Clinical decision support (CDS)

Clinical decision support (CDS) refers to software, usually embedded in an electronic health record system, that analyzes patient data and surfaces alerts, reminders, or recommendations to clinicians at the point of care — for example, flagging a dangerous drug interaction, reminding a doctor that a patient is due for a screening, or suggesting a diagnosis based on symptoms and history.

CDS ranges from simple rule-based systems, such as "alert if two prescribed drugs interact," to machine-learning models that predict a patient's risk of sepsis or deterioration from patterns in vital signs and labs. Rule-based CDS is easier to audit and explain; model-based CDS can catch subtler patterns but raises the same interpretability concerns as machine learning generally, which is why regulators and hospitals increasingly expect some form of Explainable AI before a model-driven alert changes clinical decisions.

CDS matters because it can catch errors and gaps a busy clinician might miss, and it is a key tool for both population health management and drug-safety work like pharmacovigilance. The best-documented pitfall is alert fatigue: when a system fires too many low-value warnings, clinicians start reflexively dismissing them, which can cause a genuinely important alert to be missed along with the noise.