi is driven by the mission of meaningful work and by the desire to understand both the micro and macro levels of the industry—what minute data points can teach us about global trends and how we can make business decisions informed from those findings to help improve patient health.
We spoke with Bharathi about her work with Kinetic.
How did you become interested in data science?
My journey to this field is somewhat nonlinear. As an undergrad, I studied biochemistry. At the time, I was fascinated, on a very micro level, by what things impact our living processes. To me, this was useful and important work that offered the kind of intellectual challenge I am drawn to. In practice, though, I realized being a biochemist required long hours in a lab, which wasn’t for me. But I was still very interested in health and in the quantitative components that exist in biochemistry. So after a short time, I went back to school to get my Masters in biostatistics and epidemiology. From there, I began working for health industry companies as a data scientist and have been looking at how to solve big issues in healthcare ever since. At Kinetic, of course, we’re creating data-informed solutions that help life sciences companies reach HCPs at critical moments along an asset lifecycle.
What makes the Kinetic data science team unique?
The Kinetic team is made up of PhD and Masters’ level scientists, most of whom have healthcare-industry expertise. That’s actually quite unique. Data scientists tend to be generalists who understand the methodologies and core statistical principles of data science and can apply that knowledge to any industry. But because the scientists and analysts at Kinetic, like myself, also understand the healthcare industry, we can translate data more conscientiously, knowing the problems life science companies, HCPs and patients face on a day-to-day basis. That’s an important distinction between our data science team and others.
Can you talk more specifically about how having life science expertise affects data translation?
Sure. The amount of data in the healthcare industry is incredibly vast, and it keeps growing every day. When data teams are working with that amount of data, their job is to cut through the noise, so to speak, build the right models, ask the right questions, and translate findings into meaningful insights that will help solve business problems. But when you have a team that doesn’t have industry experience, their ability to frame relevant questions and generate meaningful insights in that context is limited. You’ll see that they will translate data in ways that might tell alluring stories but that don’t have much use. In my mind, they haven’t cut through that noise. At Kinetic, our team has the industry experience to look at data from multiple healthcare-industry perspectives that other teams might not be able to do. That’s beneficial because it means we have a greater opportunity to develop more meaningful insights that lead to making better business decisions and products that will, ultimately, serve patients.
What other expertise defines Kinetic?
We are experts in building scalable AI and ML solutions for various life science problems.
How does that help life science customers looking to reach business goals?
If you are a life science company, you want a team that can showcase multiple approaches to a question or problem early on and work collaboratively to pick the right solutions that align with your goals and accommodate your business constraints. Applications of artificial intelligence (AI) and machine learning (ML) can enable faster and more agile solutioning to help answer these critical questions that can drive ROI and bring life-saving therapies to market faster.
Kinetic defines its work as building relationships with HCPs from lab to life. What specifically does that mean for the data science team?
All our solutions are geared toward reaching and optimizing interactions with those HCPs who have an increased capacity to prescribe in commercial settings, refer in clinical settings, and get educated in medical affairs settings. Specifically, we have developed an audience intelligence capability using AI and machine learning that can locate and leverage these propensities in HCPs using historical and real time behavioral data that pinpoints high-impact audiences for our customers.
The way we build HCP audiences is multidimensional and not solely based on prescribing or treatment behavior or patient volume. Rather, we profile HCPs more holistically based on various data points. For instance, we look at the behaviors of the physician themselves, their propensity to adopt or switch to new therapies, their peer network behavior, and their leadership influence characteristics based on their clinical trial and publication history. These are just a few examples. The list of HCP characteristics we leverage goes on. And, given a specific assignment, we can weigh those characteristics accordingly. This multidimensional approach increases our likelihood of engagement and of building those long-lasting relationships.
For clinical trials and med affairs relationship-building, we leverage patient journey data to identify patients upstream in their disease progression and reach out to the key HCPs treating these patients regarding suitable clinical trials or emerging science on the subject that may help these patients. There are some rare diseases, for example, that can be difficult to diagnose because patients are symptomatically similar during early stages of disease but progress to more distinct subtypes. We develop predictive algorithms that can identify these distinct patient cohorts based on key early characteristics to accelerate identification of trial eligible patients and relevant HCPs. I think that’s exciting.
In addition to audience building, data science is also being used to streamline real time interactions with HCPs, is that right?
Yes. Once we’ve identified high-impact HCPs, we predict their affinity and behavioral response across various channels to recommend custom channel outreach combinations. Based on their engagement behavior, we can dynamically refine outreach lists, recommend the next-best actions for field teams and choose the best channels to use on a never-ending data analysis loop. That’s exciting too.
What other capabilities make Kinetic unique from a data perspective?
Scalability. That’s really important. We build solutions that can be scaled across the industry. So instead of single point solutions, we think about how a model or solution that we built for a specific treatment or indication can easily be scaled across similar indications or be adopted for other business problems. That helps us solve problems for clients more quickly.
What’s next for Kinetic?
Well, we’re evolving alongside the healthcare industry, and we are operating at a high level in terms of innovation and agility. You can expect Kinetic to continue to cut through the noise and bring advanced data science-based solutions to our customers in new and effective ways.