Patient Bias in Sweden An Exploratory Study on Discrimination Against Migrant Physicians

Increasing diversity in the healthcare workforce has brought to the foreground the importance of addressing phenomena of racism, discrimination, and bias against healthcare workers of diverse backgrounds. This paper aims to contribute to the growing literature on patient bias by exploring the experiences of migrant physicians practicing in Sweden using thematic content analysis of 17 semi-structured interviews and applying microaggressions theory. Four themes pertaining to different forms of patient bias experienced by the respondents emerged from the transcripts such as refusal of care, questioning language competence, questioning medical competence, and ethnic jokes/stereotypes. Four additional themes dealing with strategies implemented by the respondents to cope with patient bias were also identified, namely confrontation avoidance, collaboration with Swedish healthcare staff, self-disclosure, and active listening. The findings elucidate the need to encourage inclusion in the workplace by providing opportunities for continuous language training and collegial support.

ethnically concordant consultations is believed to promote better patient-physician relations, hence better health outcomes (Shen et al., 2018). The second trend consists of the international migration of health workforce to meet the needs of the ageing population of high-income countries (Bludau, 2021). WHO projected a global gap of 2.4 million physicians, nurses, and midwives, with an additional 12.9 million health workers needed by 2035 (Bludau, 2021). In March 2016, the UN Secretary-General Ban Ki-moon established a High-level Commission on Health Employment and Economic Growth to monitor the global shortage of health workers and to make recommendations on how to manage their international migration in an effective and ethical manner (WHO, 2016).
These trends have dramatically brought to the foreground the importance of addressing phenomena of racism, discrimination, and bias against healthcare workers with diverse backgrounds. For instance, a recent survey among physicians in the United States found that "59% reported having heard offensive remarks from patients about their age, gender, ethnicity, race, weight, or other personal characteristics in the past 5 years and 47% had patients request a different physician" (Chandrashekar & Jain, 2020, p. S33). A similar survey in the United Kingdom indicated that around 75% of black and minority general practitioners reported having experienced some form of racial discrimination from patients (Snead, 2018).
The bulk of the scientific literature so far focused primarily on the negative impact of prejudice and bias against patients, whereas the attention on the opposite phenomenon, namely patient bias and prejudice against healthcare workers, has been scant (Chandrashekar et al. 2020;Kauf et al., 2021;Wheeler et al. 2019). In addition, the slim research on patient bias is concentrated in Anglo-Saxon countries, and in particular the United States. The aim of this paper is to contribute to the growing literature on patient bias by exploring the experiences of migrant physicians in the European context, namely in Sweden. The present research, based on a thematic content analysis (Braun & Clarke, 2006) of semi-structured interviews with 17 migrant physicians working in Swedish healthcare institutions, focuses on categorising types of patient bias as perceived by the physicians and the strategies adopted to cope with them.

Literature Review
Overt and crude forms of racism in healthcare settings, although still present, tend to be less conspicuous also because socially stigmatised (Marcelin et al., 2019); nonetheless, subtle, and sometimes unconscious forms of bias can still hinder healthcare professionals' preparedness to treat their patients effectively. Among the many dramatic examples of unconscious physician bias, Marcelin et al. (2019) include the association between the human immunodeficiency virus (HIV) and the white male gay community at the onset of the HIV epidemic. Such association cemented the initial biased belief that HIV could only be found in that specific group and blinded doctors' ability to recognise the disease not only in heterosexuals but also in the black gay community.
Though it is still imperative to address implicit and explicit bias among healthcare professionals toward patients (for a systematic review, see Dovidio et al., 2006;FitzGerald & Hurst, 2017), an increasing number of scholars call for more attention to the opposite phenomenon, namely bias against healthcare professionals, such as physicians from underprivileged or migrant backgrounds (Chandrashekar & Jain 2020;Jain, 2020;Wheeler et al., 2019).
Bias against healthcare professionals is observable both from a macro, institutional level and from a micro, patient-healthcare professional interpersonal level. Instances of bias at an institutional level include, for instance, more stringent legal requirements and skill devaluations faced by migrant doctors (Motala & Van Wyk, 2019;Salmosson, 2013), and discrimination in terms of fewer career opportunities and training admissions in the profession offered to members of minority groups (Marcelin et al., 2019;Nunez-Smith et al., 2009).
The concept of patient bias (Chandrashekar & Jain, 2020) captures, instead, forms of racism, discrimination, and prejudice against health workers at the micro level, namely when a healthcare professional interacts with a patient. Patient bias can take both explicit and implicit forms. Explicit refusals of care, for instance, when a white patient expresses the desire of not to be treated by a black physician, is the most blatant example of explicit patient bias (Kimani et al., 2016;Rosoff, 2018). In an analysis of a German physician-rating website, Kauff et al. (2021) found that general practitioners with non-German names tended to be evaluated less favourably compared to German ones.
Microaggressions, namely "brief, everyday exchanges that send denigrating messages to people of color because they belong to a racial minority group" (Sue et al., 2007, p. 273), are also common examples of explicit patient bias. The concept of microaggression captures a range of verbal and non-verbal interactions, intentional but often also non-intentional, that communicate negativity or hostility toward a target individual with an underprivileged background. Sue et al. (2009) identified three major types of microaggressions, namely microassaults (direct verbal and nonverbal act of aggressions, such a racist slur), microinsults (subtle, often intended remarks that still demean the target, such as addressing a female doctor as a nurse), and microinvalidations (remarks that negate or dismiss a target's feelings or experiences, such a calling out a black doctor as oversensitive to a racist joke). In a study based on semi-structured interviews with minority (Black, Hispanic, Native-American, and mixed ethnicity) resident physicians, participants reported having to deal in their practice with a constant stream of microaggressions, such as assumption of lower status, exotification, and feelings of alienation in their own land (Osseo-Asare et al., 2018). Female medical students with a minority background were more likely to experience microaggression on a daily basis (Espaillat et al., 2019). Wheeler et al. (2019) provided a comprehensive categorisation of seven types of problematic patient behaviours emerged during a series of focus groups with physicians, residents, and medical students that included explicit refusal of care, explicit or socially biased remarks, questioning clinician role, ethnic jokes or stereotypes, assertive inquiry into physicians' ethnic background and contextually inappropriate compliments or flirtatious remarks, and non-verbal disrespect (p. 1680).
Though implicit forms of patient bias are less obvious and, therefore, more difficult to observe, they still have concrete repercussions on the career and well-being of health workers. Using an experimental design employing a deception paradigm, Louis et al. (2010) found evidence in a sample of Australian students of European heritage of "selective discounting of credentials for foreign-born doctors trained in their countries of birth relative to native doctors" (p. 1245). "Skill discounting" (Shinnaoui & Narchal, 2010, p. 425) represents a subtle form of prejudice that involves a formal or informal devaluating of a person's acquired skills, qualifications, and expertise merely based on their immigrant status. In Louis et al.'s (2010) study, fictitious foreign-born Pakistani physicians were systematically evaluated less favourably than fictional Australian native-born candidates with similar qualifications, educational level, work experience, and personality traits. The small, yet statistically significant negative evaluations involved areas such as recommendations for an interview, levels of personal and social trust, and assessment of the relevance of education and work experience for a position.
First-world medical education, for instance, a medical degree obtained in the United Kingdom, appeared to attenuate nationality bias and to boost positive evaluations, for instance in areas such as perceived competence and trustworthiness. Greene et al. (2018) conducted an online survey on a predominantly white sample of US citizens and found that respondents displayed a statistically significant preference for a fictitious physician with a distinctly white male name (Dr. Jack Williams) with equivalent quality performance compared to other doctors with typical white female (Dr. Holly Williams), black male (Dr. Tyrone Williams), black female (Dr. Jasmin Williams), or Middle Eastern (Dr. Raja Fakraddin) names. The preference pattern was particularly distinct in white male respondents. In experimental studies of analogue patients' (i.e., individuals not in a patient role at the time of the studies) first impressions of physicians based on photographs, Hall et al. (2020) found that research participants found physicians of Indian nationality and ethnicity do not differ significantly from US. physicians (all Black or Caucasian, none of Indian origin) in terms of perceived competence. However, Indian physicians received significantly lower ratings than their US colleagues with respect to perceived patient-centredness. The studies' participants were also less likely to want to see Indian physicians again as opposed to the US physicians. Baquiran and Nicoladis (2020) tested the impact of foreign accents on perceptions of doctors' competence in a sample of Canadian undergraduates. Doctors with Chinese-accented English were found to receive statistically lower ratings in perceived competence than doctors with standard Canadian English.
Patient bias was found to have a negative impact on the mental well-being of healthcare professionals, in terms of increased perceived levels of stress (Kaltiso et al., 2021). Healthcare professionals who experience patient bias also report emotional burden (Wheeler et al., 2019), feelings of degradation (Chandrashekar & Jain, 2020) and powerlessness (Espaillat, et al., 2020). Respondents in Wheeler et al. (2019)'s study also reported decreased learning and practice and withdrawal from roles and rotation (Wheeler et al., 2019). Han and Humphreys (2005) found that perceived discrimination in the local community affected the willingness to stay in migrant physicians, assigned to rural communities in Australia. Hall et al. (2020) noted that "prejudices and stereotypes held by patients can create unfair and self-perpetuating disparities for physicians" (p. 296), and patient bias is believed to represent a contributing factor in explaining why healthcare professionals from ethnic or underprivileged backgrounds have comparable lower incomes (Greene et al., 2018) and lessened opportunities for career advancement (Chandrashekar & Jain, 2020).
Despite the extent and severity of patient bias clearly emerging in the scientific literature, it is nonetheless complex to identify effective strategies for addressing this phenomenon. First and foremost, there is a complex balance to strike between the rights of the patient and those of the health worker (Chandrashekar & Jain, 2020;Paul-Emile et al., 2016;Rosoff, 2018). For instance, in the United States, there is a paradox between on one hand the American Medical Association (AMA) Code of Ethics, which explicitly states that patients have the right to choose their physicians, and on the other hand Title VII of the 1964 Civil Rights Act, which entitles health workers to perform their duties in a workplace free of discrimination (Chandrashekar & Jain, 2020). Also, the physical and mental conditions of the patient need to be taken into consideration. As Paul-Emile et al. (2016) explained, "[p]atients with significantly impaired cognition are generally not held to be ethically responsible" (p. 710).
Also, on one hand, discrimination is not easy to pinpoint, and on the other hand, racial and ethnic patient-physician concordance foster better communication, improved trust, and better health outcomes (Shen et al., 2018). For instance, can the request from a black patient to be treated by a black physician be equated tout court with a similar request from a white patient for a racially concordant physician?
According to Chandrashekar and Jain (2020), it is critical to investigate patients' rationale on an individual basis. An understanding of a black patient's preference "may reveal an understandable distrust of healthcare professionals stemming from the medical establishment's historical exploitation of Black patients" (p. 37). At the same time, the request of the white patient may have been triggered by a posttraumatic stress disorder because the originally assigned doctor clinician share the "same ethnic background as a former enemy combatant" (p. 37). In both cases, accepted medical guidelines state that accommodation of the request is justifiable (Chandrashekar & Jain, 2020;Paul-Emile et al., 2016). However, if the investigation of a patient's motives highlights uncontroversial evidence of bigotry, a wide range of options are available, such as ignoring derogatory comments, persuasion, negotiation, and trustbuilding, but also responding to demeaning comments, and referring the patient to other hospitals or clinics. Simply bowing to discriminatory demands from patients make healthcare institutions appear complacent to bigotry (Chandrashekar & Jain, 2020;Paul-Emile et al., 2016;Rosoff, 2018). In sum, there is no easy, one-size-fits-all solution to the many paradoxes that come with patient bias.

Purpose and Motivation of the Study
Patient bias is a relatively novel area of research, and scholars have called for more studies on prejudice and discrimination against healthcare workers and effective ways to respond to patient bias (Chandrashekar et al. 2020;Wheeler et al. 2019). As most of the research on this topic originated in the United States, the current study extends the literature by focusing on the experiences of migrant physicians in Sweden, contributing to European research in this area. Like in many high-income countries, general life expectancy in Sweden is growing, hence the relative proportion of its elderly population. According to the Swedish Statistics Agency (SCB, 2018), "the share of people aged 80 and older will increase most, and in 2028, their share is expected to be 255 000 more people than today, an increase of 50 percent" (par. 1). In addition, the population of Sweden is becoming more ethnically, religiously, and linguistically diverse as a result of immigration, with 30% of the population aged 25-64 projected to be born abroad also by 2028 (SCB, 2018). The composition of the Swedish healthcare workforce reflects these demographic shifts toward a more diverse society. In 2021, 37% of physicians who got their medical licenses were educated outside Sweden (Socialstyrelsen, 2022). Physicians come to Sweden from both the EU/EEA area, primarily Poland, Denmark, Romania, and Finland, and from countries outside Europe; among the non-European doctors, Iraqis dominate due to the refugee wave to Sweden caused by the 2003-2010 Iraq War (Gevelin, 2015).
The topic of patient bias in Sweden recently gained media attention at national level after an exposé run by a leading Swedish newspaper, Dagens Nyheter (DN). Swedish national guidelines do not explicitly prohibit patients to choose physicians based on ethnic origins, and when a DN reporter, posing as a patient, called 120 public and private health centres around the country and requested to be visited by an ethnic Swedish doctor, 51 health centers agreed to meet the request. DN's exposé sparked a heated debate that led to revising national and regional guidelines (Sadikovic, 2021;Dagens Medicin, 2021). Salmonsson and Mella (2013) researched bias against migrant doctors in Sweden with a qualitative analysis of the journal of the Swedish Medical Association using a human capital perspective. According to the authors, "there seems to be a devaluation of human capital which might lead to marginalisation of immigrant specialists in the case of immigrant doctors in Sweden" (p. 8). Migrant doctors, especially those from outside the European Union, need not only to have qualifications equivalent to their Swedish peers but also to obtain an implicit "cultural authorisation" that demonstrates their ability to manage Swedish cultural codes. Salmonsson and Mella (2013)'s study explored the phenomenon of bias against immigrant health workers in the Swedish context from a broader, institutional perspective. The present study focuses, instead, on immigrant physicians' experiences with patient bias, hence at the interpersonal level. The following research questions guided the study: RQ1: Which forms of patient bias do migrant physicians experience in Sweden?
RQ2: What strategies do migrant physicians use for coping with patient bias?

Method
Interviews were collected as part of the research project Communication and Interaction in Intercultural Healthcare: Non-Swedish Physicians in Sweden, financed by the Swedish Council for Working Life andSocial Research (2005-2008) and within the frame of the project Breaking the Swedish Cultural Code (2008-2018) organised by the Region Västra Götaland and Sahlgrenska Academy for supporting professional development of the physicians educated outside Sweden.
Our data include 17 semi-structured interviews with migrant physicians working in Sweden. Representatives of healthcare centres and hospitals in the Region Västra Götaland were contacted and asked if they had any physicians born and educated outside Sweden who could be interested to participate in the project. The physicians who were interested contacted the first author who informed them about the study. It is worth mentioning that, in spite of the extended time for data collection, the project team experienced high rejection rates, primarily due to the sensitive nature of the study and physicians being uneasy about speaking Swedish, which corroborates with the experiences of Tariq et al. (2013) about challenges with recruiting participants from diverse populations. As many representatives and physicians did not respond fast, the first author followed up repeatedly until enough physicians were interviewed to reach saturation, i.e., until we were not discovering any new themes. We reached saturation at 17 interviewees.
The physicians (eight females and nine males) who agreed to participate spent between 1 and 24 years in Sweden. All respondents got their medical education and some working experience in their home countries before coming to Sweden. Nine respondents came from the EU/EEA countries such as Hungary (five physicians), Germany (two physicians), Poland and Finland (one physician per country). Eight physicians came from countries outside the EU/EEA area such as Iran (five physicians), Iraq, Russia and the former Yugoslavia (one physician per country).
Their medical specialties included anaesthesiology, radiology, geriatrics/rehabilitation, general surgery, ophthalmology, and general medicine. Work experience in Sweden ranged from 1.5 to 19 years. All the participants from EU/EEA countries came to Sweden under the recruitment program of the Region Västra Götaland. Since the European medical licenses were automatically approved due to EU/EEA regulations, they started working shortly after coming to Sweden, underwent a three-month Swedish language course for medical professionals, and passed a dedicated language test. The respondents from outside the EU/EEA started working after they completed their medical proficiency test and a mandatory internship. The test which is also referred to as "kunskapsprovet" proves that one has the clinical and linguistic knowledge to perform as a physician in Sweden (for details about the different procedures for EU/EEA and non-EU/EEA physicians, see Salmonsson &Mella, 2013, andSturesson et al., 2019).
As one can see, the participant group was diverse in terms of length of stay in Sweden, specialties, gender, cultural backgrounds, and opportunities for entrance into the Swedish labour market. The radiologists experienced less patient contact compared to ophthalmologists, and the physicians working in geriatrics met primarily elderly patients and their relatives/caregivers. The respondents who spent a long time in Sweden reported more experience with patient communication. Further, the non-European participants, compared to the European physicians, had to go through more challenges in obtaining a Swedish medical license. All these background factors could have affected the respondents' experiences of patient communication in general, and patient bias in particular.
The interviews were primarily conducted in Swedish. English, Russian, or German were also used in certain instances to ensure a correct understanding of the responses. The interviews were conducted by the first author in the places the participants were comfortable, such as private hospital rooms and participants' homes. The total interview time was 28.5 hours. Interviews were audio recorded after obtaining written consent from the participants before and after the interview. The interview guide included open-ended questions covering five areas: 1. work experiences in Sweden and in their home country; 2. communication in the workplace with patients and colleagues; 3. experiences of patient bias, prejudice, and discrimination in the workplace; 4. mitigation of challenges; and 5. additional comments. We also opened up opportunities for participants to discuss other issues relevant to communication and patient bias during the interviews.
Data were analysed using thematic analysis (Braun & Clarke, 2006) with the support of NVivo12 software (QRS, 2018). The goal of thematic analysis is to recognise themes in one or more texts. A theme represents deeper, abstract, systematic patterns of information that emerge on the surface of a text with different strings of words, strings that are nonetheless assumed to convey a similar underlying information pattern (Erlingsson & Brysiewicz, 2017). The process that leads from the surface of the text to the abstract underlying themes is called coding and involves the identification of codes, namely relevant chunks of text vis-à-vis the goals of the research. Codes are then grouped together in an increasing level of abstraction through condensation, namely the "process of shortening the text while still preserving the core meaning together" (Erlingsson & Brysiewicz, 2017, p. 94). The process of condensation ends when the researchers recognise themes, namely clear and distinct systematic patterns that cannot be reasonably further grouped together (Erlingsson & Brysiewicz, 2017).
For the development of themes, we followed a hybrid deductive-inductive approach (Fereday & Muir-Cochrane, 2006;Girardelli et al. 2020). In the deductive step, we defined the initial drafts of the templates or codebooks, which contained some preliminary themes for each of the two research questions derived from the available literature. According to Fereday and Muir-Cochrane (2006), this approach is considered rigorous as it balances insights derived both deductively from the relevant literature and inductively from the raw data under investigation (inductive step). In particular, for RQ1 (forms of patient bias) we took into consideration relevant frameworks elaborated by Osseo et al. (2018) and Wheeler et al. (2019). For RQ2 (coping strategies) we used a scheme by Wheeler et al. (2019). In the inductive step, the transcripts were analysed using the initial codebook as a starting point. In an iterative manner, we recursively refined the codebooks by either adding additional codes and themes if not already included or eliminating preliminary themes if they were not identifiable or grounded in the raw data, namely the transcripts. This continuous process of refinement required a form of "researcher triangulation" which involves the two co-authors checking each other's perspectives and perceptions and solving disagreements as a way to add rigor and depth to the thematic analysis (Erlingsson & Brysiewicz, 2017;Morris, 2017).
After the initial analysis of the interviews, a two-hour interactive workshop was organised at the university premises by the project team. The migrant physicians who participated in the study were invited to join and share their experiences of working in Sweden. The purpose of the workshop was twofold. The researchers aimed at getting feedback on their findings from the interviews. Another purpose was to provide opportunities for the participants to discuss their experiences with each other, network, and learn about patient bias.
Eight physicians joined the workshop. First, the researchers presented the initial findings from the interviews, and the participants were asked to discuss them in pairs and give their comments. Next, the physicians continued brainstorming on how the challenges with patient bias could be handled in practice. The workshop was finalised by a joint discussion on communication challenges in multicultural health care in general, and on patient bias in particular. The workshop was recorded using field notes, which contributed to obtaining a more in-depth understanding of the experiences of the physicians described in the interviews. Table 1 summarises the eight themes that emerged from the thematic content analysis. Four themes pertained to forms of patient bias experienced by the physicians and four themes to coping strategies employed by the physicians to deal with patient bias.

Forms of Patient Bias
In general, respondents reported to perceive Swedish patients as polite, friendly, kind, and quiet. Swedish patients also appear to avoid direct confrontations and conflicts. Nonetheless, two respondents mentioned that some patients were open about their refusal to be treated "by us, foreigners, and asking for a Swedish doctor instead" [Iraqi physician] (Theme 1). The respondents perceived that those patients most often express refusal to meet a migrant physician indirectly, talking to their colleagues, e.g., nurses who book appointments, assistant nurses, and other physicians. According to the respondents, often the patients get worried when they hear a foreign name. A physician originally from Russia commented: Some patients refuse to be treated by us, foreign doctors, but they rarely tell us directly. They tell the nurse before consultation and ask if there is a Swedish doctor available instead. Then that nurse or another colleague might mention it to me during lunch or coffee break. The second form of patient bias experienced by the respondents dealt with patients questioning the physicians' Swedish language skills, often even before having met them (Theme 2). Elderly patients, often with self-reported hearing problems, and their relatives were especially concerned, fearing potential problems with understanding "broken Swedish" [Iraqi physician]. Though respondents acknowledged that "a foreign accent could be difficult to understand for older people" [Hungarian physician], they still felt being singled out. Respondents who spent a long time in Sweden felt that, despite speaking fluent Swedish, passing medical license exams, and having an extensive professional experience in Sweden, they were still regarded as "not good enough" [German physician] and treated as "second-rate healthcare providers" [Hungarian physician].
Three respondents mentioned their patients showing impatience when they were "trying to find the right word or formulate what we want to say" [Russian physician] and made derogatory comments such as "have you taken any Swedish language courses at all?" [physician from former Yugoslavia]. However, also in this case some patients took an indirect route to express their dissatisfaction: They sat through a consultation and then "talked behind my back to my colleagues that I can't speak Swedish" [Hungarian physician]. Patients also filed formal complaints after the consultation and our respondents felt stressed out about never really being able to know what their patients are thinking "as the Swedes don't like showing their emotions openly" [Iraqi physician]. The physicians emphasised that, in some consultations, making just a small language mistake, saying that you do not understand what your patient says, or having a small misunderstanding during an interaction can have devastating consequences because many patients believe that "speaking good Swedish means being a good doctor, and speaking poor Swedish means being a bad doctor" [Finnish physician]. The Finnish physician commented further: Because of failing language, a patient questions the level of medical knowledge. One mixes them up. Someone who doesn't know the language can't be good at medicine either. That is why I never show that I don't understand. I tried to listen and, if there is a nurse present during a consultation, I ask her for explanations right after the consultation.
Patient bias was also expressed by questioning migrant physicians' medical competence (Theme 3). Respondents from countries outside the EU/EEA area reported this form of bias more frequently. A physician originally from Iran explained that "doctors educated in poorer countries" were perceived by their patients as not competent enough and "not able to provide proper patient care in a developed country like Sweden." A senior physician from Iraq, with over 10 years working experience as a surgeon in several countries, mentioned that a patient once asked him about working in his home country: Once a patient asked me, "do you perform surgeries in operating rooms in your country?" I was shocked. I tried to make joke out of it and said, "no, we simply put patients on the floor and cut them open!" The patient got scared, as he didn't understand that it was a joke. He thought I was serious about it. I had to apologise and tell him that it was not true. I am not sure if he believed me or not.
The fourth form of patient bias experienced by our respondents involved ethnic slurs and jokes cracked by patients even during consultations (Theme 4). For example, a physician commented that he was accused of being "an authoritarian German oppressor" when he refused to give a patient a sick note; a Russian physician was asked about her affiliation to the Communist Party; a female Iranian physician was "sick and tired of the questions why I don't cover my hair." Even if respondents acknowledged that those comments were rarely hostile and harsh, they still felt like being singled out. As a respondent from the former Yugoslavia commented: "I am a foreigner here, I am a refugee, and I still feel it every day. Especially when they [the patients] ask me about the reasons why I came here, about the war, what I am doing here in Sweden."

Strategies for Coping with Patient Bias
The vast majority of our respondents indicated that their most commonly adopted strategy when they experienced forms of patient bias was confrontation avoidance (Theme 5) interviewed physicians reported to have actually confronted patients, fearing that this may escalate conflict with both patients and their relatives. Some respondents also mentioned that they were unsure if their colleagues would take their side in case of direct confrontation, worrying of potentially being described as inflexible, too sensitive, and overemotional. A Finnish physician commented: You may not show that you are sad or angry. Sometimes, it is difficult just being quiet, even if you are treated unfairly. If you complain too much, your colleagues will not respect you. You will be singled out even more. Even if I am offended and hurt, I still treat my patients with respect and give them the treatment they need.
Some physicians admitted that they avoided approaching senior physicians and managers in their units, fearing that complaining about patients would have made things even worse and turned their colleagues against them. Uncertainty was also due to a lack of knowledge about laws and regulations about discrimination in the workplace in Sweden, which was mentioned primarily by the respondents who spent shorter time in Sweden. For instance, a respondent commented that he did not know so much about "the Swedish anti-discrimination law and he was not sure where to find the information or whom to ask" [Hungarian physician]. The same respondent mentioned that he really wanted to show that he was a good doctor and avoided complaining about anything. In particular, those physicians who did not have permanent contracts explicitly mention that the best approach was "flying below the radar" [Polish physician], in fear that a more direct approach could have a negative impact on their career development.
If the most common way to deal with patient bias was to simply turn a blind eye, some respondents tried to implement a more proactive approach, for instance by collaborating with the Swedish staff and in particular nurses (Theme 6). In many cases, nurses are, in fact, responsible for finalising a consultation appointment and handle initial inquiries and concerns from patients and relatives. One of the respondents described nurses as "the gatekeepers who have much power in Swedish healthcare" [Iranian physician]. A Columbian physician explained: Some patients are unsure… a foreign name you know, and they ask the nurse "Is this doctor good?" Most nurses say, "Yes he is good!" The nurse often tells me "This or that patient asked if you were good." Thus, you know in advance that the patient was worried and unsure. Therefore, you are prepared.
Nurses can aid during consultations, helping with language problems, e.g., in finding a word, providing clarifications, and additional explanations to patients. Respondents mentioned that being humble and asking nurses for help is essential for getting their support with patients. Some respondents referred to the so-called "Law of Jante" (Cappelen & Dahlberg, 2017), a modesty code that captures the Scandinavian mentality according to which "one should never try to be more, try to be different, or consider oneself more valuable than other people" (p. 419). Following the "Law of Jante", even if physicians have on paper a higher rank and expertise than nurses, they should nonetheless avoid "show off and tell that you are good" [Finnish physician] if they want to forge alliances with the nurses. However, nurses can also play the role of foes against migrant physicians, openly questioning their authority and decisions, sometimes in front of patients, relatives, and staff, and contributing to exacerbating patient bias and distrust: Nurses can be enormously helpful and can help you with everything in contact with patients. At the same time, they can completely destroy your relationship with patients and relatives. If they like you, trust you, and see that you are a competent doctor, trying to learn Swedish, they will help you. If not, in the best case, they will leave you by yourself. In the worst-case scenario, they will be against you [physician from the former Yugoslavia].
Another strategy employed by migrant physicians to deal with patient bias was selfdisclosure (Theme 7), for instance by sharing with the patients where they come from, the reasons for moving to Sweden, family and children, and their future plans. According to the respondents, self-disclosure can be strategically used to make patients become less suspicious. For instance, a Hungarian physician explained: "I told my patient about my education, my family, my career plans, and why I came to Sweden. I felt that the patient became more relaxed and felt more secure." Finally, some respondents mentioned active listening as a successful strategy for overcoming patient bias (Theme 8). An Iranian physician commented: I spend a lot of time with patients. I really try to understand their problems in detail and to help them. I know that I am being judged. They feel unsure if they feel I am not able to understand what they are talking about. That is why I listen well.
Active listening is also displayed, when possible, by taking extra time during consultations with patients and relatives to make sure that everything is clear. According to a Finnish physician: Patients measure time in factual time. If you take 10 to 20 minutes more of your time, then the patient perceives that you have devoted yourself 10 to 20 minutes more than another physician. That's why I think many patients say: "Yeah, my doctor is so thorough, this doctor really listens."

Discussion
This study aimed at exploring how migrant physicians in Sweden experience patient bias in their professional practices and which strategies they employ to cope with patient bias. Our findings are in line with the literature on patient bias (Chandrashekar & Jain, 2020;Espaillat et al. 2019;Osseo-Asare et al., 2018;Wheeler et al., 2019) and show that migrant physicians in Sweden also experience a variety of forms of discriminatory and racist behaviours.
Echoing previous empirical investigations (Ahmad et al., 2022;Espaillat et al. 2019;Kay et al., 2022;Osseo-Asare et al., 2018;Wheeler et al., 2019), our data points to the relevance of Sue et al. (2007)'s microaggression theory to study patient bias (Mazzula & Campón, 2018). Our respondents reported explicit microassaults in the form of refusal of care (Theme 1) and more subtle microinsults that involved questioning their competence, both linguistic (Theme 2) and medical (Theme 3), as well as ethnic jokes or stereotypes (Theme 4). At the same time, we were not able to find evidence of microinvalidations, namely dismissal of the respondents' feelings and thoughts vis-à-vis forms of discrimination and racism. Microinvalidation is recognised as the most difficult to pinpoint among the different types of microaggressions (Walker et al., 2022). Our focus on patient-doctor interactions may also provide a reason: if we had considered a broader range of interactions in medical institutions, in particular with colleagues and supervisors, microinvalidations may have more likely emerged, for instance in a situation in which a Swedish manager gives the cold shoulder to a migrant physician vent their frustration for a case of patient-bias (Bullock et al., 2021).
In line with research on patient bias with US samples (Marcelin et al., 2019), our data also suggest that patient bias in Sweden tends to be primarily expressed in a more indirect and covert manner. However, the fact that patient bias takes an indirect route does not mean that it is less damaging for the physicians' well-being. Our respondents appeared, in fact, to "walk on eggshells" in their professional practice and to deal with an additional level of stress (see Kaltiso et al., 2021), as they need to second guess conversations with their patients, overinterpret silence, and manage unexpected backstabbing, such as receiving a complain while no clear signs of dissatisfaction communicated during or in connection to consultation. This finding indicates that the immediate context of cultural embeddedness influences physicianpatient communication. Scholars (Mazzula & Campón, 2018;Torino et al., 2018) called for more research in different national and cultural contexts as most of the empirical literature on microaggressions has been conducted in the United States and Canada; hence, the findings of our study based in a Swedish context are also relevant in moving toward a cross-cultural perspective on microaggressions.
The findings present a range of strategies that migrant physicians in Sweden employ to deal with patient bias (Chandrashekar & Jain, 2020;Osseo-Asare et al. 2018;Wheeler et al., 2019). Collegial support appears to play a pivotal role for the integration of foreign healthcare workers (Eriksson et al., 2018;Heponiemi et al., 2018). Fostering effective alliances with Swedish nurses seems particularly effective in mitigating patient bias. Nurses are often called "patient advocates'' (Gerber, 2018), as in their role they mediate between patients and physicians. Nurses often spend more time with patients than physicians do and support patients and relatives in communication, making their voices heard and defending their rights (Ersoy et al., 1997). The present study additionally highlights the positive role of nurses when migrant physicians experience challenges in communication with patients, making a successful migrant physician-nurse collaboration a crucial prerequisite for mitigating prejudice and enhancing workplace inclusion (Motala & Van Wyk, 2019).
The findings also provide a different perspective on the value of physician self-disclosure in communication. While previous research showed limited or no impact of physicians' selfdisclosure on trust in physician-patient relationship (Arroll & Allen, 2015;McDaniel et al., 2007), our study shows that in intercultural consultations, self-disclosure can, instead, can make patients more comfortable, positively contributing to physician-patient relationship and lowering chances of patient bias. Still, although self-disclosure can represent an effective strategy against patient bias, we also found that some of our respondents perceived talking about their own, sometimes painful backgrounds as refugees, with their patient as too sensitive.
Finally, communication and in particular listening skills can contribute to overcoming patient bias and fostering physician-patient relationships based on trust (Alpers, 2016;Thom, 2001). As Swedish patients might experience higher uncertainty and an increased risk of misunderstanding in their interactions with a physician being a foreigner, the importance of active listening is additionally enhanced in intercultural medical consultations (Berbyuk Lindström, 2008).
Although our respondents reported a range of possible coping strategies, it is important to point out that the most frequently reported strategy is conflict avoidance. Guidelines on how to respond to patient bias (Chandrashekar & Jain, 2020;Paul-Emile et al., 2016;Rosoff, 2018)  also include a similar option but within a larger context in which physicians are called to assess the situation based on an investigation of the patient's perspective and to pick the most suitable coping strategy, including more assertive ones, such responding to demeaning comments and pointing the patient to other healthcare institutions. Based on our available data, our respondents seem not only to refrain from strategies that are more assertive but also to make decisions on how to handle patient bias on a trial-and-error basis, instead of referring to systematic guidelines. It may be understandable that the use of assertive strategies in conflict avoidance cultures like the Swedish one can be problematic and not straightforward (Holmberg & Åkerblom, 2007). Further, avoiding conflicts and even mentioning to colleagues or management that one experiences negative attitudes from patients can be explained by fearing that complaints can backfire and harm one's reputation more than help, negatively influencing career advancement (Chandrashekar & Jain, 2020, Heponiemi et al, 2018.

Practical Implications
Despite the explorative nature of the present study, some practical implications can be drawn from our findings for both immigrant physicians, who intend to address patient bias effectively, and healthcare institutions, which strive to foster better work environments for their increasingly diverse workforce. Also inspired by the literature on microaggressions in healthcare (see for instance Miller & Chen, 2022; Torres at al., 2019), we feel that an effective response to patient bias requires a wholistic approach that ranges from the individual to the institutional level. First, developing a policy that explicitly addresses patient bias as suggested by Kimani et al., (2020) and Chandrashekar and Jain (2020) is essential. Healthcare personnel, both potential targets and bystanders, should be systematically trained on (1) their rights and responsibilities as caregivers and employees and (2) whom to contact and how to respond in case of experiencing or witnessing demeaning patient behaviours. The availability of a common, accepted framework with structured responses will allow healthcare personnel to respond with confidence to occurrences of patient bias (Kimani et al., 2020). Managers should also create arenas for and initiate discussions about different types of patient bias in the workplace. Though these conversations might be difficult for some employees, they are important for raising awareness about such sensitive issue as patient bias.
Second, patients should be informed of anti-discriminatory guidelines and that diminishing/discriminatory behaviour is unacceptable. The information can be provided through leaflets and posters on-premises, websites, etc.
Finally, partnership and closer collaboration between migrant physicians and the local staff, in particular nurses, should be systematically fostered and supported by management with the aim of providing targeted additional support for migrant physicians to enhance their ability to engage with the patients effectively.

Conclusion
Migrant healthcare workers play an increasingly central role as healthcare systems in highincome countries, which necessitates finding effective solutions to support their integration in health care workplace (WHO, 2016). Patient bias represents a challenge that needs to be addressed to encourage inclusivity in healthcare workplaces and to promote successful patientphysician communication. Addressing patient bias should not be left in the hands of the individual migrant physicians with the risk of promoting a culture of silence and denial; instead, there is a need to set a supportive environment, clear guidelines, and specific procedures for how biased behaviour by patients should be effectively managed. Systematic research on patient bias can hence support healthcare institutions in balancing "their duty to provide highquality care and tend to the vulnerability of patients, with their responsibility to cultivate a supportive, respectful work environment" (Chandrashekar & Jain 2020, p. S41).

Limitations and Future Research
The present study represents an initial investigation of the phenomenon of patient bias in in a particular context, Swedish health care. This study has also gained the perspectives of migrant professionals only. Gaining the perspectives of Swedish counterparts, such as patients, colleagues, and managers, would help broaden the understanding of patient bias. The study is also based on self-reports and does not provide evidence on the actual effectiveness of the strategies to deal with patient bias mentioned by the respondents. Future research that, for instance, employs participant observations and an experimental design can shed light on what strategies are better suited to respond to specific types of patient bias in practice.