How do likelihood ratios help in updating pretest probability to post-test probability?

How do likelihood ratios help in updating pretest probability to post-test probability?

• A. They replace the need for pretest probability.
• B. They are only used to assess test reliability.
• C. They combine with pretest probability to update posttest probability.
• D. They determine the order of tests to perform.

Answer: (C)

The key idea is that a test result shifts our estimate of disease probability by multiplying the pretest odds by a likelihood ratio. Bayes’ updating is what this is about: start with the pretest probability, convert to odds, apply the appropriate likelihood ratio for the test result (positive or negative), and convert back to a post-test probability. For example, if the pretest probability is 20%, the pretest odds are 0.25. A positive result with an LR+ of 5 yields post-test odds of 1.25, which is about a 56% post-test probability. A negative result with an LR- of 0.2 yields post-test odds of 0.05, about a 5% post-test probability. This shows how the likelihood ratio scales the prior probability up or down based on the test result. So, likelihood ratios are the bridging factor that combines with pretest probability to produce post-test probability. They don’t replace the pretest probability, they don’t measure test reliability alone, and they don’t determine the order of tests.

The key idea is that a test result shifts our estimate of disease probability by multiplying the pretest odds by a likelihood ratio. Bayes’ updating is what this is about: start with the pretest probability, convert to odds, apply the appropriate likelihood ratio for the test result (positive or negative), and convert back to a post-test probability.

For example, if the pretest probability is 20%, the pretest odds are 0.25. A positive result with an LR+ of 5 yields post-test odds of 1.25, which is about a 56% post-test probability. A negative result with an LR- of 0.2 yields post-test odds of 0.05, about a 5% post-test probability. This shows how the likelihood ratio scales the prior probability up or down based on the test result.

So, likelihood ratios are the bridging factor that combines with pretest probability to produce post-test probability. They don’t replace the pretest probability, they don’t measure test reliability alone, and they don’t determine the order of tests.