Several studies have explored differences between North American and European doctor-patient relationships. They have focused primarily on differences in philosophical traditions and historic and socio-economic factors between these two regions that might lead to differences in behaviour, as well as divergent concepts in and justifications of medical practice. However, few empirical intercultural studies have been carried out to identify in practice these cultural differences. This lack of standard comparative empirical studies led us to compare differences between France and the USA regarding end-of-life decision making. We tested certain assertions put forward by bioethicists concerning the impact of culture on the acceptance of advance directives in such decisions. In particular, we compared North American and French intensive care professional’s attitudes toward: 1) advance directives and 2) the role of the family in decisions to withhold or withdraw life-support.

Cross-cultural analyses comparing the North American bioethical model to the Latin model of Southern Europe typically characterize the North American model as liberal and individualistic and the Southern European model as being a more socially oriented and responding to the norms of professional deontology.1 In the American system, on this view, the moral obligations of doctors are derived from the so-called “negative” rights of individual patients, which are based first and foremost on a respect for individual liberty. These rights ensure the non-intervention of other individuals — particularly doctors and close relatives of the patients — or of the State in the private life of patients.² In the USA, informed consent and advance directives (AD’s) –in cases where the patient is no longer able to make decisions- are seen as the principal means for ensuring that these rights are respected. As in other liberal societies, the ability to give consent in the USA is paramount and the State cannot take the place of the individual. In Latin societies, such as France, individual consent is recognised, but is sometimes tempered by the intervention of professionals, close relatives and the State, who can modify individual choice for the good of the person, in the name of a collective rule.³

According to some,^4,5 the paternalism of French healthcare professionals is much more complex and justifiable than a simple desire for power. For example, F. Pochard et al. suggest that this paternalism in intensive care is not an authoritarian desire of power by doctors, but instead reflects an acceptance of responsibility for decisions accorded to doctors by patients and their close relatives. They suggest in particular that patients “may prefer to relinquish autonomy in situations they feel are overwhelmingly painful or guilt-ridden”(p.1890).⁵ Our study compares the attitudes of French and American intensivists toward AD’s and the role of family so that differences can be better understood while avoiding mere cultural prejudice.

We explored the acceptance of AD’s by healthcare professionals in France and the US. We also tested the hypothesis that the decision to withdraw or withhold life-sustaining treatment is more dependent on the patient’s wishes and less dependent on families’ wishes in the USA than in France.

Three different clinical scenarios corresponding to incompetent patients having expressed, through an AD, a refusal to receive vital treatment were presented, during individual interviews, to 66 health professionals. In April 2003, 36 professionals belonging to three ICUs from Parisian hospitals were interviewed.A One year later, we carried out similar interviews with 30 intensivists working in two ICUs in Ohio. Only the professionals in each department likely to take part in end-of-life decision-making (doctors, nurses and interns) were interviewed.

The scenarios correspond to typical situations and shed light on a heterogeneous group of decision-making issues that have been discussed in bioethical and medical papers^15–18. These scenarios differ in their circumstances (i.e. patient diagnosis, prognosis and age, and the patient’s reasons for writing an AD), but all have one thing in common: respect of the patient’s wishes would inevitably lead to the patient’s death.

• Case 1 concerns a 40-year-old patient in a vegetative state since a road accident six months previously. In his AD, this patient refused to accept the parenteral nutrition and hydration currently keeping him alive. The work of the Multi-Society Task Force on PVS has shown that a patient in this state no longer has the mental capacity to be conscious of himself or of his environment. He has a 46% chance of “coming round” (awake with reactions to visual and auditory stimuli) and a less than 0.5% chance of recuperating his functions (communication, ability to learn, ability to adapt, mobility, ability to take care of himself and to take part in recreational activities). Furthermore, having shown no improvement in the previous three months, this patient has a significantly reduced chance of displaying functional improvement with no associated severe sequelae.¹⁹
• Case 2 concerns a 70-year-old patient who has suffered a minor cardiac event and who also suffers from Alzheimer’s disease, making it impossible to obtain informed consent. Apart from this acute episode, he clearly still has a certain quality of life. His life expectancy, if he overcomes the cardiac problem, may be several years, during which time his state of health will gradually worsen, due to his underlying degenerative disease. Prior to losing his mental capacity and while still fully competent, the patient signed an advance directive in case of his becoming incompetent and suffering from Alzheimer’s Disease. In this document he declared that he did not want to have his life artificially prolonged if he became incapable of living his current standard of life.
• Case 3 concerns a 39-year-old patient with no significant medical history. He requires a transfusion, but his life is not immediately in danger. He is mentally capable and refuses blood products due to his religious beliefs, claiming to be a Jehovah’s Witness. He has asked that all other treatments be tried to cure him. However, his situation deteriorates; he loses consciousness and a transfusion becomes absolutely necessary to save his life.

We asked them whether they would follow the AD’s in two different situations: 1) in which the close relatives wanted to follow the AD of the patient, and 2) in which the close relatives opposed the AD of the patient.

We asked health professionals to respond according to their own convictions and to not take into account any eventual legal consequences that could affect their decisions, so as to collect opinions as independent as possible from the legal context in each country.^B Three responses were possible for each situation: “I would follow the advance directives”; “I would not follow the advance directives”; and “No definitive opinion”.

For American health professionals, the script for the interview was translated by an American translator and validated by an American intensivist. The results were analysed by the group of researchers who designed the interview, together with two other external people competent in medical bioethics — one French, one American.

In US, the protocol was approved by the ethics commitee of MetroHealth System, Cleveland (Ohio). In France, ethics commitee approval for this kind of sociological research using professionals as research subjects is not required.

The two groups were homogeneous for the number of professionals, mean age, sex, post held and number of years of experience in the ICU. We found no statistical link between these variables and the responses of the professionals. We found a significant difference between French and American professionals in terms of the distribution of religious beliefs: the numbers of religious and non-religious professionals were similar in France, whereas there were more religious than non-religious American professionals (Table 1). Compared to France, US professional’s religious affiliation was more diversified. In France we found a significant number of professionals who considered themselves religious but who did not have or disclose a particular affiliation (Table 2).

The situation in France today with respect to end-of-life practices looks similar to how it looked in the early 1990’s in the US. In the US, AD’s have not been widely adopted despite their legal support and the apparent broad recognition of caregivers that patient’s wishes, as expressed through AD’s, should be respected. In France, the practice of writing AD’s is much more recent and less extended than in US. We found some important but not radical differences between American and French intensive care professionals’ attitudes towards AD’s and the family wishes. Neither French professionals ignore systematically the autonomy of the patient nor the American ones systematically oppose the wishes of the family in order to honour those of patients. Nevertheless, French intensivists clearly afford more weight to family wishes in cases of conflict with patient previous wishes, than do American doctors (at least in hypothetical case responses). That led us to remain sceptical concerning the respect that AD’s will obtain in France, especially when proxies will disagree with them and want to continue undesired or futile life sustaining treatment. Mechanisms for teaching professionals how to prevent or deal with potential conflicts between proxies’ wishes and patient’s interests might be developed, both in France or the US. The deeper issue underlying AD’s whether in the US or France is the moral and political right of patients to direct their health care even when unable to make decisions for themselves. To the extent that this right is to be taken seriously in either context, mechanisms for allowing patients to exert control over decision making must be developed. AD’s are a flawed, but defensible, way of doing this. Although AD’s cannot resolve all problems, they are often informative for end-of-life decision making and, in certain very problematic cases, the only means by which the best interests of patients who are no longer able to communicate can be decided. Perhaps other more culturally tailored approaches might be more effective. More comparative empirical research is needed in order to reach a deeper understanding of the simmilarities and diversities between North American and European models of patient-proxy-professional relationship.