Likelihood Ratio

Understanding the performance of diagnostic tests is fundamental to clinical decision making. The primary measures for assessing diagnostic test performance are sensitivity, specificity, positive and negative predictive value, and likelihood ratios. The likelihood ratio (LR) incorporates sensitivity and specificity into a single parameter and allows one to determine how much a positive or negative test result changes the likelihood that a patient has the disease of interest. Specifically,

The likelihood ratio positive indicates how much more likely a positive test result will be seen in someone with disease, relative to someone without the disease of interest. As a rule of thumb, a likelihood ratio positive above 10 provides strong evidence to rule in a diagnosis in most circumstances. Similarly, the likelihood ratio negative indicates how much ...