The phrase “medical intelligence” has been around for a while, but its meaning inside pharmaceutical companies has settled only recently. It used to be a loose label for whatever the medical team happened to know. Now it usually means something more specific: a deliberate capability, with owners and a cadence, that keeps the medical organization current on the science and turns that knowledge into decisions.
This guide covers what medical intelligence includes, how it differs from related disciplines, who uses it, and how to build it in practice. It is written for people in Medical Affairs who are being asked to set up the capability, or to make an existing one work better.
A note on the term before going further. Outside the industry, “medical intelligence” usually refers to government and military work on foreign health threats. If you search for the phrase you will find both meanings. This guide is about the pharmaceutical one.
Why medical intelligence has become a function in its own right
For most of its history, Medical Affairs was seen as a support function. It answered questions, managed publications, ran advisory boards and kept the commercial side honest. Each of those jobs needed information, and each team found its own.
Three things changed that.
The volume of evidence grew faster than any team could read it. In an active therapy area, the number of new publications, trial updates, congress abstracts and regulatory documents in a single month is more than one person can follow. Teams that rely on individual reading miss things, and the things they miss are not random. They tend to be the ones published somewhere unexpected or in a format nobody was watching.
Medical Affairs became strategic. Medical teams now shape evidence generation, launch readiness and the scientific narrative for a product. They are asked to explain what a competitor’s data means on the day it is released. That requires knowing the evidence base as a whole, not just the parts that one team works with.
The cost of duplication became visible. When medical information, publications, field medical, HEOR and drug safety each run their own searches, they repeat a large share of each other’s work. When the outputs disagree, someone has to reconcile them. Leaders started asking why five teams were producing five versions of the same landscape.
Medical intelligence is the answer to all three: one capability, run once, that the whole medical organization draws on.
What medical intelligence covers
The scope varies by company and therapy area, but most mature programs cover the same core areas.
| Area | What teams follow | Questions it answers |
|---|---|---|
| Published literature | New papers on your products, competitors, the condition and its treatment | What new evidence exists, and does it change anything we say? |
| Clinical trials | Registered trials, status changes, enrollment, results postings | What is coming, when, and who is running it? |
| Congress data | Abstracts, posters and presentations at the meetings that matter | What was presented before it was published? |
| Regulatory decisions | Approvals, label changes, safety communications, regulator meetings | What can be said, by whom, and what changed? |
| Guidelines | Updates from professional societies and guideline bodies | How are clinicians being told to treat the condition? |
| Health technology assessment | HTA decisions and the evidence requested | How are payers and assessors judging the evidence? |
| Experts | Publications, trial roles, presentations and guideline work of the people who shape the field | Who is influential, on what, and what are they saying? |
| Competitors | All of the above, for competing products | What are they doing, and what does it mean for us? |
| Field insights | What medical teams hear in scientific exchange | What are clinicians asking that the evidence does not answer? |
The last row is what makes medical intelligence more than a literature service. External evidence tells you what has been published. Field insights tell you what clinicians actually think about it, where they are confused, and which questions keep coming up. A program that combines the two can see evidence gaps that neither would reveal alone.
Medical intelligence, competitive intelligence and market research
These three disciplines overlap, and companies draw the lines differently. The distinctions below are the most common ones.
| What differs | Medical intelligence | Competitive intelligence | Market research |
|---|---|---|---|
| Main question | What does the science say, and what does it mean for our medical strategy? | What are competitors doing, and what does it mean for our business? | What do customers think and do? |
| Primary users | Medical Affairs and clinical teams | Brand, strategy, business development, leadership | Brand and marketing |
| Main inputs | Publications, trials, congress data, regulatory and guideline decisions, expert activity, field insights | Pipelines, trials, regulatory events, deals, patents, pricing and access, public company information | Surveys, interviews, prescribing and claims data |
| Typical outputs | Evidence summaries, landscape reviews, expert profiles, gap analyses, briefings | Competitor profiles, scenarios, catalyst calendars, war games | Segmentation, message testing, attitude and usage studies |
| Promotional boundary | Must stay non-promotional | Commercial by nature | Commercial by nature |
In practice, competitive intelligence and medical intelligence share a lot of raw material. A competitor’s phase 3 readout matters to both. The difference is in the question asked of it. The CI team wants to know what it does to the market. The medical team wants to know what it does to the evidence, what clinicians will ask, and whether the company’s own scientific position still holds.
Many companies now run the two from shared sources with separate outputs. That avoids duplicated monitoring while keeping medical work independent of commercial direction.
Who uses medical intelligence
Every medical function both contributes to medical intelligence and depends on it. What each one needs from it is different.
Medical leadership needs to know what changed and what it means. A good weekly summary, with the implications spelled out, is worth more to a medical director than any dashboard.
Medical strategy and launch teams need the whole evidence picture for planning: where the data is strong, where it is thin, what competitors will present next, and what clinicians are asking.
Medical science liaisons need to be current on the experts they meet and on the evidence those experts will want to discuss. For an MSL, medical intelligence shows up as preparation: knowing what an expert has published and presented since the last conversation.
Medical information needs to know about new evidence before the inquiries about it arrive, so that standard responses stay current and answers stay consistent.
Publications teams need to know where the literature stands, how their own work is being cited, and where the gaps are. Publication planning depends on it.
HEOR and value teams need to follow health technology assessment decisions and the evidence payers are asking for.
Clinical development uses it to benchmark trial designs and enrollment, and to find experienced investigators and sites.
Drug safety relies on literature screening to find reports of adverse events. In pharmacovigilance, this is a regulatory obligation, not an option.
Where medical intelligence comes from
The sources are mostly public, and most companies already have access to them. The difficulty is not access. It is coverage, consistency and time.
- Bibliographic databases and journals for published research, including the specialist journals in your therapy area that general searches tend to underweight.
- Trial registries for registered studies, status changes and posted results.
- Congress programs and abstract books for data presented before publication. In fast-moving areas such as oncology, important data is often presented months before it appears in a journal.
- Regulators’ websites and databases for approvals, labels, safety communications and meeting materials.
- Guideline bodies and professional societies for treatment recommendations.
- HTA agencies for assessment decisions and the reasoning behind them.
- Company communications, including press releases and investor materials, which are often where a competitor’s news appears first.
- The company’s own field insights, captured from scientific exchange and handled under internal policies.
A useful test for any medical intelligence program is to ask, for a recent important event, how long it took for the right person in the company to hear about it. If the answer is measured in days, the sources are probably fine. The problem is the process.
How to build a medical intelligence capability
Most companies do not start from nothing. Individual teams already monitor what matters to them. The work is usually to connect what exists, fill the gaps, and give it an owner.
1. Start with the questions, not the sources
Ask each medical team to write down the questions it needs answered every week, and where the answers come from today. “What did competitors present at the last congress?” “Which experts in my territory have published this month?” “Has anything changed in the guideline?” The list becomes the specification for the program. It also shows quickly where teams are answering the same question separately.
2. Agree the scope once
Define the products, competitors, indications, mechanisms, experts and topics the program follows. Write it down. Review it when the portfolio changes. A scope agreed once and shared means every team monitors the same ground, and a gap in one team’s coverage is no longer a gap for the company.
3. Decide the cadence
Most programs settle on three speeds. Some events need to reach the right person the same day: a competitor readout, a label change, a safety communication. Most things can wait for a weekly summary. A few deserve a deeper quarterly review, such as a full landscape update before a planning cycle. Be explicit about which is which, or everything ends up as an urgent email.
4. Give it an owner
Medical intelligence works best with a named owner, often in medical strategy or medical excellence, who maintains the scope, the cadence and the outputs. The owner does not have to do all the work. They make sure it gets done and that it reaches the people who need it.
5. Write for the reader
A medical director does not want a list of forty new papers. They want to know which three matter and why. An MSL does not want a landscape review before a meeting. They want a short brief on one expert. The same underlying intelligence should be shaped differently for each audience.
6. Close the loop
Ask the users, regularly, whether they heard about the things that mattered and whether they heard in time. Track the misses. A program that never asks this question tends to drift into producing reports that nobody reads.
The compliance boundaries
Medical intelligence sits inside Medical Affairs, which means it inherits the function’s obligations. A few deserve particular attention.
Non-promotional independence. Medical activity must not be directed by commercial objectives. Sharing public competitor information with commercial colleagues is generally fine. Steering scientific exchange to serve a sales goal is not. Most companies handle this with clear policies on what flows from medical to commercial, and in what form.
Field insights. What MSLs learn from experts is valuable and sensitive. It should be captured and shared under the company’s insight policies, usually in aggregated form, and not used to target individual clinicians commercially.
Adverse events. Anyone working with medical information may come across a possible adverse event, whether in a publication, a conversation or a congress poster. Companies require these to be passed to drug safety within defined timelines, and a medical intelligence program should make that easy rather than leave it to individual judgment.
Traceability. Statements made by a medical organization need to be supportable. A summary or briefing should make it easy to see where each claim came from. This matters more as teams use AI to draft summaries: the speed is useful only if the output can be checked.
How to tell whether it is working
Medical intelligence is hard to measure directly, but a few indicators are reliable.
- Time to awareness. How long after a significant event did the people responsible hear about it? This is the single most useful measure.
- Coverage. When an event surprises the team, was it inside the agreed scope? If so, the process missed it. If not, the scope needs to change.
- Duplication. How many teams are still running their own searches on the same topics?
- Use. Are the weekly summaries read? Are the briefs opened before meetings? Low use usually means the output is shaped for the wrong reader.
- Decisions. Can medical leaders point to decisions in the last quarter that the intelligence informed? A few concrete examples say more than any dashboard.
Common mistakes
Building a library instead of a service. A searchable archive of everything is useful, but it does not tell anyone what changed. The value of medical intelligence is in what it brings to people’s attention, not only in what it stores.
Sending everything to everyone. If every team receives every alert, people stop reading them. Route by role.
Treating it as a reporting exercise. Monthly reports that summarize the literature are easy to produce and rarely change a decision. Ask what the report is for before producing it.
Ignoring the field. External evidence without field insight misses what clinicians actually think. Field insight without external evidence lacks context. The combination is the point.
Leaving it to one heroic person. Many companies have one person who knows everything about the therapy area. When they go on holiday or leave, the capability goes with them. A program should survive the absence of any individual.
Where AI helps, and where it does not
AI is now part of most medical intelligence programs, and it is useful for specific jobs. It reads quickly. It can screen large volumes of new publications, group them by topic, summarize what they say, and draft a first version of a briefing in minutes.
What it does not do well on its own is judge significance. Knowing that a small change in a guideline matters more than a large new trial, because of how clinicians in your therapy area actually practice, still takes an experienced person. Neither can AI take responsibility for a statement. In a medical organization, every claim has to be traceable to its source and reviewed by someone accountable for it.
The programs that use AI well treat it as a fast first draft with its sources attached, and keep a person in the loop for review. The ones that struggle either avoid it entirely, and fall behind on volume, or trust it too much, and spend their time correcting errors after the fact.
Further reading
- Medical Affairs Professional Society (MAPS), the professional body for medical affairs, publishes guidance and standards on many of the practices described here.
- The FDA’s overview of real-world evidence explains how regulators think about evidence from outside clinical trials, which increasingly forms part of medical intelligence work.
- Our guide to pharma competitive intelligence covers the commercial side of the same territory.
- The glossary defines the terms used in this guide.