Demonstrating the value of generative AI in Medical Affairs

There has never been more pressure on Medical Affairs to demonstrate its value and impact than now. Medical Affairs heads are under intense pressure to prove their organization’s worth. The old “trust us, we’re valuable” approach isn’t cutting it. Enterprise leadership demands concrete numbers that show the real-world impact of their efforts, not just the volume of activities like advisory boards, posters, and MSL interactions.

Medical Affairs leaders face significant challenges, but that’s where generative AI becomes more interesting. It can help them offer the larger enterprise an additional perspective of value demonstration.

Summarized by Writer

• Generative AI has the potential to solve real problems in Medical Affairs and reinvent its importance at the enterprise level.
• Core needs for Medical Affairs include synthesizing complex information, managing medical language, and personalizing content for specific audiences ‌— and generative AI excels at meeting these requirements.
• Measuring ‌impact requires baselining. Six easy measures to start with are labor cost, vendor cost, output cost, cycle time, production cycle cost, and quality improvement.
• Successful implementation of generative AI in Medical Affairs requires designating change agents, empowering them with resources, allowing for Proof of Concept (POC) runway, involving senior leadership, enhancing human capabilities, measuring and adjusting continuously, maintaining high-quality standards, gradually converting skeptics, and focusing on scalability and sustainability.
• Generative AI in Medical Affairs amplifies human expertise by solving real problems with measurable results, but it’s important to be transparent about both the benefits and costs.

The limited impact of past digital initiatives

Throughout my years in pharma, whether working for top five pharma companies or doing my own independent consulting, I’ve watched pharma companies chase digital transformation projects — sometimes even aimlessly. Some worked out well, but most didn’t.

Initiatives like equipping Medical Science Liaisons (MSLs) with iPads, implementing medical triggered emails, creating healthcare professional (HCP) portals, and adding QR codes to posters modernized the function. Unfortunately, their impact was difficult to measure and often limited to specific areas like Medical Communications or HCP Engagement.

But here’s what’s different about generative AI in Medical Affairs: it’s not just another shiny tech toy. It has the potential to solve real problems that Medical Affairs has battled for decades — and it presents new ways for Medical Affairs to reinvent its importance at the enterprise-level.

The promise of generative AI for Medical Affairs: This time it’s different

Medical Affairs has always been that crucial bridge between pharma companies and healthcare providers. Medical Affairs is the function of translating complex science into practical insights. But let’s be honest: The function and industry are drowning in information ‌— and HCPs are, too ‌— ‌they can’t keep up. The medical literature keeps exploding and HCPs expect immediate answers. There’s also a high demand for personalized, high-quality communications, especially given the standards set by companies like Apple, Netflix, and Amazon.

The new era of digital in Medical Affairs, driven by generative AI and the future of agentic AI, addresses the entire function. Generative AI excels at synthesizing complex information, managing medical language, and personalizing content for specific audiences — core needs that are essential for Medical Affairs.

Show me the numbers

Through my years in procurement and my more recent consulting work, I’ve found that measuring impact depends on solid baselining activities done upfront. To keep it easy to understand‌ ‌and track, these measurements need to be simple.

First, we need to find a baseline that we can measure generative AI against. I usually advise teams to consider six easy and generic measures to start with:

1. Labor cost: What are we spending on people, whether they’re full-time employees or contractors?
2. Vendor cost: What are we paying a third party to do?
3. Output cost: What’s the real price tag per deliverable or widget?
4. Cycle time: How long does stuff really take? What is the right start and end date to include?
5. Production cycle cost: What’s the cost each time something has to run a cycle?
6. Quality improvement: Are we getting better? Reducing errors, increasing approvals, etc. 

Here are two examples of Medical Affairs baselining, broken down as simply as possible:

The money part of baselining people sometimes forget

Before everyone gets overly excited, it’s important to ultimately include implementation costs in the calculations.

Say your generative AI Med Info project saves ~17% per Medical Information response letter ($100 to $75) across 50,000 responses. Looks great on paper — $1.25 million in savings. But hold on.

The real costs:

• Developing the platform and build-out of generative AI model: $300,000
• Internal IT support: $60,000
• Vendor platform maintenance: $40,000
• Number of licenses: $50,000

The real savings after the dust settles? $800,000‌ — ‌not $1.25 million. Still great, but we must be transparent about the numbers.

Suggested best practices for generative AI success in Medical Affairs

Based on what I’ve seen succeed (and fail), here’s what I recommend to Medical Affairs teams looking to embark on a generative AI transformation.

Designate change agents. Medical Affairs leadership should appoint dedicated change agents to champion the AI agenda.

Empower AI champions with resources. Ensure these change agents have access to tangible resources, not just moral support.

Allow for POC runway. Give POCs sufficient time to demonstrate their value and potential.

Involve senior leadership. Engage senior medical leaders as mentors to guide and support the AI initiatives.

Enhance human capabilities. Design systems that make people better, rather than making them redundant.

Measure and adjust continuously. ​​Regularly assess the impact of AI on your processes and outcomes, and  prepare to make necessary changes.

Maintain high-quality standards. Don’t compromise on quality in the pursuit of efficiency.

Gradually convert skeptics. Resistance to change is natural, but by demonstrating the positive impact of AI in small, manageable steps, you can help skeptics see its value and become advocates for the transformation.

Focus on scalability and sustainability. Avoid the temptation to fragment projects across multiple AI vendors. Instead, plan for a cohesive and scalable approach that integrates across the entire organization. Make sure that the organization consistently realizes the benefits of AI  over the long term.

The bottom line: Generative AI in Medical Affairs amplifies human expertise

Over my career, I’ve learned a lot about digital transformation in pharma‌ — ‌especially Medical Affairs. ‌The new era of digital transformation for Medical Affairs is fundamentally different. Not just because the technology is revolutionary (though it is) but because it’s solving real problems with measurable results.

The key is being brutally honest about both the benefits and the costs. Baseline everything and baseline early. Measure relentlessly. And most importantly, remember that this isn’t about replacing human expertise — it’s about amplifying it‌ — ‌and that exact message needs organizational amplification, too.