[Research and ethics in emergency medicine. Findings of a workshop].

Prerequisites for experimental study designs are extremely difficult to realise under prehospital emergency conditions. Results obtained in animal experiments always need validation with prehospital or in hospital patient studies. Investigations related to emergency medicine are, however, an ethical obligation on behalf of the patient. A retrospective analysis of the available literature should be considered a prerequisite for prospective randomised and controlled studies. Frequently, a pilot study or feasibility trial needs to precede the actual study. Informed consent must be obtained for all patient studies. However, under emergency conditions informed consent cannot always be obtained due to unconsciousness, etc. Nevertheless, the following principles should be observed: (a) randomised studies are essential, even in emergency medicine; (b) they are ethically acceptable if the treatment provided for the study group is at least equivalent to the therapy for the control group; (c) only these preconditions guarantee that the patient always receives treatment in accordance with the standard of treatment. If the patient is unconscious or otherwise unable to give informed consent, the principle of deferred consent is acceptable. The deferred consent principle should be carefully documented. A prospective randomized study represents the gold standard for an investigation, even under emergency conditions. There are different principles of randomization: telephone or telefax randomisation, etc. Emergency medicine investigations need to be completed within a reasonable time frame which should, as a rule, not exceed 2 years. Otherwise, too many items might change without being noted. Curative treatment and the intention to treat are terms employed by ethics committees. If an investigation does not bear in itself additional risk factors, it does not necessarily have to be brought before the ethics committee. If, however, for example, the effect of obtaining blood samples is investigated separately for research purposes, the patient needs to be informed and sign the respective agreement. In Germany the ethics committees have agreed to accept their respective decisions. For multicentre and multinational studies, however, the ethics committees also need to accept decisions made by non-German ethics committees. Not all therapeutic principles can serve as the basis of prospective randomized studies because they need to be considered the present standard of care (e.g., endotracheal intubation, etc.). Nevertheless, the feasibility of a study should be assessed according to the 1-3 rating proposed by the American Heart Association. In principle, the overall emergency physician protocol does not fulfill scientific requirements. For all studies, particularly in the prehospital setting, an independent observer should be involved for documentation purposes. Scores can be of importance for investigating different treatment regimens. Biases in emergency medicine studies focus on the variability of EMS, personnel qualifications, etc. This is in part why it is extremely difficult to prove efficacy and efficiency in emergency medicine. The results of emergency medicine investigations should be published in suitable journals, i.e., journals with a reasonable rating.