Reference

Method

How clinical intelligence is gathered, assessed, synthesised, and what it can and cannot claim

The four gaps

The Value Forge operates on the premise that clinical intelligence failures cluster into four distinct gaps:

Epistemic gap — what the published evidence shows versus what clinical teams know about. The gap is not ignorance; it is the structural absence of a mechanism for turning published research into timely, domain-specific intelligence at team level.

Translation gap — what is known versus what is commissioned or implemented. Evidence does not automatically become practice. The translation gap is where well-evidenced interventions fail to reach the patients who need them.

Incentive gap — where the financial and performance architecture pulls against the clinical evidence. When tariffs, contracts, or accountability frameworks reward the wrong things, well-motivated clinicians cannot close the first two gaps alone.

Capacity gap — where even clear evidence, honest intent, and aligned incentives are defeated by workforce shortage, estate constraints, or access barriers. Knowing what to do and being unable to do it.

The Value Forge directly addresses the epistemic and translation gaps. It does not close the incentive or capacity gaps — naming that boundary is part of its intellectual honesty.

How intelligence reaches this page

The Value Forge runs on a regular analytical cadence. Three upstream stages produce the raw intelligence; everything then passes through a single translation checkpoint before publication. That checkpoint is a boundary function, not an editorial one — it changes register, never meaning. If an item cannot be published without breaking that rule, it is dropped and the gap is named.

Honest scope

The system can reach: published literature accessible via standard academic databases and preprint servers; grey literature from NHS England, NICE, parliamentary committees, professional bodies, and major charities; local administrative data where publicly available.

The system cannot reach: unpublished local audit, patient-level data, real-time operational data, or anything behind institutional paywalls that have not yielded to repeated access attempts. Access gaps are named, not elided.

How to read this publication

Everything above describes how the intelligence is produced. This key describes how to read it — the lens, the domains, the artefact types, the labels and statuses on every page, and the pattern register.

Value added time — the central lens
Value added time (VAT) is time within the health system that demonstrably improves or preserves a patient’s functional capacity, quality of life or clinical outcome — as distinct from time spent waiting, in unnecessary admission, or serving institutional convenience rather than the person receiving care.
Value added time is not a metric; it is the evaluative frame. Every piece of intelligence here is filtered through one question: does this add genuine value to patient time? On this site, gold marks VAT significance and nothing else: the gold VAT Signal callout inside an article names its single most consequential VAT implication, and the front page’s pattern strip turns gold only when a pattern is editorially designated VAT‑consequential. Gold is never decoration — its rarity is its meaning.
The five domains
D01
Acute Frailty
When and how does specialist frailty input change outcomes?
D02
Clinical Psychology
What is the relationship between psychological care and physical health trajectories?
D03
Palliative Care
How does community palliative care affect place of death, hospitalisation, and carer outcomes?
D04
Therapies & AHPs
What is the timing and sequencing evidence for AHP input across the pathway?
D05
OPAT
What are the safety signals, preservation effects, and MDT requirements for community IV therapy?

These domains are not separate bulletins. They are five perspectives on the same question — where does NHS practice stand against world best practice? — and the most consequential insights emerge at their intersections, not within any single domain alone. Every article names its domain span in the line above its title, and the domain pages collect each lens in one place.

The three artefacts — what you are reading
Signal
Weekly · 0–n
A single evidence item with national significance — published only when it clears an evidence-weight bar, so the weekly count floats with what the evidence earns. Each carries its evidence label, finding status and domain span in the masthead, and a full reference list beneath the text; PMIDs link straight to the record.
Synthesis
Weekly · 1
The week’s evidence read across all five domains: cross‑domain connections made explicit, unresolved threads carried as named open questions, contrary signals kept. The analytical core of the publication.
Monthly
Monthly · flagship
The deep cross‑domain review: domain‑by‑domain assessment, pattern movements, absence detection, and the international benchmark. Edition‑dated, featured on the landing page, and the place where the pattern register is confirmed, updated or revised.
Evidence status — how strongly to hold a finding
StatusWhat it meansTypical basis
Established
Consistent across multiple high‑quality studies
Safe to treat as a working basis for clinical and operational decisions. Not immutable — significant contrary evidence would revise it.
Meta‑analysis
Systematic review
Provisional
Promising but not yet replicated at scale
Directionally informative. Useful for shaping questions and pilots; premature as the sole basis for service redesign.
Single RCT
Cohort study
Contested
Active disagreement in the literature
Named here because it is being acted on in practice, not because it is settled. Decisions resting on contested evidence carry identifiable risk.
Conflicting
primary studies
Finding status — the pills on signals and syntheses
MarkerWhat it means
NEW
First appearance in this publication
Not tracked before, or newly significant nationally.
CONFIRMED
Prior finding reinforced by new evidence
The position held previously now has additional support.
UPDATED
Prior finding revised — direction or magnitude has changed
Read the current entry, not the previous one. The update is the finding.
DISCARDED
No longer supported by the evidence base
Named explicitly so that decisions built on it can be revisited.

These pills appear on the article masthead and on every card and list row. The front page’s briefing strip adds a trend marker (Rising, Emerging) for the direction a signal is moving.

The pattern register — the longitudinal spine

Patterns are recurrent findings that hold across domains and across editions — the most consequential intelligence the publication produces, invisible if you read any single article in isolation. Each carries a permanent id (VF‑P1, VF‑P2 …), a status, and a dated history with the references that earned every entry. Patterns enter the register as provisional and are promoted to established only by repeated observation from independent evidence; evidence that narrows or contradicts a pattern is published against it — qualified, contested, or discarded — and nothing is ever quietly rewritten.

VF‑P1 Earlier specialist clinical input generates materially better outcomes, lower cost and more value‑added patient time Established
Every pattern id in an article links to that pattern’s own page, which carries its full dated history, its evidence, and every article that bears on it. The register page holds the current state of all patterns; the evidence timeline shows how they have moved.
Evidence labels used throughout: [Meta‑analysis] · [Primary Study] · [Guidance] · [Audit/Data] · [Synthesis] · [Preprint] · [Unverified]
These appear inline and as badges wherever the type of evidence materially affects how strongly a finding should be held. A [Primary Study] and a [Meta‑analysis] warrant different levels of confidence. DISCARDED findings stay named so that decisions built on them can be revisited — the site’s public revision history is the audit trail.