Related Experiment Videos
Design and pilot evaluation of a system to develop computer-based site-specific practice guidelines from decision
G D Sanders1, R F Nease, D K Owens
1Center for Primary Care and Outcomes Research, Department of Medicine, Stanford University, California 94305-6019, USA. sanders@stanford.edu
Summary
A new web-based system, ALCHEMIST, automatically creates tailored clinical practice guidelines (CPGs) from decision models. Pilot studies show these computer-generated CPGs are superior in quality and usability compared to traditional guidelines.
Area of Science:
- Medical Informatics
- Clinical Decision Support
- Health Services Research
Background:
- Local tailoring of clinical practice guidelines (CPGs) is challenging due to the scarcity of medical experts.
- Decision models (DMs) offer a framework for guideline development.
- A computer-based system can facilitate the creation, dissemination, and tailoring of CPGs.
Purpose of the Study:
- To develop and evaluate a web-based system (ALCHEMIST) for automated CPG generation and tailoring.
- To assess the quality, usability, accountability, and accuracy of ALCHEMIST-generated CPGs compared to published guidelines.
Main Methods:
- ALCHEMIST analyzes a DM to create an annotated algorithm CPG.
- Remote users can tailor CPGs by modifying input variables.
- A pilot study involved 15 subjects comparing ALCHEMIST CPGs with published guidelines for non-small-cell lung cancer staging.
Main Results:
- ALCHEMIST CPGs received significantly higher overall quality scores (1.502) than published CPGs (0.987).
- ALCHEMIST CPGs outperformed published guidelines in usability, accountability, and accuracy (p < 0.05 for all).
- Users rated ALCHEMIST highly for ease of use (4.76/5), content usefulness (3.98/5), and presentation (4.64/5).
Conclusions:
- A web-based system can feasibly automate CPG creation as annotated algorithms.
- ALCHEMIST facilitates the development of site-specific CPGs by remote users.
- ALCHEMIST-generated CPGs demonstrate favorable quality and usability, suggesting their utility in guideline development.