Declarations of conflict of interest are still inadequate.

Declaration of conflicts of interest (COI, understood mainly as financial) in medical publications is long established. Most journals refer only to the guidelines of the International Committee of Medical Journal Editors (ICMJE) but not to those of the WAME (World Association of Medical Editors). We surveyed 17 journals and found only one (BJOG), which explicitly mentioned "religious interest" as an example of a possible COI and one other journal included "personal belief" (Journal of Obstetrics and Gynaecology of India) as a COI. Of the other 15 journals, 10 used the ICJME as their COI model. They were the general journals, NEJM, JAMA, Lancet, BMJ and JIM (Journal of Internal Medicine); the pediatric/neonatology journals Pediatrics and Journal of Pediatrics (this also mentions WAME) but not Acta Paediatrica, which mentions COPE; the obstetrics/gynaecology journals AJOG and IJOG; and the British Journal of Haematology but not Blood, which uses the American Society of Hematology's own COI model. Neither EJOG, JOG, IndianObs Gyn, nor J Obstet Gynaecol India clearly specified a COI model. Yet the ICMJE COI guidelines fail to include involvement in religious and/or secular groups which take sides on the subject being discussed, while the WAME guidelines specifically do so. Instead the ICMJE uses the vaguer phrase "intellectual beliefs". The actual ICJME COI-form does not itemise religion. To maintain their scientific credibility, medical journals must start requiring disclosure of such ties. A typical example where current ICMJE rules fall short is the ongoing heated debates over the ethics of prenatology and of physician assisted suicide.