Is Your Measurement Model Undervaluing Point of Care? Why Patient-Centric Inputs Matter

Measuring the true impact of healthcare marketing dollars can be difficult. Let’s say, for example, that a pharmaceutical brand finishes a Point of Care (POC) campaign and receives two different answers about its performance. A test-versus-control study shows strong impact. But the brand’s marketing mix model suggests a smaller contribution. Both findings arrive at the planning table, where the team must decide how much to invest next year.

Which result should guide the coming year’s investment decision? And what happens when the team can’t confidently explain the discrepancies they’re seeing?

That challenge sits at the center of a recent POC Perspectives podcast on patient-centric measurement, featuring Joanne Biscardi, SVP of Business Development & Sales Operations at ConnectiveRx and POCMA’s Measurement Committee Chair, and Joy Joseph, EVP of Analytics at MedFuse and founder of Syneractiv.

Their discussion identifies a consequential weakness in how some marketing mix models represent POC: they record whether a placement was active, without accounting for the relevant patient population it could reach. That missing information can distort the channel’s estimated contribution and leave marketers making investment decisions with an incomplete view of performance.

Conflicting Results Create a Planning Problem

In the podcast, Joseph recalls earlier industry research in which test-versus-control studies generally produced POC impact estimates around 1.5 times those from MMM for the same campaigns. He presents that discrepancy as a reason to investigate the methods and inputs behind the results.

For a brand manager, competing estimates can make a promising investment difficult to defend. An agency may recommend expanding POC based on a lift study, while the brand’s broader model points toward a more cautious allocation.

Biscardi emphasizes the purpose of resolving that tension: “it’s not about picking the better outcome.” The work is to understand the discrepancy and build a more coherent basis for decisions.

For brands building their 2027 media plans, the implication is practical. Before using a measurement result to determine POC’s budget, you need to examine the inputs that produced the result.

POC Presence Does Not Explain Patient Reach

Marketing mix modeling, or MMM, helps brands estimate how different marketing activities contribute to business outcomes.

Joseph explains that other media commonly enter models through audience-based inputs, such as impressions. POC, by comparison, has often been represented by a binary indicator: a one when a practice has an active placement, and a zero when it does not.

That approach captures presence. But it leaves substantial differences in audience size invisible.

Consider two practices participating in the same campaign. One sees a relatively small number of patients each day; another sees hundreds. If both receive the same input value in the MMM, the model cannot use that variable to distinguish the scale of their patient traffic. Joseph explains that this can overestimate lift in smaller offices and underestimate it in larger ones.

Even organizing the analysis around individual healthcare providers does not necessarily resolve the problem. Providers working in different practice environments can see different patient volumes and operate alongside different combinations of POC assets.

Biscardi describes the underlying complexity of POC placements succinctly: “[POC is] a channel, but it’s also an environment.” Multiple tactics can engage patients and providers within that environment, and the audience changes from practice to practice.

For marketers, this raises a question about the specificity of their MMM inputs: does the POC input describe the relevant audience with enough detail?

Patient-Centric Inputs Add Volume and Clinical Relevance

So, how do marketers go about understanding this audience information?

The approach Joseph discusses addresses the variation in patient traffic through a Patient Reach Index, using claims data to estimate relevant patient traffic associated with a campaign’s activated practices and providers.

The distinction between total traffic and relevant traffic matters. A family practice serves patients with many different needs. Only a portion may be relevant to a particular campaign.

Joseph explains that diagnosis codes and other clinical criteria, including comorbidities, can help define the patient population for the analysis. The input can then reflect variation in that population across participating practices and over time.

This makes the model more sensitive to the audience a campaign is intended to engage. Two offices with similar overall patient volumes may represent very different opportunities for a specific brand. A presence flag alone cannot express that difference.

As POCMA’s patient-centric measurement white paper explains, the Patient Reach Index brings together the breadth of campaign activity and the depth of patient traffic. That gives the MMM more information about the scale of POC activity than a simple count of activated locations.

Marketers should also be precise about what the input represents. An estimate of relevant patient traffic is not, by itself, confirmation that every patient saw or engaged with an asset. Brands and measurement partners should agree on how traffic relates to the campaign’s exposure assumptions.

That clarity makes the input more useful and the resulting conclusions easier to interpret.

One Campaign Shows Why the Input Deserves Testing

The podcast’s most concrete example comes from a collaboration involving MedFuse and Trinity Life Sciences on a launch brand using several POC platforms.

MedFuse conducted the test-versus-control lift study, while Trinity conducted MMM. The teams compared the model’s results using the legacy 0 vs. 1 binary input with results using the patient-centric input for the same campaign, keeping the other inputs constant.

Joseph reported a 4.4x improvement in estimated POC impact and ROI, alongside better model fit, lower error, stronger statistical significance and closer alignment with the independent test-versus-control study when using the patient-centric input.

He also explained that the estimated contributions of other channels remained stable in the comparison. The improvement came with lower model error, rather than a redistribution of their measured impact to POC.

Together, those observations offer a reason to test the approach. A higher POC estimate is more persuasive when accompanied by evidence that the model represents the campaign more accurately.

Better Inputs Can Change How POC Enters Budget Planning

The podcast discussion then moves from how marketers should evaluate a completed campaign to how they should decide what happens next.

Joseph describes how POC findings can lead to relatively directional budget decisions: increase investment somewhat when ROI looks stronger, or reduce it when the result looks weaker, whereas other channels may be incorporated into budget optimization tools that estimate how returns change across different investment levels.

A binary POC input provides limited information for understanding those relationships. Patient-based inputs introduce more variation, which can support the development of response curves and a more informed assessment of where additional investment may be productive. This creates an opportunity to evaluate the channel more fully within the brand’s planning process, subject to the strength of the data and model.

Begin With a Focused Test During 2027 Planning

To evolve the approach to POC measurement during planning, Joseph recommends a manageable starting point: test a better input within the existing model. Brands can work with measurement partners to evaluate patient-centric data or explore whether suitable in-house claims data can support the approach.

Marketers should bring brand, agency and analytics teams together around five questions to put that recommendation into practice:

1. How is POC represented today in my measurement model? Establish whether the model uses a presence flag, provider counts, patient traffic or another measure, and what variation that input captures.
2. Which patients are relevant to my campaign? Agree on the clinical criteria and timing used to define the audience before interpreting the results.
3. What would constitute a better measurement model? Evaluate fit, error, stability and other channels’ estimates alongside POC’s estimated contribution.
4. How will the MMM’s findings be reconciled? Compare the revised MMM with available lift studies and investigate remaining differences in scope or methodology.
5. What planning decision could change with better measurement? Identify how validated findings would inform allocation, placement choices or the tools used to plan investment.

The podcast also introduces a consortium study offering data at no cost to up to ten participants willing to test the approach, which is intended to help evaluate patient-centric inputs alongside existing measurement methods. Learn more about the initiative here.

Consistent Standards Make the Learning More Useful

The final topic covered by the podcast is consistency. If organizations define and construct inputs differently, a change in agency, analytics partner or data source can also change the model’s conclusions. Consistency makes it easier to assess performance over time and understand whether an apparent change reflects the campaign or the measurement process.

That’s why POCMA is collaborating with the Media Rating Council (MRC) through a Point of Care Standards Evaluation Group to examine how existing media standards should apply or be adapted to POC.

Together, we’re identifying where current out-of-home standards can apply as written, where they require modification and where the unique characteristics of POC may require new guidance or definitions. Our goal is to create a consistent, credible and practical framework that allows marketers to evaluate POC with greater transparency and confidence.

As 2027 plans take shape, marketers should be able to explain how POC’s relevant audience is represented in the evidence guiding investment. When that audience is missing from the inputs, the budget decision deserves another look.

Listen to the full POC Perspectives episode and learn about the measurement consortium and MRC working group.