This platform does not calculate clinical scores

There is no clinical calculator here. Topic search is the only input on this site.

That is three decisions, not one. No score is implemented from memory — the formula, its units, its version and its bounds must be verified against the originating publication, and unit tests must reproduce published worked examples, or nothing ships. No diagnostic tool is offered to the public, so no reader can be given a personalised risk or prognosis. The small search enhancement only filters an existing topic catalogue; it performs no clinical calculations.

The register

Every score this platform's content references, and its state. A score in this register is one this platform has decided it may not compute — the reasons differ and are recorded in full.

UKELD — 🔴 BLOCKED
MELD and variants — 🔴 BLOCKED
Child–Pugh — 🔴 BLOCKED
King's College criteria — 🟠 PARTIALLY SUPERSEDED FOR UK USE
CLIF-C ACLF — 🔴 BLOCKED
ALBI grade — 🔴 BLOCKED
BCLC staging — 🟡 NOT A CALCULATOR
West Haven criteria — 🔴 BLOCKED
ISGLS PHLF grading — 🟠 PARTIAL

What a dedicated attempt to clear these found

A retrieval round in September 2026 tried to clear four of them. None cleared, and the near misses are the interesting part.

The UK's own listing score was recovered in full — from the national body's own calculator file rather than from the paper that defined it. It did not clear, because that file applies an undocumented correction when one named transplant centre is selected, a correction that appears in no published description of the score. An implementation is not the publication, and this one demonstrably differs from it.

Child-Pugh was retrieved from four acceptable sources — two regulators, a specialist society guideline, and an NHS pharmacy service. They disagree with one another on how two of the five variables are graded. There is no single canonical table, and no published worked example exists in any of them. It is the score this platform's content leans on most heavily, and the one it is least able to compute.

MELD turns out to be four different scores sharing a name, two of which are incompatible equations both called the same thing. The variant with the cleanest evidence is the one nobody uses.

The full record, per score, is in the repository's blocker register.

What you can rely on instead

Where a UK guideline states a threshold, this platform reports the threshold and attributes it — NICE's use of a MELD score, the UK listing threshold, the discriminant function value at which NICE recommends considering corticosteroids. Those are quoted from the guideline that states them. What this platform will not do is work out which side of a threshold you are on.

No clinical values are requested, stored or transmitted. Topic search runs locally in your browser and does not request personal health information.