Likelihood Ratios

Likelihood Ratio – Diagnostic Testing

Likelihood Ratios are used to assess two things; 1) the potential usefulness of a particular diagnostic test, and 2) how likely it is that a patient has a condition/disease.

Clinicians need to know which tests will best help or rule – in or rule – out conditions when performing diagnostic tests. The decision to perform a test should be based on the initial assessment of the likelihood of the condition, and how important it is to rule-in or rule-out the condition.

The likelihood of the condition/ disease (“pre-test probability”) need to be assessed. Likelihood ratios tell us how much we should shift our suspicion for a particular test result.

Positive likelihood ratio (LR+)

Positive likelihood ratio is the ratio of positive test results in patients with the pathology (true positives) compared to positive test results in patients without the pathology (false positives). LR+ corresponds to the concept of “ruling-in disease.”

Negative likelihood ratio (LR-)

Negative likelihood ratio is the ratio of negative test results in patients with the pathology (false negatives) compared to negative test results in patients without the pathology (true negatives). LR-corresponds to “ruling-out disease.”

Likelihood Ratio Interpretation

• LR+ and LR- don’t change as the underlying probability of disease changes (predictive values do change)
• LR’s using multiple “levels” of positive (i.e. not just a simple positive/negative result) which provide much better, more useful information to you as a clinician and help take into consideration the subtleties of interpreting many physical exam procedures
• LR+ > 1 indicates an increased probability that disorder is present (bigger number is better)
• LR- < 1 indicates a decreased probability that disorder is present (smaller number is better)

Likelihood Ratio Table

• The table below contains general guidelines on the meaning of the values of Likelihood Ratios, which must be correlated with each individual clinical scenario