Protopathic bias occurs when an exposure is implicated as the cause of a problem but in reality the exposure occurred because of a manifestation of a problem which had not yet been diagnosed.
Example: aspirin is taken for fever, which is the first manifestation of influenza
Features of protopathic bias:
(1) The patient is involved in an observational study.
(2) An outcome has been ascribed to an exposure.
(3) The outcome is actually due to an underlying disease.
(4) The underlying disease had not been diagnosed when the exposure occurred.
(5) A manifestation of the underlying disease was the reason for the exposure.
Differential diagnosis:
(1) indication bias (confounding by indication)
(2) post hoc fallacy
Protopathic bias can be avoided by introducing a lag-time, during which any exposure occurring in a period of time prior to the date of diagnosis is excluded from the exposure assessment. The time period excluded would ideally only be for the time when the problem is subclinical or incubating. The risk is that an adverse effect caused by the exposure would be dismissed.
