Limits of confidentiality: a duty of care?
Are you asking questions that could reveal difficulties where you might need to act to ensure that the participant, or someone identified by them, can access help or support? That might include disclosure of abuse or neglect, or it could relate to serious physical or mental health problems, or it could relate to their support needs.
You need to consider – in advance – what could arise and at what level of concern it might be appropriate to act. You also need to make sure that you have clear procedures in place for members of the research team, in case such concerns arise.
If at all possible, individual researchers should not have to make the decision to breach confidentiality on their own – systems should include provision for the researcher to consult with a supervisor, project director, or another experienced researchers (and, if necessary, to make contact outside office hours). See our section on assessing risk and harm for further guidance.
You should check the guidelines of your own professional association to see their guidance in relation to confidentiality. However, in general, the following principles are a useful starting point:
- As part of the consent procedure, the researcher should explain that if (s)he hears or sees something that gives cause for concern, (s)he has a duty to act, but will talk with the participant (adult or child) first about what to do. That might mean that the researcher should first encourage the person to talk to someone who could help, or agree that the researcher should talk to someone else on their behalf.
- In exceptional circumstances – if someone would be put at greater risk by consulting in this way with the participant – it may be necessary for the researcher to breach confidentiality without first talking to the participant.
This means that when seeking consent – and in your consent form and information sheet – you need to explain the limits of confidentiality. For example, you might say:
’We will not tell anyone what you tell us unless we think someone might be hurt. If so, we will talk to you first about the best thing to do.’
In practice, statements such as this have prompted concern about potential damage to the integrity of the research relationship, or about the potential to cause anxiety for research participants. For example, Bostock (2002) gave an account of interviewing mothers about their young children. She described some respondents saying that others might feel that a statement (such as that above) about the limits of confidentiality might increase the pressure to take part – because to refuse would be seen as an admission of guilt, of having something to hide.
Whilst such concerns are valid, you can’t achieve fully informed consent without making clear the limitations of confidentiality in your specific project.