2025 Annual Conference - Agenda

Struggling with underperformance, missed opportunities, misaligned HCP targeting, or stalled pull-through? In this hands-on workshop, see how AI Agents enable self-service with business users and reveal the key drivers of performance—across competitor, subnational, market access, and engagement data. Join pharma leaders using Tellius to act faster and win more.

Join MathCo and industry experts in an interactive session focused on improving adoption of data and AI in real-world decision-making. Learn how to apply Design Principles, build Custom Products, and integrate Agentic AI to drive business effectiveness—drawing on proven strategies from both pharma and other industries.

This session will showcase real-world use cases where human expertise is amplified by intelligent agents within Axtria InsightsMAx.ai to drive smarter, faster actions across sales, marketing, clinical, and analytics. Attendees will gain insights into how Axtria InsightsMAx.ai’s agentic AI empowers teams to unlock value from data, improve agility, and optimize performance across both commercial and clinical domains.

Enhance your impact with greater creativity. Learn about why creativity is critical, what kills creativity in the workplace today and how you can enable creativity for you and your team.

This presentation will focus on Axtria’s experience with developing variations on the Dynamic Call Planning (or Dynamic Targeting) problem for clients. The idea of providing promotional guidance in a more timely fashion has been broadly accepted within Pharma and implemented in various ways by multiple companies. In building this guidance, there are many possible approaches, and many possible stumbling blocks to get past.

Among the many issues to consider:

• What time frame will we recommend actions over?
• How does the Dynamic Targeting intersect with an “NBA” (Next Best Action) program, or a traditional Call Plan over a longer period?
• What scope should the plan have?
• Should we try to specify “content” as well as actions to take?
• Should we generate the plan from business rules, models, or both?
• What “target” outcome(s) should we be trying to improve?
• What larger-scale constraints does the Dynamic Call Plan need to accommodate?
• What challenges will we face when implementing the recommendation?

We will discuss our experience in wrestling with all these issues and offer preliminary “best practices” (from our experience) to consider. This is still an active area of investigation and accumulating experience across our industry, with the “final form” yet to be resolved.

Focus Areas:

• Machine Learning

Translating natural language queries into SQL is essential for accessing data efficiently, yet traditional Large Language Models (LLMs) often struggle with complex real-world data due to intricate schemas and inadequate tokenization of domain-specific terms. This challenge is particularly pronounced in the healthcare sector where access to data can be restricted based on users' capability to understand database systems.

To overcome these challenges, this presentation introduces an innovative approach using Knowledge Graphs (KG) and gist tokens. Key learnings will include:

• The principal ways that KGs can be utilized in complex data environments are to enhance accuracy and reduce typical hallucinations in query generation.
• The concept of gist tokens and their role in improving performance and reducing cost while maintaining the context of the complex real-world systems.
• How integrating KGs can significantly improve the precision and comprehensiveness of SQL query generation by LLMs over non–KG approaches.
• Why this is a scalable solution for complex data environments that can broaden accessibility and usability across different systems.

This mixed-methods approach transforms data utilization to enable a wider range of users, from analysts to business leaders, to interact seamlessly with data, ultimately helping to enhance data-driven decision-making while democratizing technology.

Focus Areas:

• Generative AI

Over 50% of patients discontinue chronic therapies within the first year, significantly affecting their health outcomes. While some early therapy drop-offs may be inevitable, personalized touchpoints delivered through coordinated patient services can be crucial for a seamless patient experience, helping patients remain on therapy. ZS and Jazz have collaborated to develop an intelligent patient insight engine that utilizes diverse data sources, including proprietary pharmacy data, and employs an explainable AI model to predict weekly patient discontinuation risk. This engine also provides insights into the reasons behind therapy drop-offs, enabling personalized n=1 recommendations for interventions to be deployed to healthcare providers and patients via digital and field channels, including case managers, nurse educators, and sales representatives.

Join us to discover how an integrated patient data ecosystem, AI, and orchestration can revolutionize the patient support experience.

Focus Areas:

• Predictive Analytics
• Real-World Evidence (RWE)

In the rapidly evolving landscape of AI-driven analytics, organizations must adopt robust data strategies that not only address current challenges but also pave the way for future innovation. This session will delve into how generative AI is revolutionizing analytics-ready data (ARD) creation—accelerating, automating, and enhancing critical processes like descriptive statistics, data transformation, and modelling. By leveraging advanced AI-configured data models and cutting-edge libraries, participants will discover how to streamline data workflows to deliver actionable insights across diverse use cases.

The discussion will emphasize the importance of ensuring ARD is efficiently processed and ready for consumption by analytical systems, showcasing real-world examples of scalable, automated solutions. Attendees will gain practical insights into integrating generative AI into their analytics frameworks, enabling smarter, faster, and more impactful decision-making. Whether you're a data scientist, strategist, or technology leader, this session will equip you with the knowledge needed to harness the power of AI and drive next-generation analytics outcomes.

Focus Areas:

• Data Strategy
• Generative AI

AI for Strategic Market Protection: A Predictive-Prescriptive Framework

Focus Areas:

• Machine Learning
• Predictive Analytics

Join us to learn how to adopt an enterprise mindset that prioritizes customer needs and enhances team dynamics for better outcomes and insights. We will explore what design thinking is, its benefits, and how you can start implementing it in your life today whether you are a cross functional analytical team or planning your next Disney vacation!

Focus Areas:

• Storytelling

Everyone wants you to “think outside the box.” But to truly give yourself a competitive advantage and lead in the business world, you must approach problem-solving to invent an entirely new box. Welcome David Corsaro. For the past 25 years, David has been the "go-to" guy in the business world regarding creativity and problem-solving. Additionally, David is an accomplished magician and mentalist, performing in 4 different off-Broadway magic shows in NYC as well as the world-famous Magic Castle in Los Angeles and he even appeared on the hit TV show “Penn & Teller: Fool Us”....and yes, he fooled them. In David’s seminar, “Problem-Solving like a Magician”, you will not only learn how a piece of magic works but more importantly you will learn WHY it works and how you can take the principles of psychology and neuroscience and apply them to your business challenges.

This session will showcase how AI/ML can enhance Non-Personal Promotion (NPP) strategies for healthcare providers (HCPs). NPP strategies in the pharmaceutical industry have historically been calendar-based, relying on media vendors to deliver promotions on fixed schedules. While field strategies have evolved and are currently embracing advanced AI and analytics to empower sales reps to engage HCPs with personalized and near real-time insights, the NPP space has largely remained unchanged. Promotions are still sent at predetermined frequencies from pre-planned campaigns, often overwhelming HCPs and failing to deliver on expected value. There is a need to leverage AI-driven strategies in the digital omnichannel space.

We will present a comprehensive strategic framework for NPP orchestration that extends AI-driven engagement strategies into the digital environment. The proposed solution leverages machine learning models, enabling dynamic and personalized outreach to HCPs. This AI-powered framework automates decision-making by predicting the best channels and content based on historical interactions and real-time data, thus improving promotional targeting, frequency, and content relevance. Incorporating business rules further refines the AI models, ensuring agility and adaptability to evolving data signals. This approach demonstrates how advanced AI/ML techniques can drive automation, predictive accuracy, and more impactful, data-driven decisions in pharmaceutical marketing.

Focus Areas:

• Machine Learning
• Predictive Analytics
• Storytelling

Pharmaceutical sales representatives often face challenges in accessing and retaining complex, segment-specific information about Healthcare Professionals (HCPs) during training and customer interactions. This session introduces an innovative Generative AI (GenAI)-powered avatar/bot that addresses these challenges by acting as an interactive training and application tool for sales representatives.

Leveraging advanced prompt engineering and domain-specific knowledge, the bot simulates real-world interactions to help sales reps develop a deeper understanding of customer personas and improve communication strategies. This solution represents a significant advancement in sales training by reducing training time, increasing knowledge retention and enhancing confidence during customer interactions.

In the rapidly evolving field of healthcare analytics, deriving meaningful insights from patient data is crucial. Existing AI/ML solutions enable users to profile patients, map their journeys, and drill down into sub-cohorts, but they require significant time for data exploration and often produce too many findings, some of which may be trivial. To address these issues, Cohort Comparison Synthesis (CCS) was developed. CCS uses statistical methods to identify significant differences between patient cohorts and leverages Generative AI to provide comprehensive insights in natural language. The presentation will demonstrate CCS's capabilities in generating detailed cohort comparisons and highlight the integration of Generative AI techniques in healthcare analytics. CCS benefits pharma and biotech professionals by filtering out trivial insights and providing executive summaries of top insights, enabling data-driven decisions with greater accuracy and efficiency. The approach leverages reusable components that generalize to various use-cases, benefiting the identification of actionable healthcare insights. Key components include data sources and processing (IQVIA longitudinal Prescription Data and Office-Based Practitioners Medical Claims Data), statistical techniques (significance testing and effect size calculations), payload definition (using an extensible ontology), and Generative AI techniques (state-of-the-art prompt engineering with GPT-4 LLM).

Focus Areas:

• Generative AI

In today’s competitive pharmaceutical landscape, engaging healthcare professionals (HCPs) with precision and relevance is no longer a luxury—it’s a necessity. Traditional email outreach methods often fall short, relying on static rules that fail to adapt to the evolving behaviors of HCPs. But what if we could move beyond rigid segmentation and instead continuously learn and adapt, delivering exactly the content HCPs need, when they need it?

The talk will focus on a novel Reinforcement Learning paradigm to automate content recommendation for HCP emails. Through a combination of NLP, advanced machine learning, scenario planning, and integer optimization, the pipeline utilizes a broad array of HCP characteristics and Omni-Channel interaction history to dynamically recommend email content that is most likely to engage, while still adhering to core business rules. The pipeline also incorporates an epsilon-greedy mechanism to balance exploration and exploitation of the problem space.

The current talk will uncover through replay analysis that an RL-based recommender system for home office emails could result in a significant uplift in customer engagement. This scalable, intelligent solution can be extended across various customer touchpoints within the pharma industry.

Focus Areas:

• Machine Learning
• Predictive Analytics

More information coming soon.

This talk explores the transformative potential and challenges of Artificial Intelligence (AI). AI can analyze large datasets and predict consumer behavior, revolutionizing personalized marketing with precise targeting. The discussion covers examples of improved advertising outcomes, in both efficiency and effectiveness, due to AI. However, issues like data privacy, algorithmic bias, and loss of human interaction pose significant challenges. The talk provides a balanced view of AI's opportunities and challenges, emphasizing the need for human intelligence to maximize AI's benefits through ethics, creativity, and empathy.

In the talk we will discuss how AI/LLMs often produce wrong results because data is not representative and misleading. We propose HI approach to fix the data first before it goes into the AI/LLM systems.

Focus Areas:

• Generative AI
• Machine Learning
• Predictive Analytics
• Data Strategy

Imagine a comprehensive data and insights repository on all things CML… a digital super Disease Area Expert!

We compiled all market research content over the last two years and used genAI to synthesize the data and build a patient journey. Once tested and validated, this formed the backbone of our digital Disease Area Expert (DAE). We can now use the digital DAE to get instant answers to a range of questions that we constantly grapple with – testing of multiple TPPs, impact of new Takeda and competitor data readouts, drivers and barriers, identifying untapped opportunity areas etc. – eventually we should also be able to run sensitivity analysis on sales and marketing tactics!

We started with market research for our pilot and over time we will layer in additional primary and secondary data such as publications, market access and pricing research, conference abstracts, competitive intelligence, etc., and eventually data from legitimate external CML websites.

The Virtual super DAE is a “live” application that will continue to learn and evolve as we layer in external publications and additional work, ensuring insights are always up to date with the latest data.

Focus Areas:

• Generative AI
• Machine Learning
• Storytelling

Inherited vision loss due to RPE65 mutations affects approximately 1,000 to 2,000 individuals in the U.S., leading to conditions like retinitis pigmentosa (RP) and Leber congenital amaurosis (LCA). This study evaluates the clinical outcomes of Luxturna (voretigene neparvovec) gene therapy, approved by the FDA in December 2017, focusing on vision preservation and disease progression in patients with RPE65 mutation-associated vision loss. Luxturna delivers a normal RPE65 gene to retinal cells, restoring essential protein production and improving the visual cycle.

Using the Optum® Market Clarity Dataset, we conducted a retrospective analysis of 58 patients who received Luxturna between January 2017 and March 2024. Clinical outcomes were assessed through visual acuity, retinal function, and quality of life metrics over an 18-month post-treatment period. Advanced Agentic AI and natural language processing (NLP) techniques were employed to extract and analyze data from unstructured clinical notes, enabling precise evaluation of vision changes and healthcare utilization.

Results indicated significant improvements in vision preservation and a reduction in ophthalmology specialist visits post-therapy.

Focus Areas:

• Data Strategy
• Generative AI
• Machine Learning

Pharma companies need to plan for and manage large-scale events such as national meetings, conferences, launches, etc. These types of events present unique challenges such as involvement of multiple stakeholders, complex handoffs, moving targets and figuring out critical paths. Utilizing bleeding-edge Machine Learning techniques, we have built a toolkit that solves for all of the above challenges. In doing so, it improves adherence to plans, optimizes around impending and ongoing issues, and presents options to meet initial objectives. It also has a set of unique and highly interactive visualizations and reports that facilitate effective presentation/communication by following a storytelling approach. In the presentation, we plan on sharing all of these insights, as well as talk about our experience as we've built this unique toolkit.

Focus Areas:

• Machine Learning
• Storytelling

Agentic AI is transforming healthcare analytics, offering unprecedented opportunities to enhance decision-making, optimize workflows, and ultimately reduce the burden of disease. This session will explore how agentic AI systems, powered by advanced natural language processing (NLP), is enhancing decision-making and innovation across key areas in the pharmaceutical industry.

This presentation will break down the core concept of agentic AI—its capabilities, challenges, and strategic considerations for implementation. Through a real-world case study on MapAI, Komodo Health’s generative AI analytics assistant, discover how agentic AI can be seamlessly integrated with business logic, existing services, and industry-specific datasets—such as structured claims from one of the largest U.S. healthcare databases—to drive intelligent, precise, and scalable analytics solutions.

Attendees will gain actionable insights into the transformative power of agentic workflows, including best practices for ensuring AI accuracy, mitigating risks, and fostering collaboration between AI agents and human experts. We will also provide a forward-looking perspective on how agentic AI is shaping the next generation of healthcare analytics.

Join us to discover how agentic AI is reshaping healthcare analytics, accelerating innovation, and setting the foundation for a more intelligent, data-driven healthcare ecosystem.

Focus Areas:

• Generative AI
• Real-World Evidence (RWE)

More information coming soon.

The landscape of cardiometabolic disease management is shifting with the introduction of GLP-1 receptor agonists, SGLT2 inhibitors, and innovative weight-loss medications. However, the interconnectedness of conditions such as hypertension, diabetes, and obesity requires a holistic patient management approach. As healthcare research pivots toward personalized (n=1) strategies, traditional disease-focused methods may overlook the broader relationships between these conditions.

The primary objective of our initiative was to develop a data-driven framework using Claims and Social Determinants of Health (SDoH) data to identify and quantify patient journey archetypes among approximately 100 million patients with diabetes, obesity, and ASCVD, which reflects variations in clinical characteristics, treatment adherence, and SDoH factors.

This initiative has fostered strategic alignment among stakeholders, revealing intervention opportunities to enhance patient support, address care gaps, and improve health equity. By aligning patient and healthcare professional archetypes, we created a Market map for personalized marketing strategies, demonstrating Novo Nordisk's commitment to advancing patient-centered care in the cardiometabolic field.

Focus Areas:

• Machine Learning
• Storytelling

In this session, Amazon Web Services (AWS) and Amgen will describe how they use Generative AI, AI/ML, and analytical techniques to optimize omnichannel marketing. From recommending the timing and tactics for Next Best Experience of provider-facing campaigns to overseeing content strategy to incorporating sales feedback, we describe how to catalyze a modern omnichannel cycle powered by GenAI.

Focus Areas:

• Generative AI

GenAI agents are good at performing specific tasks, however, are limited when it comes to understanding organizational context and user needs. To overcome this challenge, we have developed a "brain" agent that will infer user needs, apply the right organizational context, and provide the right user experience through a GenAI agentic ecosystem. This has helped in the development and deployment of very persona-specific solutions like Sales Rep and Marketing AI assistants.

This presentation introduces a step-by-step framework for extracting meaningful, unbiased insights from large-scale, healthcare claims data integrated with consumer demographics and machine learning-modeled propensity data. Attendees will discover how to leverage these insights to inform brand media strategies and tactical decisions—including messaging, creative development, and channel mix—even in the absence of historical campaign data.

Brands seeking to engage this or any other population can build a foundation of campaign analytics to better understand consumer behavior, media preferences, and barriers to care. Attendees will leave with actionable insights and tools to navigate the unique challenges of healthcare advertising for their brand or therapeutic area, driving impactful engagement and better outcomes for patients.

Focus Areas:

• Predictive Analytics
• Storytelling

Deloitte has partnered with clients and key alliance partners to develop a vision for the future of Agentic AI in Commercial Life Sciences. By re-imagining the flow of work across sales, marketing, market access, patient services, medical, and decision support, we have identified the critical roles non-human agents play in supporting commercial objectives. Our investment in AI-driven agents provides data-driven intelligence for omnichannel marketing and sales, enabling continual sense and respond capabilities in dynamic therapeutic areas.

Attendees will learn how these AI agents independently perform functions such as data extraction, monitoring and alerting, decision making, content generation, channel execution, and dialogue and explanation. Discover how the interaction between these agents, through data sharing and query exchanges, enables efficient commercial analytics goals, including engagement enhancement, closed-loop measurement, and mid-campaign promotion optimization. This session will showcase how our innovative approach can transform commercial operations in life sciences, driving efficiency and effectiveness in achieving commercial objectives.

Focus Areas:

• Generative AI

• Leverage AI/ML capabilities to identify harder-to-diagnose patient sub-sets within current indications.
• Identify new patient populations and sub-populations that are difficult to map and identify for upcoming indications and brands.
• Analyze patterns in tests, co-morbidities, and misdiagnoses to develop models that surface these patients.
• Model patients at higher risk of severe events to enable timely intervention.
• Predict patient presentation to allow for intervention, education, and action.

Focus Areas:

• Machine Learning
• Predictive Analytics

GenAI Driving Business Value For MSLs In VoKOL & Strategic Portfolio Management: Leading pharmaceutical companies increasingly rely on Key Opinion Leader (KOL) feedback to shape strategies for engaging healthcare professionals (HCPs). There is a growing need for a robust solution to analyze sentiments from Voice of KOL (VoKOL) data, enabling organizations to track key performance indicators (KPIs) and derive actionable insights through an automated reporting process. In this context, Generative AI (GenAI) plays a pivotal role. Its advanced capabilities to process, understand, and interpret unstructured data—such as feedback from KOLs—empower pharmaceutical companies to uncover hidden trends, measure sentiment accurately, and inform strategic decisions with speed and precision. Also, these insights are helping companies in spearheading research and development on specific molecules/drugs etc.

Focus Areas:

• Generative AI
• Predictive Analytics

Generative AI (Gen AI) is transforming the pharma commercial space by enabling greater efficiency, personalization, and decision-making. Agilisium is having solutions built in Gen AI across multiple LLMs.

1. Generative AI driven Global Field Medical Insights Improved insights from CRM data at a global level through accurate translation of representative responses
• Better understanding at a global level of HCP sentiments, medical education needs, and adverse events in individual countries
• Enhanced compliance and regulatory adherence Increased competitiveness in global markets
2. Gen AI driven standardization to quicken MLR approvals Provides instant, actionable insights during pre-review, enabling quicker turnaround times
• Maximizes reviewer efficiency & productivity on critical evaluations
• Scales processes to accommodate document volume increases
• Reduces the time to market for materials by enhancing operational efficiency
3. Summarizes documents and acts a natural language query tool to retrieve data Ease of access, conversational interface and quick information retrieval result in (1) data-driven decision making, (2) boost in efficiency, (3) ensuring compliance with guidelines and (4) consistent high-quality performance
• These insights help accelerate drug discovery, development and customer engagement
4. Accelerating drug approval by creating clinical trial documents using Generative AI Reduce the time to create clinical trial documents by 30%.

Focus Areas:

• Generative AI
• Predictive Analytics

The rapid advancement of data analytics has transformed how industries engage with target audiences, including the healthcare sector. Physicians and healthcare providers are increasingly the focus of personalized outreach by pharmaceutical companies, medical device manufacturers, and health-tech organizations. By leveraging data such as demographics, prescribing patterns, and behavioral insights, these entities can deliver more tailored communications and recommendations. This study explores an innovative approach to segmentation and targeting using unsupervised learning and generative AI (GenAI). Central to this effort is ensuring the right treatment reaches the right patient through the right provider, a critical factor in improving outcomes and building trust in healthcare.

Focus Areas:

• Generative AI
• Machine Learning

The successful organizational adoption of artificial intelligence (AI) in the pharmaceutical sector hinges on overcoming key challenges: siloed operations, lack of synergies, and absence of a clear roadmap. Departments often operate independently, leading to redundant efforts and missed opportunities for broader organizational benefits. A unified strategy that fosters cross-departmental collaboration is essential to unlocking AI's full potential.

A defined roadmap, aligned with business goals, provides structure for AI implementation. A proposed three-step framework emphasizes: (1) Building a strong foundation—ensuring robust data governance, data quality, and adopting mature technological frameworks for scalability; (2) Prioritizing impactful use cases—focusing on high-impact, low-complexity initiatives to build momentum and ensure resource efficiency; and (3) Enterprise-wide adoption—managing expectations through leadership support, change management, and structured planning.

By adopting this comprehensive approach, pharma companies can align their AI investments with broader goals, enabling scalable, impactful, and sustainable innovation.

Focus Areas:

• Data Strategy

Application of GenAI capabilities in supporting coaching activities for the field force.

Focus Areas:

• Generative AI

The PLD data sources we use contain an alarmingly high rate of patient duplicates. In other words, 2 or more patient ids refer to the same patient. We conducted measurements on several PLD data sources for different therapeutic areas. The result is the same: The duplication rate is much higher than what the industry says it is. The extent of patient duplication is not a one-off anomaly but a systematic feature of PLD data sources.

Focus Areas:

• Data Strategy

This talk explores the transformative potential and challenges of Artificial Intelligence (AI). AI can analyze large datasets and predict consumer behavior, revolutionizing personalized marketing with precise targeting. The discussion covers examples of improved advertising outcomes, in both efficiency and effectiveness, due to AI. However, issues like data privacy, algorithmic bias, and loss of human interaction pose significant challenges. The talk provides a balanced view of AI's opportunities and challenges, emphasizing the need for human intelligence to maximize AI's benefits through ethics, creativity, and empathy.

Provider segmentation is often based on simplistic volume metrics. Our poster presents a framework for 4 categories of segmentation and tests the ability to execute that framework. We explore reimbursement trends through the use of McKesson Compile's Patient Ready Open Claims.

A critical component of any pharmaceutical revenue forecasting model is estimating product market share distribution over time. When primary market research is unavailable, share estimates are often based on a product's attributes vs. competitor products, order of entry and/or promotional share of voice. We leverage 3 distinct share estimation models and combine them to predict market share at equilibrium. We then use an innovative approach to model share evolution over time, based on the inputs used for the share estimation modules. We examine the implications of this innovative approach versus one more commonly used.

Learn how tracking product, patient, and prescriber journeys supports advanced analytics and predictive AI to inform commercial strategies.

In the pharmaceutical industry, business development is essential for driving innovation and revenue growth, especially in an environment marked by rapid scientific advancements and the expiration of blockbuster drug patents. While internal pipelines play a role, external opportunities such as licensing, mergers, and acquisitions are increasingly crucial. Emerging biopharma accounts for the majority of the global R&D pipeline, and mid-to-large pharmaceutical companies must identify and acquire promising assets to sustain and expand their portfolios. These acquisitions are pivotal in meeting the growing demand for novel therapies and addressing evolving standards of care. Poster specifically highlights various levers which help enabling accelerated commercial due diligence during the business development phase to assess viability, risks, potential value and ensure that transactions align with strategic business goals.

In this presentation, we'll showcase how integrating structured and unstructured patient data can revolutionize healthcare. Our approach focuses on leveraging diverse data such as clinical, behavioral, and socioeconomic to create detailed patient profiles, improving diagnosis, treatment, and outcomes. We’ll share examples, such as using AI to identify rare disease patients based on symptom patterns and improving intervention efficiency for clients. This session will highlight how advanced AI technologies combined with cross-functional collaboration are driving personalized patient-centric solutions that accelerate healthcare delivery and enhance outcomes.

In the era of generative AI, many analytics solutions promise automation—but deliver inconsistency. Orcana.AI takes a different path: a hybrid AI architecture that combines GenAI for natural language understanding with deterministic (Narrow AI) execution to deliver consistent, high-accuracy insights for commercial pharma teams. This poster showcases how Orcana addresses the limitations of fully agentic GenAI workflows, where compounding errors lead to <60% accuracy. Instead, by separating intent recognition from analytics execution, Orcana achieves 96%+ accuracy with complete explainability and business-rule compliance. Use cases span sales, brand analytics, and forecasting—enabling reps to identify high-impact HCPs, managers to address territory gaps, and strategic teams to uncover root causes in performance shifts. Attendees will walk away with a scalable AI blueprint that puts trust, transparency, and relevance at the center of enterprise analytics.

Pharmaceutical promotional activities have traditionally emphasized increasing customer touchpoints to enhance visibility and engagement. However, with the growing complexity and competitiveness of the healthcare landscape, merely reaching customers multiple times is insufficient. High-quality, relevant, and engaging content is crucial for effective messaging that resonates with healthcare professionals, patients, and stakeholders. Today's content must inform, inspire trust, drive action, and comply with regulatory standards. Understanding audience preferences, sentiment, and behavior helps refine messaging to meet diverse needs. Ultimately, a strategic combination of touchpoints and compelling content drives successful pharmaceutical marketing.

In the competitive landscape of pharmaceutical marketing, reaching healthcare professionals (HCPs) effectively is crucial. Carefully designed experimentation with statistically significant results can help empirically test whether marketing programs/pilots work and avoid suboptimal commercial results. To this end, we have built a sophisticated A/B testing tool specifically designed for the pharmaceutical industry.

The objective of this work is to introduce a decision engine framework designed to optimize investments at the brand level with the ultimate goal of enhancing ROI and revenue generation. The framework offers a transparent, interpretable marketing mix model that links key assumptions and marketing mix results to a broad range of potential investment scenarios. A key feature of this approach is its explicit connection between inputs and outputs enabling decision makers to understand the sensitivities and tradeoffs involved in allocating media & Field Force spend. This transparency allows for a comprehensive view of the impacts of increasing or reducing media spend, or reallocating resources on overall performance. By building a robust analytics framework, the solution serves as a one-stop-shop for portfolio level media investment decisions. It empowers users to compare outcomes across multiple optimization scenarios, providing a clear view of how different budget allocations will influence brand performance. This capability helps senior leadership to make data-driven decisions with greater confidence, aligning investments with the overall business objectives of maximizing ROI. The significance of this decision engine framework lies in its ability to dramatically improve the speed at which insights are generated, fostering quicker and more informed decision making. Additionally, it provides a high level of transparency, crucial for senior leadership to assess the direct impact of their media performance and make adjustments based on a well supported understanding of potential tradeoffs. The framework also effectively incorporates business context, ensuring that the decision-making process is not just data driven but aligned with broader strategic goals. Ultimately this approach transforms the process of media & Field Force investment optimization, ensuring that decisions are made with a clear, actionable understanding of their potential outcomes, leading to more efficient and effective media spend management at the brand level.

Real-world data sources like claims and EMRs are crucial for successful pharma launches, supporting opportunity identification, performance tracking, segmentation, and predictive analytics. However, these datasets often have gaps in clinical capture, falling into two key categories:

Breadth Gap – Coverage varies across regions and accounts, with most datasets capturing only 60–80% of the U.S. population.

Depth Gap – Even when patients are captured, key information (e.g., place of service, procedures) may be missing.

These gaps hinder granular market and brand performance insights. To address this, we developed an AI-based data enrichment framework focused on bridging breadth gaps, enabling scalable enrichment of fields like place of service, treatments, diagnoses, and biomarkers. The approach has been explored in more specialized markets (like oncology and rare) in the past, however exploration in the mass markets have been limited, primarily due to the volume of data involved. We refined and fine-tuned the data enrichment approach for a mass market (Acute Pain) and helped increase capture of different procedure types (e.g., orthopedic vs foot vs ENT, etc.) as well increase capture of Place of Service information.

This poster explores how generative AI enhances business intelligence platforms for field managers and sales representatives. The solution generates executive summaries with actionable insights tailored to specific user roles, providing both data-driven recommendations and their underlying rationale. A custom chatbot enables natural language interaction with data through text and voice commands, supporting both descriptive and prescriptive analytics while maintaining strict security protocols. Voice integration, including Siri for CarPlay compatibility, enhances accessibility for mobile users. The platform reduces cognitive load by synthesizing complex data into coherent narratives, allowing users to focus on strategic decisions rather than data interpretation. By explaining the reasoning behind recommendations, it helps sales representatives better prepare for healthcare provider meetings while ensuring sensitive information remains protected through row-level security measures. Learning objectives include understanding generative AI applications in business intelligence tools and exploring natural language interactions in data visualization contexts. The methodology integrates large language models with Power BI and Tableau frameworks, with attendee engagement facilitated through mockups, wireframes, and interactive discussions.

The pharmaceutical industry is experiencing a paradigm shift due to the high volume of data that flows in, with advanced analytics being leveraged to harness this data effectively. This transformation relies on a robust and strategic approach to data structuring and management. This abstract outline a comprehensive framework for creating a data ecosystem optimized for analytics, tailored to the unique needs of pharmaceutical companies, specifically in Affiliation Data Management. Key areas of focus include designing scalable data architectures that accommodate the complexity of multi-source data integration. This integration spans multiple levels of Pharma Affiliations, such as HCPs (Healthcare Providers), HCOs (Healthcare Organizations), Regional Parents, and IDNs (Integrated Delivery Networks), to identify which affiliations are target customers for drug commercialization. The module includes data consolidation, data standardization, data cleansing, eliminating duplicates and unwanted affiliations, and re-ranking/re-scoring the final affiliation output. The resulting affiliation data can be directly used for advanced analytics and reporting purposes.

Trinity's Generative AI Development Life Cycle (GDLC) methodology brings a modular and enterprise-scalable framework to life sciences, enabling rapid prototyping and deployment of Generative AI (GenAI) solutions. This approach ensures adaptability, efficiency, and innovation, empowering pharmaceutical organizations to achieve faster outcomes and greater flexibility. The GDLC leverages a modular design, allowing teams to efficiently adjust and scale solutions as new business needs arise. By integrating cutting-edge GenAI models, including multimodal models capable of generating outputs across text, images, and other data types, the methodology expands the range of actionable insights and enhances the versatility of AI applications. Key Objectives and Benefits:

1. Accelerated Implementation
2. Streamlined Development
3. Enhanced Productivity Across the SDLC
4. Rapid Output Delivery

By implementing the GDLC framework, pharmaceutical organizations can unlock new opportunities for innovation while maintaining agility and scalability. This approach not only accelerates timelines but also empowers diverse roles within the SDLC, enabling them to drive impactful results with GenAI.

The presentation titled "From Insight to Impact: Last Mile Analytics for Field Force Effectiveness" explores the integration of AI to enhance field force performance and market dynamics within the pharmaceutical industry. It introduces a comprehensive ML/AI framework designed to transform traditional performance analysis into strategic actions. The project emphasizes data quality assessment, performance segmentation, predictive intelligence, and prescriptive optimization. By analyzing U.S. geographies, the framework identifies performance drivers and develops machine learning models tailored to specific market segments. The expected outcomes include data-driven benchmarks, actionable geographical recommendations, and improved field force effectiveness. The presentation also addresses the importance of uncontrollable market conditions and suggests opportunities to revise strategies, aiming to bridge the gap in geographic performance and optimize market share growth. The insights derived from high-performing areas are leveraged to enhance the performance of underperforming territories, providing a roadmap for pharmaceutical leaders to maintain and enhance their market position.

An AI driven solution to Business Development and Portfolio Management is an industry-first attempt. In this presentation, we will not only demonstrate the feasibility of an AI driven solution as proof of concept but also show that it can deliver value to decision makers and make a real difference in terms of highlighting promising molecular assets for licensing and potential firms for partnership that were previously not considered.

Marketing mix insights frequently shape consequential promotional brand strategic and tactical decisions. However, marketing mix results are generated under enormous ceteris paribus* assumptions of which the most important is the delivery of those details in terms of message, delivery and even HCP perceptions of those encounters remain constant. Details that are optimized in terms of message, delivery and perceptions based on in-market response can usually be made 5% to 20% more effective.

We will show a robust machine learning approach combing deep-dive primary research data with prescribing data to provide core insights that all brands need for evaluating their sales force and thinking about sales force investments:

• Potential Upside: How much more effective can brand details be for made for driving new patient starts, switches and add-ons for our product by adjusting detail messaging, delivery and encounter perceptions?
• Brand Consequences: What is the projected incremental new patients for our brand from adjustments to details, and the potential revenue of those incremental?
• Impact on Sales Force Investment Decision: How would the returns for the sales force differ overall and by HCP segments if details are adjusted for optimal impact, and how might that affect thinking about sales force investment and tactics?

The growing complexity of data systems in the life sciences industry necessitates efficient approaches to error detection, analysis and resolution—especially in mission-critical domains such as patient safety, drug development and healthcare operations. Genentech’s data science team wanted to explore the possibility of leveraging LLMs for tracking and summarizing errors in complex pipelines. The goal was to reduce operational time spent on debugging, navigating and locating errors in complex ML pipelines/products. Debug logs of any complex pipeline often span over 60,000–100,000 lines when tracked extensively and provide critical insights into pipeline performance. They were key to solving the problem, but there were challenges associated with them in terms of sheer size, unstructured nature, etc. The methodology leverages GenAI and effective parsing techniques to mine out the relevant error information and then processes the information in effective chunks to generate a detailed and precise summary. Such automated error summaries enable faster cross-functional team response, driving operational efficiencies in time-critical workflows. This approach achieved the following outcomes:

• Effective Information Retrieval: A systematic pipeline which leverages regular expressions (i.e. Regex), effective chunking strategies and retrieval mechanisms to process information into manageable chunks of information.
• Automated Error Summaries: Prompt engineering techniques to process each chunk through the LLM, generating detailed summaries that explain root causes, error impacts and potential resolutions.
• Reduced debugging time: As a result of detailed & precise summary provided by the LLM, we were able to reduce the time taken to debug by ~35% on an average. This presentation will cover the design, development and implementation of the entire solution, including real-world examples.

This research explores a novel approach to time-series forecasting in healthcare by developing a Large Language Model (LLM) foundation model. Leveraging a massive dataset of over 100 million IQVIA time-series trends, including procedure demands, sales, and total prescriptions (TRx), alongside publicly available data spanning over 20 years, this approach aims to significantly enhance predictive accuracy across various healthcare forecasting applications. The model's transformer-based architecture, incorporating self-attention mechanisms, is designed to efficiently capture complex temporal dependencies among various historical time-series trends. In the use case demonstration, we will fine-tune the pre-trained foundation model with Novo Nordisk’s in-house data, specifically focusing on forecasting the sales of key products like Ozempic and Wegovy. The project's outputs include the pre-trained model checkpoints for broader application utilization, demonstration of forecasting use cases, benchmark results. This work aims to establish a new paradigm for time-series analysis using generative AI, offering more intelligent, adaptive, and scalable solutions for healthcare forecasting.

Understanding risk factors for chronic illness development or progression is crucial for personalized healthcare. It enables lifestyle interventions, targeted treatments, timely screenings, and precise therapy administration. While traditional observational Real-World Evidence (RWE) studies are cost-effective compared to clinical trials, they are limited by biases and a narrow focus on known factors. Machine learning (ML)-driven RWE studies address these limitations by identifying hidden risk factors, predicting outcomes, and guiding tailored treatment strategies to advance precision medicine. We developed a ML-Based Outcome Model, a 2-step process, where tree-based classic machine learning models using administrative claims data are first used to identify and interpret risk factors for use in classic statistical outcome based models. We present a case study of patients with Refractory Focal Onset Seizures (R-FOS) with a goal to identify the clinical risk factors and their impact on outcomes related to R-FOS that included increases in the number of falls, injuries, hospitalizations, ER visits and the related costs.

This poster embarks on an exploratory journey into the realm of tech-driven specialty pharmacy collaboration, with a primary focus on patients navigating the complexities of specialty and rare diseases. Currently inundated with manual operations characterized by delays and nebulous accountability, our aim is to shed light on the potential of a more tech-centered approach. Drawing upon CRM system pathways, we delve into the impact and benefits that technology could offer to specialty brands, especially in optimizing fill rates.

The effectiveness of TV creatives tends to decline over time as audiences are repeatedly exposed to the same content. TV creative wear-out analysis is essential for identifying when a creative's impact diminishes, enabling optimized media spending and effective creative rotation. This is particularly important in competitive markets like the pharmaceutical sector, where maintaining a strong presence is vital. Our approach enhances traditional wear-out analysis by incorporating innovative components. We use real-time data to track customer interactions, offering a timely and cost-effective alternative to traditional methods. Additionally, we introduce advanced metrics to quantify wear-out levels, providing deeper insights into the lifespan and effectiveness of TV creatives. Lastly through we determine the optimal frequency for rotating creatives through simulations, ensuring sustained audience engagement and maximizing return on investment. This comprehensive framework provides a practical solution for optimizing media strategies and enhancing communication effectiveness. By incorporating advanced analytics into the creative process, we can enable data-driven decisions that amplify the impact of marketing campaigns.

This poster details the approach used to develop a multi-product rare disease launch forecast to help company leadership and commercial teams understand various market forces at play during the launch, and to provide the organization a more accurate picture of the risks and launch year potential of the products.

Navigating the complexities of data management often feels like confronting a swamp. This challenge was evident for a pharmaceutical company launching a specialty drug targeting patients with a complex treatment journey, seeking to establish a unified data ecosystem to prioritize and integrate diverse data sources. This poster will tell the story of that PharmaCo, and how they employed a framework of critical questions to assess data sources and vendors. Pre-launch activities focused on gaining market insights, identifying KOLs, understanding patient access barriers, and analyzing the patient journey. Post-launch, the emphasis shifted to operational use cases such as sales analysis, segmentation, and forecasting. Practical considerations included evaluating analytical approaches, compliance with regulations like HIPAA, and linking extensive datasets. Timing was crucial, aligning data processes with product launches. A detailed vendor checklist emphasized data quality, demographics, and bias, alongside the logistics of tokenization and linking. Strong, collaborative partnerships were critical for long-term support and alignment. This case highlights the importance of strategic planning, robust evaluation, and teamwork in creating a scalable data foundation. By addressing these aspects, the PharmaCo successfully developed a future-proof ecosystem to support growth and innovation in the competitive pharmaceutical market.

OmniBERT is a transformer-based AI architecture designed to enhance next-best-action (NBA) recommendations in the pharmaceutical industry by integrating heterogeneous data sources and continuous responses. Traditional NBA models rely primarily on brand promotion data and HCP demographics, overlooking the impact of non-promotional events like patient progression and real-world occurrences. While prior transformer-based architectures like PharmaBERT and RareBERT have addressed some challenges, they struggle with temporal data and continuous responses. OmniBERT extends RareBERT by incorporating claims, marketing events, and other external data, improving HCP engagement strategies. It integrates a Constrained-based Genetic Algorithm (CBGA) to analyze multi-dimensional sequential data, optimizing NBA recommendations. Experiments show that OmniBERT outperforms CNN and LSTM models, with a 10% improvement in learning from customer journeys and a twofold increase in topline uplift opportunities. The architecture also enhances performance through improved model design and an additional improvement leveraging alternative data sources. Moreover, it reduces model training time by upto 90%, making it a more efficient solution for personalized omnichannel strategy optimization in pharma.

This poster explores the transformative role of predictive analytics in surfacing differences in healthcare payor processes for GLP-1 drugs. As the market for these drugs expands with emerging indications, understanding the adjudication process becomes crucial for timely patient care. This study delves into the factors that influence claim approval times by payer type, that in turn impact a growing patient base within the U.S. Key learnings from this research will include:

• Payer Type Influence: The different factors that drive the speed of GLP-1 claim adjudication between Medicare and commercial insurers.
• Demographic Impact: The effect of state-level demographics, including patient numbers and age ranges, on adjudication timelines between Medicare and commercial payers, revealing regional disparities in processing efficiency.
• Predictive Analytics Application: By leveraging predictive analytics on remittance data, the study identifies key drivers of GLP-1 claims processing, offering insights into enhancing efficiency and patient care outcomes.

This research provides a roadmap to how predicative analytics can be applied to remittance data to attain actionable insights for pharmaceutical manufacturers, payers, and providers to expedite time to treatment, ultimately improving patient access to vital therapies.

We focus on developing an effective and reliable framework to measure the impact of personalized promotions on Healthcare Professionals (HCPs) and identifying the key factors that drive successful personalized campaigns.

Pharma uses claims data to analyze behaviors that help them better understand the patient journey. To many pharma, that first claim for their brand is perceived as the start of the patient’s journey with their product. However, this new-to-brand first-claim is not the start of the journey, but rather the result of one. We believe the journey truly starts at the first capture of a prescriber’s interest or intent for a brand, as seen across two novel data sources, Prescriber Benefit Check and Prior Auth (PA) data. This poster quantifies the prescriber’s original intent against the influence of these processes relative to what, if anything, the patient ultimately adopts. Early investigation of these administrative processes reveals data that is noisy and complex, a true reflection of the convoluted and confusing process that prescribers face when trying to access brands. With proper methods applied to the data to prepare it for analysis, pharma may benefit from transparency into measures of prescriber challenges, abrasion and their preferences with individual brands, to better inform profiling and targeting of prescribers.

As the presence of Nurse Practitioners (NPs) and Physician Assistants (PAs) in the healthcare landscape continues to grow, accurately capturing their clinical specialty is increasingly important for pharmaceutical targeting and analytics. However, many NP/PAs are not assigned meaningful specialty classifications in traditional datasets, leaving gaps in provider engagement strategies - especially for specialty products requiring precise outreach. To address this, IQVIA developed a scalable, multi-method approach to reassign NP/PA specialties using its Longitudinal Access and Adjudication Data (LAAD). The methodology integrates address-based clustering, historical prescribing patterns, and therapeutic market concentration to inform a Random Forest classifier. This machine learning model predicts provider specialties with associated confidence scores. Results are then synthesized using hierarchical logic, combining computational outputs with business rules to finalize assignments. This approach reassigned 49% of NP/PAs with prescribing activity in 2023 to more actionable specialties, enhancing the value of downstream analytics and commercial strategies. The model is customizable, allowing tuning for different confidence thresholds and business needs. By combining AI-driven modeling with expert oversight, this solution enables more precise HCP segmentation, improving targeting strategies and aligning with the evolving dynamics of specialty care.

This poster will showcase how AI complements human expertise to transform Medical Affairs strategy and execution. In Medical Affairs, vast amounts of RWE and complex datasets often overwhelm traditional analytics approaches. Through ML, organizations can align insights from this data to unmet medical needs, while GenAI provides leadership with strategic insights and enhances accessibility to critical information. Additionally, AI empowers medical science liaisons (MSLs) by orchestrating engagement strategies across channels and delivering actionable insights to guide interactions with healthcare professionals. This integration of AI and human intelligence elevates analytics beyond data processing, enabling strategic decision-making, enhanced customer engagement and streamlined operations. The poster underscores how AI can amplify human expertise, providing practical examples of how analytics-driven innovation drives both organizational success and improved healthcare outcomes.

Leveraging diverse datasets to discover actionable insights is critical in commercial healthcare analytics. Yet, current analytical tools often provide fragmented perspectives, analyzing structured or unstructured data in isolation. This limitation hinders a holistic understanding of business challenges. Our innovative system combines agentic AI and conversational GenAI to bridge this gap by creating a unified analytical framework that integrates structured and unstructured data, delivering actionable and goal-oriented insights.

We’re surrounded by data—but what truly moves people isn’t numbers, it’s meaning. Storytelling, especially when paired with compelling visuals, transforms raw information into something emotionally resonant and memorable. Like Shakespeare’s timeless tales, effective stories explore universal themes, forging human connections that transcend complexity. GenAI-enabled storytelling enhances this by evoking emotion, humanizing insights, creating common ground, and inspiring action. Emotion acts as “cognitive glue,” helping information stick. A chart showing therapy outcomes becomes far more powerful when paired with a patient’s personal journey—suddenly, it’s not just data, it’s a life changed. Equally important, storytelling enables matching—mirroring the audience’s language, values, and worldview—and reciprocity—creating space for shared understanding and empathy. When stories align with what the audience cares about and reflect their reality, trust deepens and engagement rises. Business leaders may overlook metrics, but they’ll respond to a narrative that speaks their language and shows a shared goal. Storytelling also bridges gaps between teams, roles, and functions by building mutual respect and perspective. In the end, it’s not the data that drives action—it’s the story that meets people where they are and gives them something meaningful in return.

Language models such as GPT have demonstrated impressive performance in modelling natural language. This is because pretrained GPT models capture fine-grained representations of language from large corpora of data. Our goal in this work and ongoing research is to build GPT models on claims data which can understand the language of patient medical events. To accomplish this we train a GPT model on patient sequences with the goal of predicting the next medical event, analogous to GPT models predicting/generating next word in a sentence. Patient sequence data has a different structure than natural language, which presents unique challenges when adapting GPT to train on claims data. We focus on the asthma market for this GPT model as we can use a subset of patient codes that focus on treatment, diagnostic, and procedural pathways specific to asthma in the hope this can be a good start for a GPT model to begin learning the “language” of claims data. In addition, we incorporate multiple vocabularies and temporal dependencies in the data preparation stage and employ rigorous evaluation frameworks to measure performance. With a trained GPT model that understands patient sequences in the asthma market, we continue by fine-tuning this to predict biologic-initiation. This is accomplished by adding a classification layer on top of the GPT model. We then compare the fine-tuned classifier performance to tree-based models, LSTM models, and other baselines. Our goal is to continue this work and build better and better foundational GPT models that are market specific, such as asthma, and then possibly build larger scale general claims data models which can be used for generative tasks, as well as fine-tuning for classification in disease-detection, treatment initiation/switching, as well as a host of other use cases.

Pharmaceutical sales representatives face numerous challenges, from managing complex interactions with healthcare professionals (HCPs) to staying updated with the latest medical advancements and adapting to changing market conditions. Current AI tools, like next best action (NBA) models, chatbots, and AI rep navigators provide support by offering recommendations based on past data and set guidelines. However, these tools often struggle with real-time adaptability and autonomous decision-making, leaving reps to adjust strategies and handle intricate sales processes manually. Agentic AI effectively addresses these challenges. Unlike traditional AI, it uses individual “agents” that autonomously plan, reason, and execute complex tasks. Agentic AI can analyze data from various sources in real-time with minimal human intervention, foregoing the need for generative AI-style prompts. This ability is especially valuable in pharmaceutical sales, where swift and proactive engagement is crucial. Agentic AI enhances current AI solutions by using a multi-layered reasoning framework that continuously adapts strategies based on evolving data. This means agentic AI can offer more than just simple suggestions; it provides proactive and adaptive engagement with HCPs. This translates to better support for pharmaceutical reps throughout the entire sales process, from Pre-Visit to Post-Visit stages. In the Pre-Visit phase, agentic AI not only predicts the next best actions but also anticipates potential challenges, preferred product lines, and effective engagement tactics. This helps reps approach each interaction with a tailored strategy. During the Visit phase, agentic AI generates real-time conversation strategies, guiding reps through detailed, multi-step interactions with HCPs. This ensures reps can respond to evolving discussions and provide relevant information on the spot. In the Post-Visit phase, agentic AI automates follow-up tasks, learns from each interaction, and continuously improves its engagement approach. This reduces manual work, boosts efficiency, and allows reps to focus on building stronger relationships with HCPs. By introducing autonomous, context-aware decision-making into each step of the sales process, agentic AI significantly enhances the capabilities of existing AI tools. The result is better sales outcomes, increased personalization, and more efficient workflows. Pharmaceutical reps can now rely on a powerful AI partner that supports their efforts and drives them toward greater success in a competitive industry.

A Target Product Profile (TPP) is a crucial document that informs strategic and regulatory workstreams within the pharmaceutical industry. Traditionally, the creation and updating of TPPs demand extensive desk research and deep therapy area expertise. These processes are often time-consuming and inefficient, posing significant challenges to timely and accurate TPP development.To address these challenges, a novel solution leveraging Generative AI (GenAI) has been developed to streamline the creation and updation of TPPs. This solution utilizes GenAI to scrape and extract pertinent information from the web, drastically reducing the time and effort required for TPP development. By automating the extraction and refinement of content, the solution is anticipated to cut the time spent on TPP creation by more than 50%. Furthermore, it eliminates inefficiencies associated with the periodic updation of TPPs, ensuring that they remain current and comprehensive.

As healthcare policies evolve and screening guidelines are updated, the adoption of mammograms and colonoscopies is expected to rise, leading to earlier cancer detection. The U.S. Preventive Services Task Force (USPSTF) recently lowered the recommended starting age for colon cancer screening from 50 to 45 and revised breast cancer screening guidelines, influencing screening patterns. Additionally, the growing adoption of non-invasive, at-home colon cancer tests is expanding access, particularly in underserved regions, creating new market opportunities.This research examines how real-world data (RWD) derived from medical claims can provide insights into the effects of these guideline changes and the increasing use of non-invasive screening methods. By continuously analyzing up-to-date, real-world claims data, healthcare organizations can better track changes in screening trends, healthcare utilization, patient behavior, and resource demand. This facilitates informed decisions across clinical outcomes, regulatory submissions, and market access strategies, particularly in underserved populations.

Drug wastage in injectable and infusion treatments poses a significant challenge due to fixed vial sizes, variable patient-specific dosages, and the inability to reuse remaining medication once opened. Regulatory, medical, and legal constraints, along with the economic infeasibility of resizing vials, further exacerbate the issue. Despite increasing pressure from regulatory bodies—such as CMS mandates and penalties for wastage—manufacturers continue to face difficulties in accurately quantifying wastage due to fragmented data, incomplete patient journeys, and missing vitals. Since 2018, over 500 million units of the top 50 drugs have been discarded annually, highlighting a pressing need for accuracy in assessing wastage. In this poster presentation by Indegene and Pfizer, we explore a structured AI/ML-powered approach that integrates FDA-approved dosage cycles, strategic real-world data (RWD) selection, and multi-source triangulation to accurately calculate drug wastage in oncology.