A new approach to patient identification is beginning to challenge a basic assumption in pharmaceutical marketing: that marketers already know who the patient is. What happens when the next potential patient does not have a diagnosis, does not recognize the condition, and is not actively searching for treatment? Tarsus Pharmaceuticals and Moon Rabbit recently offered an intriguing answer with “XDEMVY: Finding the Invisible Patient,” a campaign built around identifying people who may be living with Demodex blepharitis before diagnosis. The work won the 2026 Fierce Pharma Marketing Award for Data-Driven Campaign, signaling that predictive patient identification and the search for previously unseen patient populations may be becoming an important new frontier for pharma marketers.
Table of Contents
- Finding undiagnosed patients before diagnosis
- How predictive models are changing pharma marketing
- The line between identification and intervention
- What patient discovery means for pharmaceutical market development
- Conclusion
- Frequently Asked Questions
Finding Undiagnosed Patients Moves Pharma Marketing Upstream
Traditional pharmaceutical targeting often starts after a recognizable signal appears. A patient receives a diagnosis, visits a condition-related website, searches for a treatment, or enters a claims-based audience segment. At that point, marketers have evidence that the person is somewhere within an established disease journey.
Predictive models can move that starting line much earlier.
According to Fierce Pharma, Tarsus Pharmaceuticals and Moon Rabbit developed a predictive framework for the XDEMVY campaign designed to identify people who had not yet been diagnosed with Demodex blepharitis. The campaign won the newly introduced Data-Driven Campaign category at the 2026 Fierce Pharma Marketing Awards.
The disease provides a compelling use case. Tarsus estimates that Demodex blepharitis may affect approximately 25 million U.S. eye care patients, while the company has reported roughly 1.5 million diagnosed patients. That difference suggests a substantial disease-recognition gap, although the 25 million figure is an extrapolation from prevalence research rather than a count of confirmed cases.
Meanwhile, XDEMVY, or lotilaner ophthalmic solution 0.25%, became the first FDA-approved treatment for Demodex blepharitis in 2023. The U.S. Food and Drug Administration reports that its approval was based on two trials involving 833 patients.
For marketers, therefore, the commercial challenge is not simply reaching diagnosed patients. It can also involve helping the healthcare system recognize disease that may already be present.
Data-Driven Pharma Marketing Is Finding New Signals
The concept creates an important shift in audience strategy. Instead of asking, “Where are diagnosed patients?” a predictive model can ask, “What patterns may indicate that someone belongs to an underrecognized patient population?”
Those patterns might come from combinations of permitted, privacy-appropriate data signals rather than one obvious indicator. Consequently, the value comes from connecting signals that individually may reveal very little but collectively may help define an audience for disease education.
This is different from simply expanding a targeting list. These predictive approaches can reveal gaps between disease prevalence, symptom recognition, clinical evaluation, and diagnosis.
That distinction matters because pharma marketing is already becoming more dependent on higher-quality audience signals. As discussed in Pharma Marketing Network’s coverage of authenticated HCP targeting, marketers increasingly need confidence that the audiences they reach are genuine and relevant.
Predictive approaches add another layer. Marketers are no longer only validating identity. Instead, they are using data to estimate where unmet recognition or education may exist.
As a result, media strategy begins to overlap with market development. The audience itself becomes something that analytics can help uncover.
The Invisible Patient Creates New Responsibilities
However, finding a potential patient is not the same as diagnosing one.
That boundary becomes especially important when predictive models deal with health conditions. A marketing platform can identify patterns associated with a potential audience, but diagnosis remains a clinical function. Therefore, campaigns need to distinguish carefully between encouraging disease awareness and implying that an individual has a medical condition.
Privacy also becomes central to the strategy. Healthcare marketers operate in an environment shaped by privacy regulation, platform restrictions, consent requirements, and increasing consumer sensitivity around health data.
For that reason, the future of predictive patient analytics will depend partly on how responsibly these models are designed and activated. Transparency, appropriate data governance, validation, and privacy-safe audience development cannot be treated as secondary technical issues.
The messaging matters too. A predictive model may help determine where education could be useful, while creative work still needs to give people a reasonable next step. In many cases, that means encouraging a conversation with a qualified healthcare professional rather than pushing someone toward a self-diagnosis.
When appropriate medical evaluation is needed, consumers can also use resources such as Healthcare.pro to seek professional healthcare guidance.
Predictive Analytics Could Reshape Market Development
The larger implication extends beyond one eye-care campaign.
Pharmaceutical market development has traditionally relied heavily on physician education, disease-awareness campaigns, epidemiology, claims analysis, and consumer advertising. Predictive analytics could connect these activities more closely by helping marketers identify where awareness gaps may exist before patients become visible through conventional targeting signals.
That could be particularly important in underdiagnosed conditions. However, it may also matter in diseases where symptoms overlap with common problems, patients normalize symptoms, or clinicians do not routinely screen for a particular condition.
The commercial funnel could therefore start earlier than marketers once assumed.
Instead of awareness beginning with a known patient audience, data may help identify populations that warrant broader disease education. HCP campaigns could then reinforce recognition among clinicians, while consumer communications encourage appropriate conversations about symptoms.
At the same time, pharma companies are increasingly building more connected patient journeys. Pharma Marketing Network has previously examined how direct-to-patient pharmaceutical models are connecting awareness, providers, support services, pharmacies, and fulfillment.
Predictive analytics could add something new to the beginning of that patient journey: discovery.
The XDEMVY campaign is notable because it illustrates how data strategy can move beyond optimizing media against a predefined audience. Instead, data can help marketers question whether the predefined audience represents the full market at all.
Conclusion
The next major opportunity in pharmaceutical marketing may not always be a patient already searching for a brand or condition. It may be someone who has symptoms but has never connected them to a treatable disease.
Tarsus Pharmaceuticals and Moon Rabbit’s “Finding the Invisible Patient” campaign shows how data-driven patient identification can push pharma marketing further upstream. Rather than simply competing for attention among known audiences, marketers can use responsible analytics to explore gaps in disease recognition and education.
The opportunity is significant, but so is the responsibility. Predictive models should support awareness, not substitute for clinical diagnosis. If pharma marketers can maintain that distinction while respecting privacy and delivering useful education, identifying the “invisible patient” could become an increasingly important part of market development.
Frequently Asked Questions
What is predictive patient identification?
It uses data patterns and analytical models to identify populations that may share characteristics associated with an underdiagnosed or underrecognized health condition. The goal in a marketing context is generally to identify opportunities for relevant disease education rather than to diagnose an individual.
How does predictive patient identification differ from traditional pharma targeting?
Traditional targeting often begins with known signals such as diagnoses, treatment activity, or condition-specific behavior. Predictive approaches can look further upstream for patterns that may indicate an unmet disease-awareness need.
Why did the XDEMVY campaign receive attention?
Tarsus Pharmaceuticals and Moon Rabbit’s “XDEMVY: Finding the Invisible Patient” used a predictive framework to identify people who had not yet been diagnosed with Demodex blepharitis. It won the Data-Driven Campaign category at the 2026 Fierce Pharma Marketing Awards.
Can predictive marketing diagnose a patient?
No. Predictive marketing can identify patterns or potential audience segments, but it does not provide a clinical diagnosis. Diagnosis should come from an appropriately qualified healthcare professional.
Will predictive patient analytics become more common in pharma marketing?
Its usefulness may grow as marketers gain access to better analytics and privacy-safe data. However, adoption will also depend on responsible data practices, regulatory requirements, model quality, and clear boundaries between marketing and clinical decision-making.
This content is not medical advice. For any health issues, always consult a healthcare professional. In an emergency, call 911 or your local emergency services.
