What is the difference between a knowledge-based and non-knowledge-based clinical decision support system?

Knowledge-based CDSS uses explicit rule engines drawing on clinical guidelines (e.g. NICE recommendations) — ‘if condition X, then suggest Y’. Non-knowledge-based CDSS uses machine learning models that identify patterns in clinical data without explicit rules. Hybrid systems combine both approaches. Knowledge-based systems are more transparent and explainable; ML-based systems can detect complex patterns but may be less interpretable.