Daniel Lane
Anne Marti
Sanjay Unni
Nicki Kwon
Everyone wants real-world evidence, and they want answers quickly, so having the capability to turn data around
quickly is important.”
The expectations surrounding evidence generation have changed dramatically in recent years. Organizations are being asked to deliver high-quality real-world evidence faster than ever, while navigating increasingly complex datasets, integrating unstructured clinical notes, and exploring AI-driven workflows. Success now depends not only on generating insights, but also on aligning analytics with business strategy, improving transparency across teams, and ensuring governance keeps pace with innovation.
At Panalgo’s annual Data to Insights conference, Daniel Lane, Head of U.S. Data Strategy at Novartis, interviewed Anne Marti, HEOR Data Scientist at W.L. Gore Associates, Sanjay Unni, Director, RWD at Boehringer Ingelheim, and Nicki Kwon, Senior Director, IHD Insights & Operations at Panalgo, about what organizations must do to balance speed, transparency, governance, and innovation in an increasingly data-driven and AI-reliant environment.
Q: How has evidence generation changed in the last five years?
Anne Marti: There’s a growing demand for real-world evidence from clinical organizations and market intelligence teams. Everyone wants real-world evidence, and they want answers quickly, so having the capability to turn data around quickly is important. There’s also more interest in EHR data, because our company deals with complex patients, and those answers aren’t available in claims.
Sanjay Unni: The wild west is over. Our analysis needs to align to strategy. Before, it was research for the sake of research. Now that we have capabilities to align research and strategy, it provides a level of accountability: we want to know why we did the study and have transparency on our analysis. We want to know what we’re doing is meaningful.
Nicki Kwon: Data is getting more complex. We’re integrating sources. There’s more emphasis on using physicians’ notes to find more complex patients. Treatments are more targeted and identifying patient populations is more specific.
Q: How are your organizations harnessing the power of unstructured notes and leveraging them to get evidence? What are your challenges in leveraging unstructured data?
SU: It’s important to be able to use capabilities like AI and work with your clinicians to understand what’s in unstructured notes. Because it’s real-world data, there are things coming from these notes that you’re not going to expect.
AM: It’s helpful to review raw clinical notes and compare them to what comes out in natural language processing. It makes you think about it differently when you see the raw data.
Q: Knowing how evidence is being used is a challenge. How is that changing, and are your teams getting more clarity on that?
SU: There’s nothing more frustrating than when you’re working with your analytics teams and they can’t answer questions about data. It’s not their fault because they’re not given that transparency. It’s my job to open that door. There are so many ways of capturing certain outcomes or inputs. We need to understand how data is being used—it’s about making sure that people who didn’t have a seat at the table have a seat at the table.
AM: We also need to know what level of quality control is being used. Having constant communication about what a project involves is important at the outset.
Q: How are you maintaining integrity of research as speed to insights is faster?
AM: At Gore, we’ve worked to standardize workflows in data sets to make sure we’re using the same definitions to make sure when we get these questions, we can execute high-level insights quickly and tailor analysis from there.
SU: Helping stakeholders in analytics teams move quicker and faster is all about communication. Anything we can do from a stakeholder perspective or analyst perspective to make sure we’re streamlining communication and being transparent makes our analysts’ lives a lot easier.
Q: With so many data analytics platforms out there, are your organizations making efforts to streamline this?
NK: There’s not one tool for all purposes. Maybe instead of using 22 platforms, you can identify three to five that shine the best. It’s about identifying strengths and weaknesses of each tool. Panalgo’s IHD platform makes it easier to stay with one tool instead of being on different platforms that split your focus.
SU: Stakeholders don’t often understand the amount of effort it takes for humans to keep these things running, checking data on multiple systems. The fewer systems, the better.
AM: Our team uses IHD, and having one platform that can do so much is a relief. We don’t have to learn multiple program languages or switch between programs.
Q: What are your experiences, successes, and challenges using AI?
SU: There are so many stakeholders in the AI experience. Leadership wants it to run a certain way, IT teams want it to run a certain way, and it’s not often the same. It’s about level-setting expectations on what you’re going to get from AI tools. Stakeholders have different experiences and different needs, so it’s about communication and change management.
AM: We’ve found AI most helpful in analytic planning and summarizing and communicating results.
NK: We use AI to standardize communication. It helps us become domain experts or reinforce our beliefs, and checks different results on input.
SU: The governance of AI has evolved. At first, people didn’t realize that AI needed a secure environment, so there needed to be education about that. Also, you have to consider the internal politics of AI—who owns it, who’s going to build it. Thinking about operationalizing it can slow down innovation.
NK: From the top down, everyone’s asking to use it to create efficiencies, but it can be challenging to find the right use cases for AI that can be trusted.
Daniel Lane: Organizations should focus on upskilling people and educating customers on these technologies, providing resources on using AI and teaching people how to leverage these tools. My concern is, we’ve built things so fast but haven’t put the governance in place. Building proper governance is going to be critical to leverage AI for evidence-generating activities.
NK: It’s about having open dialogue with clients and understanding what that analysis is for. Start with decision you’d like to make, find the best data, and then work backwards.
Q: What has been the biggest differentiator at your organizations in working with Panalgo and IHD?
AM: The support from the Panalgo team has been incredible, in terms of training and additional resources. It’s a one-stop shop, which has been critical for us. We’re exploring new data sources using Panalgo.
SU: Not having to worry about managing a complicated system is a big differentiator for us. I don’t have the time, energy, or budget to do what needs to be done if I have to manage it myself. Leveraging Panalgo and IHD to do that has been a huge selling point for us.
Contact us to find out how Ella AI and IHD can power your evidence generation strategies, faster.