Beyond Activity Metrics: Analytics as a Force Multiplier for Medical Affairs

Throughout its evolution from a support function, to owning therapeutic leadership within their respective organizations, Medical Affairs has always functioned in a data-centric environment. As the life sciences industry balances opportunities for next-level patient impact with dynamic market pressures, organizations are investing significantly in analytics capabilities to unlock insights held within internal and external datasets.

In 2025, the Medical Affairs Digital Strategy Council met to discuss the latest results of an ongoing survey of its membership, the Digital Excellence Maturity Assessment framework (DEMA). In parallel, the council initiated dialogue on the critical role of analytics in elevating the role of Medical Affairs. This whitepaper discusses outputs and reflections from that meeting.

While technology will play a significant role in scaling analytics capability, Medical Affairs leaders will drive the realization of business outcomes through domain expertise; identification of relevant organizational questions and their corresponding data sources, and governance and processes to implement the insights.  There has never been a more exciting time to operate at the intersection of data and expanding processing power to drive business outcomes.

Analytics and medical excellence roles within Medical Affairs are relatively new, but becoming increasingly prevalent within lifesciences organizations. The value of analytics for Medical Affairs lies in its ability to communicate in a universal language across the organization, specifically in providing unambiguous rationale for taking certain business decisions. It has been acknowledged that decisions related to investments in content, channels, and HCP engagement have historically been based on historical practices and familiarity. Today, analytics offers the capability to dynamically measure performance and impact with speed and precision. As a result, accepted practices stand to be challenged; for instance investment in field medical staff vs digital outreach. Everything is correct without data.

Pre-launch is a unique opportunity to isolate and quantify impact as Medical Affairs typically supports ph2, ph3 and market shaping activities. Analytics is the crucial mechanism which enables Medical Affairs leadership teams to make and monitor the impact of their business decisions.

What does ‘great’ look like?

An ideal picture is that Medical Affairs works in a highly integrated way with commercial teams during throughout launch and lifecycle planning. The impact of a mature analytics capability is particularly pronounced where Medical Affairs often spearheads identification of potential new clinical indications and patient subpopulations.

Since the early 2000s “Data is the new oil” has become an apt metaphor describing how the digital age will be powered by data.

While this phrase is familiar and intuitive on the surface, some refinement is required in the Medical Affairs context in 3 areas:

In line with the oil refinery metaphor, Artificial Intelligence (AI) has emerged as a catalyst for the refinement phase, which should accelerate the transition to a data-driven Medical Affairs function. It is important to emphasize that organizations need to make their data analytics/ AI ready. Refinement is a critical step (often overlooked and misunderstood) before AI can have its maximum impact.

In line with the limitless nature of data, and yet-to-be-seen impact of continuous innovation, the council offered some pragmatic approaches to AI based on current experience and capability.

Curated datasets with the right RAG (Retrieval-Augmented Generation), to boost accuracy of retrieved information, may yield higher performance compared with large enterprise-level general purpose models. For instance, to address a specific business question, AI may yield more relevant information from a subset of a database rather than accessing the entire CRM (Customer Relationship Management) data lake. While several companies have spent resources to develop the “best” LLM (Large Language Model), recent papers and most futurists predict they will become a freely available commodities, and that resources are better spent in using high-performing SLMs (Small Language Models) in agentic AI system to implement insights regardless of data source and LLM.

AI has been purported to be a panacea for being able to build an integrated data view across diverse functions within Medical Affairs. The question is whether implementation of focused datasets diminishes the ability to present an integrated view of a specific business challenge. While this remains to be seen, a practical consideration is the optimal balance between computation cost and speed to analysis. Demonstrating the impact of Medical Affairs requires integration of clinical trial data, CRM, content and channel data and possibly patient level claims data, which automatically drives towards medium-to-large language models.

Data Governance

Medical Affairs teams are cognizant of the challenge of having actual or perceived linkages between their activities and commercial metrics and outcomes. In a time where organizations must work in a highly integrated way, Medical Affairs must work collaboratively with compliance partners to identify appropriate ways of utilizing robust cross-functional datasets rather than assuming that these types of analyses inherently cross compliance boundaries.

Organizations will have to continually evolve their position on this historical stance. So long as Medical Affairs is not being incentivized by sales performance, and not promoting off-label use, claims data can be a useful data source for measuring its impact (i.e. reduction in care gaps).

Data governance will also need to consider types and quantity of data and prevalent data management regulations. In some organizations raw data are managed at the enterprise level and then made available to functions, who then define use cases and desired insights while balancing against intended usage of data, external regulations, and the temperature of internal compliance teams.

The primary reason for failure of technology projects, particularly AI, is low “AI fluency” across the organization, particularly low understanding of what the technology does, its required infrastructure, and its use cases at the leadership level. The council underscored the importance of change management with a higher emphasis on adoption of analytics into working practices, compared with technology deployment alone. Empirically, every dollar spent on AI implementation should be matched with an approximate 2-3 fold spend on change management. From the council’s experience this aspect is often overlooked or not sufficiently addressed.

Another reason AI projects fail is that point solutions are introduced from a technology standpoint, rather than end-to-end solutions that are business led. An analogy from chemistry is useful here.  If you accelerate a single part of a chemical reaction- you don’t get more of the final product. There’s a rate limiting step, so every step of the chemical reaction needs to be accelerated. Likewise a business-led workflow that defines the utility of AI will yield greater final outcome than a single point solution implemented at a single stage of a workflow.

The power of analytics lies in its ability to objectively challenge widely accepted operational decisions. Face-to-face MSL interactions with HCPs are regarded as the as the “gold standard” that drives clinical practice change.  Our industry has generated compelling data across brands which challenges this long-held belief.

Market impact is a function of Medical and Commercial activities.  A reasonable question is around the relative contribution of these two functions to that tangible outcome. The time and number of data points to statistically attribute impact to Medical or Commercial activities likely prohibitive. Incremental impacts from company interactions are likely additive and cumulative, therefore, a useful perspective is to measure impact of Medical Affairs activities during the clinical stage of a company’s development cycle, where Medical Affairs predominantly drives market engagement.

Recently organizations have explored a new strategy of shifting away from measuring activity, to measuring impact. Several organizations orient their Medical Affairs strategy towards closure of clinical care gaps; for instance the number of patients actually receiving clinical interventions (screening, dose titration, monitoring, diagnostic testing) vs evidence based recommendations . Another possibility is to select a control group to permit A vs B analytics. For instance, 2 evenly distributed cohorts of HCPs receive either a field medical-centric engagement strategy vs a digitally augmented approach, and then compare results of interest such as engagement, satisfaction, and improvements in clinical care gaps. Accurate attribution of market impact to engagement with any function within a company is unlikely, nor is it absolutely required.   

Medical Affairs leadership teams will be most interested in understanding how the sum total of evidence generation, data dissemination, and HCP engagement have positively impacted patient care and outcomes. Measuring changes in clinical care gaps and practice behaviors will take a long time, and the council recommended considering shorter-term leading indicators to predict larger definitive changes. This may include changes in digital behavior (e.g. increased views of company sponsored Medical Affairs website following congress data readout) or engagement patterns (e.g. email alerts and updates on clinical trial progress leading to increased requests for field medical interactions).

The value of analytics is context-sensitive, which considers the intended way in which the analytics will be used. The council identified the following top 3 priorities for the utility of analytics:

A potential model of analytics maturity is presented below.

The general feeling of the council is that the relative value or importance of descriptive, diagnostic, predictive, or prescriptive analytics is not absolute, nor does it follow a defined sequence. Every organization’s maturity pathway will be driven by their business challenges and priorities.

For instance, prescriptive analytics may be employed to recommend specific content/channels for individual HCPs based on historical engagement patterns. If the goal of this strategy is to increase reach beyond the capacity of field-medical teams, this may have immense value, particularly in resource constrained environment.

However, this same analytic capability will have limited value for a different organization who still does not understand the reason for observed clinical care gaps. For the latter organization, focusing on diagnostic and predictive analytics to understand the “why” about a care gap. A solid understanding of this will automatically inform next HCP, content and engagement modality.

The ability of analytics to be a force multiplier for Medical Affairs will be driven by clarity of leadership and guided by domain expertise in the form of meaningful business questions.

It is clear that AI can accelerate analysis and facilitate scale-up of data and analytics practice, the tendency to implement a technical solution should not be at the expense of establishing strong foundations.

As Generative AI is able synthesize data from diverse sources, the organization must create an environment which holds “AI-ready” data. Similar types of data should be tagged similarly, and the meaning of a datapoint in one context should retain its meaning when taken to another use case.

Technical infrastructure should be selected based on its merits and capability to surface information important to the specific organization.  Depending on the types and use cases, data stewardship becomes critical; planning for data migration, data integrity, lifecycle oversight, and understanding appropriate use rules of respective organizations will shape and define data and analytics practices.

Finally, the importance of change management cannot be underestimated. Embedding analytics into Medical Affairs workflows requires shifts in behavior, mindset, and ways of working. Training, communication, and process redesign are critical to adoption and are often underestimated. Organizations that focus on these fundamentals will be best positioned to scale analytics responsibly, adopt AI with confidence, and strengthen Medical Affairs’ contribution to patient care and organizational impact.

We thank the members of the Medical Affairs Digital Strategy Council for their active contributions towards this paper. In particular, the Council recognizes Dat Nghiem, Vice President, Head of Medical Analytics and AI, Pfizer, and Sandeep Gantotti, Associate Vice President, Medical Solutions, Indegene for their roles in planning and conducting this session on the critical role of analytics in elevating the role of Medical Affairs.