Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Implicit and explicit ethnic biases in multicultural primary care: the case of trainee general practitioners

Implicit and explicit ethnic biases in multicultural primary care: the case of trainee general... Background: General Practitioners (GPs) are the first point of contact for people from ethnic and migrant groups who have health problems. Discrimination can occur in this health care sector. Few studies, however, have investi- gated implicit and explicit biases in general practice against ethnic and migrant groups. This study, therefore, investi- gated the extent of implicit ethnic biases and willingness to adapt care to migrant patients among trainee GPs, and the factors involved therein, in order to measure explicit bias and explore a dimension of cultural competence. Methods: In 2021, data were collected from 207 trainee GPs in the French-speaking part of Belgium. The respond- ents passed an Implicit Association Test (IAT ), a validated tool used to measure implicit biases against ethnic groups. An explicit attitude of willingness to adapt care to diversity, one of the dimensions of cultural competence, was meas- ured using the Hudelson scale. Results: The overwhelming majority of trainee GPs (82.6%, 95% CI: 0.77 – 0.88) had implicit preferences for their ingroup to the detriment of ethnic and migrant groups. Overall, the majority of respondents considered it the responsibility of GPs to adapt their attitudes and practices to migrants’ needs. More than 50% of trainee GPs, however, considered it the responsibility of migrant patients to adapt to the values and habits of the host country. Conclusions: This study found that the trainee GPs had high to very high levels of implicit ethnic bias and that they were not always willing to adapt care to the values of migrants. We therefore recommend that they are made aware of this bias and we recommend using the IAT and Hudelson scales as educational tools to address ethnic biases in primary care. Keywords: General Practitioner, Discrimination, Ingroup relationship, Migrant health, Racial bias, Cultural competence, Implicit association test equity of the health care system. Nevertheless, even GPs Introduction who are well-intentioned may be vulnerable, like anyone As aspiring health care professionals, trainee General else, to implicit biases and may lack willingness to adapt Practitioners (GPs) care for a diverse and multicultural care to migrants, a willingness that is a prerequisite of population in their practice [1] and are usually the first cultural competence [2]. This could lead GPs to discrimi - point of contact for ethnic and migrant patients within nate unintentionally against their ethnic and migrant the health care system. They are, therefore, key to the patients [3, 4] especially if they are under time pressure, lack solid knowledge/information needed to make a decision, or are affected by cognitive overload or fatigue *Correspondence: camille.duveau@uclouvain.be Institute of Health and Society, UCLouvain, Brussels, Belgium [5], or due to inappropriate intercultural contact [6] less Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visithttp:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Duveau et al. BMC Primary Care (2022) 23:91 Page 2 of 10 concordant physician–patient ethnicity, which is associ- has been paid to the North African population. This ated with less favourable patient ratings care [7], stereo- applies particularly to people of Moroccan descent, who typing, prejudice, political attitudes [8], or organizational constitute the largest and fastest-growing minority group and institutional factors [9]. Implicit ethnic biases among in countries in the European Union (EU), including Bel- health care professionals and cultural competence are gium, where the Moroccan population doubled between important factors that contribute to health care dispari- 1991 and 2014. This ethnic group is also dispropor - ties [10, 11]. tionately at risk of poor health outcomes, such as more In social psychology, a “bias” is defined as a prejudicial depressive symptoms [18] and higher mortality caused by attitude towards a group (hereafter, the exogroup) and its diabetes and infectious diseases [19]. members, of which the holder may or may not be aware. To the best of our knowledge, few studies have inves- Implicit bias is often automatically activated and can tigated the extent of bias among general practitioners lead to impaired judgement and to negative evaluation of against racial/ethnic groups in relation to GPs’ willing- the exogroup [12]. Implicit ethnic bias has been shown ness to adapt to diversity. A recent study highlighted the to play a role in differential recommendations by physi - need to increase recognition and awareness of ethnic cians for managing disease [10], disparities in empathy disparities in health and health care among health care [13] and differential drug prescriptions, interaction with professionals [20]. This paper aims to expand this area of patients, and treatment decisions [10]. research by investigating (a) the level of implicit ethnic To address these differences in care for ethnic/migrant bias among trainee GPs and (b) whether this level of bias groups, the report Unequal Treatment by the Institute varies as a function of a GP’s willingness to adapt care to of Medicine recommended exploring healthcare provid- migrants or as a function of openness to cultural diversity ers’ conscious (explicit bias) and unconscious percep- which, in this paper, includes intercultural contact, politi- tions [14]. This report also recommended developing and cal opinions, and workload (time pressure). enhancing cultural competence training for health care professionals. Over the past decade, many cultural com- Methods petence programmes have been initiated in health care Design & participants systems to improve the relationships between health care This prospective quantitative study used an observa - professionals and patients with diverse cultural back- tional design. The design measures implicit ethnic biases grounds, and to enable them to work and interact more and the specific explicit attitude of willingness to adapt effectively with those patients [11]. These programmes, to diversity, a prerequisite of cultural competence. To however, are based on the assumption that health profes- encourage trainee general practitioners to respond to the sionals are motivated and accept responsibility for adapt- online survey, we offered to present the overall results ing to ethnic diversity [15], an assumption that may need to them anonymously and give them an opportunity to to be verified. This study aims to add to the literature by comment on and discuss the results in small groups. measuring implicit and explicit biases against ethnic/ Between February and March 2021, we contacted 220 migrant groups. French-speaking Belgian GPs, who were in the second A large body of research has investigated unconscious year of their three-year internship by email in one of the biases and levels of cultural competence among health important French-speaking universities in Belgium. Of care professionals. This research found significant evi - those trainees, 207 completed the online survey (a par- dence of implicit bias among care professionals against ticipation rate of 94%) (Male = 68 (33.0%); Female = 139 African Americans, ranging from slight bias to high lev- (67.0%); Mean age = 26.7 years (± 2.1); See Table 1). els of bias, but found low levels of explicit bias against this ethnic group. The research also found a weak asso - Measures ciation between self-perceived cultural competence and Implicit association test levels of implicit bias [10, 16]. Little is known, however, The Implicit Association Test (IAT) is a validated tool about these issues within the European context and a in that is used, in social psychology, to assess implicit recent study comparing clinicians’ racial biases in The biases, which can be associated with a wide range of United States and France concluded that health provid- behaviours and attitudes [21]. As bias is often activated ers’ biases differed between cultures and countries [17]. by situational cues such as a first name [3 ], respondents So far, most health care studies have been carried out took an IAT with French-language and North African in the United States and have compared African Ameri- can patients with White patients. There is a lack of data on unintentional ethnic and racial discrimination among Cf. Myriatics. (2015, décembre). Myria. https:// www. myria. be/ fr/ chiff res/ general practitioners in Europe and too little attention myria tics D uveau et al. BMC Primary Care (2022) 23:91 Page 3 of 10 Table 1 Demographic characteristics, implicit bias, and cultural D-scores according to the strength of the implicit asso- competence score of participants (n = 207) ciations: “neutral” (< 0.15), “slight” (< 0.35), “moderate” (< 0.65), and “strong” (≥ 0.65), and the reverse for scores Characteristics % or mean (std) with a negative value [22]. Age (years) 26.7 (2.1) Sex female 67.0 3 Hudelson scale Type of practice The trainee GPs completed the validated Hudelson scale, Solo 27.7 which assesses one dimension of cultural competence: Group 69.4 the perception of relative responsibility for adapting care Mixed 2.9 to migrant patients [2]. This was assessed using descrip - Ethnicity tions of five potential situations: (1) when migrants’ val - White 78.5 ues and habits differ from those of the host country, (2) Arab 8.3 when the patient does not speak the language of the host African 3.9 country, (3) when the patient expresses the wish to be Mixed or other 9.3 treated by a male/female health professional, (4) when Religion the patient cannot read the language of the host coun- Don’t want to reply 4.4 try, and (5) when the patient’s health beliefs contradict Catholic 39.1 the knowledge of the health professionals. The trainee Muslim 10.1 GPs had to state their opinion about whose responsi- Atheist 31.9 bility it is to adapt in those situations by ticking a score Other 14.5 between 1 (it is the professional’s responsibility) and 7 Political opinion (it is the migrant’s responsibility), with 4 meaning that Don’t want to reply 10.6 the responsibility is shared between the patient and the Right-wing liberal 29.5 professional. Centre 31.9 Regarding openness to cultural diversity, we expected Left-wing (e.g. socialist) 28.0 left-wing political opinions, the proportion of migrants Practice area among the patients GP cared for, more social interactions Urban 7.8 with people of North-African origin and a lighter work- Suburban 33.1 load to be associated with less unintentional discrimi- Rural 25.3 nation [23]. To compute the frequency of intercultural Missing values 33.8 contact, we created a score composed of four variables: Frequency of contact with North Africans (/4) 2.5 (1.2) (1) presence of North-African people in their neighbour- Workload (no. of patients/day) 10.2 (5.6) hood, (2) daily interaction with North-African people, (3) Proportion of patients of foreign origin (/100) 33.0 (30.4) having North-African friend(s), and (4) having North- Implicit Association Test score (-2, 2) 0.53 (0.44) African colleague(s). We obtained a score out of 4 points Hudelson score (5, 35) 18.9 (5.1) (where 0 means that the respondent has no contact with North Africans and 4 means that the respondent fre- quently interacts with North Africans). Control variables were added to the models, such as age, sex, ethnicity, first names. The IAT was designed to measure implicit and type of practice (e.g. solo, group, or mixed), to avoid ethnic bias. The IAT with North African first names potential selection bias. uses reaction times to assess the strength of automatic associations between target pairs, each consisting of an Reactions and feedback sharing ethnicity (e.g. French-language or Belgo-Moroccan first The trainee GPs were presented with feedback on the names), and a category (e.g. positive or negative). This aggregate results and then asked to comment on the test provides a D-score ranging from -2 (respondent results. We collected 204 trainee GPs’ anonymized open- has implicit associations of positive words with North ended reactions to their IAT and Hudelson scale scores. African first names) to + 2 (respondent has implicit We then divided them into small groups of 4–5 to discuss associations of positive words with French-language the experience and explore possible solutions. first names). We used standard cut-offs to classify the 2 3 see Appendix 1. see Appendix 2. Duveau et al. BMC Primary Care (2022) 23:91 Page 4 of 10 Fig. 1 Distribution of implicit associations for first names associated with different ethnic groups Statistical analysis Results Quantitative analysis Participants First, we computed descriptive statistics of variables. Two hundred and seven General Practitioners in the sec- We computed correlations between the level of implicit ond year of a three-year specialisation in general medi- bias and the Hudelson score to check whether they over- cine responded to the online questionnaire. Of those lapped. Then, linear regression was performed to relate respondents, two hundred and four took part in the dis- the two outcomes (implicit and explicit attitudes) with cussion about the presentation of their results from the the following variables: political opinions, practice area, IAT and the Hudelson scale. Most of the participants frequency of intercultural contact, workload, and propor- were of Belgian nationality (89.0%). Participants self- tion of patients of foreign origin, controlling for variables reported as White (78.5%), Arab (8.3%), or of mixed/ such as sex, type of practice, and ethnicity. We ran one other ethnicity (9.3%). A third of the respondents self- model for each of the following explanatory variables: reported as Catholic (39.1%). Nearly 70% of the respond- political opinion, frequency of intercultural contact with ents were not very religious or not religious at all. In North Africans, frequency of patients of foreign origin in terms of political opinion, the majority selected Centre consultations. Statistical analysis was performed using (31.9%) and 29.5% of the sample identified as right-wing. SAS 9.4. Most worked in suburban areas (33.1%). A quarter of the respondents worked in rural areas. A minority (7.8%) worked in the seven largest cities in the Wallonia-Brus- Reactions and feedback sharing sels Federation (Brussels, Liège, Verviers, Namur, Char- With the aid of NVIVO, we grouped similar quotes leroi, Mons, and Tournai). The area was computed based from the written reactions into main themes in order to on their medical practice’s postcode. explore the general reactions to the results and the pos- The average score for frequency of contact with North sible solutions that trainee GPs highlighted within the Africans was 2.5/4 (± 1.2), meaning that the respond- small groups. ent sometimes interacts with North African friends, D uveau et al. BMC Primary Care (2022) 23:91 Page 5 of 10 Table 2 Relative responsibility to adapt of migrants and health professionals, as perceived by trainee GPs (2021): % per item (n = 207) The responsibility The responsibility lies with The responsibility Total lies with the health both health professionals and lies with the professionals patients patients When migrants’ values and habits differ from those of 10.6 30.9 58.5 100 the host country (%) When the patient does not speak the language of 37.7 27.5 34.8 100 the host country (%) When the patient expresses the wish to be treated by 60.4 15.0 24.6 100 a male or female doctor (%) When the patient cannot read the language of the 50.7 21.3 28.0 100 host country (%) When the patient’s health beliefs contradict medical 46.4 26.1 27.5 100 knowledge (%) colleagues, or neighbours. In their practice, they visit professionals. The majority of trainee GPs considered 10.2 patients per day on average, 33% (± 30.4) of whom it to be mainly the health professionals’ responsibility are foreign patients. to adapt when the patient expresses the wish to choose the sex of the health care professional, when he/she can- Implicit association test not read the language of the host country, and when Figure 1 presents the distribution of implicit associations his/her health beliefs contradict medical knowledge. of positive/negative words with French-language first Two items, however, differed from this trend. Roughly names compared with North African first names. The 38% of respondents considered it the responsibility of IAT mean score (D-score) was 0.53 (± 0.44), indicating health care professionals to adapt to migrants when the a moderate association of positive words with French- patient does not speak the language of the host country. language first names over North African first names. The Approximately 35% considered it the responsibility of majority of respondents (82.6%, CI 95%: 0.77 – 0.88) had the patient to adapt and 27.5% considered it the respon- a positive D-score, i.e.an implicit association of positive sibility of both the professional and the patient to adapt. words with French-language first names. Several of the This result is quite surprising because, when it comes to scores for implicit associations stand out. providing written information in the patient’s language, Around 7% of the scores were negative, meaning that respondents placed the responsibility on the health care the trainees had an implicit association of positive words professional (50.7%). When the migrant’s values and hab- with North-African first names. Of the respondents with its differ from those of the host country, however, GPs negative scores, 4.3% were Arab, 3.4% were White, and considered it to be the patient’s responsibility to adapt 1.9% were of mixed/other ethnic origin. The scores varied (58.5%). significantly according to ethnicity (χ2 = 88.6, p < 0.0001). There was a significant small positive correlation In contrast, about 12% had a slight implicit association of between implicit ethnic bias and cultural competence French-language names with positive words, while 45.4% (r = 0.172, p < 0.05), suggesting that respondents for had a strong implicit association of French-language whom French-language names had more implicit posi- names with positive words. tive associations were also more likely to consider it the It is also worth noting that approximately 10% of the migrants’ responsibility to adapt. respondents showed no group preference. Table  3 provides the betas from the linear regressions for the factors associated with the IAT and with the Hudelson scale Hudelson score. Respondents’ ethnic origins influenced Table  2 presents the perception of the relative responsi- the level of implicit ethnic bias. Being of Arab ethnicity bility to adapt to migrant patients in health care among or of other or mixed ethnic origin reduced the strength trainee general practitioners. The five items of the Hudel - of implicit negative ethnic associations with North Afri- son scale were moderately correlated (Pearson coeffi - can first names (β = -0.54, p < 0.001, β =—0.29, p < 0.05 cients 0.09–0.46) and the internal consistency coefficient respectively). Having left-wing political opinions also (Cronbach’s α) was 0.62. The average score for Hudelson had a negative correlation with the IAT score (β = -0.16, scale was 18.9/35 (± 5.1), which means that adapting was p < 0.05). Nevertheless, other individual and contextual considered to be mainly the responsibility of health care variables such as age, sex, type of practice, practice area, Duveau et al. BMC Primary Care (2022) 23:91 Page 6 of 10 Table 3 Factors associated with Implicit Association Test and with the Hudelson score: betas from the linear regression models Implicit Association Test (-2, 2) Hudelson score (5–35) Covariates: β CI p-value β CI p-value 95% 95% Age (years) 0.00 (-0.03, 0.04) 0.93 0.03 (-0.09, 0.68) 0.13 Sex (ref = Male) Female -0.05 (-0.18, 0.08) 0.49 -2.02 ( -3.46, -0.57) < 0.01 Type of practice (ref = Solo) Group practice 0.07 ( -0.07, 0.21) 0.32 -1.24 ( -2.75, 0.28) 0.11 Mixed practice 0.01 ( -0.34, 0.37) 0.94 -2.68 ( -6.61, 1.25) 0.18 Ethnicity (ref = White) Arab -0.54 ( -0.76, -0.31) < .0001 -0.41 ( -2.90, 2.08) 0.75 African -0.08 ( -0.40, 0.23) 0.60 1.57 ( -1.96, 5.09) 0.38 Other or mixed -0.29 ( -0.51, -0.07) 0.01 -1.51 ( -3.96, 0.94) 0.23 Political opinion (ref = Right-wing) Centre 0.03 (-0.17, 0.12) 0.72 -0.36 (-1.97,1.24) 0.66 Left (e.g. Socialist) -0.16 (-0.32, -0.01) 0.03 -3.07 (-4.74, -1.40) < .001 Practice area (ref = Urban) Suburban -0.04 (-0.25, 0.16) 0.67 -1.92 (-4.30, 0.46) 0.11 Rural -0.05 (-0.26, 0.17) 0.65 -0.42 (-2.94, 2.09) 0.74 Frequency of contact with North Africans -0.05 (-0.10, 0.01) 0.07 0.13 (-0.49, 0.74) 0.69 Workload (no. of patients/day) -0.00 (-0.01, 0.01) 0.95 0.06 (-0.06, 0.19) 0.30 Proportion of patients of foreign origin -0.04 (-0.26, 0.17) 0.70 -0.37 (-3.00, 2.26) 0.78 R -range [0.16–0.24] [0.09–0.17] β are adjusted for age, sex, ethnicity and type of practice; bold coefficients have p-value < 0.05 Low scores indicate that the responsibility to adapt is considered to lie with the health care professionals, whereas high scores indicate that it is considered to lie with the migrants Table with all R-squared is in appendix 4 frequency of contact with North Africans, and workload “Very impactful. It reveals that even unconscious biases were not associated with the level of implicit bias. can exist. This test confirms how important it is to know Concerning the Hudelson score, being a woman that these biases exist in order to be able to fight them (β = -2.02, p = 0.01) was associated with considering it to consciously in our practice…” (Participant 165). be mainly the health professional’s responsibility to adapt We found few individual or contextual explanations to migrants. Respondents with left-wing political opin- in the regression analysis of implicit biases, aside from ions also had a significant tendency to place the respon - ethnicity. The influence of their ethnicity is also dem - sibility on health care professionals (β = -3.07, p < 0.01). onstrated in the comments: “our culture played an Age, type of practice, having liberal political opinions, unconscious bias in this test.” (Participant 36). practice area, frequency of contact with North Africans, Unlike the quantitative results, the association workload, and frequency of patients of foreign origin between political opinions and propensity to place the were not associated with considering it the health care responsibility for adapting to migrants on health pro- professional’s responsibility to adapt to migrants. fessionals was not addressed in the comments. Only The R ranged between [0.16–0.24] and [0.09–0.17] one comment referred to political opinions: “Surprised for the IAT score and the Hudelson score, respectively. by the 60% [who said that the migrant should adapt to Appendix 4 presents the details of the R for each factor. the values of the host country] … we are GPs not politi- cians… the main thing is the patient’s health.” (Partici- pant 132). Reactions and feedback sharing They also shared solutions with one another to imple - From the comments, it appears that some of the GPs ment in their practice as GPs. They suggested keeping the were a bit sceptical about their implicit ethnic associa- Hudelson scale in mind in their everyday practice. It will tion scores and questioned the method and the exist- help them to remember that in most situations they bear ence of their own biases. A number of trainee GPs also responsibility for adapting to migrant patients. GPs also highlighted the need for them to be aware of biases: D uveau et al. BMC Primary Care (2022) 23:91 Page 7 of 10 emphasized the importance of being conscious of one’s people than White respondents [16, 25]. The authors, own biases in order to fight against them. however, also found lower levels of ethnic bias among Finally, several trainees proposed “cultural openness” women, which was not the case with our results. as a way to reduce the effects of these biases and improve Our results suggest that completing an IAT and dis- the quality of care, such as working in an intercultural cussing the results could help to reduce the negative team, learning several languages and learning about dif- impact of bias on health professionals’ attitudes. Indeed, ferent cultures. it could reduce both implicit and explicit bias among health providers with higher levels of implicit bias and thereby improve clinical interaction and verbal commu- Discussion nication and make practice more patient-centred [26]. Main findings This study set out to assess the trainee GPs’ implicit biases and their willingness to adapt care to diversity. Hudelson score Our results indicate that the majority of our trainees For three out of five situations, the majority of trainee had a moderate to strong level of implicit bias favouring GPs were willing to bear the responsibility to adapt care French-language first names over North African ones. to migrants. They were more reluctant, however, to adapt This result applied to all sociodemographic characteris - to migrants’ values and habits. These findings are in line tics, except ethnicity and political opinion. On the whole, with those of Hudelson, Junod Perron [2]. In the respond- trainees were willing to adapt care to migrants in most ents’ feedback, we noticed that some were surprised by care situations, with two exceptions: when a patient did this result. One of the respondents pointed out that they not speak the language of the host country and when the were GPs and not politicians. Our statistical analysis pro- immigrants’ values and habits differed from those of the vides some explanations: adaptation is related to politi- host country. This result applied to almost all sociodemo - cal orientation and the topic of values/habits is certainly graphic groups. a political issue, also from the GPs’ perspective. In a less competitive market, health care professionals may also Implicit association test feel less responsible for adapting to the values and habits The literature has often reported implicit associations of their patients as patients are free to move to another with ingroups (and biases against exogroups) that are GP if they are not satisfied. One might thus speculate that stronger than those found in this research [16]. Our the decrease in the percentage of GPs among all doctors study, however, does not support Allport’s finding that in Belgium over the last decade has enhanced the market reduced implicit bias is associated with higher frequency power of GPs [27]. of positive contact with the outgroup [23]. One possible Paradoxically, although more than half of the respond- explanation for this is the fact that we only assessed the ents see it as the health professional’s responsibility to frequency of contact with North Africans, not quality adapt when the patient cannot read the language of the of that contact. Furthermore, the p-value for the vari- host country, opinion was spread evenly across the board able “frequency of contact with North Africans” in the when the patient cannot speak it. While Dauvrin et  al. regression (Table  3) was 0.07, which suggests that either found that physicians took responsibility for providing this measure needs to be assessed more accurately, for interpreters [28], our respondents disagreed with that. example, by including an assessment of the quality of One explanation could be the extra time and resources trainee GPs’ relationships with members of the outgroup, needed to provide an interpreter during an everyday con- or interaction with outgroup members does not reduce sultation when GPs are often already under time pres- implicit bias among trainee GPs. sure. Moreover, access to intercultural mediators remains One important finding was that biases affect everyone limited in primary care. and few protective factors exist in GPs’ working envi- A greater willingness to adapt was found among female ronments, apart from their own ethnicity and political GPs and those with left-wing political opinions. This is opinions. This result has been demonstrated previously in line with previous research which found that female by Nosek, Banaji [24], who found a relationship between providers were more engaged in patient-centred commu- right-wing ideology and stronger implicit bias. They also nication and that African-American patients were more concluded that bias could vary according to “the con- likely to be seen by female GPs [2]. Another explanation straints that culture imposes on individual attitudes”, could also be that women tend to be more altruistic than which could explain the influence of ethnicity on the IAT men, when altruism requires more resources. We could score. We found similar results in several studies, which therefore speculate that diversity might be considered showed that Black providers also had significantly weaker as a form of altruism which demands more equalitarian ethnic implicit associations for both White and Black Duveau et al. BMC Primary Care (2022) 23:91 Page 8 of 10 behaviours, behaviours more likely to be found among competence. The debriefing on the results helped to women too [29]. raise awareness and sensitize the trainee GPs to the The combination of the findings from the IAT and the potential impact of their implicit and explicit biases Hudelson scale provides some support for the suggestion in multicultural consultation. They also became aware that there is a need to increase recognition and aware- that these biases could have an impact on their rela- ness of implicit ethnic bias and willingness to adapt to tionships and the way they care for patients. diversity in order to become more culturally competent. One implication of this research is that it should be The weak correlation between the two concepts sug - possible to use the results as a basis for reflection and gests that these are two different issues, requiring differ - discussion that could motivate health care profession- ent interventions. The R was also weak for both the IAT als to tackle their ethnic biases. Another implication is scores and the Hudelson scores, meaning that our model the possibility of using the Hudelson scale as a tool to did not fully explain the variation of these scores. Despite debate the extent of cultural competence among future increased interest in providers’ biases in the literature GP. This debate could lead them to recognize and con - and in medical education, there is still a lack of effective front their own biases and to consider their responsibil- training methods. In a narrative review of the reduction ity as GP and how they could improve their intercultural of implicit bias in health care, Zestcott, Blair [30] sug- interactions. The ethical approach to care, whereby a gested that public and professional awareness is a crucial balance must be struck between patient autonomy, not starting point for efforts at reduction. doing harm, and helping could contribute to ensuring equitable healthcare in a diverse society [35]. According to Razai, Kankam [20], diversifying health care “work- Limitations forces improves the performance of the entire healthcare This study has some limitations, however, and the data system” and could improve the relationship between must be interpreted with caution because they cannot be GPs and patients with different ethnic backgrounds. extrapolated to all trainee GPs. We recruited trainee GPs Future research could be carried out among people in their second year of internship from one important with more diverse cultural backgrounds to compare French-speaking university in Belgium, accounting for the results at the level of implicit biases. We could roughly a fifth of all trainee GPs in the country. In order also interview respondents after a year of practice, to to compare the data internationally, therefore, it would see whether being aware and sensitized influenced be useful for this study to be replicated in other Belgian their practice and reduced unintentional discrimina- medical schools and in other European countries. This tion. Further research should be carried out to inves- study should be replicated among qualified GPs so that tigate the relationship between GPs’ implicit/explicit the results can be compared with those of trainees to see biases and attitudes towards migrant patients and the whether the extent of bias and willingness of trainee GPs actual perceived discrimination reported by patients in to adapt to migrants changes over time. Further qualita- Europe [4]. tive research is also required, such as in-depth interviews Another idea would be to give the IAT score and with respondents to gain a better understanding of their the Hudelson score to each respondent individually in reactions. order to investigate personal feedback in relation to the We found several criticisms of the IAT regarding its group’s average scores in more depth. construct validity and questioning the biases that it We hope that this research will provide a promising actually measures [31]. Finally, we considered it to be a starting point for working towards a more ethnically reasonable research tool for assessing the level and distri- equitable health care system. bution of implicit bias in a group [32]. Although the IAT has been criticized, most medical school programmes lack formal methods for assessing and reducing bias Abbreviations among medical students [33]. Moreover, completing an GP: General Practitioner; GDPR: General Data Protection Regulation; IAT: Implicit Association Test. IAT also helps to reduce explicit biases via less direct mechanisms such as patient interaction and non-verbal Supplementary Information behaviour [34]. The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s12875- 022- 01698-8. Conclusion In conclusion, we believe that it is feasible to use the Additional file 1: Appendix 1. List of words used for the Implicit Associa- IAT to encourage trainees to engage with the topics of tion Test. Appendix 2. Hudelson scale. Appendix 3. Correlation between diversity and discrimination. The Hudelson scale was IAT and Hudelson scale. Appendix 4. R² for each covariate. used to confront trainees with their levels of cultural D uveau et al. BMC Primary Care (2022) 23:91 Page 9 of 10 Acknowledgements 6. Te Lindert A, Korzilius HPLM, Stupar-Rutenfrans S, Van De Vijver FJR. The We gratefully acknowledge our survey respondents who took time to partici- role of perceived discrimination, intergroup contact and adoption in pate in this survey. We are also grateful for the support provide by Professor acculturation among four Dutch immigrant groups. Int J Intercultural Jean-Marc Feron. Relations. 2021. https:// doi. org/ 10. 1016/j. ijint rel. 2021. 02. 005. 7. Cooper L, Roter D, Thornton R, Ford D, Steinwachs D, Powe N. Patient- Authors’ contributions Centered Communication, Ratings of Care, and Concordance of Patient CD and VL contributed to study conception and design, data acquisition, and Physician Race. Ann Intern Med. 2004;139:907–15. methodology, data management and data analysis, interpretation of data, and 8. Shavers VL, Fagan P, Jones D, Klein WMP, Boyington J, Moten C, et al. The creation of new software used in the work; they drafted the work and substan- State of Research on Racial/Ethnic Discrimination in The Receipt of Health tially revised it. SD, MD, and BL contributed to study conception, methodol- Care. Am J Public Health. 2012;102(5):953–66. ogy, and interpretation of data and substantially revised the work. The authors 9. European-Commission. Communication from the Commission to the have read and approved the manuscript. European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions: Action plan on Integra- Funding tion and Inclusion 2021–2027. Brussels 2020. p. 26. Belgian Research Action through Interdisciplinary Networks (BELSPO), grant 10. Maina IW, Belton TD, Ginzberg S, Singh A, Johnson TJ. A decade of study- number B2/191/P3/REMEDI. ing implicit racial/ethnic bias in healthcare providers using the implicit association test. Soc Sci Med. 2018;199:219–29. Availability of data and materials 11. Alizadeh S, Chavan M. Cultural competence dimensions and outcomes: The datasets used and/or analysed during the current study are available from a systematic review of the literature. Health Soc Care Community. the corresponding author on reasonable request. 2016;24(6):e117–30. The datasets generated and analysed during this study are not publicly avail- 12. Hall WJ, Chapman MV, Lee KM, Merino YM, Thomas TW, Payne BK, et al. able due to the fact that consent was obtained from participants on condition Implicit Racial/Ethnic Bias Among Health Care Professionals and Its Influ- that their data would not be shared, but are available from the corresponding ence on Health Care Outcomes: A Systematic Review. Am J Public Health. author on reasonable request. 2015;105(12):e60–76. 13. Rossen B, Johnsen K, Deladisma A, Lind S, Lok B. Virtual Humans Elicit Skin-Tone Bias Consistent with Real-World Skin-Tone Biases. Springer, Declarations Berlin Heidelberg. 2008;5208:237–44. 14. Nelson A. Unequal treatment: confronting racial and ethnic disparities in Ethics approval and consent to participate health care. J Natl Med Assoc. 2002;94(8):666–8. Informed consent was obtained from all participants. This study did not 15. Sorensen J, Norredam M, Suurmond J, Carter-Pokras O, Garcia-Ramirez require the agreement of the ethics committee according to Belgian law, but M, Krasnik A. Need for ensuring cultural competence in medical pro- all study methods were approved by the Hospital University Ethics Commit- grammes of European universities. BMC Medical Education. 2019;19(1):1. tee of the Université catholique de Louvain on 16 February 2021. Participants 16. Sabin JA, Nosek BA, Greenwald AG, Rivara FP. Physicians’ Implicit and were provided with the legal information relating to consent. All information Explicit Attitudes About Race by MD Race, Ethnicity, and Gender. J Health related to participants’ consent and the GDPR is available on request. This is in Care Poor Underserved. 2009;20(3):896–913. accordance with the privacy legislation of 30 July 2018, the European General 17. Khosla NN, Perry SP, Moss-Racusin CA, Burke SE, Dovidio JF. A comparison of cli- Data Protection Regulation (GDPR) of 25 May 2018, and Belgian regulations in nicians’ racial biases in the United States and France. Soc Sci Med. 2018;206:31–7. effect. 18. Levecque K, Lodewyckx I, Bracke P. Psychological distress, depression and generalised anxiety in Turkish and Moroccan immigrants in Belgium. Soc Consent for publication Psychiatry Psychiatr Epidemiol. 2009;44(3):188–97. Not applicable. 19. Reus-Pons M, Vandenheede H, Janssen F, Kibele EUB. Differences in mortality between groups of older migrants and older non-migrants in Competing interests Belgium, 2001–09. Eur J Pub Health. 2016;26(6):992–1000. The authors declare that they have no competing interests. 20. Razai MS, Kankam HKN, Majeed A, Esmail A, Williams DR. Mitigating ethnic disparities in covid-19 and beyond. BMJ. 2021;372:m4921. Author details 21. Nosek BA, Greenwald AG, Banaji MR. Understanding and Using the 1 2 Institute of Health and Society, UCLouvain, Brussels, Belgium. Faculty of Psy- Implicit Association Test: II. Method Variables and Construct Validity. chology and Educational Sciences, UCLouvain, Louvain-la-Neuve, Belgium. Person Soc Psych Bull. 2005;31(2):166–80. Belgian Health Care Knowledge Centre, KCE, Brussels, Belgium. 22. Klein C. Confidence Intervals on Implicit Association Test Scores Are Really Rather Large. 2020. Received: 26 November 2021 Accepted: 8 April 2022 23. Van Ryn M, Hardeman R, Phelan SM, Phd DJB, Dovidio JF, Herrin J, et al. Medical School Experiences Associated with Change in Implicit Racial Bias Among 3547 Students: A Medical Student CHANGES Study Report. J Gen Intern Med. 2015;30(12):1748–56. 24. Nosek BA, Banaji MR, Greenwald AG. Harvesting implicit group attitudes References and beliefs from a demonstration web site. Group Dyn Theory Res Pract. 1. Horvat L, Horey D, Romios P, Kis‐Rigo J. Cultural competence education 2002;6:101–15. for health professionals. Cochrane Dat Syst Rev. 2014;5:CD009405. 25. LondonoTobon A, Flores JM, Taylor JH, Johnson I, Landeros-Weisen- 2. Hudelson P, JunodPerron N, Perneger TV. Measuring Physicians’ and Medi- berger A, Aboiralor O, et al. Racial Implicit Associations in Psychiatric cal Students’ Attitudes Toward Caring for Immigrant Patients. Eval Health Diagnosis, Treatment, and Compliance Expectations. Acad Psychiatry. Prof. 2010;33(4):452–72. 2021;45(1):23–33. 3. Blair IV, Havranek EP, Price DW, Hanratty R, Fairclough DL, Farley T, et al. Assess- 26. FitzGerald C, Hurst S. Implicit bias in healthcare professionals: a systematic ment of biases against Latinos and African Americans among primary care review. BMC Med Ethics. 2017;18(1):19. providers and community members. Am J Public Health. 2013;103(1):92–8. 27. OECD. Health at a Glance 20152015. 4. Hanssens LGM, Detollenaere JDJ, Van Pottelberge A, Baert S, Willems 28. Dauvrin M, Lorant V. Adaptation of health care for migrants: whose SJT. Perceived discrimination In Primary Healthcare in Europe: evidence responsibility? BMC Health Serv Res. 2014;14:294. from the cross-sectional QUALICOPC study. Health Soc Care Community. 29. Andreoni J, Vesterlund L. Which is the Fair Sex? Gender Differences in 2017;25(2):641–51. Altruism. Q J Econ. 2001;116(1):293–312. 5. Dehon E, Weiss N, Jones J, Faulconer W, Hinton E, Sterling S. A Systematic 30. Zestcott CA, Blair IV, Stone J. Examining the presence, consequences, and Review of the Impact of Physician Implicit Racial Bias on Clinical Decision reduction of implicit bias in health care: A narrative review. Group Process Making. Acad Emerg Med. 2017;24(8):895–904. Intergroup Relat. 2016;19(4):528–42. Duveau et al. BMC Primary Care (2022) 23:91 Page 10 of 10 31. French D. Unconcious bias training a study in junk science. Arkansas Democrat Gazette. 2018. https:// www. arkan sason line. com/ news/ 2018/ apr/ 29/ uncon cious- bias- train ing-a- study- in- jun/. 32. Greenwald AG, Poehlman TA, Uhlmann EL, Banaji MR. Understanding and using the Implicit Association Test: III. Meta-analysis of predictive validity. J Per Soc Psych. 2009;97(1):17–41. 33. Ruben M, Saks NS. Addressing Implicit Bias in First-Year Medical Students: a Longitudinal. Multidisciplinary Train Prog Med Sci Educ. 2020;30(4):1419–26. 34. Cooper LA, Roter DL, Carson KA, Beach MC, Sabin JA, Greenwald AG, et al. The Associations of Clinicians’ Implicit Attitudes About Race With Medical Visit Communication and Patient Ratings of Interpersonal Care. Am J Public Health. 2012;102(5):979–87. 35. Aambø AK. Ethics in cross-cultural encounters: A medical concern? Med Humanit. 2020;46(1):22–30. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Primary Care Springer Journals

Implicit and explicit ethnic biases in multicultural primary care: the case of trainee general practitioners

Loading next page...
 
/lp/springer-journals/implicit-and-explicit-ethnic-biases-in-multicultural-primary-care-the-KuiIUS8V6Q

References (44)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2022
eISSN
2731-4553
DOI
10.1186/s12875-022-01698-8
Publisher site
See Article on Publisher Site

Abstract

Background: General Practitioners (GPs) are the first point of contact for people from ethnic and migrant groups who have health problems. Discrimination can occur in this health care sector. Few studies, however, have investi- gated implicit and explicit biases in general practice against ethnic and migrant groups. This study, therefore, investi- gated the extent of implicit ethnic biases and willingness to adapt care to migrant patients among trainee GPs, and the factors involved therein, in order to measure explicit bias and explore a dimension of cultural competence. Methods: In 2021, data were collected from 207 trainee GPs in the French-speaking part of Belgium. The respond- ents passed an Implicit Association Test (IAT ), a validated tool used to measure implicit biases against ethnic groups. An explicit attitude of willingness to adapt care to diversity, one of the dimensions of cultural competence, was meas- ured using the Hudelson scale. Results: The overwhelming majority of trainee GPs (82.6%, 95% CI: 0.77 – 0.88) had implicit preferences for their ingroup to the detriment of ethnic and migrant groups. Overall, the majority of respondents considered it the responsibility of GPs to adapt their attitudes and practices to migrants’ needs. More than 50% of trainee GPs, however, considered it the responsibility of migrant patients to adapt to the values and habits of the host country. Conclusions: This study found that the trainee GPs had high to very high levels of implicit ethnic bias and that they were not always willing to adapt care to the values of migrants. We therefore recommend that they are made aware of this bias and we recommend using the IAT and Hudelson scales as educational tools to address ethnic biases in primary care. Keywords: General Practitioner, Discrimination, Ingroup relationship, Migrant health, Racial bias, Cultural competence, Implicit association test equity of the health care system. Nevertheless, even GPs Introduction who are well-intentioned may be vulnerable, like anyone As aspiring health care professionals, trainee General else, to implicit biases and may lack willingness to adapt Practitioners (GPs) care for a diverse and multicultural care to migrants, a willingness that is a prerequisite of population in their practice [1] and are usually the first cultural competence [2]. This could lead GPs to discrimi - point of contact for ethnic and migrant patients within nate unintentionally against their ethnic and migrant the health care system. They are, therefore, key to the patients [3, 4] especially if they are under time pressure, lack solid knowledge/information needed to make a decision, or are affected by cognitive overload or fatigue *Correspondence: camille.duveau@uclouvain.be Institute of Health and Society, UCLouvain, Brussels, Belgium [5], or due to inappropriate intercultural contact [6] less Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visithttp:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Duveau et al. BMC Primary Care (2022) 23:91 Page 2 of 10 concordant physician–patient ethnicity, which is associ- has been paid to the North African population. This ated with less favourable patient ratings care [7], stereo- applies particularly to people of Moroccan descent, who typing, prejudice, political attitudes [8], or organizational constitute the largest and fastest-growing minority group and institutional factors [9]. Implicit ethnic biases among in countries in the European Union (EU), including Bel- health care professionals and cultural competence are gium, where the Moroccan population doubled between important factors that contribute to health care dispari- 1991 and 2014. This ethnic group is also dispropor - ties [10, 11]. tionately at risk of poor health outcomes, such as more In social psychology, a “bias” is defined as a prejudicial depressive symptoms [18] and higher mortality caused by attitude towards a group (hereafter, the exogroup) and its diabetes and infectious diseases [19]. members, of which the holder may or may not be aware. To the best of our knowledge, few studies have inves- Implicit bias is often automatically activated and can tigated the extent of bias among general practitioners lead to impaired judgement and to negative evaluation of against racial/ethnic groups in relation to GPs’ willing- the exogroup [12]. Implicit ethnic bias has been shown ness to adapt to diversity. A recent study highlighted the to play a role in differential recommendations by physi - need to increase recognition and awareness of ethnic cians for managing disease [10], disparities in empathy disparities in health and health care among health care [13] and differential drug prescriptions, interaction with professionals [20]. This paper aims to expand this area of patients, and treatment decisions [10]. research by investigating (a) the level of implicit ethnic To address these differences in care for ethnic/migrant bias among trainee GPs and (b) whether this level of bias groups, the report Unequal Treatment by the Institute varies as a function of a GP’s willingness to adapt care to of Medicine recommended exploring healthcare provid- migrants or as a function of openness to cultural diversity ers’ conscious (explicit bias) and unconscious percep- which, in this paper, includes intercultural contact, politi- tions [14]. This report also recommended developing and cal opinions, and workload (time pressure). enhancing cultural competence training for health care professionals. Over the past decade, many cultural com- Methods petence programmes have been initiated in health care Design & participants systems to improve the relationships between health care This prospective quantitative study used an observa - professionals and patients with diverse cultural back- tional design. The design measures implicit ethnic biases grounds, and to enable them to work and interact more and the specific explicit attitude of willingness to adapt effectively with those patients [11]. These programmes, to diversity, a prerequisite of cultural competence. To however, are based on the assumption that health profes- encourage trainee general practitioners to respond to the sionals are motivated and accept responsibility for adapt- online survey, we offered to present the overall results ing to ethnic diversity [15], an assumption that may need to them anonymously and give them an opportunity to to be verified. This study aims to add to the literature by comment on and discuss the results in small groups. measuring implicit and explicit biases against ethnic/ Between February and March 2021, we contacted 220 migrant groups. French-speaking Belgian GPs, who were in the second A large body of research has investigated unconscious year of their three-year internship by email in one of the biases and levels of cultural competence among health important French-speaking universities in Belgium. Of care professionals. This research found significant evi - those trainees, 207 completed the online survey (a par- dence of implicit bias among care professionals against ticipation rate of 94%) (Male = 68 (33.0%); Female = 139 African Americans, ranging from slight bias to high lev- (67.0%); Mean age = 26.7 years (± 2.1); See Table 1). els of bias, but found low levels of explicit bias against this ethnic group. The research also found a weak asso - Measures ciation between self-perceived cultural competence and Implicit association test levels of implicit bias [10, 16]. Little is known, however, The Implicit Association Test (IAT) is a validated tool about these issues within the European context and a in that is used, in social psychology, to assess implicit recent study comparing clinicians’ racial biases in The biases, which can be associated with a wide range of United States and France concluded that health provid- behaviours and attitudes [21]. As bias is often activated ers’ biases differed between cultures and countries [17]. by situational cues such as a first name [3 ], respondents So far, most health care studies have been carried out took an IAT with French-language and North African in the United States and have compared African Ameri- can patients with White patients. There is a lack of data on unintentional ethnic and racial discrimination among Cf. Myriatics. (2015, décembre). Myria. https:// www. myria. be/ fr/ chiff res/ general practitioners in Europe and too little attention myria tics D uveau et al. BMC Primary Care (2022) 23:91 Page 3 of 10 Table 1 Demographic characteristics, implicit bias, and cultural D-scores according to the strength of the implicit asso- competence score of participants (n = 207) ciations: “neutral” (< 0.15), “slight” (< 0.35), “moderate” (< 0.65), and “strong” (≥ 0.65), and the reverse for scores Characteristics % or mean (std) with a negative value [22]. Age (years) 26.7 (2.1) Sex female 67.0 3 Hudelson scale Type of practice The trainee GPs completed the validated Hudelson scale, Solo 27.7 which assesses one dimension of cultural competence: Group 69.4 the perception of relative responsibility for adapting care Mixed 2.9 to migrant patients [2]. This was assessed using descrip - Ethnicity tions of five potential situations: (1) when migrants’ val - White 78.5 ues and habits differ from those of the host country, (2) Arab 8.3 when the patient does not speak the language of the host African 3.9 country, (3) when the patient expresses the wish to be Mixed or other 9.3 treated by a male/female health professional, (4) when Religion the patient cannot read the language of the host coun- Don’t want to reply 4.4 try, and (5) when the patient’s health beliefs contradict Catholic 39.1 the knowledge of the health professionals. The trainee Muslim 10.1 GPs had to state their opinion about whose responsi- Atheist 31.9 bility it is to adapt in those situations by ticking a score Other 14.5 between 1 (it is the professional’s responsibility) and 7 Political opinion (it is the migrant’s responsibility), with 4 meaning that Don’t want to reply 10.6 the responsibility is shared between the patient and the Right-wing liberal 29.5 professional. Centre 31.9 Regarding openness to cultural diversity, we expected Left-wing (e.g. socialist) 28.0 left-wing political opinions, the proportion of migrants Practice area among the patients GP cared for, more social interactions Urban 7.8 with people of North-African origin and a lighter work- Suburban 33.1 load to be associated with less unintentional discrimi- Rural 25.3 nation [23]. To compute the frequency of intercultural Missing values 33.8 contact, we created a score composed of four variables: Frequency of contact with North Africans (/4) 2.5 (1.2) (1) presence of North-African people in their neighbour- Workload (no. of patients/day) 10.2 (5.6) hood, (2) daily interaction with North-African people, (3) Proportion of patients of foreign origin (/100) 33.0 (30.4) having North-African friend(s), and (4) having North- Implicit Association Test score (-2, 2) 0.53 (0.44) African colleague(s). We obtained a score out of 4 points Hudelson score (5, 35) 18.9 (5.1) (where 0 means that the respondent has no contact with North Africans and 4 means that the respondent fre- quently interacts with North Africans). Control variables were added to the models, such as age, sex, ethnicity, first names. The IAT was designed to measure implicit and type of practice (e.g. solo, group, or mixed), to avoid ethnic bias. The IAT with North African first names potential selection bias. uses reaction times to assess the strength of automatic associations between target pairs, each consisting of an Reactions and feedback sharing ethnicity (e.g. French-language or Belgo-Moroccan first The trainee GPs were presented with feedback on the names), and a category (e.g. positive or negative). This aggregate results and then asked to comment on the test provides a D-score ranging from -2 (respondent results. We collected 204 trainee GPs’ anonymized open- has implicit associations of positive words with North ended reactions to their IAT and Hudelson scale scores. African first names) to + 2 (respondent has implicit We then divided them into small groups of 4–5 to discuss associations of positive words with French-language the experience and explore possible solutions. first names). We used standard cut-offs to classify the 2 3 see Appendix 1. see Appendix 2. Duveau et al. BMC Primary Care (2022) 23:91 Page 4 of 10 Fig. 1 Distribution of implicit associations for first names associated with different ethnic groups Statistical analysis Results Quantitative analysis Participants First, we computed descriptive statistics of variables. Two hundred and seven General Practitioners in the sec- We computed correlations between the level of implicit ond year of a three-year specialisation in general medi- bias and the Hudelson score to check whether they over- cine responded to the online questionnaire. Of those lapped. Then, linear regression was performed to relate respondents, two hundred and four took part in the dis- the two outcomes (implicit and explicit attitudes) with cussion about the presentation of their results from the the following variables: political opinions, practice area, IAT and the Hudelson scale. Most of the participants frequency of intercultural contact, workload, and propor- were of Belgian nationality (89.0%). Participants self- tion of patients of foreign origin, controlling for variables reported as White (78.5%), Arab (8.3%), or of mixed/ such as sex, type of practice, and ethnicity. We ran one other ethnicity (9.3%). A third of the respondents self- model for each of the following explanatory variables: reported as Catholic (39.1%). Nearly 70% of the respond- political opinion, frequency of intercultural contact with ents were not very religious or not religious at all. In North Africans, frequency of patients of foreign origin in terms of political opinion, the majority selected Centre consultations. Statistical analysis was performed using (31.9%) and 29.5% of the sample identified as right-wing. SAS 9.4. Most worked in suburban areas (33.1%). A quarter of the respondents worked in rural areas. A minority (7.8%) worked in the seven largest cities in the Wallonia-Brus- Reactions and feedback sharing sels Federation (Brussels, Liège, Verviers, Namur, Char- With the aid of NVIVO, we grouped similar quotes leroi, Mons, and Tournai). The area was computed based from the written reactions into main themes in order to on their medical practice’s postcode. explore the general reactions to the results and the pos- The average score for frequency of contact with North sible solutions that trainee GPs highlighted within the Africans was 2.5/4 (± 1.2), meaning that the respond- small groups. ent sometimes interacts with North African friends, D uveau et al. BMC Primary Care (2022) 23:91 Page 5 of 10 Table 2 Relative responsibility to adapt of migrants and health professionals, as perceived by trainee GPs (2021): % per item (n = 207) The responsibility The responsibility lies with The responsibility Total lies with the health both health professionals and lies with the professionals patients patients When migrants’ values and habits differ from those of 10.6 30.9 58.5 100 the host country (%) When the patient does not speak the language of 37.7 27.5 34.8 100 the host country (%) When the patient expresses the wish to be treated by 60.4 15.0 24.6 100 a male or female doctor (%) When the patient cannot read the language of the 50.7 21.3 28.0 100 host country (%) When the patient’s health beliefs contradict medical 46.4 26.1 27.5 100 knowledge (%) colleagues, or neighbours. In their practice, they visit professionals. The majority of trainee GPs considered 10.2 patients per day on average, 33% (± 30.4) of whom it to be mainly the health professionals’ responsibility are foreign patients. to adapt when the patient expresses the wish to choose the sex of the health care professional, when he/she can- Implicit association test not read the language of the host country, and when Figure 1 presents the distribution of implicit associations his/her health beliefs contradict medical knowledge. of positive/negative words with French-language first Two items, however, differed from this trend. Roughly names compared with North African first names. The 38% of respondents considered it the responsibility of IAT mean score (D-score) was 0.53 (± 0.44), indicating health care professionals to adapt to migrants when the a moderate association of positive words with French- patient does not speak the language of the host country. language first names over North African first names. The Approximately 35% considered it the responsibility of majority of respondents (82.6%, CI 95%: 0.77 – 0.88) had the patient to adapt and 27.5% considered it the respon- a positive D-score, i.e.an implicit association of positive sibility of both the professional and the patient to adapt. words with French-language first names. Several of the This result is quite surprising because, when it comes to scores for implicit associations stand out. providing written information in the patient’s language, Around 7% of the scores were negative, meaning that respondents placed the responsibility on the health care the trainees had an implicit association of positive words professional (50.7%). When the migrant’s values and hab- with North-African first names. Of the respondents with its differ from those of the host country, however, GPs negative scores, 4.3% were Arab, 3.4% were White, and considered it to be the patient’s responsibility to adapt 1.9% were of mixed/other ethnic origin. The scores varied (58.5%). significantly according to ethnicity (χ2 = 88.6, p < 0.0001). There was a significant small positive correlation In contrast, about 12% had a slight implicit association of between implicit ethnic bias and cultural competence French-language names with positive words, while 45.4% (r = 0.172, p < 0.05), suggesting that respondents for had a strong implicit association of French-language whom French-language names had more implicit posi- names with positive words. tive associations were also more likely to consider it the It is also worth noting that approximately 10% of the migrants’ responsibility to adapt. respondents showed no group preference. Table  3 provides the betas from the linear regressions for the factors associated with the IAT and with the Hudelson scale Hudelson score. Respondents’ ethnic origins influenced Table  2 presents the perception of the relative responsi- the level of implicit ethnic bias. Being of Arab ethnicity bility to adapt to migrant patients in health care among or of other or mixed ethnic origin reduced the strength trainee general practitioners. The five items of the Hudel - of implicit negative ethnic associations with North Afri- son scale were moderately correlated (Pearson coeffi - can first names (β = -0.54, p < 0.001, β =—0.29, p < 0.05 cients 0.09–0.46) and the internal consistency coefficient respectively). Having left-wing political opinions also (Cronbach’s α) was 0.62. The average score for Hudelson had a negative correlation with the IAT score (β = -0.16, scale was 18.9/35 (± 5.1), which means that adapting was p < 0.05). Nevertheless, other individual and contextual considered to be mainly the responsibility of health care variables such as age, sex, type of practice, practice area, Duveau et al. BMC Primary Care (2022) 23:91 Page 6 of 10 Table 3 Factors associated with Implicit Association Test and with the Hudelson score: betas from the linear regression models Implicit Association Test (-2, 2) Hudelson score (5–35) Covariates: β CI p-value β CI p-value 95% 95% Age (years) 0.00 (-0.03, 0.04) 0.93 0.03 (-0.09, 0.68) 0.13 Sex (ref = Male) Female -0.05 (-0.18, 0.08) 0.49 -2.02 ( -3.46, -0.57) < 0.01 Type of practice (ref = Solo) Group practice 0.07 ( -0.07, 0.21) 0.32 -1.24 ( -2.75, 0.28) 0.11 Mixed practice 0.01 ( -0.34, 0.37) 0.94 -2.68 ( -6.61, 1.25) 0.18 Ethnicity (ref = White) Arab -0.54 ( -0.76, -0.31) < .0001 -0.41 ( -2.90, 2.08) 0.75 African -0.08 ( -0.40, 0.23) 0.60 1.57 ( -1.96, 5.09) 0.38 Other or mixed -0.29 ( -0.51, -0.07) 0.01 -1.51 ( -3.96, 0.94) 0.23 Political opinion (ref = Right-wing) Centre 0.03 (-0.17, 0.12) 0.72 -0.36 (-1.97,1.24) 0.66 Left (e.g. Socialist) -0.16 (-0.32, -0.01) 0.03 -3.07 (-4.74, -1.40) < .001 Practice area (ref = Urban) Suburban -0.04 (-0.25, 0.16) 0.67 -1.92 (-4.30, 0.46) 0.11 Rural -0.05 (-0.26, 0.17) 0.65 -0.42 (-2.94, 2.09) 0.74 Frequency of contact with North Africans -0.05 (-0.10, 0.01) 0.07 0.13 (-0.49, 0.74) 0.69 Workload (no. of patients/day) -0.00 (-0.01, 0.01) 0.95 0.06 (-0.06, 0.19) 0.30 Proportion of patients of foreign origin -0.04 (-0.26, 0.17) 0.70 -0.37 (-3.00, 2.26) 0.78 R -range [0.16–0.24] [0.09–0.17] β are adjusted for age, sex, ethnicity and type of practice; bold coefficients have p-value < 0.05 Low scores indicate that the responsibility to adapt is considered to lie with the health care professionals, whereas high scores indicate that it is considered to lie with the migrants Table with all R-squared is in appendix 4 frequency of contact with North Africans, and workload “Very impactful. It reveals that even unconscious biases were not associated with the level of implicit bias. can exist. This test confirms how important it is to know Concerning the Hudelson score, being a woman that these biases exist in order to be able to fight them (β = -2.02, p = 0.01) was associated with considering it to consciously in our practice…” (Participant 165). be mainly the health professional’s responsibility to adapt We found few individual or contextual explanations to migrants. Respondents with left-wing political opin- in the regression analysis of implicit biases, aside from ions also had a significant tendency to place the respon - ethnicity. The influence of their ethnicity is also dem - sibility on health care professionals (β = -3.07, p < 0.01). onstrated in the comments: “our culture played an Age, type of practice, having liberal political opinions, unconscious bias in this test.” (Participant 36). practice area, frequency of contact with North Africans, Unlike the quantitative results, the association workload, and frequency of patients of foreign origin between political opinions and propensity to place the were not associated with considering it the health care responsibility for adapting to migrants on health pro- professional’s responsibility to adapt to migrants. fessionals was not addressed in the comments. Only The R ranged between [0.16–0.24] and [0.09–0.17] one comment referred to political opinions: “Surprised for the IAT score and the Hudelson score, respectively. by the 60% [who said that the migrant should adapt to Appendix 4 presents the details of the R for each factor. the values of the host country] … we are GPs not politi- cians… the main thing is the patient’s health.” (Partici- pant 132). Reactions and feedback sharing They also shared solutions with one another to imple - From the comments, it appears that some of the GPs ment in their practice as GPs. They suggested keeping the were a bit sceptical about their implicit ethnic associa- Hudelson scale in mind in their everyday practice. It will tion scores and questioned the method and the exist- help them to remember that in most situations they bear ence of their own biases. A number of trainee GPs also responsibility for adapting to migrant patients. GPs also highlighted the need for them to be aware of biases: D uveau et al. BMC Primary Care (2022) 23:91 Page 7 of 10 emphasized the importance of being conscious of one’s people than White respondents [16, 25]. The authors, own biases in order to fight against them. however, also found lower levels of ethnic bias among Finally, several trainees proposed “cultural openness” women, which was not the case with our results. as a way to reduce the effects of these biases and improve Our results suggest that completing an IAT and dis- the quality of care, such as working in an intercultural cussing the results could help to reduce the negative team, learning several languages and learning about dif- impact of bias on health professionals’ attitudes. Indeed, ferent cultures. it could reduce both implicit and explicit bias among health providers with higher levels of implicit bias and thereby improve clinical interaction and verbal commu- Discussion nication and make practice more patient-centred [26]. Main findings This study set out to assess the trainee GPs’ implicit biases and their willingness to adapt care to diversity. Hudelson score Our results indicate that the majority of our trainees For three out of five situations, the majority of trainee had a moderate to strong level of implicit bias favouring GPs were willing to bear the responsibility to adapt care French-language first names over North African ones. to migrants. They were more reluctant, however, to adapt This result applied to all sociodemographic characteris - to migrants’ values and habits. These findings are in line tics, except ethnicity and political opinion. On the whole, with those of Hudelson, Junod Perron [2]. In the respond- trainees were willing to adapt care to migrants in most ents’ feedback, we noticed that some were surprised by care situations, with two exceptions: when a patient did this result. One of the respondents pointed out that they not speak the language of the host country and when the were GPs and not politicians. Our statistical analysis pro- immigrants’ values and habits differed from those of the vides some explanations: adaptation is related to politi- host country. This result applied to almost all sociodemo - cal orientation and the topic of values/habits is certainly graphic groups. a political issue, also from the GPs’ perspective. In a less competitive market, health care professionals may also Implicit association test feel less responsible for adapting to the values and habits The literature has often reported implicit associations of their patients as patients are free to move to another with ingroups (and biases against exogroups) that are GP if they are not satisfied. One might thus speculate that stronger than those found in this research [16]. Our the decrease in the percentage of GPs among all doctors study, however, does not support Allport’s finding that in Belgium over the last decade has enhanced the market reduced implicit bias is associated with higher frequency power of GPs [27]. of positive contact with the outgroup [23]. One possible Paradoxically, although more than half of the respond- explanation for this is the fact that we only assessed the ents see it as the health professional’s responsibility to frequency of contact with North Africans, not quality adapt when the patient cannot read the language of the of that contact. Furthermore, the p-value for the vari- host country, opinion was spread evenly across the board able “frequency of contact with North Africans” in the when the patient cannot speak it. While Dauvrin et  al. regression (Table  3) was 0.07, which suggests that either found that physicians took responsibility for providing this measure needs to be assessed more accurately, for interpreters [28], our respondents disagreed with that. example, by including an assessment of the quality of One explanation could be the extra time and resources trainee GPs’ relationships with members of the outgroup, needed to provide an interpreter during an everyday con- or interaction with outgroup members does not reduce sultation when GPs are often already under time pres- implicit bias among trainee GPs. sure. Moreover, access to intercultural mediators remains One important finding was that biases affect everyone limited in primary care. and few protective factors exist in GPs’ working envi- A greater willingness to adapt was found among female ronments, apart from their own ethnicity and political GPs and those with left-wing political opinions. This is opinions. This result has been demonstrated previously in line with previous research which found that female by Nosek, Banaji [24], who found a relationship between providers were more engaged in patient-centred commu- right-wing ideology and stronger implicit bias. They also nication and that African-American patients were more concluded that bias could vary according to “the con- likely to be seen by female GPs [2]. Another explanation straints that culture imposes on individual attitudes”, could also be that women tend to be more altruistic than which could explain the influence of ethnicity on the IAT men, when altruism requires more resources. We could score. We found similar results in several studies, which therefore speculate that diversity might be considered showed that Black providers also had significantly weaker as a form of altruism which demands more equalitarian ethnic implicit associations for both White and Black Duveau et al. BMC Primary Care (2022) 23:91 Page 8 of 10 behaviours, behaviours more likely to be found among competence. The debriefing on the results helped to women too [29]. raise awareness and sensitize the trainee GPs to the The combination of the findings from the IAT and the potential impact of their implicit and explicit biases Hudelson scale provides some support for the suggestion in multicultural consultation. They also became aware that there is a need to increase recognition and aware- that these biases could have an impact on their rela- ness of implicit ethnic bias and willingness to adapt to tionships and the way they care for patients. diversity in order to become more culturally competent. One implication of this research is that it should be The weak correlation between the two concepts sug - possible to use the results as a basis for reflection and gests that these are two different issues, requiring differ - discussion that could motivate health care profession- ent interventions. The R was also weak for both the IAT als to tackle their ethnic biases. Another implication is scores and the Hudelson scores, meaning that our model the possibility of using the Hudelson scale as a tool to did not fully explain the variation of these scores. Despite debate the extent of cultural competence among future increased interest in providers’ biases in the literature GP. This debate could lead them to recognize and con - and in medical education, there is still a lack of effective front their own biases and to consider their responsibil- training methods. In a narrative review of the reduction ity as GP and how they could improve their intercultural of implicit bias in health care, Zestcott, Blair [30] sug- interactions. The ethical approach to care, whereby a gested that public and professional awareness is a crucial balance must be struck between patient autonomy, not starting point for efforts at reduction. doing harm, and helping could contribute to ensuring equitable healthcare in a diverse society [35]. According to Razai, Kankam [20], diversifying health care “work- Limitations forces improves the performance of the entire healthcare This study has some limitations, however, and the data system” and could improve the relationship between must be interpreted with caution because they cannot be GPs and patients with different ethnic backgrounds. extrapolated to all trainee GPs. We recruited trainee GPs Future research could be carried out among people in their second year of internship from one important with more diverse cultural backgrounds to compare French-speaking university in Belgium, accounting for the results at the level of implicit biases. We could roughly a fifth of all trainee GPs in the country. In order also interview respondents after a year of practice, to to compare the data internationally, therefore, it would see whether being aware and sensitized influenced be useful for this study to be replicated in other Belgian their practice and reduced unintentional discrimina- medical schools and in other European countries. This tion. Further research should be carried out to inves- study should be replicated among qualified GPs so that tigate the relationship between GPs’ implicit/explicit the results can be compared with those of trainees to see biases and attitudes towards migrant patients and the whether the extent of bias and willingness of trainee GPs actual perceived discrimination reported by patients in to adapt to migrants changes over time. Further qualita- Europe [4]. tive research is also required, such as in-depth interviews Another idea would be to give the IAT score and with respondents to gain a better understanding of their the Hudelson score to each respondent individually in reactions. order to investigate personal feedback in relation to the We found several criticisms of the IAT regarding its group’s average scores in more depth. construct validity and questioning the biases that it We hope that this research will provide a promising actually measures [31]. Finally, we considered it to be a starting point for working towards a more ethnically reasonable research tool for assessing the level and distri- equitable health care system. bution of implicit bias in a group [32]. Although the IAT has been criticized, most medical school programmes lack formal methods for assessing and reducing bias Abbreviations among medical students [33]. Moreover, completing an GP: General Practitioner; GDPR: General Data Protection Regulation; IAT: Implicit Association Test. IAT also helps to reduce explicit biases via less direct mechanisms such as patient interaction and non-verbal Supplementary Information behaviour [34]. The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s12875- 022- 01698-8. Conclusion In conclusion, we believe that it is feasible to use the Additional file 1: Appendix 1. List of words used for the Implicit Associa- IAT to encourage trainees to engage with the topics of tion Test. Appendix 2. Hudelson scale. Appendix 3. Correlation between diversity and discrimination. The Hudelson scale was IAT and Hudelson scale. Appendix 4. R² for each covariate. used to confront trainees with their levels of cultural D uveau et al. BMC Primary Care (2022) 23:91 Page 9 of 10 Acknowledgements 6. Te Lindert A, Korzilius HPLM, Stupar-Rutenfrans S, Van De Vijver FJR. The We gratefully acknowledge our survey respondents who took time to partici- role of perceived discrimination, intergroup contact and adoption in pate in this survey. We are also grateful for the support provide by Professor acculturation among four Dutch immigrant groups. Int J Intercultural Jean-Marc Feron. Relations. 2021. https:// doi. org/ 10. 1016/j. ijint rel. 2021. 02. 005. 7. Cooper L, Roter D, Thornton R, Ford D, Steinwachs D, Powe N. Patient- Authors’ contributions Centered Communication, Ratings of Care, and Concordance of Patient CD and VL contributed to study conception and design, data acquisition, and Physician Race. Ann Intern Med. 2004;139:907–15. methodology, data management and data analysis, interpretation of data, and 8. Shavers VL, Fagan P, Jones D, Klein WMP, Boyington J, Moten C, et al. The creation of new software used in the work; they drafted the work and substan- State of Research on Racial/Ethnic Discrimination in The Receipt of Health tially revised it. SD, MD, and BL contributed to study conception, methodol- Care. Am J Public Health. 2012;102(5):953–66. ogy, and interpretation of data and substantially revised the work. The authors 9. European-Commission. Communication from the Commission to the have read and approved the manuscript. European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions: Action plan on Integra- Funding tion and Inclusion 2021–2027. Brussels 2020. p. 26. Belgian Research Action through Interdisciplinary Networks (BELSPO), grant 10. Maina IW, Belton TD, Ginzberg S, Singh A, Johnson TJ. A decade of study- number B2/191/P3/REMEDI. ing implicit racial/ethnic bias in healthcare providers using the implicit association test. Soc Sci Med. 2018;199:219–29. Availability of data and materials 11. Alizadeh S, Chavan M. Cultural competence dimensions and outcomes: The datasets used and/or analysed during the current study are available from a systematic review of the literature. Health Soc Care Community. the corresponding author on reasonable request. 2016;24(6):e117–30. The datasets generated and analysed during this study are not publicly avail- 12. Hall WJ, Chapman MV, Lee KM, Merino YM, Thomas TW, Payne BK, et al. able due to the fact that consent was obtained from participants on condition Implicit Racial/Ethnic Bias Among Health Care Professionals and Its Influ- that their data would not be shared, but are available from the corresponding ence on Health Care Outcomes: A Systematic Review. Am J Public Health. author on reasonable request. 2015;105(12):e60–76. 13. Rossen B, Johnsen K, Deladisma A, Lind S, Lok B. Virtual Humans Elicit Skin-Tone Bias Consistent with Real-World Skin-Tone Biases. Springer, Declarations Berlin Heidelberg. 2008;5208:237–44. 14. Nelson A. Unequal treatment: confronting racial and ethnic disparities in Ethics approval and consent to participate health care. J Natl Med Assoc. 2002;94(8):666–8. Informed consent was obtained from all participants. This study did not 15. Sorensen J, Norredam M, Suurmond J, Carter-Pokras O, Garcia-Ramirez require the agreement of the ethics committee according to Belgian law, but M, Krasnik A. Need for ensuring cultural competence in medical pro- all study methods were approved by the Hospital University Ethics Commit- grammes of European universities. BMC Medical Education. 2019;19(1):1. tee of the Université catholique de Louvain on 16 February 2021. Participants 16. Sabin JA, Nosek BA, Greenwald AG, Rivara FP. Physicians’ Implicit and were provided with the legal information relating to consent. All information Explicit Attitudes About Race by MD Race, Ethnicity, and Gender. J Health related to participants’ consent and the GDPR is available on request. This is in Care Poor Underserved. 2009;20(3):896–913. accordance with the privacy legislation of 30 July 2018, the European General 17. Khosla NN, Perry SP, Moss-Racusin CA, Burke SE, Dovidio JF. A comparison of cli- Data Protection Regulation (GDPR) of 25 May 2018, and Belgian regulations in nicians’ racial biases in the United States and France. Soc Sci Med. 2018;206:31–7. effect. 18. Levecque K, Lodewyckx I, Bracke P. Psychological distress, depression and generalised anxiety in Turkish and Moroccan immigrants in Belgium. Soc Consent for publication Psychiatry Psychiatr Epidemiol. 2009;44(3):188–97. Not applicable. 19. Reus-Pons M, Vandenheede H, Janssen F, Kibele EUB. Differences in mortality between groups of older migrants and older non-migrants in Competing interests Belgium, 2001–09. Eur J Pub Health. 2016;26(6):992–1000. The authors declare that they have no competing interests. 20. Razai MS, Kankam HKN, Majeed A, Esmail A, Williams DR. Mitigating ethnic disparities in covid-19 and beyond. BMJ. 2021;372:m4921. Author details 21. Nosek BA, Greenwald AG, Banaji MR. Understanding and Using the 1 2 Institute of Health and Society, UCLouvain, Brussels, Belgium. Faculty of Psy- Implicit Association Test: II. Method Variables and Construct Validity. chology and Educational Sciences, UCLouvain, Louvain-la-Neuve, Belgium. Person Soc Psych Bull. 2005;31(2):166–80. Belgian Health Care Knowledge Centre, KCE, Brussels, Belgium. 22. Klein C. Confidence Intervals on Implicit Association Test Scores Are Really Rather Large. 2020. Received: 26 November 2021 Accepted: 8 April 2022 23. Van Ryn M, Hardeman R, Phelan SM, Phd DJB, Dovidio JF, Herrin J, et al. Medical School Experiences Associated with Change in Implicit Racial Bias Among 3547 Students: A Medical Student CHANGES Study Report. J Gen Intern Med. 2015;30(12):1748–56. 24. Nosek BA, Banaji MR, Greenwald AG. Harvesting implicit group attitudes References and beliefs from a demonstration web site. Group Dyn Theory Res Pract. 1. Horvat L, Horey D, Romios P, Kis‐Rigo J. Cultural competence education 2002;6:101–15. for health professionals. Cochrane Dat Syst Rev. 2014;5:CD009405. 25. LondonoTobon A, Flores JM, Taylor JH, Johnson I, Landeros-Weisen- 2. Hudelson P, JunodPerron N, Perneger TV. Measuring Physicians’ and Medi- berger A, Aboiralor O, et al. Racial Implicit Associations in Psychiatric cal Students’ Attitudes Toward Caring for Immigrant Patients. Eval Health Diagnosis, Treatment, and Compliance Expectations. Acad Psychiatry. Prof. 2010;33(4):452–72. 2021;45(1):23–33. 3. Blair IV, Havranek EP, Price DW, Hanratty R, Fairclough DL, Farley T, et al. Assess- 26. FitzGerald C, Hurst S. Implicit bias in healthcare professionals: a systematic ment of biases against Latinos and African Americans among primary care review. BMC Med Ethics. 2017;18(1):19. providers and community members. Am J Public Health. 2013;103(1):92–8. 27. OECD. Health at a Glance 20152015. 4. Hanssens LGM, Detollenaere JDJ, Van Pottelberge A, Baert S, Willems 28. Dauvrin M, Lorant V. Adaptation of health care for migrants: whose SJT. Perceived discrimination In Primary Healthcare in Europe: evidence responsibility? BMC Health Serv Res. 2014;14:294. from the cross-sectional QUALICOPC study. Health Soc Care Community. 29. Andreoni J, Vesterlund L. Which is the Fair Sex? Gender Differences in 2017;25(2):641–51. Altruism. Q J Econ. 2001;116(1):293–312. 5. Dehon E, Weiss N, Jones J, Faulconer W, Hinton E, Sterling S. A Systematic 30. Zestcott CA, Blair IV, Stone J. Examining the presence, consequences, and Review of the Impact of Physician Implicit Racial Bias on Clinical Decision reduction of implicit bias in health care: A narrative review. Group Process Making. Acad Emerg Med. 2017;24(8):895–904. Intergroup Relat. 2016;19(4):528–42. Duveau et al. BMC Primary Care (2022) 23:91 Page 10 of 10 31. French D. Unconcious bias training a study in junk science. Arkansas Democrat Gazette. 2018. https:// www. arkan sason line. com/ news/ 2018/ apr/ 29/ uncon cious- bias- train ing-a- study- in- jun/. 32. Greenwald AG, Poehlman TA, Uhlmann EL, Banaji MR. Understanding and using the Implicit Association Test: III. Meta-analysis of predictive validity. J Per Soc Psych. 2009;97(1):17–41. 33. Ruben M, Saks NS. Addressing Implicit Bias in First-Year Medical Students: a Longitudinal. Multidisciplinary Train Prog Med Sci Educ. 2020;30(4):1419–26. 34. Cooper LA, Roter DL, Carson KA, Beach MC, Sabin JA, Greenwald AG, et al. The Associations of Clinicians’ Implicit Attitudes About Race With Medical Visit Communication and Patient Ratings of Interpersonal Care. Am J Public Health. 2012;102(5):979–87. 35. Aambø AK. Ethics in cross-cultural encounters: A medical concern? Med Humanit. 2020;46(1):22–30. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions

Journal

BMC Primary CareSpringer Journals

Published: Apr 21, 2022

Keywords: General Practitioner; Discrimination; Ingroup relationship; Migrant health; Racial bias; Cultural competence; Implicit association test

There are no references for this article.