Key points
- Evidence-led format.
- Use explicit denominators.
- No invented composite score.
Detailed explanation
A real release needs population, selection period, exclusions, and measurement method. Missing disposable or catch-all assessments must not count as clean. Coverage changes must accompany trends.
Example
Duplicate rate uses source rows; an invalid-outcome rate uses a clearly defined checked population.
How Refresh List handles it
All preview measurements are marked not collected. There is no manufactured chart.
Limitations
Any score should show its denominator, date, source context, and coverage. It should not be read as a market-wide prevalence or ranking.
Related topics
Methodology and sources
Product behavior is described from the current Refresh List implementation and provider contract. Protocol context should be checked against the applicable standards and provider documentation. This page does not create a benchmark or certification.
