Related Experiment Videos
When is a diagnostic test result positive? Decision tree models based on net utility and threshold.
Archives of Pathology & Laboratory Medicine
|September 1, 1986
Summary
This study introduces decision tree models to determine positive diagnostic test results. Providing raw test performance data aids in selecting appropriate cutoff levels for clinical decision analysis.
Area of Science:
- Medical Decision Making
- Biostatistics
- Diagnostic Test Evaluation
Background:
- Determining when a diagnostic test result is positive is crucial for clinical practice.
- Existing methods for setting cutoff levels may lack a systematic approach.
Purpose of the Study:
- To develop and present simple decision tree models for selecting appropriate diagnostic test cutoff levels.
- To emphasize the importance of providing raw test performance data for robust analysis.
Main Methods:
- Development of two decision tree models: a net utility model and a threshold model.
- Incorporation of these models into a user-friendly software program for desktop computers.
Main Results:
- The developed models offer a structured approach to defining positive diagnostic test results.
- The software facilitates the application of these models for practical use.
Conclusions:
- Clinical decision analysis, using decision tree models, can effectively address the question of positive diagnostic test results.
- Availability of raw test performance data is essential for identifying appropriate and optimal cutoff levels.