Related Experiment Videos
The clinical algorithm nosology: a method for comparing algorithmic guidelines.
S D Pearson1, C Z Margolis, S Davis
1Clinical Guidelines Program, Harvard Community Health Plan, Brookline, Massachusetts.
Summary
A new method, Clinical Algorithm Nosology (CAN), objectively compares clinical algorithms. CAN reliably measures complexity and similarity between medical guidelines, aiding healthcare choices.
Area of Science:
- Health Services Research
- Medical Informatics
- Clinical Decision Support
Background:
- Rising healthcare costs and quality concerns drive the development of numerous clinical practice guidelines.
- Objective methods are lacking to compare the similarity of alternative guidelines for identical clinical issues.
Purpose of the Study:
- To introduce and validate the Clinical Algorithm Nosology (CAN) as a novel method for comparing clinical algorithms.
- To assess CAN's ability to measure algorithm complexity and clinical similarity objectively.
Main Methods:
- The CAN method was developed to measure clinical algorithm design complexity independently of content.
- CAN qualitatively assesses clinical differences and quantitatively scores similarity between algorithms.
- Interrater reliability was tested on five pairs of clinical algorithms across three medical topics.
Main Results:
- CAN complexity scores showed high correlation with clinician estimates (r = 0.86).
- The CAN clinical-similarity scoring achieved 80% interrater agreement (kappa = 0.73).
Conclusions:
- The CAN is a valid and reliable tool for assessing clinical algorithm structural complexity, differences, and similarity.
- CAN can support the evaluation of guideline development and aid providers and purchasers in selecting among clinical guidelines.