# Clinical decision support

> Clinical decision support is software that provides clinicians with patient-specific information and recommendations to inform a clinical decision. The defining constraint is that it assists rather than replaces judgement — it issues no orders and makes no diagnoses, and the treating clinician remains responsible for every decision taken.

*Source: https://getpathways.ai/glossary/clinical-decision-support*

Also known as: CDS, CDSS, clinical decision support system

- **Abbreviation**: CDS / CDSS
- **Produces**: Patient-specific recommendations for clinician review
- **Does not produce**: Diagnoses, orders, or treatment directives
- **Responsibility**: Remains entirely with the treating clinician

## Recommends versus decides

The distinction sounds like a legal formality and is in fact an architectural
one. A system that recommends must be able to show its working — the evidence
that triggered a recommendation, the rule or guideline behind it, and the
version of that rule — because a clinician cannot exercise judgement over a
conclusion they cannot inspect.

That requirement is what pushes safety-critical logic towards deterministic,
versioned rules rather than model output. A rule that fires whenever its
conditions are met can be tested, audited and reproduced months later. A
probabilistic answer that is usually right cannot be examined the same way, and
so cannot be reviewed in the sense clinical governance means by review.

## In perioperative practice

Pathways Perioperative operates as clinical decision support: it produces
severity-tiered risk flags with the triggering evidence, a guideline reference
and a recommended action. It places no orders, clears no patients, and routes
nobody to theatre on its own authority.