Assess before-and-after improvement evidence across a declared population and metric without claiming causation or closure.
