Spectrum effect (or spectrum bias) refers to distortions in a diagnostic test's performance (sensitivity, specificity, etc) caused by a distortion in the study population. Testing is not done across a population with the expected distribution of disease severity, but rather limited subsets.


Conditions impacting test performance:

(1) A mild or inactive disease may test negative if the test result reflects disease activity.

(2) The presence of other conditions that can cause a false positive result.

(3) The distribution of disease severities (mild, moderate and severe).


This should be suspected:

(1) when real world performance does not match that reported in the literature

(2) when the study group section describes a skewed selection of subjects

(3) exclusion of patients with conditions that may cause a false positive test

(4) exclusion of patients with mild disease


This may be done accidentally or intentionally. A diagnostic test manufacturer may compare performance of the test in patients with severe disease vs people who are healthy, thereby enhancing the sensitivity and specificity for marketing purposes.


To read more or access our algorithms and calculators, please log in or register.