A systematic error that occurs when the follow-up of two groups does not begin at strictly comparable times.  This occurs especially when one group has been diagnosed earlier in the natural history of the disease than the other group.