Consent for data sharing in sensitive qualitative research: burden or benefit?
About the research
These two studies from Jane Seymour demonstrate some of the issues raised when considering consent for data sharing in sensitive qualitative research. This researcher’s experiences, a decade apart, of archiving data and documentation from projects focusing on end of life illustrate multiple perspectives on a perennial question in qualitative research with vulnerable participants: whether key processes involved are burdensome, even distressing, or can be potentially empowering. The discussion shows how different research ethics committees can take different approaches to consent for data sharing and that professional participants may experience themselves as more “vulnerable” than lay participants. Both studies show how participants can appreciate the opportunity to share their experiences with others, beyond the original researchers, in the context of trusting the researcher and the significance of the research topic. The potential for clinical researchers to revert to their clinical role and seek to “protect patients” from the “burden” of thinking through consent for data sharing in research is also highlighted by Jane Seymour.
The first study (SN 4840), funded by the Economic and Social Research Council (ESRC), was part of the Innovative Health Technologies Programme and focused on how older people understand technologies used in end of life care. It combined perspectives from gerontology and palliative care. Over seventy people from three diverse areas of Sheffield were selected and divided into three cohorts according to age: 65 – 74, 75 – 84, 85 years and over. The ethical demands of this topic required the development of new methodologies that reflected awareness of its sensitive nature and the advisory group for the study also included participants.
The UNBIASED (UK Netherlands Belgium International Study) project (SN 850749) looked at deep sedation until death. It was developed as part of the European Association of Palliative Care Research Network and the UK component was funded by the ESRC. Palliative or continuous deep sedation is offered at the very end of life for those suffering intractable pain and distress that has not been addressed by other means. This qualitative study aimed to explore how decisions are made in applying continuous sedation until death from the experiences of both clinical staff and the informal care-givers of the deceased. This addressed a widespread and inexplicable variation in continuous sedation shown in previous surveys and allowed the ethical complexity surrounding this practice to be more clearly understood. In the first phase, focus groups were held with clinicians and bereaved informal care-givers, along with a preliminary review of clinical case notes. Phase two involved qualitative interviews with staff and informal care-givers of cancer patients who had received continuous sedation until death.
About the data
Data and documentation from the technology and end of life care study (SN 4840) include transcripts and field notes from qualitative interviews and focus groups, as well as consent and information sheets. Field notes recorded discussions with groups of potential participants in front of gatekeepers and how they were used as opportunities to model how safe and respectful conversations could be opened up on sensitive topics. The data reveal the consideration given to how technologies should be offered in ways that respect values and dignity, and that care at the end of life could involve obligation and burden, comfort and love.
For the later, multi-site project on deep sedation data from the UK component consisted of semi structured interviews with doctors, nurses and informal care givers. Interviews focused on recollections of how the deceased in question was cared for, why and how sedation was used and the processes of decision making. They were conducted in hospitals, hospices and the community, ultimately informing 22 case studies. In the hospital context, only 2 nurse interviews were deposited with the UKDA. In the hospice setting 8 interviews with doctors, 8 interviews with nurses and 8 with relatives are deposited. In the community 4 interviews with doctors, 3 with nurses and 8 relatives are deposited. This reflects the problem of gaining retrospective consent for sharing data, as discussed below.
The key archiving challenges for these studies were the differing perceptions of consent for data sharing. The range of formations of NHS research ethics processes in the two studies gave two different responses to the timing of seeking consent for data sharing. The earlier study was allowed to request consent for data sharing before the interviews and focus groups, whereas the later study was required to seek retrospective consent because it was felt that to request consent for both participating in the research and data sharing before the field work would be confusing and burdensome for participants.
In the technology and end of life care study, the research team’s understanding of research ethics and data sharing was formed by the prevailing, pre National Research Ethics Service (NRES), NHS approach that involved multiple ethics committees. This process was generally viewed as burdensome and broadly unsympathetic to qualitative studies. At the same time, Jane Seymour and her colleagues were coming to terms with the 1998 Data Protection Act and were using documents from the UK Data Archive to help them think through questions around data management on the ethics committee forms. The UK Data Archive also provided a workshop to the research team to support them in the process of data archiving and the ESRC offered advice and direction. This allowed the research team to support ethics committees in gaining a better understanding of what data sharing would involve. The result was that the team were able to request consent for data archiving and sharing at the same time as seeking it for participation in the research.
The ESRC funded most of the UK element of the UNBIASED study on deep sedation, but sources funding the Dutch and Belgium work did not stipulate data sharing. Therefore, the material prepared for archiving was from the UK study alone. The NHS NRES was generally supportive and offered advice, although they stipulated retrospective consent for data sharing. They felt that participants would not necessarily understand the implications of their interviews being available to other researchers without knowing the content. A simultaneous dual consent process was also perceived to be burdensome, particularly for the relatives.
In an interview reflecting on her experiences with depositing these studies some years later, Jane Seymour felt that the attitudes of participants to data sharing also appeared to differ in some instances. In the study on technology and end of life care the participants were all retired, elderly people sharing their direct thoughts and experiences of interventions in end of life care. Once consent for data sharing had been discussed with them, the impression was that they were generally enthused by the idea that what they were offering the study would continue to be of use for other researchers. The involvement of some participants in the advisory body may have enhanced this.
The deep sedation study sought the experiences of professionals: nurses and doctors, as well as bereaved relatives. Data from the interview sets with health professionals indicated that nurses felt more exposed ethically in terms of decisions around deep sedation. Jane Seymour speculated that this might also have brought up concerns around data sharing that the doctors, who generally receive more training in ethics and research, did not have. However, the healthcare setting may also have influenced this. Looking at the breakdown of interviews archived, it is interesting to note that there are two with nurses and none with doctors from the hospital setting. In the community setting, interviews with four doctors and three nurses were archived, while interviews with eight nurses and eight doctors in the hospice were made available. This may have been because staff retention in the hospice allowed more participants to be contacted for data sharing consent once the fieldwork had ended, as demanded by the NRES. There were also no interviews uploaded from bereaved relatives from the hospital setting, which again suggests that retrospective consent was harder to achieve here. This can be contrasted with the success in achieving consent for data sharing with relatives in both community and hospice settings: eight in both cases.
In retrospect, Jane Seymour felt that it is best to seek consent for research participation and data sharing at the start of engaging with potential participants to prevent the problem of losing contact with participants after the study. She also identified the danger of clinicians as qualitative researchers feeling ambivalent about their dual roles as clinician and as researcher. She considered that this may result in a reversion to their clinical roles in order to “protect” vulnerable participants from what could be perceived as an additional “burden” of asking for permission to share data. The need to maintain the focus on research and explain objectively what the possible outcomes of data sharing might be counteracts the potential for paternalism and allows participants to feel further empowered by the increased reach of their voices.
She reflected particularly on this when thinking about the openness of the bereaved participants: “I was remarkably struck by their generosity, given the material, and actually - those people had nothing left, apart from their memories [and] the drive to help other people…”.
Reuse publications and outputs
Extracts from interviews as well as the participant information sheet and consent form from (SN850749) have been used in UK Data Service workshops, focusing on the ethics of consent for data sharing and anonymising in sensitive qualitative research areas.
In this brief extract from a longer interview, Jane Seymour discusses the process of anonymising the interview transcripts as well as the value of depositing field notes and other background documentation. She also offers an interesting account of how researchers from the different nationalities involved in the UNBIASED study (SN850749) came to clarify their definition of palliative sedation. Background information such as this helps in demonstrating transparency of process in qualitative research.