Statistical Explanation of the Probability Calculator
[ Return to: Probability Calculator ]
The calculator computes a post-test probability given a pre-test probability by using the likelihood ratio. The likelihood ratio is a ratio of probabilities, the probability that an ELISA test is positive when the patient does not have Lyme disease (i.e. the rate of false positives). If T+ = ELISA test is positive, T- = test is negative, L+ = the patient has Lyme disease, and L- = the patient does not have Lyme disease, then the likelihood ratio for a positive test can be represented as:
And likewise, the likelihood ratio for a negative test is as follows:
The likelihood ratio of a positive ELISA test for patients with non-specific myalgias would be:
which means that a patient with Lyme disease is 5 times as likely to have a positive ELISA than a patient without Lyme disease. This relatively low number is due to the limited sensitivity and specificity of the test.
Here are the likelihood ratios used in the nomogram:
Patients with non-specific myalgias:
Likelihood ratio for positive test = 0.95/0.19 = 5.0
Likelihood ratio for negative test = 0.05/0.81 = 0.06
Patients with possible EM (erythema migrans):
Likelihood ratio for positive test = 0.59/0.07 = 8.4
Likelihood ratio for negative test = 0.49/0.93 = 0.44
Given a pre-test probability of Pr(L), prior odds are calculated as follows:
Prior Odds = Pr(L)/1-Pr(L)
Post-test Odds = Prior Odds × Likelihood Ratio
The post-test probability is calculated as follows:
Post-test Probability = Post-test Odds / (1+Post-test Odds)
Search PIER® - Decision Support
ACP Members Only. Decision support for over 460 clinical topics.
Introducing ACP Clinical Shorts - Expert Education on Your Schedule
 |
|
10-minute videos give you answers to challenging clinical dilemmas seen in practice and are a terrific way to earn CME credit on-the-go. See more. |
ACP JournalWise: Reviews of the World's Top Medical Journals-FREE to ACP Members!
|
ACP JournalWise is mobile optimized with optional email alerts! Get access to reviews from over 120 of the world's top medical journals alerting you to the highest quality, most clinically relevant new articles based on your preferred areas of specialty. ACP Members register your FREE account now! |
