Claims Data vs. EHR Data vs. Lab Data for Campaign Measurement

Most measurement vendors talk about "real-world data" as if it were one thing. It is at least three different things, collected for different purposes, with different gaps. Claims exist so someone gets paid. EHRs exist so clinicians can document care. Lab systems exist to run and report tests. None of them were built to measure advertising, which is why each one sees part of the picture and misses the rest.

This article stays vendor-neutral. Many providers combine more than one type, and the broader platform view is in the guide to pharma outcomes measurement platforms.

Claims data for campaign measurement

Claims come in two main forms. Pharmacy claims record dispensed prescriptions. Medical claims record billed services, with diagnosis and procedure codes. Data providers usually source them from switches, clearinghouses, payers, or pharmacies.

For HCP media, pharmacy claims are the workhorse. They carry the prescriber NPI, so they support prescriber-level measurement of new-to-brand and total prescriptions. That is why most Rx lift studies for programmatic HCP media run on claims.

The gaps are predictable. Cash-pay prescriptions, some specialty pharmacy volume, hospital dispensing, and free samples can be partly or fully missing depending on the source. Medical claims arrive with meaningful lag and may take months to be complete. For a buy-and-bill oncology drug, pharmacy claims may barely see the product at all.

EHR data for campaign measurement

EHR data comes from provider systems: problem lists, diagnoses, orders, blood pressure and weight readings, notes in structured form. It is closer to the clinical decision than claims, and it can show things claims never will, such as a diagnosis that was recorded but not treated.

That makes it useful for campaigns aimed at diagnosis rates, screening, or treatment consideration. If your media is meant to get rheumatologists to identify more patients earlier, EHR data can measure that change before any prescription is written.

The weaknesses: coverage is limited to the health systems and practices in the network, and those are not a random sample of US care. A prescription written in the EHR may never be filled. Data structure varies by system, so normalization matters, and clinical notes are rarely usable for measurement without heavy processing.

Lab and diagnostic data for campaign measurement

Lab data records test orders and, in some cases, results. For precision medicine, biomarker testing, and screening programs, it is often the most direct outcome available. If the campaign goal is more EGFR or HER2 testing, measuring prescriptions months later is too indirect.

Lab data is narrow by design. A test result does not equal a treatment decision, and coverage depends on which lab networks feed the data. Reference labs, hospital labs, and specialty labs each see different patients. Judging that coverage is the subject of how to judge the accuracy of lab and diagnostic data, and the commercial use of testing data is in diagnostic data in precision medicine launches.

Claims vs. EHR vs. lab data: side-by-side comparison

How to choose a data provider for your campaign

Start with the behavior, not the vendor. Write down, in one sentence, what the HCP or patient should do differently because of the campaign. Then pick the data that records that behavior most directly.

1. Name the outcome. Test ordered, diagnosis recorded, new start, refill, or switch.
2. Map the channel. Retail pharmacy, specialty pharmacy, buy-and-bill, or hospital. This decides which claims sources matter.
3. Request coverage against your data. Ask each provider to estimate the share of your actual volume they capture, using your sales or hub data as the benchmark.
4. Check prescriber match. What share of your target list NPIs appear in their data in the last 12 months?
5. Ask about lag and completeness. How many weeks until 90 percent of the period's claims or records are in?
6. Review provenance. Where the data comes from, under what rights, and how it was de-identified. Health data provenance questions is a good checklist.

For the targeting side of the same decision, see how to compare HCP targeting data providers.

When to combine data types

Combining data types makes sense when the patient path crosses them. A precision oncology brand might measure testing lift in lab data and treatment starts in claims. A chronic disease brand might use EHR data for diagnosis and pharmacy claims for fills. The cost is complexity: different lags, different coverage, and two sets of match rates to explain. Only combine when the second source answers a question the first cannot.

Privacy rules apply to each source separately, and linking datasets can raise re-identification risk. Ask how the linked data is de-identified as a whole, in addition to piece by piece. This is a question for your privacy team, not legal advice from a media article.

Practical takeaway

Take your current measurement report and write the outcome it measures next to the outcome your campaign brief says you are trying to change. If they are different (for example, the brief says "increase testing" and the report measures TRx), you are probably using the wrong data type, and that mismatch is worth raising before the next renewal.

Editorial note. Analysis and frameworks are the author's own and do not represent Acxiom or any current or former employer, client, or named platform. Examples labeled hypothetical or illustrative are not results from real campaigns. Nothing here is legal, regulatory, or medical advice.