Recruitment and characteristics of informants

Seven GPs were recruited for individual interviews. Recruitment was performed through requests to the Norwegian Christian Medical Association (NKLF), to which we believed that many objectors would belong. Two informants were recruited through the investigators’ own professional acquaintances. Ideally we would have wanted more informants, but repeated attempts did not yield more willing participants. However, it was decided not to include the GPs that had been most active and vocal in the public debate.

Of the seven informants four were women and three men. Their age ranged from 30 to 55, and they practiced in a variety of locations, ranging from cities to rural areas in western and central Norway. All informants were interviewed by the first author (EMKN) in their offices.

A question guide was used when conducting the interviews and modified after the initial interviews. In particular, informants were asked to describe in detail occasions on which they had refused to refer, their reasons for objecting and the perceived impact on physician-patient relationships. The interviews lasted about an hour, were taped and subsequently transcribed. Five interviews were conducted in the winter of 2012, whereas the remaining two took place in the first half of 2013.

Content analysis

Research ethics

Informants were informed in writing and orally, and gave written consent. The project and the handling of recordings and transcriptions was approved by Personvernombudet (The Data protection office). According to the Norwegian system the study did not need institutional review board approval.

Investigators’ preconceptions

As conscientious objection is a highly normatively charged issue, we believe that scientific investigators in this field do right in declaring their normative preconceptions. EMKN had no settled view on the morality of conscientious objection at the outset or throughout the study. HS supports GPs’ limited moral right to conscientious objection, with a pragmatic view on how these questions should be solved, so as not to interfere with the patients’ rights and the doctor-patient relationship. MM has authored academic papers on normative aspects of conscientious objection[12, 13], supporting a limited moral right to conscientious objection that includes a GP’s objection to referrals for abortion. He has defended this view in the Norwegian public debate[7, 14].

How conscientious objection is carried out in practice

Among the respondents two main approaches emerged. The first group of GPs attempted to prevent all consultations with women seeking abortion from taking place. This they did by having their secretaries schedule all such consultations for one of the GP’s colleagues. The colleagues in question had explicitly agreed to this general arrangement.

For this first group of GPs the issue of referral for abortion nevertheless sometimes came up in the consultation, mainly because not all patients disclose their agenda to the secretaries. In such cases the GPs then immediately conveyed that they would not refer for abortion. The GPs expressed this in different ways, but were united in placing the emphasis on themselves, rather than on the (morality of) the act of abortion or factors pertaining to the patient. The GPs told the patient that they were unable to comply with the patient’s request for a referral. These GPs would then help patients set up an appointment with a colleague.

In this first group, only one GP had openly declared his conscientious objection in the local community. Subsequently this doctor had not had any consultations where a referral for abortion was requested. An additional two doctors had gained acceptance for their practices by the municipality at the time of appointment.

The second group of GPs wanted to have the consultations with the women requesting referral, and the doctors did nothing to prevent these consultations taking place. These GPs performed the physical examination, history taking, and gave information, but ultimately would not provide the referral itself. One doctor would immediately inform the patient that she would not provide the referral, whereas the others in this group did not convey their refusal before the end of the consultation. The doctors in this group expressed their refusal in ways similar to the first group. The patients were then informed about how to obtain a referral; typically the GP would arrange for contact with a colleague, who would then refer the patient.

Two of the informants in this second group worked in areas where the local gynecological department did not require a referral for abortion, but could accept patients directly. One of these GPs nevertheless made a point of stating to patients that she could not provide a referral. The other GP only disclosed her objection in the case that the patient requested referral documents. This GP maintained that her objection would be considerably more practically difficult in other parts of the country where a written referral is required.

Several of the informants pointed to another possible solution: a written statement that the patient was pregnant, as an alternative to a referral letter. The patient could then bring this statement to the gynecological department. These informants saw this alternative as morally preferable to writing a referral – a practical compromise that would satisfy both the patient’s rights and needs, and the GP’s own need for not contributing in the abortion process in a morally problematic way. However, none of the informants used this method more than sporadically.

Common to all informants was the emphasis on not unduly obstructing the fulfilment of the patient’s legal right to abortion. The informants also claimed that the requests for referrals for abortion were uncommon occurrences in their practices. With one exception the doctors had not informed their constituency about their objections, and neither did they see any need for this.

Justification for conscientious objection, and associated emotions

Relations to colleagues

The informants found their relations with their local gynecological departments to be unproblematic. As for their GP colleagues, the informants all had experienced understanding and respect for their views of abortion and their refusal practice. Even though some colleagues had spoken critically of conscientious objection in the media, our informants had not experienced negative reactions from colleagues they cooperated with daily. Two of the informants stated that their conscientious objection had been an entry point to good conversations with their colleagues about ethical dilemmas pertaining to abortion. The informants had all found colleagues who were willing to take over the consultations with the patients requesting abortion.

Consistency

The informants were asked whether they saw their refusal practice as morally consistent. Informants saw challenges to their consistency along three axes: the reason for abortion, other issues regarding early human life, and degrees of cooperation in referral.

Burdens to patients and challenges to the physician-patient relationship

Consistency, ambivalence, and absolutism

According to a 1999 study on moral attitudes, Norwegians tend to separate between the questions "what is morally right and wrong", and "what should be allowed in a legal sense"[2]. E.g., in 2004, more than 50% of Norwegian women who considered themselves conservative on religious questions, supported the right to abortion on demand[15]. The 1999 study further showed that Norwegian doctors are more permissive toward abortion (which is legal) than the general population, but less permissive regarding euthanasia (which is prohibited). This could indicate that Norwegian doctors subscribe to a separation of law and morality to an even greater extent than the general population.

In a study among health professionals working in prenatal screening, Farsides et al. introduced the labels ‘facilitators’ , ‘tolerators’ , and ‘absolutists’[16]. Whereas facilitators separate their own moral views from their professional duties, and tolerators accept the perceived wrongdoing of another, absolutism ‘entails having a moral belief about something which is fixed and non-negotiable’.

One would perhaps expect conscientious objectors to be absolutists in this respect, but the label does not fit our informants entirely. Most displayed ambivalence towards situations in which an abortion was thought to be especially well justified, and some did in fact refer for abortion in such situations. In addition, many informants thought that their practice was not fully consistent when other issues pertaining to early human life were considered, and that their practices amounted to a degree of cooperation that was a necessary compromise.

Thus, even though the category of the absolutist is helpful for theoretical delineation of different attitudes, the typical absolutist – at least among our informants – is not absolutely absolutist, but rather marked by a striking ambivalence and a willingness to make certain compromises (e.g., prescribing contraceptives that have post-fertilization effects). A certain pragmatism seems to prevail, even among most doctors not willing to make referrals to abortion.

Diversity of refusal practices, lack of legal regulation, and christian denominations

The study uncovered a diversity of practical arrangements among conscientious objectors. Informants differed as to whether they attempted to avoid the consultations with the patients seeking abortion, as to at what stage in the consultation they communicated their objection, and as to how they ensured the patient’s subsequent access to abortion. They also differed as to whether they had informed officials in the municipality in which they practiced, and whether they had informed the patient population they served – something only one of the informants had done.

We identify two factors that we think can contribute to an explanation of this observed diversity: the relative lack of a developed legal regulation of conscientious objection; and the lack of guidance from Christian denominations. First, the relative ‘legal vacuum’ (until 2011) gave the GPs an incentive to conduct their refusals ‘silently’. If these practices were to surface and become a public concern, this could potentially lead to legal repercussions and explicit societal disapproval. The legal vacuum meant that no guidelines for how referrals ought to be conducted were available. The silent refusals also meant that conscientious objectors were unlikely to know many like-minded colleagues. In sum, each physician was left to figure out for themselves whether and in which situations to object, and how to go about objecting.

Second, Protestant and other non-Catholic churches, to which five of the seven informants belonged, typically do not issue detailed guidance in bioethical questions. For instance, the Church of Norway (the former state church) has no settled view on how physicians should deal with abortion. If no guidance was forthcoming from legal rules, then, neither was there any from official church bodies. Again, conscientious objectors were left to shape their practices on their own.

However, the two remaining informants were Roman Catholics. For one of these, but not for the other, the Catholic Church’s official stance on cooperation in abortion was important. This informant stated that obligations towards the church’s teaching were crucial in guiding and shaping the line of action chosen.

As a consequence of refusal practices surfacing and becoming a public concern, public attention has led to debate on how refusals ought to be carried out in practice. If official regulations are issued, these may prescribe the practical arrangement for refusal practices that the public and the authorities deem most appropriate.

Well-established physician-patient relationships and the acceptability of refusals

In the normative literature on conscientious objection it is sometimes stated that conscientious objection is only – or at least more – morally acceptable if it takes place before a physician-patient relationship has been fully established[17]. The main reason is that the physician in such cases has not yet promised the provision of health care services.

However, a novel argument to the opposite effect may be constructed from the present findings. Many of the informants had the impression that a long-standing physician-patient relationship was relatively impervious to deleterious effects of disagreements and moral differences. In a well-established relationship the physician and the patient know each other well and are, seemingly, likely to be more tolerant of each others’ idiosyncrasies. Patient-doctor relationships are generally more trustful when they have lasted for a few years[18]. The Norwegian ‘regular GP scheme’ facilitates the establishment of long-running physician-patient relationships.

By and large the informants did not think that their practice involved significant burdens to patients. However, the objecting GPs’ own reports of course do not settle this issue (see Limitations below). Many patients seeking abortion are young women who may not have formed long-running relationships with their GPs. They may feel more abandoned by the GPs’ refusals than the GPs themselves claim in our interviews.

In the GP’s refusal may lie the potential for negative patient experiences of rejection and moral condemnation. In the Norwegian debate, this potential for negative experiences has figured prominently as a weighty argument against conscientious objection. This is in contrast with the international academic debate, in which this argument typically is not mentioned or lent much weight[17, 19, 20]. We believe this observation can be partly explained by objections in Norway often occurring ‘silently’. The practice is not transparent, and so when patients confront a GP who refuses to refer, they may be wholly unprepared. Another factor that can help explain the significance given to the potential for rejection experiences is that abortion is considered to be a settled issue by most (i.e., a right to abortion on demand is accepted by a large majority); patients may thus be unprepared for encountering a moral dissident at the doctor’s office.

Performing versus referring, and varieties of referrals

Granted that refusals to perform abortion are ethically justified, how strong is the corresponding justification for refusals to refer? Our informants do not address this question directly, but are unanimous in perceiving referrals as ethically highly problematic active participation in the abortion process. Still, the informants saw it as a duty not to obstruct the continuity of care. They had all arranged for colleagues to take over their consultations.

Our findings include nuances about referrals that the normative literature often does not take into account. Typically, the literature discusses the moral complicity involved in referring to an abortion clinic (‘vertical referral’), and in referring to a colleague who then refers for abortion (‘horizontal referral’). In addition, Chervenak and McCullough discuss ‘indirect referrals’ , thereby designating the act of merely ‘providing patients with referral information’[21]. Authors have differing views on the moral acceptability of the complicity involved in various kinds of referrals[17, 21, 22].

However, our findings indicate that horizontal referral can be carried out in different ways. For instance, horizontal referral can be both the act of personally ensuring that the patient is offered a consultation with a colleague who then refers; and developing systems that will ensure that patients go directly to a colleague, thus bypassing the conscientious objector entirely. In addition, a written confirmation that the patient is pregnant was seen by some as a (morally preferable and distinct) alternative to a direct referral. It becomes a question for normative analysis whether different varieties of horizontal referral and the actions involved therein constitute different degrees of moral complicity, and whether any differences are of moral import.

Different refusal practices may make a moral difference

A normative analysis of the diverse refusal practices uncovered in this study may find that the practices are not necessarily morally consistent. This is a discussion we will not enter into here, other than pointing out some possible questions for subsequent normative analysis. For instance, Wicclair argues that the moral justification for conscientious objection is strengthened to the extent that the doctor seeks to minimize burdens to patients[17]. Which practical arrangement will burden patients the least – the advance disclosure to the patient population, the secretary’s direct arrangement of a consultation with a colleague, the GP’s disclosure of their refusal at the outset of the consultation, or the disclosure at the end of the consultation? And are the differences of moral significance?

One upshot is that there may be room for professional guidelines for how conscientious refusals should take place in practice. That is, if society deems that one way of conducting the refusal is more acceptable than others, and that this way is in fact acceptable, then a practical guideline for conscientious objectors could shape the way refusals take place, ultimately minimizing burdens to patients. For instance, a guideline could demand advance notification of the GP’s objections for the patient population.

The study’s weaknesses

The study includes fewer informants than would be ideal. However, the study has merit in that each participant provided rich descriptions, and in shedding light on an under-researched topic. All informants were Christians; however, it is generally believed that most if not all Norwegian GPs who object to referrals for abortion are in fact Christians. The interviews gave the informants the opportunity to present their practices in a positive light. The informants can have had an incentive for a positive self-representation, especially considering the public controversy surrounding conscientious objection. When interpreting the informants’ reports it is vitally important to keep in mind that these reports, however illuminating, do not give the full picture of the practice of conscientious objection. Crucially, the patients’ own views are missing here. How do patients seeking referral for abortion experience their GP’s conscientious refusal? Very little is known about this, and there may be methodological obstacles to the empirical study of these experiences. For the assessment of patients’ experiences we currently only have indirect sources such as media stories and complaints to the health authorities that have been made public.