Introduction to: Discussion, questions and answers
The same research questions may be answered by the use of a variety of different research methodologies. There is no single ‘correct’ research method to answer a research question. Indeed the ‘triangulation’ of methods is sometimes used to try and answer the same research question, so that the answer can be demonstrated not to be an artefact of the method used. Just as in land surveying, in which several reference points are used to fix better the position of the object being surveyed, researchers may study the same question or phenomenon using a variety of research techniques. The results may then be compared to get a better ‘fix’ on the question. The use of a single reference point or methodology gives one perspective of the object under study.
To illustrate this point it is proposed to consider a research issue of some importance and much contention in modern health care – the allocation of donated organs to transplant recipients. Despite the advances in transplantation techniques, there is a chronic worldwide under-supply of donor organs, to the extent that, in some countries, over one-third of patients accepted for transplantation die while on the waiting list due to the unavailability of suitable organs. A consequence of the chronic under-supply of donor organs is the necessity for transplantation teams to choose the most ‘worthy’ candidate to receive the donated organ from a panel of potential recipients. How these decisions are made and how they ought to be made have been the subjects of substantial research.
Many transplantation units around the world use a points system to assist with decisions concerning the priority of the people on their waiting lists. The points are typically based on tissue and blood type match (methods of measuring donor–recipient compatibility), time on the waiting list, ‘urgency’ (the extent to which the patient can wait) and prognosis (the quality of the expected outcome). This is in addition to certain technical requirements that vary with the organ system being replaced. For example, the size of the donor’s and recipient’s organ systems is a factor in heart–lung transplants. However, in recent years many commentators have presented data that indicate that ‘social’ criteria (such as the age, sex, education, race and wealth of the potential recipient) as well as ‘medical’ criteria are highly influential in the allocation of organs to recipients. These findings have generated considerable controversy.
How might one discover what criteria are used in the allocation of organs? One method might be to conduct a survey of those people involved in the decisions. The survey could take the form of an individual or group interview, or a written questionnaire. Alternatively, one could examine the written policies of various transplantation units or examine the medical records of the potential recipients for transplantation in the units and follow up the actual outcomes for each candidate.
Let us consider the questions we may wish to ask of people involved in the organ transplant decisions. They might include:
In your opinion, what criteria should be used in deciding who should receive an organ transplant?
In your opinion, what criteria are actually used in deciding who should receive an organ transplant?
These questions could be asked directly of the respondents in person and their comments recorded on tape and/or in the form of notes for subsequent analysis. In other words, the respondents could be interviewed. Excerpts of typical responses to these questions when the authors have asked them of members of transplant teams include:
We have an unwritten policy that we won’t transplant into patients over 60. We think they have had their turn. Also their prognosis is often poorer than younger patients.
A consequence of the HLA emphasis is that people with unusual tissue types never get a transplant. I have a problem with this. In my area, there is only slim evidence that it makes much difference. Fifteen per cent max (sic) between best and worst case match. Most people fall in the middle so the actual difference would be even smaller. This means a lot of people miss out. I think it should be a strict queue.
Another method of collecting the data would involve getting people to use a structured response question in a questionnaire. The following might be an example:
In your opinion, what criteria should be used in deciding who should receive an organ transplant. Please circle as many as you like.
| Age | Urgency of need |
| Sex | Prognosis |
| Education level | Whether recipient is a parent |
| Race | Citizenship status |
| Occupation | Other, please specify |
| Wealth | |
| Blood type match between donor and recipient | |
| HLA match between donor and recipient tissues | |
A follow-up question might include something like:
When you have selected the criteria you consider should be used in allocating organs to recipients, please write a number alongside each one to indicate your opinion of its importance in the decision. For example, the one you consider the most important should have a ‘1’ placed alongside it, the second a ‘2’ and so on until you have exhausted all the ones you selected.
In this type of exploratory study, the use of fixed-response questions may send a clear signal to the respondent as to what the researcher considers to be acceptable criteria. Therefore, respondents may not be inclined to reveal their true answer if it is not included in the list of available answers, because of fear of disapproval or perhaps disinclination for the effort to do so. Thus, before a structured response question is used, the clinical researcher needs to ensure that all possible and likely answers are included. The current authors advocate the use of unstructured questions with a small sample prior to the use of structured questions in a large-scale survey.
It is interesting to note that, in a study of 310 Australian people from the community recently conducted by the authors, the respondents consistently nominated the age of recipients, their prognosis, the length of time they spent on the waiting list and whether they had children as the four factors they considered ought to be used to determine recipient priority in transplantation decisions.
Thus, clinical researchers could use interview or questionnaire approaches in the conduct of their survey. A further approach would involve the auditing of potential recipient medical files over a period in the various transplantation units in order to check whether those who actually received the transplants were representative of the potential recipient pool. For example, if it were found that women were less likely to receive transplants than men, taking into account their relative proportions in the recipient pool, then this would be of major concern. Unfortunately, as mentioned above, there is strong evidence in the US (where most such research has been performed) that social factors have an important impact upon the actual transplantation decisions. These findings reflect the situation in the delivery of health care in areas other than transplantation.