Establish the service-line cohort
Define procedures, settings, facilities, clinicians, market, and observation period before reviewing results.
Keep evidence domains separate
Price, public quality, access, utilization, capacity, and affiliations answer different questions and should not disappear into an unexplained score.
Inspect relationships
Retain the source and effective period behind facility, system, group, and clinician relationships. Affiliation is not proof of employment or network participation.
Show conflicts and missingness
A candidate can be strong on access and weak on evidence completeness. The comparison should expose that tradeoff.
Record the decision
Preserve alternatives, reviewer comments, assumptions, source versions, and the rationale for advancement or exclusion.
See the method in context