>he refers to the clinical trial equivalent of "survivorship bias”< That’s exactly correct (e.g. <a href='read_msg.asp?message_id=20623841'>#msg-20623841</a>). You may ignore what ocyan posted, which is just disinformation.