The results indicate that consent for long-term storage of excess blood samples for possible use in future studies was neither rare nor universal among these urban Aboriginal and Torres Strait Islander research participants. Consent for blood storage was strongly related to consent for other study components, although it was the component for which consent was least likely. It was associated with some but not all of the socio-demographic and cultural factors examined. Consent to store blood was not associated with substantial differences in previous blood test results, which suggests that any bias resulting from future use of stored samples is unlikely to be large, at least for analyses of similar substances.
A few studies have examined people's willingness to allow storage and use of their biological samples, with higher proportions among research participants and patients than among the general public. In one study of the general public, 43 percent of respondents indicated they would be willing both to donate blood for research "to find genes that affect people's health" and to have their blood stored for later use [:20]. The impact on the results of an explicit reference to genetic research is unclear. In a study examining actual consent among research participants, 87 percent agreed to allow their samples to be used for future research. Similarly high proportions were seen for patients, family members and healthy volunteers, although a lower proportion (75 percent) of African American participants consented . Other studies of research participants and clinical patients have reported similarly high levels of consent – much higher than in the present study – for future use of stored samples [15, 16].
In the present study, cultural factors appeared to play a role in participants' decision-making. Although the exact nature of that role remains highly speculative, differing levels of trust – whether of institutions, processes, social groups and/or individuals – may be a common factor. For example, the greater likelihood of consent among those whose witness was Indigenous may reflect increased trust due to shared membership of a salient social group. Higher levels of consent among participants with non-Indigenous grandparents may reflect greater familiarity with mainstream health services, and a possible reduction in distrust of health-related institutions, including research organisations. Shavers and colleagues have suggested that lower research participation by African Americans is related to their lower levels of trust of medical research . It is important to understand that "trust" is not a binary variable; it is possible to have varying degrees of trust, and to have trust in some things (or people) but not others. Trust in one dimension does not necessarily guarantee trust in other areas. In the present study, there were a number of interactions and processes in which the level of trust could be relevant, including: trust in the staff member administering the consent form; trust in the study staff collecting information, taking blood and/or taking clinical measurements; trust in the researchers named in the participant information materials; trust in the various institutions involved in the study; trust that blood samples and other participant information would be handled appropriately and respectfully; trust that participants' decisions about what should be done with excess samples would be carried out; and trust that stored samples would not be accessed in a way that might bring harm to participants. The results of this study suggest that there were varying degrees of trust relating to the storage and future use of excess blood samples among participants, with 45 percent choosing to have excess samples destroyed, 32 percent allowing storage but requiring permission for further use, and 23 percent agreeing to storage and future use of samples provided their use was approved by the study's Indigenous Steering Group. It is notable that, even among this last group, there is a requirement for community input and agreement. That is, participants were not asked to trust the researchers to decide unilaterally what could be done with excess samples, although they obviously needed to have sufficient trust in the integrity of the researchers to expect that they would follow the processes outlined in the consent form.
Participants who identified as Torres Strait Islanders were more likely to agree to long term blood storage than those who identified as Aboriginal, although the 95 percent confidence interval was relatively wide. Whether this reflects cultural differences is not known. It is important to note that Aboriginal Australians are an extremely heterogeneous group with respect to culture, language, and tribal affiliation, as well as social and economic circumstances. Unfortunately, it was not possible in this study to stratify the "Aboriginal" group into more meaningful sub-groups, in part because almost 4 in 10 participants indicated that they did not identify with a clan, tribal or language group.
Smokers were less likely to agree to storage than non-smokers. Indigenous Australians are about twice as likely to smoke as non-Indigenous Australians, and smoking is a major contributor to the ill health of Indigenous Australians . Many Indigenous Australian groups used native tobacco and similar plants prior to contact with Europeans, and tobacco continues to play a ceremonial role in some areas, especially in Northern Australia [18, 19]. It has been suggested that tobacco's role in ensuring social cohesion may be greater among Indigenous Australians than among other Australians , and that tobacco is now "embedded in the sociability and exchange of everyday life for thousands of Aborigines and Torres Strait Islanders" [:120]. Thus smoking may represent a tension between maintaining a healthy lifestyle and enhancing one's social and cultural connectedness.
Blood has been described as carrying "a heavy cultural freight" [:3011]; it is "inherently powerful" and "a topic of great cultural sensitivity" among at least some Indigenous Australian groups [:25–26]. Although it is certain that some people did not participate in the study because they did not wish to have blood taken, it is not clear to what extent any blood-related non-participation was due to cultural considerations rather than other factors such as an aversion to needles or "bad veins".
Despite such potential sensitivities, the collection of blood was considered necessary to meet the main aims of the DRUID Study. By contrast, the long-term storage of blood was not essential, and the decision to proceed was a considered one. There are widespread perceptions that Indigenous Australians as a group have been "over-researched", but relatively little is known about the health and wellbeing of Indigenous people living in urban areas. The DRUID Study was, to our knowledge, the first large study of urban Indigenous adults ever undertaken in Australia, and the potential to answer future research questions without imposing additional burdens on members of the population was seen as attractive, albeit not without ethical challenges.
Decisions about requesting or giving consent for storage and later use of tissue samples must recognize the potential tension between two central ethical principles, namely justice and respect . Justice includes maximising the usefulness of an individual's participation and sharing the research burden equitably. Respect includes supporting participants' ability to determine what happens with information collected from or about them. Maximising the scientific return for an individual's participation helps maximise community research benefits, but collecting biological materials may compromise participation, thereby potentially introducing bias and ultimately limiting a study's scientific value. In this study, we tried to take both justice and respect for autonomy into account by allowing retention of excess samples while giving participants a choice regarding their own samples. No information was available on how participants made their decision or on the criteria they used to determine whether or not to provide consent; further investigation in this area is warranted.
The ethical principle of respect for autonomy was clearly an important consideration in this study. Autonomy, defined as "the ability to make informed choices about what should be done and how to go about doing it" [:53], is considered by Doyal and Gough to be one of two basic universal needs that must be met to enable individuals to "flourish" as human beings . The other basic need is survival/physical health. Importantly, physical health and autonomy are considered to have equal priority and to apply across cultures. To the extent that health research improves physical health, participation in research could be viewed as being morally desirable. However, according to this framework, any such action could not be undertaken in a way that reduces autonomy without compromising the ability to meet an individual's basic needs.
A tension between justice and respect for autonomy is not necessarily inevitable. In the Australian National Health and Medical Research Council (NHMRC) guidelines on ethical conduct in Indigenous health research, these two principles are framed as complementary rather than competing. The NHMRC approach is based not on compliance with prescriptive rules, but on six core values: spirit and integrity; reciprocity; respect; equality; survival and protection; and responsibility . These values are intended to guide researchers in establishing ethical relationships of trust with Indigenous communities. By facilitating the creation of meaningful opportunities for Aboriginal and Torres Strait Islander people to contribute to all facets of research, from identification of research priorities to development and implementation of research processes to analysis, interpretation and translation of data, this type of approach has the potential to maximise community benefit through both higher participation and higher scientific return. Such an approach shifts the focus away from the often paternalistic protection of "vulnerable" individuals towards a model of community partnership and inclusion [24, 25]. In theory, cooperation and collaboration between communities and researchers can increase both the representation of "vulnerable" people in research and their access to the benefits of that research. In practice, most research involves unequal power relationships ; although this can make the development of true partnerships challenging, it is critical to find mutually agreeable ways to overcome such obstacles.
The present study was undertaken in a context of ongoing suspicion of the research enterprise on the part of many Indigenous Australians and their community leaders. Despite a shift over the past decade to greater Indigenous control of and involvement in research (as exemplified by the creation of the Indigenous-led Cooperative Research Centre for Aboriginal Health  and the NHMRC's Aboriginal and Torres Strait Islander Health Forum ), there continues to be a perception that Indigenous people are subjected to research that does not address their needs but which provides researchers with substantial benefits, such as prestige, fame, employment, higher degrees, etc. In such a context, researchers need to prove themselves worthy of trust over a sustained period if they are to build respectful and productive partnerships with communities. However, the requirements of this long-term process may compete with the research team's ability to meet the scientific aims of a particular project in the shorter term. In the present study, for example, we needed to balance autonomy against the risk of bias. Although we believed that it was "right" to have participants decide what would happen to their excess blood samples, we could not be sure whether such a process would result in a sufficient number of stored samples to be useful in the future, nor did we know whether consent for storage would be related to blood biochemistry. Although it appears that we have been reasonably fortunate in both respects, whether this is an indicator of the "correctness" of the original decision depends on one's underlying ethical approach.
As in other endeavours, what is "right" in research may be determined on the basis of a variety of factors, such as duties or rules, outcomes, the character of the people involved, or on ethical principles, such as justice, respect for autonomy, beneficence and non-maleficence . Sponsors, researchers, community leaders and participants may operate using different ethical frameworks, and this may be especially relevant in a cross-cultural setting (although it could be argued that all research is cross-cultural to a greater or lesser degree). Even within research teams, more than one approach may be used concurrently, especially when multiple disciplines are involved, such as in the present study. The resolution of conflict can be difficult is such situations, even when the points of contention are clearly articulated, which is not always the case. When values are incommensurable – that is, when there is no common standard by which to evaluate options – it is not possible to make choices in a rational way , and the resulting compromises may be less than satisfactory to those involved.
The NHMRC guidelines on ethical conduct in Indigenous health research do not specifically address the issue of tissue storage and use, but they clearly indicate the responsibility of researchers to "do no harm to Aboriginal and Torres Strait Islander individuals or communities" [:16]. This language reflects another important ethical consideration: the potential for harm – including stigma and discrimination – to non-participants and to social groups [7, 8, 31]. As part of working with the community to anticipate and address potential harms , any future use of stored samples must be considered and approved by the DRUID Study's Indigenous Steering Group, as well as relevant ethics committee(s), regardless of individual participants' prior consent.
The potential for harm is only one side of the ethical equation, however; it must be assessed against the potential for benefit, both to individuals and communities. As is the case with harms, however, different stakeholders, such as sponsors, researchers, community leaders and individual participants, may have different views about whether something is actually a benefit, and about the magnitude and relevance of any such benefit.
There has been considerable discussion in recent years about how to prevent exploitation in research, particularly in relation to clinical research undertaken in developing countries by researchers from developed countries. For example, a Fair Benefits Framework has been developed by researchers and ethicists as a means of assessing whether the distribution of benefits is "fair" in relation to the level of burdens borne . Considerations about fairness and exploitation are relevant not only in the developing world but also for research in marginalised communities in developed countries, such as Indigenous peoples. The World Health Organisation has developed a guide to preparing research agreements between Indigenous peoples and research institutions; the aim of such agreements is to ensure transparency, an appropriate balancing of interests, and a shared understanding of a range of important issues, such rights, responsibilities and expectations . Although we did not have a formal agreement in the present study, we tried to address what we – the researchers and the Indigenous Steering Group – considered to be the key issues of concern. For example, the Steering Group was involved in the development of study protocols and materials, in the recruitment and selection of staff, and in the development and implementation of the recruitment strategy; we developed agreed terms of reference for the Steering Group and the Chief Investigators Group and had cross-membership on the two groups to ensure communication and transparency; and we attempted to provide benefits at several levels, such as the provision of individual results and targeted health information for all participants, employment and training opportunities for local Indigenous people, and the provision of a part-time diabetes educator for the local Aboriginal community-controlled health service. Whether such putative benefits were sufficient to prevent exploitation in the context of this study remains an open question.
One of the challenges for researchers working with Indigenous communities is a commonly experienced lack of clarity about what the relevant "community" is, and who or what represents it. Indigenous communities are not homogeneous entities that speak with one voice. To the contrary, there are likely to be multiple perspectives and competing interests within a given community, and it can be difficult for researchers and research institutions to determine the appropriate body with which to negotiate. In Australia, as elsewhere, there are a large number of Indigenous community controlled organisations, such as Aboriginal Community Controlled Health Services, which have valuable experience in representing the often diverse and conflicting interests of the people they represent (although their representative status is not always uncontested) and in interacting with outside agents. As a result, such organisations are well-placed to be able to broker agreements between researchers and communities. However, because these organisations are often already under-resourced, it may be necessary for research institutions to provide practical support to enable them to undertake an active role in research brokerage .