How to Integrate Diagnostic Testing Data Into HCP Targeting

Most brand teams already have a target list when diagnostic data shows up. The list came from claims, was cut into deciles, and has been argued over in three meetings with sales leadership. The question is not whether to throw that away. It is how to add testing behavior without making the list unreadable or breaking the field alignment that took months to get.

This article is part of the series on diagnostic data in precision medicine launches. It assumes you have, or are about to license, some form of HCP-level testing data.

Step one: get the data to the NPI level correctly

Diagnostic data reaches you as aggregated counts or flags per HCP, attributed from de-identified test records. The attribution method matters more than anything else in the file. Ask the vendor which NPI each record is attached to:

• The ordering provider on the requisition
• The treating provider, if the vendor infers it from linked data
• A group or facility identifier that then gets distributed to individuals

In oncology, tests are sometimes ordered by pathologists or under a group NPI. If the file credits all of those tests to the pathologist, your medical oncologist list will look like it does not test. Validate NPIs against the NPPES registry for specialty and status, and read what an NPI can and cannot tell a pharma marketer before you lean on taxonomy codes.

Step two: join and check the join

Join the testing file to your target list on NPI. Then look at three numbers before doing anything else:

1. Join rate: share of target-list NPIs that appear in the testing file at all.
2. Zero rate: share that appear with zero tests in the period.
3. Coverage split: of the zeros, how many sit in health systems whose in-house lab may not be in the data.

A zero is ambiguous. It can mean "does not test," or it can mean "tests, but through a lab the vendor does not see." Flag accounts with in-house labs and treat their zeros as unknown, not as low testers. The article on judging the accuracy of lab and diagnostic data goes further on coverage.

Step three: tier by testing behavior

Keep your existing opportunity decile. Add a testing tier. Then cross them.

"Peer median" should be within specialty and practice setting. Comparing a community oncologist to an academic thoracic specialist on raw test counts is meaningless. The broader tiering principles are in HCP tiering without overfitting to volume.

Step four: set media triggers you can actually execute

Trigger-based media sounds good in a pitch: "when an HCP orders a test, serve the message." In practice, diagnostic feeds usually arrive weekly or monthly, records are aggregated before they reach you, and the DSP audience refresh adds more delay. By the time the signal lands, the decision about that particular patient has been made.

Triggers work better as prioritization rules:

• HCP moves from zero to non-zero testing in the latest refresh: raise frequency for four weeks and add a field alert.
• HCP switches from single-gene to panel testing: switch creative from test education to product.
• Account testing volume drops two refreshes in a row: route to field before adding media.

Write the rule, the refresh date, and the expected delay into the brief so nobody expects patient-moment timing. Triggers must also never imply knowledge of a specific patient in the creative. "Your patient tested positive" is off limits. "For patients with [marker]-positive disease" is product messaging that goes through MLR the usual way.

Step five: orchestrate with the field

The fastest way to lose trust in diagnostic data is to have media target a list the field has never seen. Push the same tiers into CRM, with the same definitions and the same refresh date. Give reps the testing tier but not raw counts if counts could tempt them toward patient-level conversations the data does not support.

Agree in advance who owns each segment. A common split: field owns the testing-gap segment because those conversations need a person, programmatic and email carry efficient testers, and both share the core. The general mechanics of field and digital coordination are in coordinating HCP outreach across field, email, and programmatic, and the plumbing is in integrating HCP data into CRM, CDP, and DSP.

Where integrations usually go wrong

• Stale tiers. The media audience refreshes monthly, CRM quarterly. Different lists, same brand.
• Over-precise thresholds. Tiers based on one or two tests a quarter flip back and forth every refresh. Use rolling windows.
• Specialty mismatch. Pathologists land in the prescriber media buy because they order tests.
• No holdout. Everyone in the testing-gap segment gets media, so you cannot tell whether testing moved because of it.

Practical takeaway

Before activating anything, run the join on your current target list and produce one slide: join rate, zero rate, and the count of zeros at accounts with likely in-house labs. If the unknown group is more than a small share of your top deciles, fix coverage or route those accounts to field confirmation before you build media tiers on top of the data.

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.