The starting point for us was the finding that all twelve patients in the study expressed a will to be well informed throughout the disease, which is congruent with the findings of many other studies [[14–16], and ]. If we take these statements by patients seriously, our aim ought to be to find out what could possibly be done to reduce the lack of knowledge and/or awareness of these patients. Improved disclosure and patients' right to obtain information about their condition is usually supported by the principle of autonomy, and non-disclosure by the principle of beneficence . However, both practices can be ethically justified depending on the situation . According to the Kantian way of thinking, autonomy is the basis for making a man able to act ethically as a person with responsibility. Autonomy is what is needed for man to remain human. For our purpose a less sophisticated interpretation of autonomy will do, according to which the normative value of information and awareness only exists when it is supported by the expressed will of patients or the goals of patients. Thus neither information nor awareness are intrinsic values . If the patients instead had expressed an unwillingness to be informed, we then would have suggested that they should be less informed. But what we found was that the patients did want to know, but were not always informed or aware. What could then be done to reduce the lack of knowledge and/or awareness in patients?
How to deal with the obstacles due to the physician
According to the interviews with the patients, the physicians seem to have been eager to inform them of all results of tests taken. Only one exception was found. Why there was a delay in the informing of patient V we do not know, we only know that she had to plead for information before she got the sad news, and that is what we question. Then there was the finding of a physician withholding a potentially confirmatory bone marrow test. Without a test there was of course no test result information to give the patient. However, withholding of a test might be ethically justified by the principle of non-malificence, that is, that the potential harm the test would cause the patient should not outweigh the benefit it would bring the patient. If you can not do any good, at least do no harm to the patient. However, questions arise how as to communicate this to the patient and how to prevent withholding of a test from making the patient (more than necessarily) less aware.
Some patients in the study regarded themselves as informed, albeit non-verbally. When the non-verbal communication worked properly, as for patient X, we have no ethical objections so far, as we think that the need of this patient was met. Unfortunately, this was not the case for patient V, and the question that arises is how to be cautious in the practising of non-verbal communication. Maybe the message understood without any words will function best as a predictor of what the physician has come to tell the patient .
When the communication concerned the end of life, silence became notable. The physicians seem to have left the hard questions without an answer. If the physicians do not provide space for such communication with the patients about their approaching death, who should?
How to deal with the obstacles due to the patients
The patients developed different coping strategies. But their coping strategies (not even the strategy of denying) were not really found to be an obstacle to patients being aware of their health status. On the contrary, we found that the coping strategies all presupposed that information had been given to the patient, even though the information then was sometimes transformed into more bearable messages. According to psychological theories the coping strategy of denying functions as a defence, not to be removed or questioned [22, 23].
There are patients who genuinely wish to have limited information, but their number is far smaller than many physicians assume. Good communication is required in order to be able to differentiate patients who need more information from those who do not want information.
How to deal with the obstacles due to collusion between the physician and the patient
The collusions mostly consisted of focusing on the treatment calendar and at the same time ignoring the long-term perspective. To reduce the uncertainty of progressing disease and approaching death, could it be possible for the physicians to have a more proactive attitude (instead of waiting for the patients to ask for information)? On the other hand, could the patients be encouraged to be more proactive by asking?
The collusion was also of the kind that the patient could, more or less, leave it to the physician to decide what would be in the best interest of the patient. This could be ethically justified if this is in accordance with the expressed will of the patient. Otherwise, unless the patient is unconscious, such collusion is paternalistic and may be questioned.
How to deal with the obstacles due to neither the physician nor the patient
Finally, we have come to what we found in this study to be a major obstacle to patients' awareness of their disease, namely the insidious way that lung cancer (mostly) and haematological malignancies (occasionally) progress in a slow and hidden way behind other diseases. In order to help these patients to become more aware during the trajectory of their disease, the question arises whether or not it would be possible to inform these patients what most probably is to come. If the physicians know, from experience with other patients, when the disease most probably is turning lethal, is there an obligation to inform the patient even before the malignancy is manifested in the test results? We believe so and once again we would like to refer to the principle of autonomy. It is the will of the patient that should be the determining factor. We believe in balancing the pros and cons for what seems best for each patient, even when the malignancy is not yet manifested in test results. Important for this standpoint is that the survival rates of patients with the diseases included in this study are low, as noted. The information would not have been that urgent, had the survival rates been higher. But as they are very low we believe that at least some of these patients would have gained from receiving more information about the fatal course of the disease.
How is the physician to inform the patient when the manifested test result is not yet at hand? The problem of communicating probabilities and uncertainties deserves to be explored, but this would lead us too far from the aims of this study and to do this more research is needed. It will have to be left for another study.
Limitations, trustworthiness and validity
The limitation and trustworthiness of this study, as with other qualitative interview studies, are based on the small number of interviews. Besides, out of the twelve patients included in the study from start, only seven for various reasons were followed into the terminal phase (see Table 2). Patient B died four days after the first interview and three other patients (A, D, F) for various reasons were excluded from the study after three interviews each and we do not know what the awareness of those four patients would have been like in later phases of their disease. Nor do we know about patient G's awareness in later phases of the disease. She was followed for two years, but did not relapse during the period.
Another limitation was that the intention to follow the patients in a regulated way could not always be realized. Some prearranged interviews had to be shortened in time or postponed for days, or even weeks, because of the patient's current state of health.
Finally, even if the intention was to follow the words of the patients as closely as possible, there was always the possibility of misinterpretations or over-interpretations. To limit such risks confirmatory questions were used, the process of coding followed and checked by a co-reader, and critical researchers, physicians and co-readers have been invited to continuously read and check the interpretations made. According to S. Kvale, valid knowledge emerges where the consistency and internal logic of statements as well as the research procedure is tested: "valid knowledge claims emerge as conflicting interpretations and action possibilities are discussed and negotiated among the members of a community" .