Initially identified in 1978, spectrum bias refers to the phenomenon that the performance of a diagnostic test may change between different clinical settings owing to changes in the patient case-mix thereby affecting the transferability of study results in clinical practice. In the statistical sense it is not truly a bias and has led some authors to refer to the effect(s) as 'spectrum effects', whilst others maintain it is bias if the true performance of the test differs from that which is 'expected'.
详细百科解释