The Next Decade of Patient Engagement: A New Operating Model

We're planning a clinical development program in a specific disease area. Which patient communities should we engage, and what relationships do we already have?

Ask that inside almost any pharmaceutical company today and you will get an answer eventually. It will take weeks. It will involve three or four people forwarding spreadsheets, one person who happens to remember a partnership from two years ago, and an agency deck that was accurate when it was written. The answer will be partly right, and nobody will be able to say which part.

Swap clinical development for a market access dossier, a policy position, a launch plan or a post-approval support program and the shape of the problem does not change. Somebody needs to know which patient communities matter here, and what the company's history with them already is.

The gap is not a people problem. Patient Engagement and Advocacy (PE&A) teams are among the most capable and most committed people in our industry. It is an infrastructure problem, and it is the defining constraint on what these teams can do next.

Why this is getting harder, not easier

Three things have changed at once.

1. The patient ecosystem is larger and more sophisticated than it has ever been. Patient communities educate patients, run registries, fund research, publish evidence, convene expert panels, and hold real influence with regulators and payers. The number of organizations worth knowing about has grown faster than any team's capacity to know them, especially at regional and local level.
2. PE&A teams now support the entire lifecycle, not a single moment in it. In development, that means research priorities, protocol design, endpoint selection and recruitment. In evidence and regulatory work, it means patient experience data, submissions and advisory committees. In access, it means HTA dossiers, reimbursement cases and the argument for what a treatment is worth to the people who live with the condition. In advocacy itself, it means policy positions, awareness and disease education, health equity programs, community partnerships and the long tail of post-launch support. The internal customer base has multiplied, and each of those customers asks in a different vocabulary and on a different timeline.
3. Knowledge remains scattered across functions and therapeutic areas. The same community may be engaged by clinical, medical, market access and advocacy teams in the same year, with no shared record of any of it. When someone changes role or leaves the company, the relationship history leaves with them.

Two capabilities, neither owned

Answering that opening question, and others of that nature, requires two things at once.

The first is external intelligence: which patient communities exist in this disease area, which ones are credible and active, what research and evidence they hold, which patient registries they run, how they connect to each other. This is a landscape that changes constantly and is documented nowhere in one place.

The second is institutional memory: what relationships your company already has, who owns them, which projects and engagements have run, what patients told us in them, and what was promised in return. This lives in people's heads and inboxes.

Neither capability is exotic. Both are simply absent as infrastructure, and it is worth being concrete about what that means. It does not mean the work is not happening. It means there is no system of record for it. No single place where a colleague in another therapeutic area can see which communities matter in a disease and why. No standing source that stays current between programs, so each new program starts the landscape work from scratch. No default path for an insight to travel from the meeting where it was heard to the team that needs it. The capability lives in projects and in particular people, not in the company.

The right patient communities for a given program sit at the intersection of the two, which is exactly where most organizations have the least to work with.

What the current state actually looks like

Over a period of 12 months, we assessed 34 pharmaceutical companies against our framework for how patient community work gets done today, scoring three of its four dimensions. Scores were assigned by us, not self-reported.

Three things stand out.

• Community discovery scores highest, and that is not good news. The typical large and mid-size company sits at "agency-supported or led". The capability exists, but it is rented rather than owned, and it resets with every new program.
• Relationship management and insight capture sit between 1.0 and 1.7. In plain terms: spreadsheets, email threads and individual memory. This is the institutional memory half of the problem, and it is where the industry is weakest.
• The same pattern was visible across company sizes. Larger companies scored higher, helped by the internal systems they have built. Those systems do not move the insight picture: insight capture sits near the bottom of the scale at every size.

“We scaled back our CRM deployment due to low user adoption”

Multiple responding PE&A executives

Generic CRMs and other relationship tools have been tried, and in many cases they have not held. The reason usually has little to do with the software being bad. These tools tend to ask teams to do data entry and give back little the team can use, and capture that costs more than it returns rarely survives contact with a busy quarter. The failure is not evidence that patient community relationships cannot be managed systematically. It is evidence that borrowing a system built for selling to physicians does not work for partnering with patient communities.

We have seen this transition before in Medical Affairs

A decade ago, Medical Affairs looked very similar.

The scientific landscape was explored manually. Knowledge lived with individuals rather than with the organization. Resources and processes were disconnected. Visibility and reporting were limited. And the consequence, which will sound familiar, was that Medical Affairs found it difficult to demonstrate its value.

Medical Affairs has spent the past decade building its way out of that state. Scientific intelligence operates at scale. Institutional memory is a system, not a person. Teams increasingly work in a shared operational environment with standardized workflows and reporting. And Medical Affairs is widely recognized as a strategic function with budget and a key voice in development decisions.

What is worth being precise about is the order in which that happened. Plenty of forces pushed Medical Affairs up the agenda, the separation of medical from commercial among them. But the function could only hold that position once it could show its work, and the capability to do that had to be built before the case could be made.

Patient Engagement and Advocacy is standing at the same inflection point, with the same fundamentals, roughly ten years later.

The impact problem is an infrastructure problem

Every PE&A leader we speak with is under pressure to show impact. It is the most common question they get asked and the hardest one to answer well.

The instinct is to treat this as a measurement problem and go looking for the right metrics. It is not. It is an upstream problem. You cannot measure the impact of engagement activity that was never systematically recorded. You cannot show that patient input changed a protocol if the insight was delivered verbally in a meeting and lives in a slide somebody made once. You cannot demonstrate the value of a relationship portfolio you cannot describe.

Counting activities is what teams fall back on when the underlying record does not exist. Number of partnerships, number of advisory boards, number of touchpoints. Those numbers are real, and they persuade nobody, because they say nothing about what changed as a result.

The measurable version of this function requires three things to be captured as a matter of course: which communities we currently engage and why, what we learned from them, and which decisions that learning informed. None of those are reporting features. They are byproducts of having an operating model in the first place. Build the model and the impact story assembles itself from the record. Skip it and absolutely no dashboard will save you.

What a strong operating model looks like

Three questions define it.

How well do we understand the external patient and advocacy ecosystem?

Turn community intelligence into faster, more confident action, instead of commissioning a fresh landscape study every time a program starts.

How effectively do we capture and preserve what we learn as an organization?

Capture relationship history, engagements and insights with minimal effort and friction. If capture is expensive, it will not happen. Tools that ignore this tend to end up unused, which is what the failed CRM implementations are really telling us.

How do we make community knowledge accessible and activated wherever work happens?

A shared operating layer of structured data, connected systems and agentic AI, so that collective knowledge is visible and actionable at the point where decisions get made, not filed somewhere it has to be looked for. This is the part most organizations underestimate.

In daily work, this changes three things.

• From searching to askingPut a question anywhere and get a dynamic answer grounded in relevant internal and external context.
• From individual knowledge to shared organizational memoryRelationship history, decisions and lessons learned are shared, so teams build consistently on one another's experience even as roles change.
• From one-off work to on-demand executionMuch manual work is replaced with agentic workflows, repeatable frameworks and reports that run when you need them.

Technology can enable the model. Change management is what makes it real.