To the Editor: The recent article by Scott and colleagues1 described a retrospective analysis by postal questionnaire of the attitudes of family members to participation (about a year earlier) in a face-to-face interview about their child's diagnosis of Ewing's sarcoma. This was accompanied by an editorial exposing the complexities of research participation, including the potential risks of interviews as well as the role of altruism.2

Although results derived from the questionnaire have only recently been published (November 2002), the questionnaire was distributed in November 1997, before introduction of the National statement on the ethical conduct of research involving humans.3 Some ethical uncertainties and questions of historical interest arise. First, what was the nature of the original consent obtained for the initial interviews? Presumably it involved written informed consent in which the risks of participation were clearly mentioned, including the possibility of distress associated with the interview. Did it mention a procedure for aborting or complaining about the interview? Second, did the patients (aged up to about 35 years) and their families give permission to be contacted again by the same research group? Third, for the follow-up on research participation, was consent implied simply by return of the questionnaire? Finally, how would the conduct of the initial interviews and the follow-up questionnaire differ in the light of recent developments in the ethics of research involving humans?

Importantly, the respondents (84% of those surveyed) indicated that participation in the original study had not "upset them".1 While the attitude of non-responders is unknown, this would seem to confirm that the original process had been sensitive and appropriate. This is supported by the finding that families whose child had died after the initial interview were more likely to respond to the questionnaire.

As a long-time member of a university human research ethics committee, I have often been required to evaluate research protocols that involve potentially threatening or distressing interviews. This has occurred more frequently since introduction of the National statement, as much qualitative research previously conducted under different jurisdictions (such as quality control or clinical audit) has been submitted for formal ethical review. The risk of harm to participants in qualitative research cannot be trivialised. Its impact can be minimised by wording the consent form to warn of possible adverse psychological reactions to interviews and questionnaires, using trained interviewers and providing counselling support if needed.