Few people would disagree with the need to conduct acute clinical research to develop and improve methods of treating patients admitted to hospital with life-threatening and debilitating conditions. In these circumstances obtaining informed consent can be problematic. Patients may initially be unconscious, in shock or have some cognitive impairment, and if the study intervention has a short therapeutic window there is little time to locate a proxy for consent. It follows, therefore, that for research of this kind to proceed with patients who have impaired decision-making capacity, in some patients it must do so without their consent. This illustrates the key ethical dilemma of allowing acute research in which society as a whole will benefit, whilst maintaining respect for the individuals who contribute to that research1. Given the turn towards greater patient and public involvement in health care decisions 2 following the events of the 20^th century relating to human subject research, research without consent clearly raises concerns about unethical practices 345 and the loss of individual autonomy6 that is "respect for the patient's capacity of self-determination, and exercise of personal choice"7. The subject of considerable debate among researchers, medical ethicists and philosophers for a number of years, the overriding objection to waiving individual informed consent is that it erodes patient autonomy. However, some point out autonomy is not the sole applicable ethical principle and those involved in research have a duty – in respect to the principle of justice – to develop potentially beneficial therapies that are available to all populations89. Others have questioned the equity of excluding patients where individual or proxy consent is not possible and denying them the right to participate and to any potential benefits1011. Opinion regarding informed consent ranges from the view that (in certain circumstances) the process is 'needlessly cruel'12 and violates the principle of beneficence for some patients13, to concerns that it should always be obtained1415 and permitting waiver of consent is a slippery slope toward unethical practices16. At the same time, emergency researchers are frustrated by a somewhat paradoxical situation where physicians are free to prescribe, without ethical approval, a medication for their own patients that has not been tested scientifically with that particular medical condition 17181920.

Research without consent (RWC) is currently permissible in certain situations under specific conditions in the US, EU member states, Canada, and Australasia. In the US it has been argued that these regulations (Food and Drug Administration. 21 CFR 50.24 – Exception from Informed Consent) have led to a reduction in the number of emergency medicine studies implemented 21. Although a published review of journal articles (up to 2004) and FDA information (up to 2003) identified only five studies as being conducted under the US regulations since their implementation in 1996 22, at a Food and Drug Administration hearing held in October 2006 it was reported that 21 studies had been conducted, were underway or about to commence23.

In Europe, RWC in emergency situations was permissible in many countries but in 2004 the European Union Directive 24 on clinical research became law across the (then) 25 EU member states. The Directive which was introduced to 'simplify and harmonise the regulation of clinical trials across the EU' 25 makes no provision for RWC in emergency situations. Although at variance with the Directive, Austria, Belgium, France, The Netherlands, Germany and Spain retained their existing national legislation for emergency research: Denmark, Italy, Ireland, Portugal and the UK attempted to adhere to the rules of the Directive 26. At the end of 2006 the UK issued an amendment to the Clinical Trials Directive 27 to permit RWC in emergency situations with ethical committee approval.

Legislation has been implemented to protect patients and facilitate acute research with adults who are unable to give consent. Public involvement in health, social policy and practice is already well established in the UK 28 and researchers are increasingly expected to involve the public in all stages of the research process. Yet to date the views of patients and the public have not been considered when developing research legislation29. This paper reviews research examining public opinion regarding RWC in emergency situations, and whether the rules and regulations permitting research of this kind are in accordance with the views of those who ultimately may be the most affected.

Seven electronic databases – Medline, Embase, CINAHL, Cochrane Database of Systematic Reviews, Philosopher's Index, Age Info, PsychInfo, Sociological Abstracts and Web of Science – were searched up to the end of May 2007. The search terms are listed in Table 1. Searches were conducted by JL and articles were individually reviewed by both JL and MM. Any queries over the suitability of studies for inclusion were resolved by a third person (HR).

Only those articles pertaining to the views of the public in the US, Canada and EU member states regarding research in an emergency setting involving adults were included. Opinion pieces and those not published in English were excluded. We recognise the important issues relating specifically to obtaining consent from parents to enrol their children in paediatric research but feel they are beyond the scope of this review.

Using the terms relating to emergency research and patient/public opinion (Table 1) we elicited 7298 and 13224 articles respectively in the Medline search. Combining these terms resulted in a total of 28 papers; from the titles 24 were deemed to be unsuitable for the review. The search was exploded to ensure greater coverage and a further seven papers were identified using the terms 'emergency research', 'emergency exception' and 'exception from informed consent'. Two further relevant papers were identified from searching the references lists of key articles. Searches in the other databases listed did not elicit any new papers suitable for this review.

The available data from the US reveals around half of those surveyed do not agree generally with research without consent in emergency situations 31343335. Paradoxically, however, a higher percentage would personally take part in such a study. Unfortunately the studies were not designed to investigate in any depth individuals' views regarding participation in studies where consent was waived. The association between level of risk, or invasiveness of study intervention, and support for research without consent is understandable, but that with study design is not. Patients were more likely to accept an experimental treatment 'outside' of a research protocol without giving consent than in a randomised controlled trial, but again it is not known why this is so. In reality, patients who participate in an organised research study that has undergone scientific and ethical review are more protected than those whose doctor prescribes a treatment unproven with their condition32. However, it is not clear whether the public are aware of the protections in place. One explanation is that the use of randomisation may be problematic for the lay person. Concern about randomisation and a preference for the doctor to choose treatment have been most commonly cited as reasons for refusing to enrol in one study53. It is not clear whether this stemmed from a lack of, or a full understanding of randomisation. In another trial the authors felt that some of their study respondents fully understood but disliked and resisted concepts such as placebo and randomisation, for example54. They believe this may be because the idea that the physician does not know – or knows and does not randomise them to – the 'best' treatment, sits uncomfortably with some people and conflicts with their views of, and trust in, medical care.

Considering the wealth of literature on the perspectives of professionals, ranging from bio-ethicists to physician investigators, there is relatively little information about public attitudes toward research without consent or of their views on how to engender the development of new treatments whilst protecting vulnerable participants. Reasons for a lack of support for research without consent may be due to a general lack of awareness of how few treatment options available there are for some conditions and that many current therapies are poorly supported by evidence or have poor outcomes23, although research in this area is scarce. This is an interesting point, however, considering the backlash against protectionism with some AIDS and cancer sufferers demanding the right to participate in clinical trials of new medications55.

Additionally, though there are very few data from other countries with which to draw comparisons, the American public may be more suspicious and less supportive of medical research. A US focus group study published in 2005 reported that participants were aware of incidences of unethical research but felt these were now "less common" and that "research ethics had improved"35. Participants in the same study also believed that medical staff often used any surplus blood for research without informing the patient or seeking their permission. A focus group study of ED patients and visitors reported that almost half (49%) equated research participants to 'human guinea pigs'56.

There are too few data to evaluate whether the US community consultation requirement protects – or is acceptable to – study participants and there is a need for research in this area. The only concern in relation to research without consent identified from communications with the public was a system to opt out. There is some public support for patients' wishes regarding research participation to be recorded on cards or wristbands32.

Professionals believe that too much emphasis is placed on informed consent, which does not in itself guarantee ethical research575859. Those advocating a move from a reliance on informed consent as a protection for research subjects suggest improved study design to ensure ethical clinical research60. The ethics of clinical research should be judged on: the value of the research; scientific validity; fair subject selection; favourable risk-benefit ratio; independent review; informed consent; and respect for participants 5861. Although still part of the ethical framework, when circumstances preclude obtaining informed consent there should be clinical equipoise which will ensure 'the subject is not worse off by enrolling'58. Clearly these should be standard requirements of every research study but they may not go far enough to satisfy those opposed to, or uncomfortable with, research without consent.

For research without consent to be accepted as a necessity in certain circumstances we suggest there is a need for public involvement in the debate. Any programmes of involvement need to attend to the apparently paradoxical findings of some of the current surveys. For example, if people are more likely to accept experimental treatments outside of a research environment than within one, we clearly need to increase public trust in medical research. This may be bolstered by providing information about the rationale for certain study designs, and highlighting the ethical assessments that clinical trials currently undertake to be approved. We may also need to increase awareness of the standard and efficacy of current therapies and of the necessity to evaluate and develop new ones within the framework of randomised controlled trials. Additionally, a system that enables people to opt out of clinical research might be another measure to reassure the public and protect patient autonomy but this warrants further investigation. However, any attempts to engage the public in such a programme of the 'public understanding of research', should take place in the context of findings from further basic research. We need to understand the factors that promote and inhibit trust in experimental treatments and the potential concerns that patients and the public have around such treatments in randomized controlled trials.

The findings from the survey research raised other interesting issues. Firstly a survey in the UK found that two-thirds of people felt that community consultation could be a practical solution but significantly less of them were willing to attend such meetings. So actually undertaking a community consultation that produces 'democratic' and 'representative' directives for research without consent appears problematic from the outset. Secondly, a survey in the US found that more people were willing to take part in a study when speaking about 'themselves' – when thinking about whether they would personally take part – than when speaking in general terms. When people think about experimental treatments in relation to a potential future moment in their own illness biography they are a lot more accepting of research without consent. We need to be aware of how we frame our research questions and how respondents are making sense of them – are we asking them to offer their views about research without consent in relation to different potential illness episodes involving them or a significant other, or at a more abstract level, about their 'faith' in the idea of research without consent. Additionally, how does this relate to the ideals of experimental research, that people should not be encouraged to take part for personal benefit per se but rather to enable the future benefit of future patients? Finally, no research has focused on the experiences and views of patients who have actually taken part in research without consent. Rather than solely relying on people engaging in hypothetical 'thought experiments', it is essential to generate additional evidence from patients enrolled in such studies to help to inform the continuing ethical dilemma of research without consent.

The question as to whether current rules and regulations permitting research without consent is acceptable to, or protects, the public remains unanswered. We suggest there is a need for accessible and comprehensive public information about medical research and for public involvement in the debate surrounding research without consent. Before this happens, however, we must understand the factors that influence patient and public trust in medical research.

HR is Deputy Director of the UK Stroke Research Network where she is responsible for patient, carer and public involvement. She is a principal investigator for acute stroke trials. GF is Director of the UK Stroke Research Network. He is a steering committee member and grantholder for commercial and academic acute stroke trials and an advisory board member for acute stroke treatments.

MM, LS, HR, GF and SL designed the main study and obtained funding. JL and MM developed the search strategy for the review. JL conducted the review and MM acted as second reviewer. JL wrote the first draft which was revised by all the other members of the study team. All authors have read and approved the final manuscript.