Guidelines for conducting ethical research in psychosocial issues in palliative care
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0973-1075.25914
Source of Support: None, Conflict of Interest: None
While it is unethical to introduce services for the terminally ill and their families that are not well founded or evaluated there are special problems in research conducted with this population. This has deterred some from carrying out research in this field and has caused others to place obstacles in the way of would-be researchers. This paper describes the ethical difficulties and provides guidelines that should enable worthwhile research to be carried out without harm to those who offer their help and without vitiating the scientific value of the research.
Keywords: Guidelines, ethics, research, palliative care
Systematic research has lent credibility to several of the interventions in palliative care today. One could go so far as to say that it is unethical to introduce any kind of new service in palliative care without carrying out research to find out if it does good or harm. Even so, this is a sensitive area in which there is danger that researchers may cause distress and may even do harm. Guidelines are needed for researchers involved in this area of scientific pursuit.
All plans for research should be scrutinised by an ethics committee.
Access to patients receiving palliative care and their families
Access can be obtained in a variety of ways ranging from hospital records, hospice records or referral from GPs/other doctors. Guardians of these sources are right to be cautious in checking the credentials of the researcher and the ethical validity of the research but they often err on the side of being over-zealous, depriving terminally ill or bereaved people of the opportunity to help with well-found research and forcing researchers to make use of non-random methods of sampling.
It is impossible to say how many researchers have been put off by obstacles of this kind but the number of well conducted projects in this field is still small. Special difficulties exist in obtaining access to terminally ill children. Parents are understandably protective and it is necessary for the researcher to spend time with them in the first instance before they are likely to trust us to talk to the children.
Individuals who are dying and their families are often emotionally upset; they find it difficult to take in explanations and may be easily persuaded to agree to procedures or interviews which they will subsequently regret. Alternatively they may refuse to help with a project which, in a calmer frame of mind, would present no problems to them. These problems are less likely to arise if respondents are sent a clearly written explanation of the project which they can study at their leisure and discuss with their family. In this the researchers should establish their own bona fides by citing their qualifications for carrying out work of this kind and reassuring the client that they are backed by a responsible and trustworthy institution to whom reference can be made if any doubt remains in the respondent's mind. Ethics committees should always expect to read and endorse this written explanation. Of course, not every respondent will read or understand written explanations and it is essential that this be followed up by a telephone call or explanatory visit at which the explanation can be given by mouth and any questions answered.
Possible harm to respondents
Individuals at the end of life and their families are vulnerable. Strong emotions can impair their judgment, their confidence and their ability to concentrate and appraise risks. Some are so desperate for help that they will accept any person who approaches them in a friendly way; others are deeply hurt and unduly suspicious. The credulous are vulnerable to exploitation and need to be protected from unscrupulous or potentially harmful intervention be it by researchers or therapists, the suspicious may need the reassurance of an impartial and trusted person before they will accept any interview.
A properly constituted ethics committee attached to a reputable organisation can reduce these risks and constitutes a source of reassurance to clients. How are the members of such committees to recognise potentially damaging research? In my view, the following criteria should be satisfied in all cases:
a) The researcher should obtain no personal gain from a particular research result, eg researchers employed by a counselling organisation to evaluate their own counselling must have the right to make public and to publish the results of their evaluation regardless of the outcome of the research. Permission for publication should be written into any contract in advance of the research. While it is not unreasonable for an employing body to have the right to see and to comment on research results before publication they should not have the right to prevent publication if those results do not suit their own commercial interests or prejudices.
Therapists and counsellors are often motivated to study the results of their own therapy, they are well able to collect data and their special knowledge of the field may be considerable. They are, however, faced with serious ethical problems if the results of their research throw doubt on the value of their service and for this reason alone, any project which sets out to test the value of a service should be carried out by a truly independent organisation. Even then it is important that all concerned recognise that any benefits to the researcher, be they financial or meritorious, derive from the quality of the research rather than the results obtained.
b) The participants should not be subjected to any pressure, financial inducement or emotional blackmail to take part or to respond in a particular way and should be informed that they are free to withdraw from the research at any time should they wish. Financial payments should be limited to expenses incurred. It should be made clear to all participants that there are no "correct" answers to any of the questions. This can be difficult when the researcher is identified in the mind of the informant with a particular viewpoint or service. Thus gratitude to a doctor or hospital can make it difficult for informants to criticise any aspect of the service provided, particularly if the research is being conducted by the doctor in question or by other representatives of the institution. Similarly, individuals receiving palliative care who are under the care of a particular counsellor or therapist should not be asked by that person to evaluate the care they are receiving. These considerations add further weight to the need for any evaluation of services to be carried out by individuals and /or organisations that are truly independent and are seen to be independent by the subjects of the research.
c) While researchers can not be responsible for solving all the problems of the people who assist them with their research they are responsible for any problems that arise as a result of the research itself. Interviewers should have received sufficient training in the support of dying individuals and their families to ensure that they will do no harm. If, in the course of an interview, the subject becomes distressed the needs of the respondent should take priority over the needs of the research. It is not always desirable to interrupt a research interview if the subject becomes distressed. Expressing grief can be a very therapeutic experience and the interviewer need not attempt to block or inhibit such spontaneous expressions. Individuals at the end of life expect the discussion of emotionally distressing events to be painful and may resent being told to "calm down". The decision to interrupt or to discontinue the interview will depend upon the response to the support given by the interviewer and the wishes of the participant. This is no place for scientific detachment in the sense of the interviewer being a 'fly on the wall'. If the respondent is in need of help, the interviewer should not hesitate to provide it. Interviewers should also be aware of other sources of help for terminally ill people and their families that are available in their area and should provide this information to any respondents who are thought to need it.
d) All interviews should be closely and regularly supervised by someone with a sound knowledge of palliative care and counselling as well as a clear understanding of the purpose and method of the research and the ethical issues involved. Supervisors need to be particularly aware of the problems that can arise when an interviewer is either emotionally over-involved (in which case they may aggravate the subject's emotional response and bias the accuracy of the data) or too detached (in which case the subject may feel unsupported and may conceal emotionally charged information).
Given these safeguards it is probable that the research will do good rather than harm to respondents.
e) The confidentiality of the sources of all information obtained in the research should be protected a) in all publications or other public communication arising from the research and b) in communicating with supervisors and other colleagues about the research. Thus names, addresses and other information that could enable the subject to be identified should be changed in ways which disguise the identity of the subject without distorting the essential findings of the research. This can be difficult if, for instance, the details of a critical life event are widely known. For example a psychologist who gave a public lecture on the psychological aspects of euthanasia might inadvertently enable a newspaper reporter in the audience to identify one of the cases under discussion. In such cases it may be necessary to report results in general rather than specific terms.
If it is the intention of the researcher to exhibit videotaped interviews confidentiality presents a particular problem. The usual editorial methods of doing this by concealing the face of the respondent may not be appropriate if the research is concerned with emotional expression. In such cases potential subjects need to be fully aware of this implication. They should see any such material before it is released and give full consent. Any limitation on audiences (e.g. that the tape should only be used for training doctors) should be strictly adhered to but it is better not to make such restrictions in the first place as it is very difficult to ensure that they will be met.
Respondents should be informed in writing that confidentiality will be respected in the information that they are given before consent to the research is obtained.
f) The interview should be organised in such a way as to minimise the stress on the respondent and maximise the accuracy of the information obtained. These two objectives seldom conflict because subjects are more likely to give accurate information if they are in a calm frame of mind than when they are upset or unsure of themselves. Thus questionnaires dealing with emotional issues and physical procedures such as taking blood samples, should, where possible, be left to the end of the interview. All interviews should allow time and opportunity for respondents to express their own needs as well as to meet those of the researcher. Verbatim comments often explain and enhance the value of the questionnaire findings and researchers would be wise to record both.
Audio- and videotape recordings are less often seen as intrusive today than they used to be but interviewers should always inform the subject beforehand that they are willing to turn them off if requested. There is no longer any need to take special lighting and camera teams into a person's home but researchers should take care that equipment is working properly and that interviewers know how to use it (There can be few more annoying experiences to a researcher or subject than to find that a lengthy interview has been "lost").
Audio- or videotapes obtained for research purposes are the property of the researcher or their employing organisation. Copies are sometimes given to research subjects as a reward for their help but researchers should be aware that they will not thenceforth be able to control how that material is used. In one case a bereaved patient who had been given a copy of a videotaped abreaction sold this to a television company!
g) While it is not the job of the ethics committee to teach science they should protect the public from sloppy methodology. Research that can only confirm the prejudices of the researcher does more harm than good. Despite the technical difficulties in obtaining adequate samples and objective data it is still important to be as rigorous as the circumstances permit and to include control groups whenever they are appropriate. In one bereavement study, for instance, it was found that there was a large difference between widows and widowers with women reporting far more symptoms than men. It was tempting to conclude that women are more vulnerable to bereavement than men. However, a control group of married women, carefully matched with the widowed sample, also reported more symptoms than the bereaved men. 2-4 years after bereavement the widows had no more symptoms than the married women controls. Whereas the widowers still had more symptoms than control group of married men.
Many of the instruments developed for use in psychiatric settings are inappropriate for use with community samples. Data which measure the frequency of rarely occurring events (such as death rates) require large samples if findings are to reach statistical significance. Conversely block-buster studies (such as the American National Hospice Study) which attempt to gather data from huge samples often lose the fine tuning and sensitivity that is possible when small or moderate-sized samples are interviewed in depth. Small sample methodology will answer a different set of questions to those that can be answered by studying large samples.
Examples of questions that can only be answered by means of large (1000+) samples would be:
'Does awareness of diagnosis hasten death?'
'Can bereavement cause cancer, suicide, psychosis, duodenal ulcers, angina, ulcerative colitis, alopecia, etc.?'
Examples of questions that can best be answered by detailed study of small (150) samples would be:
'How do men differ from women?'
'What are the common effects of awareness of diagnosis on the immune response system?'
Consideration of this kind should be born in mind by those planning research, funding bodies and ethical committees, all of whom, must satisfy themselves that the research will achieve its intended objectives.
Checklist of ethical palliative care research
Ethics committees and others concerned with research using terminally ill or bereaved subjects must satisfy themselves that the following ethical criteria are met:
1. All research proposals will be approved by a properly constituted ethics committee.
2 All people who are invited to take part in research will receive a document which:
• explains the purpose and method of the research to them,
• reassures them of the qualifications and good faith of the researcher,
• informs them that the research has been approved by a particular ethics committee,
• identifies the organisation under whose auspices the research is conducted,
• introduces any person who will be contacting them to arrange an interview,
• explains the precautions that will be taken to ensure the anonymity of the respondent in any publication or reports,
• explains their right of access to any records made about them,
• assures them of their right to withdraw from the research or to ask for recording equipment to be switched off at any time,
• reassures them that they will not be penalised in any way if they should choose to withdraw and
• gives them the name, address and telephone number of the Project Officer, inviting them to contact him or her if they have any questions, comments or criticisms.
• to ensure that the document has been read and understood the respondent will be invited to sign one copy indicating this and be given a second copy in return.
3. If a child or mentally impaired person is the object of the research similar information will be given to parents or guardians who must give their consent. In most cases a face-to-face interview with the parent or guardian should be carried out before any attempt is made to contact the child. This given, the researcher will also explain the research to the subjects themselves in language appropriate to their age, educational background and state of mind and must take seriously their views regarding participation.
4. No financial inducement or other pressure will be put on potential respondents.
5. Those who are to interview people for the purposes of research have sufficient prior training in counselling to ensure that they will do no harm. They will receive regular supervision of their work from someone with advanced level training and experience of counselling the dying or bereaved.
6. When a respondent becomes distressed interviewers will be guided by the respondent and by their own understanding in deciding whether or not to interrupt the interview. If tape or video recorders are used the respondent will be reminded that they will be turned off on request.
7. Researchers will obtain no financial gain from a particular result to the research and are independent of any commercial pressures. Evaluation of the quality or effectiveness of care will not be carried out by the people providing the care.
8. Apart from giving emotional support interviewers will confine themselves to the research and will not proselytize, advertise or advocate particular treatments. On the other hand, when help is likely to be needed they should be able to point the bereaved person to possible sources of impartial assessment and advice.
9. If videotapes of respondents are to be exhibited particular care will be taken by the researcher(s) to ensure that respondents are aware that confidentiality cannot be preserved and to satisfy themselves that consent to show the tape has been freely given.
10. The objectives of the research are worthwhile and likely to be met by the methods proposed.
11. Out of courtesy, the Project Officer will write a letter of appreciation and thanks to all respondents who have helped with the research.1 Drug trials require additional safeguards that are not covered here.
Adapted from Parkes CM, Guidelines for Conducting Ethical Bereavement Research. Death Studies 1995;19:171-81.