00;00;05;04 

MAPS 

Welcome to this episode of the Medical Affairs Professional Society podcast “Elevate”. The views expressed in this recording are those of the individuals, and do not necessarily reflect on the opinions of MAPS or the companies with which they are affiliated. This presentation is for informational purposes only and is not intended as legal or regulatory advice. And now for today’s “Elevate” episode. 

00;00;33;14 

Jennifer Riggins 

Welcome to Elevate the Medical Affairs Professional Societies podcast. As a series within this podcast, we focus on digital first communications, how digital is transforming Medical Affairs. I’m Jennifer Riggins, your co-host for today’s episode. 

00;00;49;28 

Stephen Casey 

And I’m Steve Casey. Today we’re continuing to explore scaling generative AI in Medical Affairs. The future is clear. We’ll all be using AI in our businesses in one way or another. But AI requires data to work well from chat bots and content generation to insights mining and decision support. Gen AI is quickly moving from use-case to scale up. 

00;01;14;01 

Jennifer Riggins 

So, in our previous podcast, we talked with folks from Pfizer about their internal facing AI use cases that they’ve been exploring. Today we’re joined by Sandy Kaur of AstraZeneca to share her insights on some external facing gen AI projects that they’ve done. 

00;01;31;18 

Stephen Casey 

Yes, Jen, Sandy is the Director, Digital Acceleration, Global Oncology, Medical Communications and Information at AstraZeneca. So, let’s get to it. Sandy, welcome and thank you for joining us today. We’re so pleased to have you on the podcast. Could you give us a brief overview of your background and introduce us to your current role and responsibilities at AstraZeneca. 

00;01;52;24 

Sandy Kaur 

Hi Jen and Steve. Thank you for having me on the podcast. Really happy to be here and part of this brilliant, vast practice sharing forum for the Medical Affairs community. And of course, I’m Sandy and I’ve worked in the medical communications industry for the last seven years, six of which have been with AstraZeneca, and I’ve performed various roles at AZ, most recently for the past four years, I’ve had the pleasure of leading our AI programs in global oncology, medical communications. 

00;02;22;26 

Jennifer Riggins 

Fantastic. Thank you so much for joining us today and providing us with your background. So, in our previous podcasts, we’ve jumped right into how people are using AI, but I think it would be helpful for the audience today to hear about some of the decision steps prior to getting into your use cases. So, in your role as Director of Digital Acceleration, how do you decide what use cases to move forward with and which ones you’re going to wait on? 

00;02;48;27 

Sandy Kaur 

I have two areas of focus in my role. So, the first is how can we leverage AI to be more efficient in our ways of working? And secondly, how we can leverage AI to meet HCP needs when it comes to their preferences for engaging with our medical content. Prioritization is key since the list of opportunities is unless. That said, I think starting broad is really important, but being able to get that list down to 1 or 2 things that are really going to move the needle is key. It’s important to focus on the top priorities that will create the most amount of impact. 

00;03;24;23 

Stephen Casey 

Very cool. Thanks for that. I’m curious, can you share what inspired you to explore the use of AI when communicating with HCP? 

00;03;33;03 

Sandy Kaur 

Sure. So, the data tells us that for 80% of HCP, the biggest challenge they face is a busy clinical workload. HCP’s don’t have the time to scroll through multiple resources to get the answers they are looking for. And furthermore, over 95% of oncologists express the highest demand for content in a short form, bite size, fast facts and downloadable. So, against this backdrop, our mission was and was and is to make our content as discoverable as possible. Shifting from pushing resources to solving their questions and meeting their needs. I was a key enabler to delivering on our mission. 

00;04;14;23 

Jennifer Riggins 

So that’s really important market research, I think, that you just gave us and it’s really critical to building trust and credibility. I really like that focus of solving HCP questions and meeting their needs. I’d like to shift us now, shift us just a bit more into some of the details of the AI work. And some of the external AI use cases that you’ve been working on. So, can you give us an overview of what you’re doing with the Medical Affairs AI to communicate externally? 

00;04;44;06 

Sandy Kaur 

Yeah, so we designed a tool that is created for the conversational AI search world because HCPs are already there. 77% of HCPs are already using these tools for search. So, giving them the same user experience when they visit our website is essential. Our conversational AI approach delivers our impactful medical content in a concise, guided and interactive manner for HCPs, allowing them to get the quick answers we know they desire, whilst also enabling them to engage deeper on the topic as needed. So this approach ensures that HCPs still have the choice, a quick approach to find what they need, and then the freedom to get dig further within that topic. 

00;05;29;03 

Stephen Casey 

Oh, that sounds like a great shift in approach. Before we talk to you and talk to you about the engagement you’ve seen with this tool, let’s step back and make sure everyone understands what conversational AI is, and also what you had to do, if anything, to the content to move from a traditional search to a conversational AI approach to search. 

00;05;50;22 

Sandy Kaur 

Definitely. So traditionally, as a medical information team, we would create long text heavy documents that would sit on our medical information website. Now, given the data we just discussed, we know that HCPs don’t have the time to sift through these documents to find the answer to that question. On the other hand, our conversational AI approach enables HCPs to specific questions and get specific answers. The approach is more engaging as it creates the dynamic of a conversation, enabling the HCP to ask follow-up questions as needed, ensuring they have the right level of detail based on their needs. So, think of it as it is on Google. Instead of having hundreds of search results and pages to scroll through, you can now have an AI summary that answers your direct question. Essentially, we’re reducing HCP search time from minutes to seconds. 

00;06;45;17 

Stephen Casey 

AI, thanks so much for that dive. That explanation really, really helps. Have you observed good engagement with the tool? 

00;06;52;13 

Sandy Kaur 

Absolutely. And something we’re very proud of. So we’ve observed tremendous engagement. Multiple engagements per HCP, visit, a 200% increase in engagement with our medical content compared to the website without the conversation AI tool, and also a 4.7 out of five star feedback rating. So very proud to have this engagement. 

00;07;15;12 

Jennifer Riggins 

Yay. Congratulations. That’s fantastic. Conversational AI, you know, has really been a great tool. But how do we balance innovations like it with compliance when these technologies are being used externally? You know, because of the nature of AI and Gen AI, are the conversational answers continuously changing? 

00;07;35;15 

Sandy Kaur 

I’m so glad you asked this question, because implementing AI tools like this one isn’t a one and done. It’s not a question of being compliant or innovative. It’s really about how we can do both. Ensuring compliance and trust is critical to long term success. In a HCP facing project like this is critical to ensure rigorous monitoring and subsequent continuous improvement processes are in place. Governance is key, and it’s important that priority and focus is given to compliance during both the implementation and the maintenance phase. 

00;08;11;21 

Jennifer Riggins 

I think that’s really so important for others to hear that need for accuracy is incredibly important when communicating externally. So are there any other considerations when implementing HCP facing conversational AI tools like these? 

00;08;25;15 

Sandy Kaur 

Yeah, there’s a few more examples that I’d love to share with the audience. The tool must be scalable across your portfolio, channels, and markets. It can be easy to deliver a digital experience in one channel for one brand or a therapy area, but the ability to do this at scale is where we can deliver the greatest impact. Therefore, it’s important to build it with scalability in mind, considering where configurations will be needed for market or channel needs. Another example I can share is the need for the tool to be integrated. It must be easy to maintain to ensure maximum efficiencies, and ideally it needs to be integrated with your enterprise content library and strategy. And finally, it’s got to be relevant to hit KPIs. So as we think through the ability to add new data, new content, we need to be seeking maximum automation coupled with simple processes. 

00;09;19;14 

Stephen Casey 

So you have shared a use case for external use to Jen and I that you’ve been working on. Do you see other external use cases that you want to try and tackle in the future? 

00;09;30;09 

Sandy Kaur 

Currently, we’re working on scaling this tool and enhancing it so we can make the tool even better for using it and ensuring we continue to meet that needs. 

00;09;40;11 

Jennifer Riggins 

Great. So? So, Sandy, if you don’t mind, let’s spend a little time discussing what you’re doing with internal gen AI use. So, what type of internal use cases have you been pursuing that you can share? 

00;09;52;05 

Sandy Kaur 

Similar to your previous guests on the podcast, we’re also using Gen AI to accelerate our medical content authoring processes. We began last year very much training the team and encouraging them to use the tool, and now it’s very much a standard way of working for us. Last year was a fantastic year, improving adoption, and now we’re really thinking about how we can perform AI assisted content authoring in a much more integrated manner, gaining even more efficiencies for the team. 

00;10;22;15 

Jennifer Riggins 

Excellent. That’s, that’s what we’re starting to hear from, from a lot of individuals. So, so glad that’s working well for you. So one final question. If you could give three takeaways from your learnings for our listeners, what would they be? 

00;10;37;26 

Sandy Kaur 

So this is that such a great question and a great way to end a good discussion. So my top three takeaways would be number one, prioritize solving for HCP needs. Shift from delivering content to delivering conversations. HCP aren’t asking for more assets. They’re asking for our data to be delivered in a quick and concise manner in order to answer their questions. Number two, align early on on strategic principles to guide the project’s approach, some of which we’ve discussed today. And then the last one would be expect setbacks. This is technology and things will go wrong. So be resilient and be prepared to solve problems. 

00;11;20;11 

Jennifer Riggins 

I love those three. Sandy, thank you so much for sharing your practical insights. This kind of open dialog really is how we can move the Medical Affairs field forward in the future. 

00;11;32;05 

Stephen Casey 

And thank you to our listeners. Be sure to follow Elevate, Digital First Medical Affairs wherever you get your podcasts. Until next time, stay curious and stay innovative.