What do Likelihood Ratios fundamentally describe?

What do Likelihood Ratios fundamentally describe?

• A. The ratio of true positives to false positives.
• B. The ratio of false negatives to true negatives.
• C. The ratio of disease prevalence among tested individuals.
• D. The statistic that quantifies how likely a test result will change the probability that a patient has the condition.

Answer: (D)

Likelihood ratios describe how much a test result changes the probability that a patient has the disease. They compare how likely the result is in people with the disease versus those without it, so they directly quantify the shift in diagnostic probability after a test result. A positive result increases certainty about disease by multiplying the pretest odds by the positive likelihood ratio (LR+ = sensitivity / (1 − specificity)). A negative result decreases certainty by multiplying the pretest odds by the negative likelihood ratio (LR− = (1 − sensitivity) / specificity). This approach is independent of how common the disease is in the population, making LRs useful for updating probability across different settings with Bayes’ reasoning or a nomogram. For example, if your pretest probability is 30% (pretest odds ≈ 0.43) and the test’s LR+ is 8, the post-test odds become about 3.43, corresponding to a post-test probability around 77%. If the test is negative and LR− is 0.2, the post-test odds drop to about 0.086, or a probability around 8%. The other ideas don’t capture this updating role: a ratio of true positives to false positives relates more to precision or PPV concepts, a ratio of false negatives to true negatives isn’t a standard diagnostic update measure, and disease prevalence corresponds to pretest probability rather than how a test result reshapes that probability.

Likelihood ratios describe how much a test result changes the probability that a patient has the disease. They compare how likely the result is in people with the disease versus those without it, so they directly quantify the shift in diagnostic probability after a test result.

A positive result increases certainty about disease by multiplying the pretest odds by the positive likelihood ratio (LR+ = sensitivity / (1 − specificity)). A negative result decreases certainty by multiplying the pretest odds by the negative likelihood ratio (LR− = (1 − sensitivity) / specificity). This approach is independent of how common the disease is in the population, making LRs useful for updating probability across different settings with Bayes’ reasoning or a nomogram.

For example, if your pretest probability is 30% (pretest odds ≈ 0.43) and the test’s LR+ is 8, the post-test odds become about 3.43, corresponding to a post-test probability around 77%. If the test is negative and LR− is 0.2, the post-test odds drop to about 0.086, or a probability around 8%.

The other ideas don’t capture this updating role: a ratio of true positives to false positives relates more to precision or PPV concepts, a ratio of false negatives to true negatives isn’t a standard diagnostic update measure, and disease prevalence corresponds to pretest probability rather than how a test result reshapes that probability.