Data Triangulation in Pharma Market Research: How to Read an ATU and Chart Audit Readout

A combined readout puts two kinds of evidence in one report. An awareness, trial and usage (ATU) survey describes the market. A patient chart audit describes individual decisions. Read well, the pair shows things neither source can show alone. Read carelessly, it produces numbers that look comparable and are not.

These six rules use a real example: a ZoomRx hyperlipidemia pulse fielded from 31 August to 9 September 2026, about seven weeks after the first oral PCSK9 inhibitor reached the US market.

What data triangulation means here

Data triangulation is checking one finding against two or more independent sources, so that agreement adds confidence and disagreement points to something worth explaining. In pharma market research, a common pairing is a physician ATU survey with a patient chart audit from the same panel and wave.

Start with the base ledger

Before reading any figure, list every base in the readout. In this study there were six.

Rule 1: Name the question each source answers

The ATU answers “what do physicians say and do across the market?” The chart audit answers “what happened for this patient, and what else was considered?” Write the question next to each figure before you compare them.

Rule 2: Keep every figure attached to its base

“78% aware” means 78% of 46 physicians. “72% asked for an oral option” means 72% of 25 patients on one brand. Dropping the base makes these look like the same kind of number.

Rule 3: Label stated, described, and expected data differently

Three kinds of data appear in this study:

• Stated: physicians gave route of administration about 7 of 100 importance points.
• Described: in 25 patient charts, physicians described 6 patients who moved from an injectable to the oral brand, 5 of them for route or preference.
• Expected: across 60 stable injectable patients, physicians rated 7% very or extremely likely to switch within six months.

The case study labels the expected figure as an expectation in the text. Nobody followed those 60 patients forward.

Rule 4: Never add or average across bases

The two switching figures above cover different patients and answer different questions. Adding them, or putting them in one forecast line, would be easy and wrong. Report them side by side, each with its own base.

Rule 5: Check how much one patient moves a small base

On 10 charts, one patient moves a percentage by 10 points. The study reports 40% of Leqvio patients as commercially insured; that is 4 patients. Treat figures like this as hypotheses to test, and say so in the readout.

Rule 6: Treat disagreement between sources as a finding

When the ATU and the charts disagree, that is often the most useful result. In this study, route of administration ranked low in stated importance but appeared among the top reasons for 44% of choices on the new brand. The case study reads that gap as patient preference shaping far more decisions than physicians report.

See every base in the original

The case study, Hyperlipidemia treatment decisions in 2026, reports each figure with its base.

Read the case study →

To get a readout built this way for your brand, explore ATU brand tracking and Patient Chart Audits, fielded on one panel in the same wave.

 LIPFENDRA is a registered trademark of Merck Sharp & Dohme LLC. Repatha is a registered trademark of Amgen Inc. LEQVIO is a registered trademark of Novartis AG. ZoomRx is not affiliated with, and this paper is not sponsored or endorsed by these companies.

Frequently Asked Questions