Talk:Likelihood ratios in diagnostic testing

complexity
This is still too complex - more examples might help perhaps building up to the rather complicated example used. — Preceding unsigned comment added by 194.176.105.144 (talk) 18:31, 30 August 2014 (UTC)

This
This article now begins thus:
 * In evidence-based medicine, a likelihood ratio is a method of assessing the value of performing a diagnostic test.
 * In evidence-based medicine, a likelihood ratio is a method of assessing the value of performing a diagnostic test.

Rather like starting an article titled Atlantic Ocean by saying
 * In telecommunications, the Atlantic Ocean is a body of water across which the transatlantic cable was laid.
 * In telecommunications, the Atlantic Ocean is a body of water across which the transatlantic cable was laid.

Or an article titled Italy by writing this:
 * Italy is a country whose team won two gold medals in cycling in the 1948 Olympics. [Then the article could go on for a hundred paragraphs about those two events, and that would be the whole story of Italy.]
 * Italy is a country whose team won two gold medals in cycling in the 1948 Olympics. [Then the article could go on for a hundred paragraphs about those two events, and that would be the whole story of Italy.]

I'll be back. Michael Hardy (talk) 04:03, 9 March 2010 (UTC)


 * OK, I've moved the article to likelihood ratios in diagnostic testing and deleted the POV tag. Next I'll redirect likelihood ratio to a more appropriate target than this article. Michael Hardy (talk) 04:10, 9 March 2010 (UTC)

LR+ is definitely not (true positives)/(false positives). Consider the worked example where these two numbers are 2 and 18, respectively. The wrong definition implies that LR+ is 2/18 = 0.11 where it in reality is (2/3)/(1-182/200) = 7.4. Consequently I have removed (true positives)/(false positives) and (false negatives)/(true negatives) from the article. —Preceding unsigned comment added by 90.185.26.224 (talk) 09:28, 13 March 2010 (UTC)

This sentence doesn't make sense as written, and could cause unnecessary alarm in readers:
 * A positive result for a test with an LR of 8 adds approximately 40% to the pre-test probability that a patient has a specific diagnosis.[3]

As a counter-example, if someone has a 1% chance of a certain condition, and the test has identifies the condition 80% of the time, and falsely identifies it 10% of the time, that's a likelihood ratio of 8 but it only raises the pre-test probability from 1% to a 7.5% post-test probability. — Preceding unsigned comment added by Tunesmith (talk • contribs) 19:05, 18 November 2012 (UTC)

Excel Instructions
Hi. I removed the instructions for calculating likelihood ratios in excel, which are definitely helpful, but generally not something that is included in wikipedia articles. (See WP:NOTGUIDE.) If you'd like to make information like this available to the world (which I certainly would encourage you to do), one way to do it is to post it on another site like wikiHow or on a personal blog (which is what I usually do) and then create an External links section at the bottom of page with a link to the how-to. Some people might still object to it, but such objections are a lot less frequent. Thanks a lot to the original creator of the information for his/her contribution. mcs (talk) 20:31, 20 December 2010 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Likelihood ratios in diagnostic testing. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://archive.is/20020820071706/http://www.poems.msu.edu/InfoMastery/Diagnosis/likelihood_ratios.htm to http://www.poems.msu.edu/InfoMastery/Diagnosis/likelihood_ratios.htm

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 06:24, 23 December 2017 (UTC)

Wrong interpretation of LR+
In the calculaition section, the phrase:
 * On the other hand, an LR+ < 1 would imply that an individual with a positive test result is more likely to be non-diseased than diseased.

is factualy mistaken. Take for example: Which would give LR+ = 0.83 (<1), but P(D+|T+) = 0.71 > P(D-|T+) = 0.29. I will replace it with:
 * On the other hand, an LR+ < 1 would imply that non-diseased individuals are more likely than diseased individuals to receive positive test results.

(Indeed in the above example, P(T+|D-) = 0.80 > P(T+|D+) = 0.71. Tenthkrige (talk) 11:03, 15 July 2020 (UTC)


 * Good catch. Thank you Tenthkrige! All the best  - Mark D Worthen PsyD   (talk)   (I'm a man—traditional male pronouns are fine.)  02:19, 16 July 2020 (UTC)