According to a posting on the Journal's Web site, editors at the world's top medical journal found by examining a computer diskette that relevant data about Vioxx's risk to the heart compared with an older pain drug, naproxen, were deleted two days before the paper was initially submitted to the Journal.