Redelmeier and colleagues1 bring to attention an important issue. The authors1 define their outcome as a crash that results in a visit to an emergency department, identified using International Classification of Diseases codes related to vehicle crashes. However, unmasking bias (also referred to as surveillance bias)2 should be considered because of its potential contribution to the observed effect. Sackett3 explains unmasking (detection signal) bias:
An innocent exposure may become suspect if, rather than causing a disease, it causes a sign or symptom which precipitates a search for the disease.
Given the potential adverse outcomes of trauma during pregnancy,4 pregnant women may be more likely than nonpregnant women to be transported to hospital or admitted to hospital following a collision, regardless of their injury severity.5 The authors’1 finding that pregnant women were less likely to be admitted to hospital than at baseline may suggest that women are visiting the emergency department out of caution for less severe injuries that may or may not require admission. In this study, pregnant women were more likely than at baseline to attend the emergency department following motor vehicle collisions even as passengers, further supporting this notion.