Main findings
Most (60%) of the surveyed physicians, acknowledged an obligation to become exposed to COVID-19 infection to provide care to some or a large degree. When the physicians experienced a scarcity of PPE, 42% agreed. The odds of acknowledging these obligations increased with age.
Almost half of the respondents were to some or to a large extent concerned about being infected by COVID-19 themselves, and about spreading the virus to their patients. Almost 64% worried about spreading the virus to their families. The odds of concern regarding spreading the virus to patients or family increased with younger age and with female gender. A scarcity of PPE was experienced by 55% of the respondents, and experiencing scarcity decreased the odds of acknowledging obligation to treat, and increased the odds of concern about contagion.
Perception about duty to treat
In this study, 60% of the physicians acknowledged a duty to treat, despite risking contagion. Because the survey was sent during a pandemic, many respondents had recent experiences to draw on when they answered. Other studies have used hypothetical scenarios. In a study from 2003 in the US, a lower percentage—about half of the physicians—reported a duty to treat in a hypothetical outbreak of a potentially deadly illness [11]. In a 2006 survey of employees in a German university hospital, 24% of physicians agreed that it was ethical to refrain from providing care; thus 76% thought there was a predominant duty to treat during a hypothetical influenza pandemic [12].
Balancing the duty to treat and the duty to protect healthcare workers and their families
Conflicts between physicians´ duties have been discussed concerning previously threatening epidemics or pandemics, such as those caused by HIV/AIDS (Human Immuno-deficiency Virus/Acquired Immuno Deficiency Syndrome), Ebola or SARS (Severe Acute Respiratory Syndrome) [7, 8, 10]. On the one hand, there is concern that the duty to treat patients during epidemic or pandemic outbreaks has eroded [10]. On the other hand, there is an increasing focus on the duty to care for carers´ own health, which is both intrinsically important and instrumentally necessary for them to provide their patients with good care in the longer term [19, 22].
A US study found the concern for one´s family safety to be the most significant barrier that would prevent health personnel (including physicians) from signing in to work in an influenza pandemic [23]. In a German survey, about a quarter of physicians found it ethical to refrain from providing care to protect themselves and their families [12]. Among the Norwegian physicians in this study, we found a difference between older and younger physicians. The odds were significantly higher for older physicians to report the traditional obligation or "duty to treat". Younger physicians and female physicians had higher odds of experiencing concern about spreading the virus to patients and to their own families.
These distinctions between age groups and genders could be stable between generations (due to life phases and professional experience). Conversely, it might also result from sociocultural changes in medicine, as in society in general. If the traditional medical duty to treat to a larger degree becomes challenged by the duty to protect healthcare workers and their families from infection, this can imply, over time, that fewer physicians will consider the duty to treat their top priority.
Do physicians from all specialties feel equally obligated?
Physicians working in "COVID-19-exposed specialties" had significantly lower odds of concern about spreading the virus to their families. These physicians did not have higher odds of concern about being infected themselves, even though they actually experienced more scarcity of PPE than others and reported that this led to higher risks for health personnel. Thus, one could surmise that these physicians were less stressed by the situation or that they were more accepting of these kinds of risks. However, physicians in these specialties do not acknowledge a duty to treat patients more often than others.
The findings are important regarding whether certain specialties or groups of physicians should have a stronger duty to treat than others. Malm et al. question the extent to which physicians, in general, feel bounded by the relevant provisions in ethical codes or the Geneva Declaration [8]. There can be an implied consent to abide by such general codes when starting one’s career as a physician. However, the authors claim that many physicians will not have internalised and thought through what this means in different scenarios. They further argue that one might establish special contracts or codes detailing a special duty to treat for physicians who work in specialties that are more regularly exposed to such risks and, subsequently, have more training in handling them. One example of such experience and training could be treating contagious patients in departments for infectious diseases [8].
PPE and the employers´ duties
Our study has demonstrated a link between scarcity of PPE and less support for the duty to treat. The scarcity of PPE is also linked to more concern about the infection of self and others. Cowper et al. emphasise the potential tension between the employer's dual duties in a pandemic situation: to provide adequate treatment for patients and to provide physicians and other employees with a safe work situation, in this case, by providing adequate PPE [14]. The responsibility to provide healthcare workers with PPE can be the employer´s, but Schuklenk emphasises that governments are also responsible for securing an adequate stock of PPE. This is especially the case when a pandemic has been anticipated, as was the case with COVID-19 [15]. Thus, insufficient pandemic preparedness on the part of the governments and institutions arguably weakens physicians’ moral obligation to care. Johnson and Butcher argue that physicians who make sacrifices to provide care during a pandemic are owed reciprocal obligations from their institutions and society [17]. The provision of adequate PPE could be such an obligation. The authors claim that opting out of a high-risk procedure because adequate PPE is unavailable could be justified.
Britain´s General Medical Council has acknowledged this dual duty in relation to the present pandemic: "We do not expect physicians to leave patients without treatment, but we also don´t expect them to provide care without regard to the risks to themselves or others" [9]. In the Norwegian Medical Association's ethical code for physicians, the obligation to treat patients is emphasised, but nothing is stated about how this obligation should be balanced against potential risks to physicians´ own health [24]. A recent paper by Gamlund et al. argues that neither the Declaration of Geneva, the Norwegian Medical Association's ethical guidelines, nor Norwegian laws give a precise answer to whether physicians have a moral obligation to provide medical care, without adequate PPE, during a pandemic [18].
Frameworks for balancing duties
The fact that about 40% of the physicians in this study, and a substantial number of physicians in previous studies [11, 12], are hesitant to expose themselves to the risk of infection, underscores the importance of finding ways to handle this dilemma. Different frameworks have been proposed for balancing duties and interests that might conflict.
McDougall et al. [20] describe a structure for individual reflection, staff discussions and decision making. They stress the importance of transparency and accountability for the decisions made. Ethically challenging dilemmas must be verbalised and discussed, and healthcare workers must understand and participate in deliberations. Participation and knowledge will increase the acceptance of difficult decisions.
McConnell developed a deontological framework (defining actions that are good or bad according to a clear set of rules) for evaluating health professionals’ duty to provide care [13]. He argues that professionals are morally justified to refrain from care when their duty to treat is outweighed by the risks and burdens to themselves. The obligation to protect family members can be a significant part of such burdens. He suggests that healthcare workers exposed to a higher risk of infection should be compensated, for example with prioritised healthcare for their family. Meanwhile, a healthcare worker relocated from a higher risk situation might be morally required to compensate society for this, for example financially or with free labour after the epidemic. Deontological frameworks like McConnells could inspire to thorough and broad normative discussions and deliberation processes with relevant stakeholders. Our study can be a useful empirical contribution into these processes.
Strengths and weaknesses
A strength of this study is the closeness in time between the start of the COVID-19 pandemic and the measurements of physicians´ attitudes and concerns pertaining to this period. This can contribute to results that are closer to the practical dilemmas experienced in healthcare during a pandemic and not just measure theoretical ideals of what physicians´ duties to patients entail.
The relatively high response rate (70%), which is higher than for other surveys of the medical profession [25], and the fact that the sample is representative of practising physicians in Norway in key aspects [21], provide a good basis for generalisation to the population of doctors. There were also no significant disparities regarding age or gender between the respondents and the non-respondents. This does not, however, rule out the possibility of nonresponse bias.
The pandemic laid great stress on large parts of the healthcare system, particularly its personnel. There is a possibility that the most stressed physicians opted out of answering the questionnaire. Conversely, the most burdened physicians might have wished to document the situation in which they found themselves. Although the scarcity of PPE and perceived risk were high in the first months of the pandemic in Norway, as in other countries, it soon improved. Compared to other countries, the infection rates, hospitalisation and lack of resources have been low. Therefore, the results in this study could be difficult to generalise to other countries.
Because attitudes and concerns can also vary with personality and coping style [26] it could be important to include such co-variates in future analyses. Another limitation could be that we have only self-reported data in this study. However, when investigating attitudes and concerns, this is a plausible method.
We did not used validated questions in this study, as we could not find good examples that explored what we wanted to study.