Despite the annual movement of large numbers of doctors in the NHS and healthcare systems worldwide, little is known about the new doctors’ perspective on their integration into their new working environment. Many will undergo a formal induction to the work environment, but there is also a need to introduce newcomers to the culture and practices of a specific group, build relationships and a sense of belonging and acceptance over a longer period. Most prior research has focused on preparing doctors for a single aspect of acute care, or the evaluation of initial training in acute care skills.1 The efficient delivery of the technical and required aspects of hospital induction has also received attention.2 However, an ED is an ‘ultra-adaptive environment’ where risk is embraced and managed, and in such an environment people’s skills and capabilities take precedence over rigid processes.3 Little is known about the perspectives of the rotating junior doctors about induction and socialisation1 4

Understanding early socialisation is important because of its rapid influence and lasting effects on newcomers as they move from outsider to ‘integrated and effective insider’. Socialisation affects job performance and satisfaction, organisational commitment and retention.5 For the newcomer, socialisation is a problem of uncertainty reduction, making the work environment more predictable for both task content and social relations6 In the ED, full of ‘diagnostic uncertainty, high decision density and cognitive load, time and resource limitations, interruptions, frequent transitions of patient care, and poor feedback’, the challenge of uncertainty reduction for newcomers is acute.7

Failure to guide new doctors effectively through early socialisation in the ED risks increasing uncertainty, with negative consequences for patient safety and the well-being of both new and supervising doctors.8 We aimed to understand early socialisation from the perspective of new doctors.

This qualitative study took place following the 2019 August induction of new doctors at the Emergency Department of Addenbrooke’s Hospital, Cambridge, seeing around 120 000 patients per year. During the study, the ED employed 42 junior doctors and 23 full-time equivalent consultants. The first author, who conducted the observations, visited the department 14 times between August and November, including three full induction days, for a total of 62 hours. The research objective of trying to understand the perspective of new doctors about induction and early interactions was explained to participants. All participants were guaranteed anonymity, and consent was reconfirmed.

The first author performed observations and informal interviews with new doctors, experienced ED doctors, nurses and consultants, across all shifts. Specifically planned, semistructured interviews were not used. Observed interactions between, for example, new doctors and consultants as they discussed patients, were often used as a trigger for interviews, often with a question such as: ‘how did that conversation go?’ after a new doctor talked to a consultant. However, open questions were also used, such as asking a consultant ‘How do you find working with the new doctors?’ with more specific follow-up questions as required. A consultant’s meeting to review new doctor progress was observed, as well as formal departmental and hospital induction sessions. Nurses were interviewed to determine their impression of consultant and new junior doctor behaviours. Contemporaneous handwritten notes were taken on conversations and observations.

The research is informed by grounded theory: an approach which does not offer a single research method and exists in several variants. All versions, including the current study, share four core ideas. Emergence (‘follow the data’ rather than impose preconstructed categories): no formal interview schedule was constructed in advance of conversations. The interviewer has no medical training and to minimise bias, no literature on EDs or medical induction and socialisation was consulted. Constant comparison (iterate between research data, both existing and emerging, and the literature, to ‘ground’ (inform and support) the analysis). Theoretical sampling (select data sources for relevance to the emerging model (validating, adjusting or extending); emphasise quality and relevance of the source rather than frequency). Finally, theoretical saturation (stop data collection and analysis when themes are repeated but no new themes appear). ‘In vivo’ codes, terms used by informants during conversations, were used to stay close to their experience.

The perspectives of the three types of participants were triangulated to help develop themes. Following Gioia et al,9 we explicitly demonstrate how theoretical concepts were derived from interview data (see figures 1 and 2). Coding was carried out by the first author and then assessed by two coauthors.

Unhelpful consultant behaviours.

Unhelpful consultant behaviours: data display.

No patient involvement.

We aimed to understand socialisation of new doctors in an ED. Junior doctors rotate in most medical training environments worldwide and their integration into an ongoing, complex work environment has not been well studied. Despite having undergone a formal induction, we found that both negative and positive experiences turn on the desire of new doctors for conditions which support both professional learning and patient safety. These are more set by the behaviours and attitudes of the consultants than the formal induction programme.

Five consultant qualities were identified as helpful: creating a learning space, empathy, interest, admitting fallibility and being approachable. The unhelpful behaviours mirrored these—lack of interest, overconfidence, disregarding junior doctor input, being unapproachable, or scaring or humiliating them. Unhelpful behaviours were seen as unsafe.

Our findings align well with research exploring psychological safety in teams10–12: a shared belief ‘that the team is safe for interpersonal risk taking’. This increases the chances of ‘effortful, interpersonally risky, learning behavior such as help seeking, experimentation, and discussion of error’, the behaviours that need to be rapidly established in an effective ED. In hospital settings, psychological safety is associated with ‘learn-how’ activities such as ‘experimentation and collaborative problem-solving’. A meta-analysis covering 22 000 individuals and around 5000 groups found significant and positive associations between psychological safety and learning behaviour, work engagement, information sharing, creativity, commitment, satisfaction and task performance: conditions which support the technical and social uncertainty reduction effort required during socialisation.13

This safety must be balanced by a respect for challenge, for not spoon-feeding new doctors. It is important here to avoid a false dichotomy between supporting and challenging. Cumulatively, the five behaviours in figure 3 help create an environment which facilitates quick and safe learning. They mitigate power imbalances and help demonstrate ‘leader inclusiveness’,14 as well as minimising worries newcomers might have about being ‘humiliated or rejected’. They also facilitate new doctor contributions.15

However, learning behaviour, such as sharing information, discussing errors, asking questions, seeking help and asking for feedback, rather than projecting confidence, carries risks. Admitting fallibility can be seen as ignorant (you ask questions), incompetent (you ask for help and make mistakes), negative (you point out mistakes) or disruptive (you ask for feedback and insist on sharing information).

The concern that calling for help indicates failure to cope is borne out in other US and UK studies.16 17 In medicine, ‘the perceived need for impression management to protect one’s professional image is extremely high’, partly because the impression projected can affect respect, promotion and rewards. These factors reinforce the value of consultants explicitly sharing both their evaluation criteria and their own vulnerabilities with new doctors.

In the ED, formal power differences exist which inhibit newcomers from speaking up. While the positive behaviours support ‘structure without rigidity’,11 the unhelpful behaviours and attitudes reinforce hierarchal difference, close learning space, discourage information seeking and put patient safety at risk.

The unhelpful consultant behaviours that we identified are consistent with other research in healthcare settings. ‘Scare and humiliate’ behaviours block ‘mutual self-disclosure’ and hinder the process of ‘team learning’.11 The willingness of newcomers to join in problem-solving also weakens in hostile conditions,18 disengagement is more likely19 and learning behaviour is reduced.20 Autocratic behaviour, inaccessibility or a failure to acknowledge vulnerability all can contribute to team members’ reluctance to incur the interpersonal risks of learning behaviour.10 21 22 Achieving a switch to ‘leader inclusiveness’ is known to have a positive effect in underperforming hospital units.15 Creating a safe learning environment and maintaining patient safety are symbiotic.

A meta-analysis of socialisation research5 emphasises the importance of three key indicators of newcomer adjustment: self-efficacy (learning tasks, confidence in the role), social acceptance (feeling liked and trusted) and, more technically, role clarity (understanding tasks, task priorities and time allocation).

Our work, together with existing literature, allows us to make some tentative recommendations about how to quickly and safely induct and socialise new doctors to an ED. We accept that the benefit of these is not quantified or proven by our work, we also acknowledge that early integration of new doctors must also include aspects of mandatory training (eg, fire safety) and administration (information technology training, etc).

Consultants should talk explicitly about inherent uncertainty in clinical practice: unpredictability; the necessity of deciding with incomplete, shifting information; repeated novelty; weighing of probabilities; trapped risk; and the challenge of managing yourself in these conditions. These behaviours would lead to ‘a more affirmative attitude towards doubting’12 and the sharing of doubt between hierarchical levels. An explicit distinction needs to be made between technical, scientific knowledge, and clinical judgement, with an emphasis on the difference between knowing facts and the challenge of acquiring experience, confidence and skill over time. Acknowledgement of consultant fallibility is likely to be seen by new doctors as supportive and helpful. Confidential mortality and morbidity meetings are a good vehicle for this. There is a fundamental distinction to be made between knowledge, ‘all sharp edges, confident certainty, and clean light’, and wisdom, where we are ‘learned about our ignorance’ and share the craft and its ways of thinking through uncertainty.23

Involving new doctors in decisions should facilitate rapid learning. Team leaders actively seeking out knowledge from lower status team members encourages speaking out.

Clarify consultant support for new doctors and explain how learning spaces will be created through challenge, questions, opportunities to try and ‘live’ explanations of why and how consultants make decisions. Share consultant values and objectives with new doctors. Conversely, explicitly state what the consultant body aim to avoid: a lack of empathy; being invisible, unavailable, unhelpful; being unpredictable in our responses (though we get tired and stressed, too); scaring and humiliating; creating unsafe work conditions by not listening, or by discouraging dialogue. Consultants should reason out loud, talking through their analysis and decision processes and inviting comment in order to facilitate learning.

Explicitly facilitate social relations. Social acceptance is an important indicator of newcomer adjustment and associated with other positive work outcomes. It may be helpful to consider an ice breaker social event when new doctors arrive and using a photo board of staff members. Identifying staff levels explicitly using lanyards or badges may also be helpful.

Clarify and set expectations at induction. Reducing uncertainty could include clarifying development targets over time, distinguishing, where possible, between technical knowledge and experiential learning over time (help seeking; situational awareness; differential diagnosis under pressure). Saying how new doctors will be evaluated in relation to, for example, situational awareness and help-seeking behaviour versus overconfidence is also important. Clear staged assessment criteria and a strong value attached to help seeking are important. So, ‘In the first weeks we welcome a lot of help-seeking. After that we will give guidance on where and how it can be modified.’

This is a small qualitative study in a single department. We did not formally assess the degree to which new doctor perceptions of successful socialisation might align with, or diverge from, those of consultants and this would be a useful follow-up investigation. Nor did we link consultant behaviours to specific trainee needs, such as coping with uncertainty in diagnosis. The attitudes of new doctors to formal induction and its relationship to motivation and integration were not fully explored.