AI output must always be reviewed by qualified healthcare professionals before use in clinical communication or decision-making.
1Advisory Use Only
AI-generated or automated output is informational and assistive only. It must not be treated as a diagnosis, order, treatment recommendation, or substitute for licensed professional review.
2Human Verification
Users must verify completeness, accuracy, patient identity, clinical context, recipients, and urgency before relying on or sending generated content.
3Known Limitations
Automated systems may produce incomplete, inaccurate, outdated, biased, or contextually inappropriate output and may fail to identify emergency conditions.
4Emergency Care
SBARAlerts and any AI feature must not replace emergency protocols, 911, urgent physician escalation, or facility-specific emergency procedures.
5Data Use
PHI should be submitted to an AI provider only when authorized, contractually protected, technically appropriate, and covered by required HIPAA arrangements.