Tooluniverse Clinical Guidelines

What it does

This skill allows AI agents to access and summarize clinical guidelines from MIMS and Harvard resources, specifically targeting patient care decisions. It prioritizes guideline sources based on their rigor and update frequency, favoring NICE and WHO guidelines first, followed by society guidelines (AHA, ADA, NCCN, SIGN), aggregator databases (GIN, TRIP, OpenAlex), and finally literature databases (PubMed, EuropePMC). The skill also enables the agent to perform computational analysis of retrieved data using Python.

When to use it

• When needing evidence-based recommendations for patient care.
• To quickly find guidelines related to a specific medical condition or treatment.
• For research purposes, requiring access to and summarization of clinical guidance.
• When needing to compare guidelines from multiple sources.
• When computational analysis (statistics, data processing) is needed on retrieved guideline data.

Key capabilities

• Guideline Source Prioritization: Ranks guidelines based on evidence grade and update frequency (NICE/WHO > Society > Aggregator > Literature).
• Multi-Source Search: Queries NICE, TRIP, GIN, AHA, ADA, NCCN, CPIC, PubMed, and EuropePMC.
• Search Strategy Refinement: Starts with narrow searches then broadens to disease categories if needed.
• Publication Date Emphasis: Highlights the publication date of guidelines and notes potential newer updates.
• Computational Analysis: Supports Python-based data analysis (pandas, scipy, statsmodels, matplotlib) for retrieved guideline information.

Example prompts

• "Find clinical guidelines for managing type 2 diabetes."
• "What are the NICE guidelines on treating hypertension?"
• "Summarize the ADA's recommendations for gestational diabetes management and include the publication year."
• "Retrieve guidelines related to heart failure, specifically HFpEF with SGLT2 inhibitors, from at least three sources."

Tips & gotchas

• Publication Date is Crucial: Always check the guideline’s publication date as older guidelines may be superseded.
• Source Verification: When using aggregator databases (GIN, TRIP, OpenAlex), verify the original source of the guideline.
• Computational Analysis Requires Python Proficiency: The skill leverages Python for data analysis; familiarity with libraries like pandas and matplotlib is needed to utilize this capability fully.