Gen AI - A Boon or Curse for Healthcare & Pharma

A Brief on Generative AI

Generative AI, a type of artificial intelligence technology that can produce myriad of content which includes but not limited to imagery, synthetic data, text or audio. Generative Adversarial Networks(GANs) were introduced in 2014, a type of machine learning algorithm that generative AI could create near authentic videos/images/audio of real people. With rapid advancements in LLMs(large Language Models) containing millions parameters, AI models produce engaging texts, paint photo realistic images and even create short videos on the fly across multiple media platforms.

To process content Generative AI models combine various AI algorithms to generate text, various natural language processing techniques transform raw characters to speech, entities and actions which are represented as vectors using multiple encoding techniques. For generating realistic humans faces techniques such as GANs and variational autoencoders (VAEs) are used for training the dataset.

AI & Healthcare / Pharmaceuticals

Gen AI can create value across the entire business chain through its ability to synthesize both structured and unstructured data, & generate bespoke visual, textual, and even molecular content. Although the technology will affect all industries, it will have a particularly strong impact in pharmaceuticals.

Most specific use cases will fall in one of four main categories: knowledge extraction, content and compound generation (for instance, generative chemistry), customer engagement (such as services for healthcare providers and patients), and coding / software generation.

AI is not the sole solution but an enabler of efficiency, robustness and efficacy.

Enable faster enrollment through AI-driven models which can predict underlying drivers for a given indicator

Over the past decade demand for trial participants has increased almost 10%-16% because of larger trials; comparing 2019 to 2022, the total target enrollment of trials starting in those years grew 18 percent, from 2.2 million to 2.6 million. However, the actual participant growth has reduced by 6%-10% in the same period.

The enrollment challenge is further increased by an influx focus on precision medicine, whereby study groups are tightly defined with the goal of developing medicines that are highly effective in subsets of participants. For example, as of 2022, more than half of oncology trials are estimated to have a biomarker-defined target subpopulation which makes enrollment rate much slower.

By using AI-models that analyze Electronic Medical Records (EMR) & Electronic Health Records (EHR), suitable patient types can be assessed faster. This targeted approach enhances trial success rates by focusing on those most likely to respond to treatments. Additionally, AI streamlines the vast data handling involved in tests, aiding in study arrangements, and speeding up the consent process. These models typically can identify opportunities by predicting sites that will recruit the fastest (vs the sites that won’t) an acceleration of 15 to 20 percent. These models better account for factors such as site congestion (how many trials are currently occurring at a given site) and can expand the work area of available sites.

Enhancing day-to-day analytics digitally

The fast-evolving landscape of data, analytics, and AI is already revolutionizing the way medical affairs operates. Analytics can be integrated into the daily decisions of everyone in medical affairs to generate and test hypotheses, inform prioritization, and rapidly measure impact. In predictive analytics, AI models are indispensable in forecasting treatment outcomes. They can potentially reduce reliance on animal testing and hasten the preclinical phase. Post-trial, AI-driven text summarization tools analyze extensive data for key insights. This accelerates the analysis and supports strategic decision-making, showcasing Gen AI’s profound impact on the efficiency of clinical trials.

For example, data that quantifies the education needs of individual HCPs can be used to focus medical engagement on those who most need it and monitor its impact on care delivery. Medical affairs can focus on a data infrastructure that allows timely access to the right data by the right user for each use case. Significant change management will be needed to build a culture of data-driven decision making within medical affairs.

When looking at graphs that show genes that interact with certain substances, engaging a conversational AI chatbot makes their task easy and user friendly. They can ask the chatbot how many genes are in a particular graph, find associations between genes and diseases faster, provide information, and ask even more detailed questions to get information instantly.

Expanding existing rudimentary data sets

Traditional analytical AI models, such as those currently used to promote stakeholder engagement and help diagnose diseases, will continue to capture value. The difference is that new gen AI applications will significantly enhance their capabilities. Most of the work involves adapting models to a company’s internal knowledge base and use cases because Gen AI models account for only about 15 percent of a typical project effort. Given the complexity and uniqueness of the data this is particularly true in the pharmaceutical industry. To succeed with gen AI, companies must integrate it across complex workflows to promote adoption and impact—a reality that highlights the need for effective change management.

Around 65-70 percent of digital transformations fail leaders ignore the importance of managing change. To understand issues such as molecular structures, clinical operations and patient data , Companies will need to build an intelligence layer. To create data infrastructure that can run and internal & external data sets, a multipronged approach will be necessary. Data scientists will need to collaborate closely with leaders in business strategy, medical affairs, and legal and risk to set priorities and execute strategies.

Enable automating repetitive non-core activities

At the grassroot level, typical medical affairs strategy is siloed, driving companies to revisit the collaboration model across medical, R&D, and commercial. Successful medical affairs teams can commit to developing integrated medical strategies for each therapy and collecting and sharing medical insights about external stakeholders.

With expanded expectations for medical affairs, Teams will need to prioritize the high-impact activities while deprioritizing but still delivering traditional activities, such as medical review and medical-information standard responses. Automation with recent breakthroughs in gen AI, could significantly increase efficiency across all of medical affairs. For example, gen AI can scan a broad range of sources, including external interactions, social media, surveys, trial data, and real-world data, to instantaneously produce insights and recommendations for evidence generation plans, medical strategies, and other applications. An emerging gen AI capability could also observe and generate actions in the digital or physical world to optimize medical processes such as medical review.