Does research into sensitive areas do harm? Experiences of research participation after a child's diagnosis with Ewing's sarcoma

Surveys in which people are asked to give information about their health, preferences and behaviours are fundamental to some areas of health research, such as public health. Such research raises ethical considerations that are not always as clear-cut as those of biomedical research or clinical trials, where adverse effects may be more obvious and easier to quantify. In survey research, the possible risks and benefits are essentially of a psychological nature and therefore less tangible or observable. Interview situations can be threatening, particularly when there is uncertainty about what the questions will involve, anxiety about reawakening painful memories or disclosing sensitive information, or concern about the credibility of the interviewer.1,2 However, a number of researchers have suggested that some people participate in survey research for reasons of altruism1,2 and that participation can be of personal benefit to the participants.3,4

Researchers, ethics committees and research "gatekeepers" frequently judge the likely impact on participants of a particular study on the basis of intuition, past experience (professional and personal) and cultural expectations. Such decisions are seldom easy and there may be diversity of opinion — if this were not the case, there would be no need for ethical review.5

Our article describes the experience of participating in a research interview from the perspective of families whose child has had Ewing's sarcoma. Unlike many other studies, which have asked people's views on participation at the time of the survey interview, our study provides a longer-term perspective, allowing people to reflect on experiences some time after their actual study participation.

Some demographic characteristics of participants and non-participants in our study are shown in Box 2. Of 97 questionnaires sent out, 81 were returned (83.5%), eight families did not return the questionnaire despite a telephone prompt, and eight could not be contacted. Respondents and non-respondents did not differ significantly in terms of socioeconomic status, maternal education, patients' sex or patients' age at diagnosis. However, families of children who had died were significantly more likely to participate in our study than those whose child was living (P < 0.001). Almost all participants (n = 79 [97.5%]) had entered the previous study with the belief their participation would be beneficial to others (Box 3). Most were pleased to be involved in the current study (n = 76 [93.8%]), despite almost half anticipating the interview would be painful (n = 38 [46.9%]).

Parents of a child who had died before the interview were significantly less likely to expect that answering questions about their child's life before diagnosis would be painful than those whose child was alive at interview (1 of 47 [two responses missing] compared with 17 of 32; P = 0.02). Some responses to open-ended questions provided insight into this finding: "The interview was painful, but if it saves one child from suffering as my daughter had to her death will not have been in vain."; "We particularly liked the interview, for the chance to sit and talk face to face with a wonderful lady who was professional, compassionate, understanding and answered all our questions honestly."

Most participants had felt in the previous study that they could refuse to answer questions that made them uncomfortable. Nine (11.1%) agreed some questions had made them feel uncomfortable; seven (8.6%), including five whose child was alive at the previous study interview, had found the interview in general, rather than any particular question, was disquieting.

Most participants believed their participation in the previous study benefited others; two-thirds felt participation was personally beneficial because they could talk about their child's illness. Some (n = 6 [7.4%]) had found the interview more painful than expected. Participants who had been interviewed one year or less before our study (n = 15 [40.5%]) reported finding the interview more painful than those who had been interviewed more than a year before our study (n = 7 [16.7%]) (P = 0.02). No participant disagreed with the statement, "I am glad to have participated in the interview".

Almost half the participants (n = 35 [43.2%]) believed the interview had produced some good from an otherwise bad situation, although a substantial number (n = 22 [27.2%]) felt the study had not affected them. Others (n = 9 [11.1%]) said the interview had encouraged discussion or that they had found it thought-provoking (n = 3 [3.7%]). No family felt that participation in the study had "upset them".

The results of a small number of studies addressing related issues of grief and loss are generally consistent with our findings.3,4,8-12 Neugebauer et al,3 investigating depressive symptoms in women who had miscarried, discovered that women who had completed interviews at two weeks, six weeks and six months after miscarriage had lower depression levels than those who had completed an interview only at six weeks and six months. This was attributed to "unintentional therapeutic and test effects of study interviews".3 Kitson et al,8 investigating the effects of violent deaths on families, noted that the majority of participants in their study felt it was useful to talk to someone, even though participation was distressing. In research on how terminally ill people changed their lives once they discovered they were dying, Kellehear4 wrestled with the fact his questions seemed to leave participants in tears. He was concerned that this may be harmful; however, many said being able to review their lives was beneficial. Similar sentiments were found in a study examining next-of-kin attitudes of people participating in a case–control study of adult leukaemia in Seattle, USA.9

Our results support these findings, despite some participants expecting and finding the original interview to be painful. Almost all were glad to have participated, as they felt that the research gave them the opportunity to discuss their child's illness and that their participation would benefit others.

The findings of our study provide useful information for members of ethics review boards who consider whether a study should be approved, and for other "gatekeepers" who decide whether researchers should be granted access to specific populations (ie, patients and families). It is also useful for researchers designing and conducting studies into sensitive and painful areas, who must consider the needs and expectations of those who participate in the research. It is important to note that interviewers need to be skilled and properly trained, as results may be affected by the skills of interviewers. Finally, our study provides potentially useful information for members of the community who may be invited at some stage to participate in research.

The risks and benefits of participation in survey research, particularly when it deals with sensitive or painful topics, are real but often intangible. Many people accept these risks and participate in research they feel is worthwhile and beneficial to others1 and these results have been echoed in other studies.3,4,8-12 Overall, people are pleased to participate, despite anticipating possible distress. Professionalism, interpersonal skills, compassion, and awareness of the potential for negative experiences from participants are essential components in ensuring that participation in research "does no harm". Provided these and other established ethical guidelines are met, there is a strong case for giving people the opportunity to decide whether the benefits outweigh the risks in their own particular situation.