Multi-site dashboards: compare definitions before comparing percentages

Define the decision behind the record

Decide what the comparison is intended to support. Separate completion from results and corrective actions. Document how cancelled tasks, local variations and late entries affect the figures instead of leaving managers to infer the denominator.

Compare denominators and exception status before percentages

Define what counts as required, completed, late, missing and not applicable for each reported period. Identify site scope and relevant template versions. Equal displayed percentages may describe different numbers of observations or different task boundaries.

Retain the underlying counts and make unresolved exceptions or decisions visible alongside completion. For example, 18 of 20 and 9 of 10 both equal 90%, but do not establish the same coverage. Separate submission status from implementation and verification rather than combining them into a single reassuring indicator.

Test the dashboard with one site using a local variation and another with complete submissions but an open decision. A reviewer should trace the summary back to its source definitions and records. Preserve changes in reporting logic so a trend does not silently compare measures that changed meaning.

Build a usable record

Use the fields below to capture the facts needed for this particular decision. Keep the actual observation separate from any later correction or approval.

• Site and reporting period
• Scheduled and applicable task definitions
• Completion and late-entry treatment
• Exceptions and open actions
• Local variations and limitations

Test the review, not just the form

Reproduce one displayed figure from the underlying records. Compare its definition across sites and check how a schedule change affects historical reports. Do not equate a higher score with a lower risk without an appropriate basis.