Attributed Prescriptions vs. Incremental Prescriptions
Understand why attributed prescriptions and incremental prescriptions answer different questions, and how to use each in pharma media decisions.
The short answer
An attributed prescription is an observed prescription connected to media exposure or engagement under a stated identity, eligibility, and timing rule. An incremental prescription is the estimated difference between what happened with media and what would have happened without it. The first describes an association; the second requires a credible counterfactual.
This distinction changes investment decisions. Optimizing to attributed volume alone can favor people already likely to prescribe or fill.
Ask what the method can support
| Output | What it can say | What it cannot say alone |
|---|---|---|
| Exposed Rx count | Prescriptions observed among matched exposed people | Media caused those prescriptions |
| Attributed Rx | Prescriptions assigned under a defined rule | All assigned prescriptions are incremental |
| Lift estimate | Difference versus a comparison after adjustments | Result is unbiased if design assumptions fail |
| Incremental Rx estimate | Additional prescriptions under a credible design | Estimate is exact or transferable forever |
Build the counterfactual
Random assignment is the clearest design when feasible. Geographic tests, matched controls, or observational models can be useful but rely on stronger assumptions. Review pre-period balance, contamination, concurrent activity, sample attrition, identity coverage, and sensitivity to modeling choices.
NIST describes confidence intervals as ranges intended to cover an unknown population parameter at a stated rate over repeated samples under the method's assumptions. A point estimate without its interval hides decision risk (NIST).
Keep denominators consistent
State whether results cover new-to-brand prescriptions, total prescriptions, refills, patients, prescribers, or another outcome. Document claim lag, eligibility, lookback, attribution window, and match funnel. Never compare rates built on different eligible populations as if they were equivalent.
Hypothetical example
Among 100,000 matched exposed people, 2,000 have an attributed prescription. A comparable control suggests 1,700 would have prescribed without exposure. The estimated incremental count is 300, subject to uncertainty and design assumptions, not 2,000. These numbers are illustrative.
Use both measures appropriately
Attributed outcomes can support operations, journey analysis, and partner diagnostics. Incremental estimates are better aligned to causal investment questions. If incrementality is unavailable, say so and avoid relabeling attribution.
Practical takeaway
Every Rx result should carry a method label. The next step is to add four fields to scorecards: outcome definition, attribution rule, counterfactual method, and uncertainty interval.
Sources
External guidance and platform documentation change. Links were current at publication; check them again before relying on them for a decision.
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.
New pharma programmatic breakdowns, occasionally
One email when I publish something worth reading. Benchmarks, measurement teardowns, and case studies with the caveats attached. No cadence promises, no reselling your address.
Unsubscribe any time. See the privacy policy.
Working through this decision on a real plan?
I work on health and pharma data, identity, and activation, after five years running HCP and DTC programmatic agency-side. Happy to talk through how this applies to your situation.