In most international legislation, consent by legal representatives is considered valid and the most preferable and ethically valid alternative for patient informed consent in emergency neurosurgery clinical trials (Alves 2006). Proxy consent is the substituted judgment by a close relative or legal representative about inclusion in a clinical trial. Theoretically, the proxy is supposed to act as the patient, if competent, would have decided. Proxy is essentially synonymous with substitute decision-maker and surrogate.

It is assumed that family members are best suited to make decisions that maximize the patient’s best interest as proxies are supposed to know what is in the patient’s best interest. Theoretically, this is an attractive alternative for patient informed consent, but practically there are some drawbacks. First, would the patient want to be represented by the particular relative? In one study only 41 % of 1,089 patients would want their spouse to be their representative, whereas 28 % wanted to be represented by the physician in charge for inclusion into the clinical trial (Roupie et al. 2000). A further question to ask is if legal representatives know the wishes of the now incapacitated patient. A particular proxy may know little about a patient’s values and wishes for medical treatment in case of severe neurological conditions.

Agreement about health decisions between patients and their proxies varied between 57 and 81 % in one study (Sulmasy et al. 1994). Others found a false-positive consent rate up to 20 % in surrogate decision-making for critical care research (Coppolino and Ackerson 2001). Another found that no less than 45 % of the proxies refused consent in a study involving incapacitated patients (Mason et al. 2006). In the same study, in comparison, only 18 % of patients approached declined participation in the trial. Discrepancies between patient consent and proxy consent were found between 32 and 42 % in a study searching for the ability of family members to predict patient’s consent to critical care research (Ciroldi et al. 2007). Significant discrepancy was found between patients and surrogates regarding consent for critical care research (Newman et al. 2012). In a systematic review evaluating the accuracy of surrogate decision-makers, accuracy was found between 61 and 72 % in consent for several critical interventions (Shalowitz et al. 2006). There was less effect of prior discussion of patient’s treatment preferences and values (Ditto et al. 2001; Matheis-Kraft and Roberto 1997).

The process of obtaining proxy consent in an emergency situation contains three phases. First, information about the emergency critical care trial is provided. Second, the investigator or physician in charge asks the proxy for consent. Third, the proxy consents or refuses (Sugarman 2000). Several authors state that the emotional nature of the emergency situation limits the validity of surrogate consent. Given the complexities of informed consent documents, a larger proportion of proxies might fail to comprehend an actual protocol for an emergency trial (Hsieh et al. 2001). Given the time pressure and the emotionally charged situation, comprehension may be less than optimal (Kucia and Horowitz 2000). Patients enter critical care in physiologic crisis, whereas their relatives enter it in a psychological crisis (Woolley 1990). Uncertainty as to whether the patient will survive also has a profound influence on the proxy’s reactions, actions, and strategies (Jansen et al. 2009). A questionnaire soliciting opinions rather than current practice mailed to 148 European Brain Injury Consortium-associated neurotrauma centers revealed that 48 % of the 78 respondents believe that relatives were not able to make a balanced decision under the emotional and stressful emergency conditions (Kompanje et al. 2005).

A solution to overcome some of the problems mentioned above is using deferred proxy consent. Deferred proxy consent, or retrospective proxy consent, is an alternative for informed patient consent and for prior proxy consent in the situation in which it is impossible to obtain consent from the patient, or from his/her representatives. Mostly this is due to extreme time pressure (Kompanje et al. 2007). Furthermore, proxies can be too overwhelmed to make a balanced decision. With deferred proxy consent, inclusion into the emergency research involves randomization at the discretion of the investigator or the physician in charge. After inclusion of the patient into the study and after starting the study procedures, the proxy is informed and subsequent consent for continuation in the study is requested.

Deferred proxy consent has been used in several randomized controlled neuro-emergency trails. The psychological distress may prohibit a valid understanding of the information about the trial, which is necessary for reliable consent by a surrogate. Relatives of a patient in a life-threatening condition can be temporarily incompetent. For this reason it has been proposed to approach relatives with information only when the physician in charge thinks it is ethically and psychologically valid to do so (Jansen et al. 2007, 2009). In order to prevent investigators abusing this ongoing circumstance, a time limit of 72 h after start of the study procedures for seeking consent is suggested (Kompanje et al. 2005).

More ideal than deferred proxy consent, at least judged from a theoretical point of view, is deferred patient consent. Deferred patient consent, or retrospective patient consent, is an alternative for informed patient consent, for proxy consent, and for deferred proxy consent in the situation in which it is impossible to obtain consent from the patient ab initio. Inclusion in the emergency research involves randomization at the discretion of the investigator or the physician in charge. After inclusion of the patient into the study and after starting the study procedures, the patient is informed and subsequent consent for continuation in the study is requested. In theory, this approach is, judged from an ethical point of view, most ideal. However, in practice, it can seldom be achieved. Patients suffering from an acute and life-threatening neurological condition, who are included in an emergency trial, seldom reach a state in which they have full cognitive abilities of full comprehension. It is common that patients with acute neurological conditions face, after the first stabilization of the condition, cognitive impairment with aphasia, anosognosia, drowsiness, or coma that make the consent process impossible (Ciccone 2003). In the Third International Stroke Trial, in patients with dysphasia, 92 % were randomized into the trial with assent of a representative. However, 69 % of the patients with a motor deficit were able to consent themselves (Kane et al. 2005). Patients presenting in a coma seldom reach sufficient cognitive abilities for deferred patient consent within the time of the study procedures. Nevertheless, if possible, deferred patient consent must be sought, even after deferred proxy consent.