AI glossary · Basic · Language models · Classic ML, recsys & search · Healthcare & bio
CDSS
A CDSS (Clinical Decision Support System) is software that shows alerts or recommendations based on patient information and medical knowledge when doctors and nurses make diagnostic or prescribing decisions.
About this term
- Level
- Basic · Technical terms you often meet in AI news
- Category
- Industry AI
- Industries
- Healthcare & bio
- Added
- 2026-10-10
How it works
A CDSS is usually connected to a hospital's electronic health record (EHR) and reads the patient's diagnoses, test results and current medications. The traditional approach is to load rules in advance, such as 'warn if these two drugs are used together,' and show an alert when the conditions are met. More recently, there are also more approaches in which machine learning models learn from past data to predict in advance the risk of sepsis or deterioration. The results appear as alerts or recommendations on the screens clinicians use to prescribe or review records.
Rule-based vs. AI-based CDSS
Rule-basedPredefined conditions
- Drug interaction and allergy alerts
- Clear rationale, easy to explain
- Rules need ongoing human maintenance
AI-basedPredictions learned from data
- Predicting sepsis and deterioration risk
- Can find hidden warning signs
- Needs re-validation at each hospital
공통 Both connect to the EHR to support clinicians' decisions, and clinicians make the final judgment
Why it matters
Clinicians see many patients in a short time, and it is hard to remember every drug interaction or the latest clinical guidelines. A CDSS serves as a safeguard that reduces errors such as drug allergies, duplicate prescriptions and dosing mistakes. It is also used to make care quality more consistent by reminding clinicians not to skip tests or treatments recommended by clinical guidelines. With the addition of AI, it is also expected to detect early warning signs that were previously hard to notice.
Things to know
When there are too many alerts, 'alert fatigue' sets in and clinicians habitually dismiss them. AI-based predictions can lose accuracy in settings different from the hospital where they were trained, so on-site validation is needed. If the reasoning behind a recommendation cannot be shown, clinicians find it hard to trust and use. Also, a CDSS is only a tool that supports decisions; final responsibility and judgment rest with clinicians.
Example
A typical CDSS is the warning window that pops up when a doctor prescribes a drug in the electronic health record and it contains an ingredient the patient is allergic to, or forms a dangerous combination with a drug the patient is already taking. AI prediction features that continuously monitor inpatients' vital signs and lab values and notify the care team when sepsis risk rises are also built into many hospital systems.