The content of research paper

Professionalism discourses dominate health professions education (HPE) literature; however, recognising and responding to unprofessional behaviour remains a global challenge for educators.　Professionalism lapses and poor self-awareness are the most prevalent performance issues amongst medical students, and the most difficult to remediate, with repercussions for patient safety¹⁾. A contributing factor to this complex problem is the lack of an agreed definition of medical professionalism^2,3); as a result, unprofessionalism also remains undefined.　Current professionalism frameworks are behaviour-based^4,5) , but lack the detail, context, and empirical evidence to support the consistent identification of unprofessional behaviours in practice^3,6,7).　Furthermore, formal positions on medical professionalism involve consensus statements by professional bodies⁸⁾, which risks the exclusion of multiple viewpoints within professional and lay communities and widens the gap between frameworks and their meaningful application in local teaching and subsequent medical practice.

Compounding the situation further is the Western origin of leading professionalism frameworks, such as the UK General Medical Council (GMC)’s framework ‘Good Medical Practice’⁹⁾ and ‘The Physician Charter’ created by the American Board of Internal Medicine (ABIM) in collaboration with the American College of Physicians Foundation and the European Federation of Internal Medicine¹⁰⁾. Although such frameworks dominate current discourses, applying such principles across cultures leads to translation and ethical challenges¹¹⁾. This recognition forms part of a broader movement of decolonisation within HPE, where HPE professionals working in non-dominating cultures are encouraged to explore key issues through their own sociocultural lenses¹²⁾. Within Japanese medical education, Nishigori and colleagues have explored the tenets of ‘The Physician Charter’ against the concepts and values of Bushido, the Japanese ancient Samurai code of conduct, finding areas of consistency alongside key areas of cultural diversity and intrinsic differences¹³⁾; however, to date there has not been a detailed examination of current understanding of what constitutes unprofessional behaviour in either a Japanese or a UK setting.

A key contributing factor is the subjective nature of unprofessionalism, which we feel can now be addressed using a person-centred factor analysis method called Q-methodology.　Rapidly gaining in popularity within HPE, Q-methodology provides a robust way of exploring in-group and between-group variability, allowing for the identification of viewpoints missed through variable-centred approaches¹⁴⁾. Q-methodology inverts traditional by-variable factor analysis by using participants themselves as the test variable and the behaviours as the population sampled, effectively transposing a correlation matrix table by factorising the rows instead of the columns¹⁵⁾. The final factors extracted represent groups with shared perceptions of a given topic.　Participants rank a set of heterogeneous statements onto a pyramid-shaped distribution grid, termed a Q-grid.　This provides a standardised set of data that captures individual decision-making relative to each participant, enabling systematic and rigorous quantitative analysis^14,15).　The process of sorting and arranging the statements onto the Q-grid is termed ‘sorting’ and the completed Q-grid is referred to as a Q-sort.　The interpretation of factors is then supported via qualitative data from a post-sort questionnaire.　Q-methodology has previously been applied to explore diverse topics in HPE that are confounded by subjectivity, such as attitudes to teaching and learning¹⁶⁾, professional identity¹⁷⁾, and the modernisation of postgraduate training¹⁸⁾, but has not been used to examine attitudes towards unprofessional behaviours.

Thus, we anticipate that a Q-methodology study would provide empirical evidence to further our understanding of unprofessional behaviours through differing cultural lenses, uncover viewpoints missing from current professionalism discourses, and generate new insights to support the meaningful integration of local professional values into teaching programmes and national guidance.　Our research question was therefore to examine how healthcare professionals, students, and HPE staff, in Japan and the UK, conceptualise the weighting of unprofessional behaviours, with a secondary focus on identifying cultural perspectives within those constructions.

Study context and ethics approval

Japanese and UK cultural contexts were selected owing to their opposing positions on Hofstede’s cultural dimensions of ‘power-distance’, ‘uncertainty avoidance’, and ‘individualism versus collectivism’,¹⁹⁾ and the study builds on an existing research partnership between the institutions involved.　This selection also offers the opportunity to compare ideas of unprofessionalism in a ‘dominating’ and a ‘non-dominating’ culture within HPE.　Furthermore, significant differences in how hierarchy impacts interpersonal behaviours towards others between Western versus East Asian contexts have been well described²⁰⁾, making both cultures ideally suited to exploring how socio-cultural norms exert environmental pressures.　These pressures, in turn, influence what is considered unprofessional, mediating further actions taken on encountering such behaviours to align with socio-cultural expectations.

The study was conducted at two large public medical schools: a central-London medical school in the UK, and the prefectural medical school in Fukushima, Japan.　Ethics approval was obtained from Fukushima Medical University’s General Ethics Committee (approval number 2020-009) and Imperial College London’s Education Ethics Review Panel (1920-053).　The research team comprised Japanese and British education specialists and clinicians, all of whom are experienced qualitative researchers in HPE and hold higher degrees in education.　Awareness of researcher reflexivity was paramount, with the varied personal experiences of both cultures within the team informing our assumptions and beliefs.

Study design and setting

Our study is grounded in a social constructivist paradigm, which takes the stance that individuals operating in a defined cultural context form their understanding of unprofessional behaviours through interaction with others and subsequent reflection^21,22). This is further supported by social construction carried out by participants in real time during the Q-method process, through sorting the statements and constructing meaning to their ranked responses.

Q-set and grid development

Literature review supported the initial generation of the statements, which were mapped against the taxonomy of unprofessional behaviours described by Mak-van der Vossen et al⁷⁾, and the GMC framework, ‘Good Medical Practice’ (Table 1)⁹⁾. Statements underwent a process of peer review and revision within the research team to generate a balanced and evidence-based set of statements reflective of unprofessional behaviours encountered in practice that were relevant to both cultural contexts.　Statements each focused on a single behaviour carried out by a medical student, with extreme or unlawful behaviours excluded.　Native Japanese speakers within the team translated the statements for the Japanese cohort, which were professionally back-translated.　The statements were successfully piloted with no changes made to the statements, giving a final statement set—termed a ‘Q-set’—of 48 statements (Table 1).　We used a forced-choice pyramid distribution grid, scaled from −5 (least unprofessional) to +5 (most unprofessional), which ensured all statements were ranked and supported standardisation between the completed grids.　This attention to standardisation is an important quality marker with a by-person approach because incomplete grids would not be suitable for factor analysis^15,23). An overview of the Q-set development process is included in Fig. 1.

Table 1.

The 48 statements mapped to the four domains of unprofessional behaviours as described by Mak-van der Vossen et al 2015 and against the relevant professionalism domains according to the General Medical Council (GMC) professionalism framework ‘Good Medical Practice’ (UK).

Fig. 1.

A flow diagram of the research process including the development of the Q-set, the Q-sort procedure and the analysis.

Participant sample

We used strategic sampling to recruit medical students, education faculty members, and administrators; i.e., groups identified by the research team as key stakeholders with respect to unprofessionalism.　Using Stainton Rogers as a guide²⁴⁾, we aimed for between 40 and 60 participants at a 2:2:1 ratio of medical students, medical faculty and clinical educators, and administrators, respectively.　The final sample consisted of 58 participants (39 Japanese and 19 from the UK), which fits within expert guidance on sample size respective to statement number for a suitably powered Q-method study¹⁵⁾. All participants in the Japanese cohort identified as Japanese, whereas the UK sample included participants of East Asian background, although none were Japanese.　Participation was voluntary following informed consent.　No reward was offered for participation in the Q-sort or post-sort questionnaire.

Q-sort procedure

Informed consent and data collection were completed online via Q-methodology software²⁵⁾ between May and November 2021.　Participants accessed the Q-sort anonymously remotely via QR code and completed the Q-sort and post-sort questionnaire independently.　A flow diagram of the Q-sort procedure and overall research process is outlined in Fig. 1 and an example of a completed Q-sort is shown in Fig. 2.　The UK cohort undertook the study in English whilst the Japanese cohort participated in Japanese, with materials professionally back-translated for quality assurance.

Fig. 2.

An example of a completed Q-sort, where all 48 statements of unprofessional behaviours have been ranked on the Q-grid according to perceived severity from −5 (least unprofessional) to +5 (most unprofessional).

Data analysis

Factors were extracted using the open-source software KenQ²⁶⁾. Primary components analysis and varimax rotation were discarded on ontological compatibility principle²⁷⁾ as they result in a single ‘mathematically best’ solution that maximises factor separation to reveal the ‘simplest structure’ to the data^28,29). Instead, we used centroid factor analysis with judgemental rotation of factors to provide an infinite number of statistically valid solutions²⁹⁾, allowing the qualitative data and researcher interpretation to derive a solution best representative of reality.　This approach, advocated by Q-methodology experts, ensured that decisions were based on theoretical grounds²⁹⁾.

As per Q-method convention, we sought solutions for up to seven factors, choosing the final number of factors guided by our pilot data and the validated statistical acceptance criteria (Table 2)¹⁵⁾. The threshold for significance of factor loadings at p<0.05 for each Q-sort was calculated as 2.58 (1/√(no of statements)), with factors rejected if they had fewer than two significantly loading Q-sorts or fewer than two Q-sorts with loadings greater than the standard error, calculated as 1/√(no of statements)¹⁵⁾. We applied the Kaiser–Guttman criterion, retaining factors with eigenvalues greater than 1²⁷⁾, and Humphrey’s rule, which supports the retention of factors where the product of the two highest factor loadings is greater than twice the standard error^15,27). Factor arrays were created from the arising factor estimates excluding confounded Q-sorts (i.e., Q-sorts that significantly loaded onto more than one factor)¹⁵⁾. Resulting factors were iteratively interpreted alongside the post-sort questionnaire.　The questionnaire data were semantically and then latently coded, with themes reflexively interpreted with member checking.　Awareness of researcher perspectives in the interpretation of results was maintained throughout, with meaning negotiated between team members at regular meetings, supported by memo-taking to ensure rigour in the qualitative arm of the study³⁰⁾.

Table 2.

Factor solutions generated and acceptance criteria.

The Kaiser–Guttman criterion is the retention of factors where the eigenvalue >1.

**Humphrey’s rule is the retention of factors where the product of the two highest factor loadings is greater than twice the standard error (with standard error calculated as 1/√number of statements).

***Factor loadings calculated as 2.58 (1/√number of statements).

Solutions up to seven factors were calculated and a four-factor solution explaining 53% of the variance accepted.　Table 2 shows how the statistical acceptance criteria were applied, with the four-factor solution emerging as the optimum balance between total variation explained and factor retention.　Confounding sorts (i.e., sorts which did not distinguish between two factors) were excluded, leaving 30 Q-sorts that contributed to the factor arrays (20 from Japan and 10 from the UK).　Table 3 describes the participant demographics contributing to each factor.　The four factors arising were interpreted as lenses through which participants conceptualised unprofessional behaviours and labelled according to arising themes as agreed by all research team members: clinical responsibility, personal responsibility, moral responsibility, and group responsibility.　This solution produced a single consensus statement—‘ignores emails from teaching/administrative staff’—that was ranked similarly across all factors.　All remaining statements distinguished between at least two pairs of factors at p<0.05 or p<0.01, meaning that they were ranked in a significantly different position in one factor relative to another.　Table 4 shows the ranking of statements within each factor.

Table 3.

Factor breakdown by participant contribution.

Table 4.

Factor arrays showing the ranking of statements within the factors.

Asterisk (*) indicates significance at p<0.01; otherwise p<0.05.

D denotes a distinguishing statement, where the statement in that position has been ranked in a significantly different position compared with its position in other factors.　C indicates a consensus statement, where the statement has been ranked similarly across all factors.

Factor 1: Clinical responsibility

Clinical responsibility was the largest and sole international factor extracted, accounting for 18% of the variance with nine contributors (five from Japan——one administrator, three clinical educators and one medical student—and four from the UK—three clinical educators and one medical student) (Table 3).　Participants conceptualising unprofessional behaviours under this factor used a clinical lens to sort the statements, ranking behaviours perceived as most harmful to patients as the most unprofessional.　Patient safety was the priority amongst this group in distinguishing between behaviours, regardless of whether the participant had a clinical background.

“Offering clinical advice or information they are unsure is correct [is most unprofessional] because it directly harms the patient.” Educator (Japan)

“Faking signs or faking the history [is most unprofessional because it] can lead to terrible consequences for the patient.” Medical student (UK)

“Of all the statements, in terms of medical safety, [falsifying test results] is the behaviour that should most be avoided.” Administrator (Japan)

Statements affecting patient interactions and care delivery—notably ‘reports examination as normal when the test has not been done’ and ‘offering clinical advice without appropriate knowledge or supervision’—were ranked as more unprofessional in this factor compared with the others and were considered foundational tenets in medical practice in both cultures.　UK participants additionally mentioned regulatory organisations and professional consequences in justifying their statement ranking, citing fear of investigation by a professional body as an influence on their choice of most unprofessional behaviour.

“Fabricating the patient’s history [undermines] the foundations of the relationship with the patient.” Medical Student (Japan)

“Examining a patient with insufficient consent could lead to a malpractice investigation.” Administrator (UK)

Statements describing unprofessional behaviours involving authority figures in an education context were ranked as least unprofessional, such as ‘arguing with a tutor about the relevance of the teaching session’ and ‘communicating with a tutor in the middle of the night to ask non-urgent questions’.　Participants from both cultural contexts felt that behaviours challenging power structures could be viewed positively in a professional setting.

“Arguing with a tutor [is less unprofessional] because asserting your own opinion is an important skill.” Educator (Japan)

Factor 2: Relational responsibility

Relational responsibility was the next largest, and a Japanese-only factor, accounting for 13% of the variance and derived from the views of seven contributors consisting of one administrator, one clinical educator and five medical students (Table 3).　Group characteristics for this factor were noted to differ from the other factors as the viewpoints came from medical students and two early career professionals (i.e. within the first 5 years of work).　Relational in this context represents a wide range of social bonds, from transient connections where social spaces intersect to deeper transactional relationships as part of a group.　Participants applying this lens were more aware of the impact of their actions on a wider group, choosing behaviours with the potential to affect image or cause disruption to social cohesion as most unprofessional (e.g., ‘smelling of alcohol on a clinical placement’ and ‘fabricating part of the patient history’).　Statements describing other behaviours perceived to negatively impact group dynamics were also ranked as more unprofessional in this factor compared with others; this included ‘frequent late attendance’, ‘does not attend without notice’, ‘uses phone in front of patients’, and ‘fails to appreciate the value of clerking patients over book learning’.

“Unauthorised absence and frequent late attendance are basic rules for living in our society and cause a great deal of inconvenience to those around you [if broken].” Medical student (Japan)

Behaviours that maintained harmonious social relationships through avoiding disruption to others or preventing others from getting into trouble were felt to be least unprofessional.

“Although it’s not an entirely good thing to do, there is little trouble caused directly to others.” Educator (Japan)

Behaviours felt to be less unprofessional in this factor also followed this pattern, including ‘presenting work including the name of someone who did not contribute’ and ‘witnessing cheating and not taking appropriate action’.　Through this lens, behaviours felt to be unavoidable or due to personality were considered as less unprofessional.

“I felt that the least unprofessional behaviours might be down to the personality of that individual.” Educator (Japan)

Factor 3: Moral responsibility

Moral responsibility accounted for 12% of the variance and was a predominantly UK-based factor, comprising six viewpoints made up from one Japanese administrator, one UK administrator, two UK clinical educators and two UK medical students (Table 3).　Using this construction to understand unprofessional behaviour, participants ranked statements describing breaches in trust between doctors and patients as the most unprofessional; for example, ’examining a patient with insufficient consent’ and ‘disclosing clinical information to family without consent’.

“Offering clinical advice without appropriate supervision and [falsely] reporting findings as normal go directly against the principles and foundations of medicine.　It seems that a person who would undertake such behaviours would need to consider their core motivation for medicine.” Educator (UK)

Participants also clearly demarcated dishonest and colluding behaviours as more unprofessional—including ‘signing a document under someone else’s name’, ‘accepts gifts from patients without considering possible motives’ and ‘signing in for absent peers’—ranking these statements in higher positions relative to their position in other factors.

“Signing off with someone else’s signature is not only unethical but also worthy of [having your] medical licence suspended.” Medical Student (UK)

Behaviours such as ‘appearing disinterested’, ‘not giving feedback to others’, and ‘wearing casual attire in a clinical setting’ were amongst the least unprofessional.　This view of dressing casually contrasts strongly with opinions from Japan, where participants felt that personal presentation was a more important aspect to professionalism.

“People involved in medical care cannot [wear casual attire or an unkept lab coat] as this could be seen as unclean or lacking in hygiene.” Administrator (Japan)

A distinguishing feature from Factor 1 (clinical responsibility), where the clinical lens weighted behaviours affecting interactions with patients as more unprofessional, is that behaviours influencing patient interactions—namely ‘fails to show empathy’, ‘treats patients as symptoms or diagnoses’, and ‘fails to appreciate the value of clerking patients over book learning’—were felt to be less unprofessional under a lens of moral responsibility and thus ranked lower in this factor.

Factor 4: Personal responsibility

Personal responsibility was a predominantly Japanese factor, constructed from eight viewpoints generated from three Japanese clinical educators, four Japanese medical students and one UK medical student (Table 3).　Through this lens, participants ranked statements that disrupted personal relationships as most unprofessional, such as ’blaming the patient for own history taking deficiencies’ and ‘writing on Facebook about a patient encounter’.　This differs from Factor 2 (relational responsibility), where the focus was on maintaining group cohesion, as the behaviours isolated in this factor focused on one-to-one relationships.　Statements ranked as more unprofessional in this factor relative to others included ‘laughing at a colleague’ and ‘arguing with a tutor’; least unprofessional were behaviours that only affected the individual, such as ‘does not take initiative’ and ‘seeks minimally acceptable level of performance’.

“Because humanity is important in medical care.　Even beyond illnesses there are many things that can be treated by words alone.” Educator (Japan)

“I felt that [these behaviours were less unprofessional] because they cause little disadvantage to others”.　Medical Student (Japan)

In this lens, participants spoke of a concern for how actions are perceived by others.

“I thought that the clearest examples [of unprofessional behaviour] were those that [would be] recognised as unprofessional behaviours, even if judged by non-healthcare professionals.” Educator(Japan)

In this paper, we examine cultural differences between Japanese and UK stakeholders in the perception of medical unprofessional behaviours via the innovative approach offered by Q-methodology.　Our findings reveal an internationally shared understanding of unprofessional behaviour through the lens of patient safety, which operates alongside culturally constructed unique viewpoints related to relational, personal, and moral responsibilities.　Crucially, these are viewpoints currently lacking or missing from leading professionalism frameworks.

Our findings align with our pilot data, which demonstrated an equivalent four-factor solution, and the lenses identified overlap with the professionalism domains of ‘respect for patients’, ‘responsibility for actions’, ‘ethical practice’, and ‘social responsibility’ described by Hilton and Slotnick³¹⁾. Additionally, our results further the literature in providing empirical evidence for the ideas of Martimianakis et al³²⁾, supporting a socio-cultural understanding of [un]professional behaviours.　Such a model contrasts with a purely educational view on unprofessional behaviours as a remedial problem, with the cultural viewpoints identified making clear the extent of the translation gap between professionalism frameworks and current practice in both contexts.

Significant in our findings are the Japanese viewpoints of relational and personal responsibility in distinguishing unprofessional behaviours, which are likely to be drawn from Japanese interdependence in identity formation relative to independence (a Western characteristic found in the UK context), leaving the Japanese sense of self incomplete without others³³⁾. Such a difference brings stronger ties to a collective, an increased focus on the maintenance of harmonious relationships, anticipation of the behaviour of others, and moderation of personal behaviours to prioritise in-group relationships³⁴⁾. Understanding these social norms leads to an appreciation of how behaviours that deviate from relationship preservation, cause perceived inconvenience to others, or can be anticipated to appear a certain way hold deep-seated cultural connotations of unprofessionalism that extend to medical practice.　It is significant to note the younger demographic that contributed to the factor of relational responsibility, including the persistence into early career professionals in both clinical and non-clinical settings.　This highlights earlier influences from education in the Japanese school system which may represent a more readily accessible construct of unprofessional behaviours for medical students to drawn on at this career stage.　Persistence of the viewpoint into early career indicates its importance as a foundation for professional development in the Japanese context, raising the possibility that this represents a stage of pre-professional development similar to a Western concept of ‘proto-professionalism’ proposed by Hilton and Slotnick³¹⁾.

Within the UK cohort the factors arising were narrower by comparison, reflecting guidance from professional regulatory bodies such as the GMC’s Good Medical Practice, which has been enshrined in national training and professional consciousness since 1995⁹⁾. However, the lens of moral responsibility was statistically distinct, highlighting a viewpoint considering personal motivation and ethical reasoning in conceptualising unprofessionalism, rather than using a purely clinical lens.　This difference may represent the shift towards professional values that encompass physician wellness and intrinsic motivation, which has been recently described in HPE literature¹²⁾.

Our research has several key strengths.　In using Q-methodology, we have a robust and validated method to extract and examine unprofessionalism constructs across contrasting cultural contexts, strengthened by judgemental rotation, standardisation of Q-sorts, and qualitative data to support factor interpretation.　The extracted factors are unconscious to participants and cannot be predicted by researchers^15,18), reducing the potential for response and expectation bias.　Our Q-sort was suitably powered¹⁵⁾, and the extracted factors conceptually well supported.　However, Q-methodology is not generalisable outside of our cohort¹⁵⁾, and although a process was implemented for rigour in statement generation, all unprofessional behaviours may not have been represented in the study.　This opens the possibility that other lenses conceptualising unprofessionalism may co-exist.　In addition, our study recruited from a range of HPE stakeholders but excluded patients.　Furthermore, differences in education and healthcare systems influence the sociocultural environment and are therefore likely to contribute to how unprofessional behaviours are distinguished and considered in practice; however, such evaluation was beyond the scope of this project.　Therefore, further studies are needed to validate and elaborate on our identified lenses and search for additional perspectives not captured from our cohorts, including those of patients and relatives.

Given the shared problem of medical unprofessionalism in healthcare contexts globally, sociocultural examination of the translation gap between professionalism frameworks and perceptions of unprofessional behaviour offers valuable insights and unique cultural perspectives, worthy of integration into teaching and national frameworks for local cultural relevance and reflective of professional practice.　This may help to appropriately set professionalism guidelines relative and responsive to the needs of our students and patients locally, and importantly, guide educators in assessing professionalism concerns arising within these populations.

We express our deepest thanks to Mr Oliver Stanyon for conducting the Japanese-English translation and back-translation for this study, and for language-editing the manuscript.

This study was funded by a Japan Society for the Promotion of Science Kakenhi category C grant, number 21K10377 and the NIHR Applied Research Collaboration Northwest London.　The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.

The authors have no conflicts of interest to declare.

1.Please indicate which country you are based in

◯UK

◯Japan

2.Please indicate your occupation

◯Medical Student

◯Clinician Educator

◯Non-Clinician Educator

◯Administrator

3.If you are a medical student, please indicate your year of study

◯1

◯2

◯3

◯4

◯5

◯6

◯N/A

4.If you are an educator or an administrator, please indicate how many years you have worked in education

◯0-5 years

◯5-10 years

◯10-15 years

◯15-20 years

◯>20 years

◯N/A

5.Do you have any international experience (for example living or working in another culture)? If so, please give brief details below. This may include a description of your role, duration of the experience, the countries/cultures involved, etc.

6.Which 2 statements did you feel are MOST unprofessional? Please explain WHY you felt these were the most unprofessional of the behaviours listed. You can view your Qsort by clicking the yellow “view Qsort” button at the bottom of the screen.

7.Which 2 statements did you feel are LEAST unprofessional? Please explain WHY you felt these were the least unprofessional of the behaviours listed. You can view your Qsort by clicking the yellow “view Qsort” button at the bottom of the screen.