Read below the overall star
The overall rating combines distinct domains, so a movement may not originate in medication-related measures. Open the component data and technical guide before drawing a pharmacy conclusion.
Public data also have defined periods, exclusions, and update schedules. They should not be presented as a real-time resident census or current prescribing list.
Turn one measure into three questions
Ask what the measure actually counts, whether the facility sees a similar pattern in current internal data, and which resident-level process could plausibly affect it. Keep alternative explanations visible.
For an antipsychotic measure, for example, current review work may examine indications, target symptoms, exclusions, documentation, and follow-up rather than setting an arbitrary resident target from the public rate.
Define a bounded improvement proposal
Offer a sample review, a definition check, or a recurring reconciliation between facility and pharmacy data. State what the work can and cannot conclude.
Avoid promising a star-rating change. Public results depend on specifications, lagged data, other domains, and factors outside the consultant's control.