wiki.html

<!doctype html><html lang="en"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">
<title>NHS Scotland Hospital Strain Console — Field Guide</title>
<style>
:root{--bg:#071018;--surf:#0e1c2b;--tx:#eaf2fb;--dim:#9fb6cf;--nhs:#41b6e6;--nhsd:#005eb6;--ok:#2bb673;--amber:#e0a020;--orange:#e8702a;--red:#e24b4a;--bd:rgba(255,255,255,.10);--bd2:rgba(255,255,255,.22)}
*{box-sizing:border-box}body{margin:0;background:var(--bg);color:var(--tx);font-family:system-ui,-apple-system,'Segoe UI',Roboto,Arial,sans-serif;font-size:15px;line-height:1.6}
a{color:var(--nhs);text-decoration:none}a:hover{text-decoration:underline}
header.top{padding:20px 28px;border-bottom:1px solid var(--bd);background:linear-gradient(180deg,#0a1828,#071018)}
header.top h1{margin:0;font-size:22px}header.top .sub{color:var(--dim);font-size:13px;margin-top:3px}
.honest{margin:16px 28px;padding:13px 16px;border:1px solid var(--bd2);border-left:3px solid var(--nhs);border-radius:8px;background:#08131f;color:#cfe1f4;font-size:13.5px}
.honest b{color:var(--nhs)}
.wrap{display:flex;gap:26px;padding:8px 28px 60px;align-items:flex-start}
nav.toc{position:sticky;top:14px;flex:none;width:290px;border:1px solid var(--bd);border-radius:10px;background:var(--surf);padding:14px;max-height:calc(100vh - 28px);overflow:auto}
nav.toc h3{margin:0 0 8px;font-size:12px;text-transform:uppercase;letter-spacing:.6px;color:var(--dim)}
nav.toc ol{margin:0;padding-left:20px}nav.toc li{margin:5px 0;font-size:13px}
main{flex:1;min-width:0;max-width:860px}
section{background:var(--surf);border:1px solid var(--bd);border-radius:12px;padding:20px 24px;margin-bottom:18px}
section h2{margin:0 0 12px;font-size:20px;scroll-margin-top:14px}
section h3{font-size:15.5px;margin:15px 0 5px;color:var(--nhs)}
p{margin:9px 0}ul{margin:9px 0;padding-left:22px}li{margin:4px 0}
.bdg{font-size:10px;font-weight:800;padding:1px 6px;border-radius:5px}
.bdg.M{background:rgba(43,182,115,.16);color:#6fe0a8;border:1px solid rgba(43,182,115,.5)}
.bdg.C{background:rgba(65,182,230,.16);color:#9fd8f4;border:1px solid rgba(65,182,230,.5)}
code{background:#06101a;border:1px solid var(--bd);border-radius:4px;padding:1px 5px;font-size:13px}
pre{background:#06101a;border:1px solid var(--bd);border-radius:8px;padding:12px;overflow:auto;font-size:12.5px;color:#bcd2e8}
.chip{font-size:12.5px;padding:7px 11px;border-radius:7px;margin:4px 0;border:1px solid var(--bd)}
.chip.read{background:rgba(65,182,230,.08);border-left:3px solid var(--nhs)}
.chip.warn{background:rgba(226,75,74,.08);border-left:3px solid var(--red)}
table{border-collapse:collapse;width:100%;font-size:13px;margin:10px 0}
th,td{border:1px solid var(--bd);padding:6px 9px;text-align:left}th{color:var(--dim);font-weight:600}
@media(max-width:900px){.wrap{flex-direction:column;padding:8px 14px 40px}nav.toc{position:static;width:100%;max-height:none}main{max-width:100%}}
</style></head><body>
<header class="top"><h1>NHS Scotland Hospital Strain Console — Field Guide</h1>
<div class="sub">how to read the A&amp;E / beds / delayed-discharge console · companion to <code>NHS_Scotland_Strain_Console.html</code></div></header>
<div class="honest"><b>REAL DATA, descriptive — not predictive.</b> Every number on the console is a real figure published by <b>Public Health Scotland</b> (opendata.nhs.scot), badged <span class="bdg M">MEASURED</span>. The only exception is the England comparator, which is <span class="bdg C">CITED</span> context. The console is a near-real-time situational read (weekly to quarterly cadence), <b>not</b> a live monitor and <b>not</b> a forecast.</div>
<div class="wrap">
<nav class="toc"><h3>Contents</h3><ol>
<li><a href="#s1">What this is — and why NHS, not VA</a></li>
<li><a href="#s2">The data &amp; its cadence</a></li>
<li><a href="#s3">System strain index</a></li>
<li><a href="#s4">A&amp;E 4-hour performance</a></li>
<li><a href="#s5">ED strain ranking</a></li>
<li><a href="#s6">Long waits</a></li>
<li><a href="#s7">Bed occupancy</a></li>
<li><a href="#s8">Delayed discharges — the real cause</a></li>
<li><a href="#s9">Scenario sandbox</a></li>
<li><a href="#s10">England comparator</a></li>
<li><a href="#s11">Refreshing the data (CKAN)</a></li>
<li><a href="#s12">Honest limits &amp; appropriate use</a></li>
</ol></nav>
<main>
<section id="s1"><h2>1 · What this console shows</h2><p>Real, published Public Health Scotland A&amp;E, bed-occupancy and delayed-discharge reports populate the console automatically. Their reporting periods differ. The composite strain score is an interpretive planning tool, and scenario changes are illustrative.</p><div class="chip read"><b>Use it for:</b> questions about service pressure and the ten-parameter response checklist.</div><div class="chip warn"><b>Limits:</b> these aggregate reports cannot establish current bed availability, patient-level transfers or a forecast of harm. Current care decisions use official NHS guidance.</div></section>

<section id="s2"><h2>2 · Public sources and reporting cadence</h2><p>The collector discovers downloads through the <a href="https://www.opendata.nhs.scot/">official PHS portal</a>, checks each complete CSV against an independent CKAN row count, and publishes a dated package. Each panel displays its own report period.</p><table><tr><th>Panel</th><th>Definition</th><th>Publication cadence</th></tr><tr><td>A&amp;E</td><td>Type-1 EDs; All attendance category counted once; latest 30 reported weeks</td><td>Weekly</td></tr><tr><td>Occupancy</td><td>All Specialties, territorial-board aggregate occupied/staffed bed days</td><td>Quarterly</td></tr><tr><td>Delayed discharge</td><td>All Delay Reasons, ages 18+, territorial-board census counts</td><td>Monthly</td></tr><tr><td>Site names</td><td>Exact published treatment-location code lookup</td><td>Reference updates</td></tr></table><p>A source check cannot make an older report more recent. Report dates and package capture time remain separate. The latest available bed quarter can lag the weekly attendance data substantially.</p></section>

<section id="s3"><h2>3 · System strain index <span class="bdg C">INTERPRETIVE</span></h2><p>The 0–100 composite uses declared planning weights: 40% A&amp;E reference gap, 20% over-12-hour share, 15% mean board occupancy and 25% delayed-discharge count. These weights and scaling bounds are illustrative, not fitted clinical risk estimates. Mean board occupancy gives equal weight to each board.</p><p>Labels are Normal (&lt;40), Elevated (40–57), High pressure (58–73), and Severe (74+). Read the underlying values and periods before interpreting the label; it does not activate an incident or assess patient safety.</p></section>

<section id="s4"><h2>4 · A&amp;E 4-hour performance <span class="bdg M">MEASURED</span></h2><p>Within-four-hour attendances divided by total attendances across the reported Type-1 sites, for each of 30 weeks. The 95% blue line is a planning reference. <a href="https://www.opendata.nhs.scot/dataset/weekly-accident-and-emergency-activity-and-waiting-times">PHS explains</a> that weekly ED figures cover a subset of A&amp;E activity and should not be used to assess the full four-hour A&amp;E access standard.</p></section>

<section id="s5"><h2>5 · Emergency-department comparison <span class="bdg M">MEASURED</span></h2><p>Sites are ordered by their published within-four-hour percentage for the latest report week. The count waiting over 12 hours appears alongside it. Case mix, location, service scope and attendance volume differ; this is a descriptive comparison rather than a quality league table.</p></section>

<section id="s6"><h2>6 · Long waits <span class="bdg M">MEASURED</span></h2><p>Counts above eight and twelve hours come from the same selected Type-1 reports. The over-twelve-hour group is contained in the over-eight-hour group, so the counts must not be added. They support service-planning review; the console does not calculate excess deaths or individual risk.</p></section>

<section id="s7"><h2>7 · All-specialty bed occupancy <span class="bdg M">MEASURED</span></h2><p>For each of the 14 territorial boards: occupied bed days ÷ staffed bed days ×100. The collector selects each board's own aggregate and exact All Specialties category once, excluding site-level, All Acute and individual-specialty rows. The 85% and 92% color bands are illustrative planning references, not universal clinical safety thresholds. No live available-bed count is inferred.</p></section>

<section id="s8"><h2>8 · Delayed-discharge context <span class="bdg M">MEASURED</span></h2><p>Monthly census counts use ages18+ and All Delay Reasons once per territorial board. Age or reason subgroups are not added again. These reports help identify questions about discharge pathways, but their separate reporting period and aggregate scope cannot establish that a backlog caused a particular A&amp;E wait.</p></section>

<section id="s9"><h2>9 · Scenario sandbox</h2>
<p>Pressure-test the response plan against the <em>real</em> baseline. Four what-ifs adjust the inputs and recompute the strain index: <strong>winter surge</strong> (4-hour −15pp, occupancy +5, delays +15%), <strong>norovirus bed closures</strong>, <strong>staffing shortfall</strong>, <strong>discharge gridlock</strong> (delays +40%). These are clearly labelled as what-ifs, never forecasts — they exist to rehearse decisions, e.g. "a winter surge tips us from High pressure to Severe; is the escalation plan ready for that?"</p></section>

<section id="s10"><h2>10 · England comparator <span class="bdg C">DATED REFERENCE</span></h2><p>The England panel retains explicitly dated May2026 context and links to <a href="https://www.england.nhs.uk/statistics/statistical-work-areas/ae-waiting-times-and-activity/">NHS England's original releases</a>. It is not part of the automatic Scotland update. England's all-type figure and Scotland's Type-1 subset have different scopes and are not a like-for-like country ranking.</p></section>

<section id="s11"><h2>11 · Automatic refresh and failure behavior</h2><p>The <a href="https://github.com/patrickiben/public-data-consoles/actions/workflows/update-nhs.yml">scheduled workflow</a> checks PHS every Tuesday at13:43UTC. It validates source identities, complete downloads, mutually exclusive category selections, counts and reporting periods before publishing. GitHub schedules are best effort.</p><p>The hosted page checks its published package at startup, every six visible hours, on foreground return when due and through <b>Check published reports</b>. Invalid, partial, older or failed responses retain the last good package. Offline files use embedded reports. Scenario selection is preserved, and updated reports recompute the summaries and response checklist automatically.</p><p>No visitor key or provider credential is required. A maintainer can run <code>python3 scripts/refresh_nhs.py</code> and verify portable/embedded agreement with <code>python3 scripts/refresh_nhs.py --check</code>. Source URLs, hashes and coverage counts are in <a href="data/latest.json">the public report package</a>.</p></section>

<section id="s12"><h2>12 · Honest limits &amp; appropriate use</h2>
<ul>
<li><strong>Cadence, not real-time.</strong> Weekly (A&amp;E) to quarterly (beds). This is a situational read, not a live operational monitor.</li>
<li><strong>Descriptive, not predictive.</strong> No figure forecasts future performance; the scenarios are rehearsals, not predictions.</li>
<li><strong>One week is noisy.</strong> Always read a single ED's week against the national trend.</li>
<li><strong>Definitions matter.</strong> Type-1 (major ED) figures differ from all-types; cross-board and cross-country comparisons carry caveats.</li>
<li><strong>Snapshot vintage.</strong> The console states its data dates; refresh per Section 11 for current figures.</li>
</ul>
<div class="chip read"><b>Appropriate use:</b> situational awareness and planning support for system, board and operational leaders, and a public-facing, fully-sourced picture of unscheduled-care pressure. Because it's built only on already-published PHS open data, it carries no confidentiality constraint.</div>
</section>
</main></div></body></html>