Having worked through the specific limitations of single-letter evidence grading throughout this series, it is worth asking directly what a more complete, more honest version might actually look like for a UK clinical audience specifically.
Why a single A to D grade is unlikely to capture everything a UK clinician needs
A single letter necessarily compresses several genuinely distinct dimensions, evidence hierarchy, methodological certainty, guideline status, recommendation strength, and jurisdiction-specific applicability, into one signal. Each of these can point in a different direction for the same underlying clinical question, and compression inevitably loses information a clinician might specifically need for the decision in front of them.
Separate fields worth displaying instead
A more complete display might show evidence hierarchy, the type of study or synthesis the claim rests on; certainty of evidence, the GRADE-style confidence in the estimated effect itself; guideline status, whether and how the relevant UK body has addressed this question; recommendation strength, how firmly that guidance is stated; date of last review, so a clinician knows how current the underlying assessment actually is; UK licensing status, whether the medicine or intervention is actually licensed for this indication; NHS availability, whether it is realistically accessible through the standard pathway; and patient applicability, whether the underlying evidence population resembles the patient actually being treated.
A worked example of what this might look like in practice
Consider a display reading: NICE-recommended; supported by systematic review of randomised controlled trials; moderate certainty; conditional recommendation; evidence last reviewed in 2024; newer trial published in 2026 not yet incorporated. This is visually more complex than a single letter, and it preserves genuinely important information a single letter cannot.
Why hierarchy and quality must remain visibly distinct
A meta-analysis can occupy a high position on the evidence hierarchy while still being methodologically weak, for the reasons covered elsewhere in this cluster. A carefully conducted cohort study can, for certain questions such as rare harms, represent the best available evidence despite sitting lower on the general hierarchy. Collapsing hierarchy and quality into a single figure obscures exactly this kind of important exception.
Why no display should average unrelated outcomes
As covered directly elsewhere in this series, a single answer's efficacy claim, safety claim and applicability claim can rest on entirely different bodies of evidence of entirely different strength, and any honest display needs to preserve that separation rather than blending them into one number.
The iatroX principle underlying this proposal
Favour the strongest appropriate study design for the specific question being asked, prefer systematic reviews and meta-analyses where they genuinely apply, and make limitations and any lower-level evidence being relied upon visible rather than smoothed over.
An invitation, not a finished answer
This article is offered as thought leadership from a practising UK clinician building a UK and EU clinical-knowledge platform, not as a finished specification. Whether certainty is best shown per claim, per outcome, or per overall recommendation remains a genuinely open design question worth discussing openly rather than resolving unilaterally.