Clinical Ethics · Governance-grade decision support
ClinicalEthica
For the cases that don’t resolve cleanly.

A structured ethical evaluation engine for clinicians and ethics committees. Grounded in published bioethical literature, fully traceable, and built to produce reasoning that holds up under scrutiny.

Build a case
Clinical ethicsGovernance traceThree frameworksExplainableAudit-ready
Frameworks
3

Consequentialist, deontological, and virtue-ethics reasoning combined in one evaluation.

Concern levels
5

None to critical, with recommended next steps. The tool informs the decision; it never makes it.

Traceability
Full

Every score, threshold, and citation is recorded in a governance-ready audit trail.

Scope: clinical ethics only

Decision support for the clinician who has to answer for it.

Fully traceable, framework-grounded ethical reasoning for clinical decisions, without compromising values, evidence, or accountability.

Built for clinicians, ethics committees, and governance teams.

Open case builderHow it works

ClinicalEthica is an early-stage prototype of a structured ethical reasoning layer for clinical decision-making. It provides heterogeneous ethical evaluation of proposed clinical actions by combining consequentialist, deontological, and virtue-ethics reasoning, with Beauchamp and Childress’s four principles as its default analytical framework. The output is explainable and cited, and is designed to support clinical deliberation, ethics committee review, and governance documentation.

ExplainableDeterministicCitedTraceableCommittee-ready

Built for the people who bear responsibility for clinical ethical decisions.

Clinicians
How do I make a difficult decision under time pressure and produce reasoning I can defend in a governance review?
Ethics committee members
How do I apply a consistent framework across committee members, rather than six people bringing six different intuitions to the same dilemma?
Governance and risk teams
How do I document that an ethical framework was applied, with specific provisions cited, and not just a note saying the committee reviewed and approved?
Medical educators
How do I make ethical reasoning visible to students in a way that invites challenge rather than compliance?

The reasoning behind clinical decisions is invisible. Until it needs to be.

Most clinical ethical reasoning happens silently. In the corridor, in the registrar’s head, at the bedside between two consultants. The decision gets recorded. The reasoning rarely does. When a case is reviewed weeks or months later, the trail that led to the action is reconstructed from memory, not from a record.

This works until it doesn’t. A complaint, a coroner’s inquiry, a governance audit, or a teaching review can each ask the question that the original decision-maker never had to answer in writing: why this action, and not the alternatives?

The obligation is not merely to reach the right decision but to reach it in the right way, through a process that is transparent, consistent, and capable of surviving scrutiny.
Beauchamp TL, Childress JF. Principles of Biomedical Ethics. 8th ed. Oxford University Press; 2019.
Ethics committees exist not because individual clinicians cannot reason morally, but because moral reasoning in high-stakes situations benefits from structure, consistency, and shared accountability.
UK Department of Health. Governance Arrangements for Research Ethics Committees. 2012.

ClinicalEthica does not replace that reasoning. It makes it visible at the moment of decision, in a form that survives the review.

Clinical ethical decisions are being made faster, under more pressure, with less structure than ever before.

01
From individual decisions to systemic patterns
Clinical decisions do not occur in isolation. A decision about capacity assessment today sets a precedent for the next case. A triage protocol adopted under pressure becomes standard practice. The ethics of the individual case and the ethics of the pattern it creates are not the same question.
02
From deliberation to reflex
Emergency conditions, resource scarcity, and institutional pressure have compressed the window for genuine ethical deliberation. If the framework cannot be applied at speed, it is structurally absent at the moment it is most needed.
03
From undocumented decisions to auditable reasoning
Governance reviews, legal challenges, and patient complaints increasingly demand not just what was decided but why. Decisions made without a documented reasoning trace leave clinicians and institutions without defence.

Three frameworks. One evaluation. Every decision traceable.

CASE INPUT
CLINICAL DILEMMA · ACTORS · CANDIDATE ACTIONS
EVALUATION ENGINE
DEONTOLOGICAL
Jurisdiction instrument pack
CH: Oviedo · ZGB · SAMS
E&W: MCA 2005 · Montgomery
CONSEQUENTIALIST
Local resource framework
NICE only where NICE binds
VIRTUE ETHICS
Pellegrino · MacIntyre
Jurisdiction-independent
PRINCIPLIST ARBITRATION
Beauchamp & Childress · Autonomy · Beneficence · Non-Maleficence · Justice
DISPOSITION
PROCEED · CAUTION · ESCALATE · HALT
TRACE · CITATIONS · AUDITABLE

Deterministic evaluation. No black box. Every score, threshold and citation reproducible.

See it evaluate.

A cycle through three cases. Same engine, three outcomes: Proceed, Escalate, Halt. Each with its own trace and citations.

LIVE DEMO · TRK-001 · --:--:--
Capacity dispute, adult refusing transfusion
Jurisdiction · England & Wales
Adult patient, post-RTA, declining transfusion on religious grounds
Capacity assessment performed: intact
Advance decision documented, witnessed
Candidate: respect refusal, optimise non-blood pathway
01 · Deontological screen
E&W pack · MCA 2005 ss.1, 24–26
PASS · valid advance decision
02 · Consequentialist screen
NICE (applies in E&W)
PASS · score 0.72
03 · Virtue ethics
Pellegrino
PASS · score 0.81
04 · Principlist arbitration
Beauchamp & Childress
NO CONCERN · autonomy decisive
DETERMINISTIC · RULES-BASED · TRACE ✓

How it works

Three stages, fully traceable end to end. Click through each stage to see what happens under the hood.

01 · INPUT
Build a case
Describe the clinical dilemma in plain language and let the builder turn it into structured input.
Dilemma
Free-text description of the situation, actors, and the clinical question in dispute.
Jurisdiction
Required, not optional. The jurisdiction determines which legal instruments bind the case and which resource framework applies, so it is captured before anything is scored.
Context
Urgency band and resource scarcity level, which decide which triage protocols activate alongside the jurisdiction's own guidance.
Candidate actions
Two to five candidates, each with projected benefits, harms, and a short rationale.

Your evidence, always within reach.

A case is only as good as the evidence behind it. The repository lets you store the documents that matter to your practice: trust policies, NICE guidance, committee minutes, journal articles, and past case reviews. When you build a case, type / anywhere in the dilemma or clinical context to pull in a document and ground the reasoning in your own materials.

The engine can restrict retrieval to the documents you have referenced, or treat them as the most important evidence while still searching across the broader framework library. Every referenced document is shown in the results trace, so the answer is never just an opinion; it is an opinion tied to something you can cite.

Upload

Add PDFs, pasted text, and clinical notes into folders you control.

Reference

Type / while writing a case to search your repository and insert a document.

Trace

Retrieved evidence is badged in the results, so reasoning stays auditable.

The same reasoning architecture applies wherever high-stakes decisions need to survive scrutiny.

Clinical ethics is the first domain because the frameworks are unusually well codified. The engine, however, is general. Any field in which a decision must be defensible against multiple, sometimes conflicting, normative standards can adopt the same structure of layered evaluation, non-derogable prohibitions, and arbitrated composite output.

LEGAL
Legal and regulatory review
Weigh statute, case law, and professional codes against proposed action, with a written trace ready for appellate or regulator review.
RESEARCH
Research ethics
Screen proposed studies against consent norms, welfare thresholds, and disciplinary codes before an IRB or REC sees the paperwork.
AI
AI and algorithmic governance
Test model deployments against safety thresholds, fairness constraints, and organisational red lines before release.
PUBLIC POLICY
Public policy and resource allocation
Compare distributive options under scarcity with an explicit ledger of who gains, who loses, and which prohibitions bind.
CORPORATE
Corporate ethics and compliance
Evaluate business decisions against codes of conduct, stakeholder impact, and jurisdiction-specific duties in a repeatable format.
DEFENCE
Defence and dual-use
Stress-test operational choices against IHL, rules of engagement, and proportionality, with the trace preserved for later review.
EDUCATION
Ethics education
Turn any case study into a live worked example, with the reasoning steps visible to students and open to challenge.
JOURNALISM
Editorial and journalism ethics
Apply a consistent standard to publication decisions on harm, consent, and public interest, and keep the reasoning on record.

In every case, the pattern is the same. Structured input, layered evaluation, arbitrated composite output, full trace. The frameworks change with the field. The architecture does not.

No. It is a decision-support tool. It does not prescribe treatment, replace clinical judgement, or substitute for ethics committee deliberation. Clinicians retain full responsibility.
The evaluation engine is deterministic and rules-based. AI is used only to help interpret case inputs into structured parameters. Every ethical computation is reproducible arithmetic and defined logic.
Beauchamp and Childress’s principlist framework arbitrates in every case. The deontological layer is jurisdiction-specific: each case is screened against the instrument pack that actually binds it, so Switzerland is read against the Oviedo Convention, ZGB Art. 370–373 and SAMS guidance, while England and Wales is read against the Mental Capacity Act 2005, Montgomery and GMC guidance. Oviedo is never applied to a jurisdiction that has not ratified it. The consequentialist layer likewise uses the local resource framework, so NICE cost-per-QALY thresholds appear only on cases where NICE applies.
No patient-identifiable information should be entered. Case data is stored in session only by default. The governance log is exportable as a signed PDF.
Start with a real case.

Apply a framework to a clinical dilemma. See a fully traceable analysis in seconds.

Open case builderRequest access →

Opens your email client. No account required.