[Biomedical research, the market, clinicians, safety and corporate social responsibility post-phase III: maintaining...

Randomised controlled trials (RCTs) are the gold standard in the western world for decision making, as much for the clinicians as for the agencies or managers for community policies. In this powerful deployment of investigative effort there are variable degrees of conflict of interests, and the clinicians, not foreign to this, are entering a sea of doubts on safety, a dimension that emerges like a diacritical, inalienable element. The aim of the study was to select and ctically evaluate editorials, clinical trials and/or meta-analyses published on physical support, or the more important internationally credited websites, which has patient safety as their primary objective.

Evaluative study performed in the Torrecardenas Hospital, of the Andalucian Public Health Service (SSPA), in the context of an accredited training unit (Mejora_F), so called "transference of the biomedical knowledge from the bibliography" from 2008 to 2009.

Analyzed articles: 170 (75 % in English, with predominance of N Eng J Med >50 %; followed by Journal Of the American Medical Association, The Lancet, British Medical Journal, Journal of American Geriatrics; Med Clin, Rev Clin Esp.; and Infectología). From 170 eligible articles we selected 5 key topoi due to their impact as the most representative owing to their citation frequency. They represented, in our judgment, "a red flag" of safety: long-acting beta-stimulators or LABS and increase in mortality in asthma; neuroleptics in the elderly and extension of the QT interval associated with sudden death; thiazolidinediones in type II diabetes and negative cardiovascular effects; promotion of statins-ezetimibe and the debatable association with major mortality for cancer, and intensive treatment in diabetes and probable increase in mortality.

What really maters in biomedicine is that it leads to a given strategy in real patients, not the intermediate points. Clinicians should not support partial results of designs based on intermediate information - undoubtedly less costly and more rapid - without adequately knowing the safety of the strategy built to reach them. Only in this way we can be real guarantors of safety, and only in this way, and in absence of conflicts of interests we will be able to support the given confidence.