Pathology review in cancer research.

Pathology observer agreement studies are clearly important in the development of new pathology assessments and in the quality control of those assessments in common use. Setting up such studies, and reporting and interpreting their results requires careful thought and statistical expertise. Investigators are advised to seek collaboration with a statistician before embarking on these studies. Pathology reference panel reviews in multicentre studies are useful for checking eligibility when there is a high level of disagreement on the eligibility criterion between local pathologists and the reference panel members, but good agreement between members of the panel. However, such situations are uncommon.