Clinical Scorecard: Harnessing AI Literacy in Infectious Disease Management: Navigating the Agentic Era

At a Glance

Key Highlights

• Agentic AI advances enable real-time synthesis of epidemiology, microbiology, resistance patterns, and clinical risk factors to support ID decisions.
• AI literacy is essential for ID clinicians to actively guide AI tool use, ensuring appropriate automation boundaries, equity, and safety.
• AI-powered workflows can reduce clinician burden by automating repetitive tasks and democratize access to ID expertise across varied care settings.

Guideline-Based Recommendations

Diagnosis

• Incorporate AI tools such as antimicrobial resistance algorithms and sepsis predictive models to enhance diagnostic accuracy.
• Use AI systems that integrate structured and unstructured data including clinical notes and microbiology reports.

Management

• Adopt agentic AI workflows to augment antimicrobial stewardship and infection control programs.
• Ensure human oversight in critical decisions to mitigate risks of AI errors and biases.
• Advocate for high-quality data inputs and continuous validation of AI tools in clinical practice.

Monitoring & Follow-up

• Monitor AI tool performance for false positives, generalizability, and equity across diverse patient populations.
• Establish governance policies and interdisciplinary collaboration to oversee AI integration and impact.

Risks

• Be aware of alert fatigue caused by high false positive rates in AI alerts.
• Recognize potential documentation inaccuracies from AI-generated text errors.
• Address algorithmic biases that may perpetuate healthcare inequities if AI tools are used uncritically.

Patient & Prescribing Data

Patients with infectious diseases receiving AI-augmented clinical care

AI tools support antimicrobial prescribing by enforcing guidelines and providing decision support but require clinician oversight to avoid errors and biases.

Clinical Best Practices

• Develop foundational AI literacy among ID clinicians to interpret, validate, and optimize AI tool use.
• Set clear boundaries on automation to maintain human control over critical clinical decisions.
• Promote interdisciplinary education and governance frameworks for responsible AI deployment in infectious disease practice.
• Leverage AI to automate repetitive, searchable tasks to reduce clinician burnout and improve workflow efficiency.
• Utilize AI-enabled hub-and-spoke and mobile-first models to expand access to ID expertise in underserved settings.

References

• Agentic AI systems in clinical environments
• Antimicrobial resistance deterministic algorithms
• Sepsis predictive models
• AI challenges in healthcare including alert fatigue and biases
• Large Language Models and natural language processing in healthcare