How to use and interpret interval likelihood ratios

Area of Science:

• Medical Diagnostics
• Biostatistics
• Clinical Decision-Making

Background:

• Likelihood ratios offer advantages over sensitivity and specificity for diagnostic test characterization.
• Unlike sensitivity and specificity, likelihood ratios can account for the magnitude of test result abnormalities.
• This is crucial as many diagnostic tests yield continuous or ordinal data.

Purpose of the Study:

• To demonstrate the advantages, application, and interpretation of interval likelihood ratios.
• To highlight how interval likelihood ratios can inform clinical decisions more effectively than traditional metrics.
• To illustrate these concepts using a clinical case of a febrile child.

Main Methods:

• Utilizing interval likelihood ratios for diagnostic test evaluation.
• Comparing calculations of posttest probabilities using interval likelihood ratios versus sensitivity and specificity.
• Applying these methods to a clinical scenario involving a young child with a high fever.

Main Results:

• Interval likelihood ratios provide a more detailed assessment of diagnostic test performance.
• Calculated posttest probabilities may differ when using interval likelihood ratios compared to sensitivity and specificity.
• This difference can lead to alternative clinical decision-making pathways.

Conclusions:

• Interval likelihood ratios offer a superior method for evaluating diagnostic tests with continuous or ordinal scales.
• The use of interval likelihood ratios can lead to more informed and potentially different clinical decisions.
• Understanding and applying interval likelihood ratios is essential for accurate diagnostic interpretation.