What this supports:
- Board and directorate assurance on MH access performance
- Separating genuine demand pressure from definition artefact (154 vs 134 referrals)
- Confidence-labelled headlines — withhold unsafe figures until sign-off
Definition risk: Mar 2026 referral KPI chain not fully signed off — dashboard +2% misleading.
Bottom line
March referral performance appears stable on the executive dashboard (+2% headline) but agreed definition reconciliation shows an 11% fall versus operational reality once exclusions are applied. Separate genuine demand pressure (backlog 264) from definition artefact before Board narrative.
Headline figures (with confidence)
| Measure | Dashboard | Agreed | Confidence |
|---|---|---|---|
| Referrals Mar 2026 | 154 | 134 | Low until signed off |
| Backlog W21 | 264 | 264 | Medium |
Recommended actions
- Hold definition sign-off workshop — service, BI, IM&T, performance
- Do not publish dashboard referral figure to Board without reconciliation note
- Review demand/capacity and agency spend alongside waiting risk