System strain index INTERPRETIVE · composite from published signals
A 0–100 read of unscheduled-care pressure, built only from the real published figures below (A&E 4-hour performance, long waits, bed occupancy, delayed discharge). Focuses attention; it does not rank or triage patients.
–
A&E 4-hour performance — national trend MEASURED
Share of attendances at major (Type 1) emergency departments seen, admitted, transferred or discharged within 4 hours. The blue line is a 95% planning reference. This Type-1 subset cannot assess the full A&E access standard. Last 30 published weeks.
Emergency-department strain ranking MEASURED
Every major (Type 1) ED, worst first by 4-hour performance for the latest week. Bar = % seen within 4 hours; the blue tick marks the 95% planning reference. The number of patients waiting over 12 hours is additional service-pressure context.
<50%50–69%70–84%≥85%95% reference
In plain English — what this shows
Auto-written from the real numbers, for a non-specialist.
Scenario sandbox
Pressure-test the response plan. Adjusts the strain index off the real baseline — clearly a what-if, not a forecast.
Long-wait counts MEASURED
Patients waiting over 8 and over 12 hours in major EDs in the latest week. These counts support service review; they do not estimate individual risk or excess deaths.
Bed occupancy by health board MEASURED
All-specialty occupied bed days divided by staffed bed days, for each territorial board. The 85% and 92% color bands are planning references; these aggregate reports do not assess current patient safety or available beds.
Delayed discharges — flow blockers MEASURED
Patients clinically ready to leave but unable to (the "no criteria to reside" backlog) at the monthly census, by board. Separate reporting periods cannot establish a causal link to A&E waits.
England comparator CITED — context, not live
Data health & provenance
Public sources and automatic updates
Public Health Scotland reports are checked weekly and published here automatically without a provider key. The page checks the published package on opening, every six visible hours and on return. Each measure retains its own reporting period; a new download does not make old observations current. Failed or incomplete updates preserve the last good reports. Offline copies use their embedded package. Source queries: