The most commonly used drug testing methods for detecting marijuana exposure are unreliable when used on infant babies and “should not be reported” without additional testing and followup by hospital staff and social workers, according to new research published recently in the scientific journal Clinical Chemistry.