Evidence stays separate
The profile grade and underlying human-study count are displayed independently from outcome mappings and safety flags. Multiple publications are collapsed only when explicit trial, cohort, dataset, or parent-study identity proves they are not independent evidence.