Results

Results of the review

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram (Figure 2) reports the number of studies identified, included and excluded. A description of the characteristics of the included studies is available in Appendices 3–7. The main search returned 4554 articles. Of these articles, 114 were included in the review. A further 27 articles were returned from the additional searches (i.e. citation searches, grey literature searches, requests to stakeholders and supplementary searches).

FIGURE 2 The PRISMA flow diagram. Reproduced with permission from Price et al.1 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure. (Figure not reproduced here; see the original article.)

Study characteristics

Of the 141 studies included in the review, 64% related to North America, with 14% coming from the UK. The majority of studies were published between 2008 and 2018 (72%). Forty per cent of the articles were commentaries, 37% were research papers and 31% were case studies. Of the research papers, 64% were quantitative, 19% were literature reviews and 14% were qualitative. Forty per cent of the articles were about junior doctors/residents and 31% were about practising physicians, whereas 17% were a mixture of both (with some including medical students). Forty per cent of studies focused on remediating all areas of clinical practice, including medical knowledge, clinical skills and professionalism. Twenty-seven per cent of studies focused on only professionalism and 19% focused on knowledge and/or clinical skills. Thirty-two per cent of studies described a remediation intervention, 16% outlined strategies for designing remediation programmes and 11% outlined remediation models. Appendices 3–7 provide a more in-depth overview of the characteristics of the included studies.

Conceptualisation of remediation and implication for the review

The purpose of this review is to gain a better understanding of what is working in remediation programmes to produce their effects. There is already an extensive literature on how doctors or trainees learn at different stages of their career, but remediation is a particular type of intervention, describing the process by which doctors are returned to safe levels of practice after having fallen below accepted standards. We decided in the early review meetings that the most useful approach to focus our review was to start to identify what was unique about remediation compared with, for example, medical education or, indeed, education more generally.

Remediation is, by definition, concerned with rectifying underperformance or poor behaviour. In the case of medicine, a remedial intervention is linked to a judgement on an individual’s fitness to practise medicine safely (i.e. relating to threshold standards of competence). This link between performance standards and patient safety associates remediation with failure, because the safe practice of medicine is the sine qua non of the medical profession.8,40,41 However, at the same time, implicit in the term ‘remediation’ itself and the process it describes is the potential for remedy.8 Therefore, remediation is conceptually nebulous. A study of stakeholders involved in remediation in Canada revealed that they held a dual conceptualisation of remediation (i.e. remediation was understood as both an educational and a regulatory intervention).13 This latter conceptualisation directly links to the inherent association between remediation and the minimum standards required for practice. Remediation may be seen as a direct threat to professional practice, which is a particular challenge as it discourages doctors from self-identifying as needing help and engaging with remediation processes at an early stage.

In this sense, these interventions ‘work’ when they can create the opportunities and resources to overcome the barriers that are unique to remediation. Some of the more theoretically rich literature, combined with early discussions with stakeholders, led us to conceptualise remediation not so much as an educational intervention, but a change in practice and behaviours that would influence the learning experience. Reference to behaviour change theory early in the review, combined with stakeholder and review team discussions, enabled us to tease out the various outcomes that were relevant to CMOc building.

This conceptualisation has informed the development of the programme theory and, therefore, the structure of the way in which the results are reported here. Although remediation is by no means a linear process, there is a logical sequence whereby certain outcomes would precede others. Accordingly, the results are grouped into three categories, based around the kinds of outcome that occur at these stages in a remediation process: (1) insight, (2) motivation and (3) practice change. Within these three broad categories, the results are organised around the intervention strategies that are used to facilitate these outcomes.

Despite the logical sequence of the outcomes, it is important to emphasise that, early on, we identified remediation as a non-linear and an iterative process. For example, although a degree of insight into the nature of a problem may be an important outcome that precedes a remediating doctor from being motivated to change, insight can further develop all the way through a remediation programme.

The results are presented here in such a way as to ensure that they are both accessible and academically robust. Each section begins with a narrative explanation of how the different parts of the remediation process work to produce a particular outcome. This is then followed by a realist explanation in the form of a series of CMOcs, alongside a comprehensive explanation of the evidence base underpinning the CMOcs. The realist explanations for the CMOcs will include reference to substantive theory, where necessary. The purpose of structuring the findings in this way is so that the findings can be read and understood without having to read the realist explanations. The realist explanations of how we arrived at those findings are important for transparency. To provide a full and coherent explanation of our findings, there are elements of discussion in this chapter.

Summary of CMOcs

Table 2 provides a summary of the 29 CMOcs identified through our research in their three main clusters. This constitutes our final programme theory. The programme theory is also presented graphically in Figure 3.

Facilitating remediation: developing insight

Based on our analysis and interpretation of the data, a remediating doctor must have insight for remediation to be successful. If they do not have insight going into remediation, then the remediation programme must work to help them develop insight. Insight is mentioned in a number of the returned papers,7,40,42–50 but two papers,42,43 in particular, have been useful in conceptualising insight in the context of remediation. Hays et al.43 note that the term insight has been poorly defined in the setting of medical education, and that, in this setting, insight may be a combination of awareness of one’s own and others’ performance and the capacity to reflect. Brown et al.42 discuss insight as a readiness to explore, intellectually and emotionally, our own and others’ thoughts and behaviours. Taken together, and considering the unique circumstances surrounding remediation, insight can be understood as comprising two different aspects, which we have used to help delineate some of the outcomes of the remediation processes in this review:

1. a readiness to explore others’ perceptions of one’s own behaviour or performance, and to validate those perceptions
2. acceptance that one’s own performance or behaviour is divergent from accepted standards.

Providing safe spaces and using advocacy to develop trust in the remediation process

As remediation is a threat to a doctor’s professional identity, and may also be perceived as a threat to their career, feelings such as anger and mistrust may prevent a doctor from even engaging with negative feedback on their performance or behaviour.51 A recurring theme in the literature is that a remediating doctor is more likely to be ready to explore and to validate others’ perceptions of their behaviour or performance if they have trust in the people involved in remediation and trust in the process of remediation itself.

The actual word ‘trust’, although used on occasion,52,53 does not appear frequently in the returned literature. However, the idea of developing trusting relationships is implicit in a number of papers.49,50,52,54–62 Trust also emerged as an important theme in the stakeholder meetings. To develop this relationship, the literature points to features of remediation programmes that might bring about trust. In particular, the literature noted the importance of developing safe spaces and using advocacy to engender trusting relationships.

The literature describes the need to create a ‘safe space’ in which a confidential discussion can take place, without fear of sanction or judgement.49,52,54,55,57–62 A safe space is important because remediating doctors usually fear the consequences of being identified as needing remediation. One aspect of creating a safe space is to provide assurance of confidentiality. The literature refers to conversations in a remediation process that are not only confidential, but allow the remediating doctor the opportunity to express negative emotions, such as anger and shame, that might be invoked by being identified as needing remediation.44,54 Our stakeholders suggested that this provision of confidentiality may also facilitate the remediating doctor to express any mental health issues that may not have come to light.

Trust can also be developed through building personal relationships during the process of remediation. Sometimes, such a relationship is described in the context of a coaching model53–56,59,60,63–66 and elsewhere in terms of a mentoring programme.44,48,49,55,63,67–75 Regardless of the term used, a key focus in the literature is that someone in the remediation programme takes on the role of being an advocate for the remediating doctor.

Advocacy is described as offering ‘unconditional positive regard’76 in the context of a one-to-one relationship in a remediation programme. Empathy and validation are key to this role,75 which has been described as a ‘therapeutic relationship’.77 In practice, this may mean empathising on specific issues, such as working conditions being very challenging, and acknowledging that others have similar struggles,55 or that behavioural norms have changed since the remediating doctor did their training.76 Shapiro et al.53 point to the importance of validating a doctor’s good motivations, even if the ultimate goal is to reach a point where the remediating doctor understands that they were not acting in a way that was compatible with this motive.53

The person performing this advocacy role may be a senior doctor76 or a peer55,71 in a role labelled as either coach, mentor or other. A key point noted in a number of papers is that advocacy is deemed to be most effective when the individual fulfilling this function has no role in the summative evaluation of the remediating doctor or the final outcome of the process.54,56,76 The advocate does not necessarily have to be intimately involved with the delivery of the remediation programme, but should be familiar with the case.78 Our stakeholders suggested that this individual should be someone who is chosen by the doctor. Having this supportive relationship and non-judgemental encouragement allows for trust to develop, which, in turn, means that the remediating doctor is more likely to engage in the process and be ready to explore concerns about their performance or behaviour.

Realist analysis of providing safe spaces and using advocacy to develop trust in the remediation process

Extract 1 (Box 2) is typical of the way in which insight is addressed in the literature (i.e. a lack of insight is a barrier to remediation, but is potentially remediable). The emphasis on the importance of insight was mirrored in the stakeholder group discussions, in which a lack of insight was unequivocally associated with failure in remediation. The link between establishing safe spaces because of the reactions and vulnerability of the remediating doctor is drawn from data such as those in extracts 2–6 (see Box 2) and was verified explicitly in the stakeholder group discussions. The stakeholder group, both within meetings and in later e-mail exchanges, also emphasised the importance of trust as an important outcome when someone feels psychologically safe.

1. Paice:46

The question of insight is one that tends to come up whenever underperformance is discussed . . . studies on this phenomenon have shown that when offered extra training in the activity concerned the poor performers improve . . . This research suggests that lack of insight . . . [is] not an unsurmountable obstacle to remediation.

2. Egener:76

Confidentiality is critical to the physician/client’s honest disclosure of events and his or her personal reactions to those events.

3. Kalet et al.:54

It [coaching relationships] allows a safe space for a learner to express vulnerable emotions such as anger, shame, sadness, or to explain cultural issues that may have arisen. These issues generally must be recognized and validated before the majority of remediation can occur.

4. Kimatian and Lloyd:44

A compassionate and empathetic approach to the remediation process should include an opportunity for the resident to have a frank, honest, and private discussion with a trusted mentor, advisor, or objective coach or counselor.

5. Samenow et al.:79

The priority of the program’s first day is to create a safe environment conducive to transformative learning. Ground rules, including confidentiality, are established.

6. Shapiro et al.:53

In developing a system for reporting, evaluating, and responding to professionalism lapses, we created a process that is confidential, centralized, clear, and respectful.

7. Shapiro et al.:41

Denial is often based on a profound fear of the external and internal consequences of accepting the judgement of others. Thus it is important to create an environment that is as psychologically safe for the resident as possible.

8. Whiteman and Jamieson:49

We start by discussing their situation and exploring their narrative account, trying to harness the work done in their last GP appraisal – often viewed in a positive light as a piece of work done by a non-judgemental peer.

9. Papadakis et al.:58

The core message of this [first] meeting is: ‘We have important feedback that you need to hear . . . We value you, and we want to support you in your efforts to change’.

10. Paglia and Frishman:80

For a concerned programme director, denial that there is even a problem is often the major barrier to assisting a resident in the process of remediation. One way to circumvent denial is to provide a safe environment.

11. Egener:76

While the process is framed as education, in truth it is a therapeutic relationship, not unlike the patient–physician relationship. Empathy, genuineness, and unconditional positive regard are the most transformative elements of such relationships.

12. Mar et al.:71

The events leading to the decision for the need of and the conduct of a remediation plan are likely to involve some element of controversy. The resultant stress may be mitigated by the presence of an advocate at any meetings with the resident, by ensuring a fair process.

13. Smith et al.:78

The resident is under stress during this process and benefits from general encouragement in efforts to improve from the cheerleaders.

14. Sargeant et al.:55

To his surprise, as he talks with the facilitator and shares his dismay at these results and his frustration about not being able to do anything, he finds that the facilitator understands his perspectives and confirms that he is not alone – others feel this way, too.

15. Egener:76

The physician/consultant [person acting as coach or mentor] validated his sense of unfairness at having to change behavior that had previously been tolerated.

16. Kalet et al.:54

It is worth the effort to identify a ‘neutral party’ to conduct at least some of the remediation so that program leaders are freer to make difficult decisions if need be . . . it ensures that the learner can develop in a safe environment without threat of high stakes reprisal.

17. Warburton and Mahan:56

We advocate use of the term coaching to signify a process by which a struggling trainee receives individualized mentorship and guidance from an individual who is committed to his/her success and is not directly involved in his/her reassessment.

18. Egener:76

The consultant assumes the roles of coach, mentor, or teacher, but for several reasons the referring source must remain the sole evaluator of success.

19. Shearer et al.:47

Most institutions implemented some form of mentorship program for residents in remediation. It was commonly specified that a mentor . . . must be someone who would not play an evaluative role with the resident.

Therefore, CMOc 1 establishes the context that remediation is feared by those who engage in the process and that they will not engage if they do not feel psychologically safe (Table 3). CMOcs 2–4 refer to the role of psychological safety as a generative mechanism that leads to outcomes of trust and, because there is trust, a readiness to explore performance or behavioural concerns. Only one paper41 that was returned in the remediation literature search explicitly mentions psychological safety (see Box 2, extract 7). Psychological safety was examined as a concept at the suggestion of a member of the review team. We subsequently conducted a supplementary literature search to examine whether or not the concept was appropriate to help understand the causal mechanisms at work in building trusting relationships.

The literature returned in the supplementary search showed that psychological safety was specifically concerned with developing trust within the contexts where perceptions of risk and fear are prevalent. Extract 3 (Box 3), from the most recent study83 we could find on psychological safety in medical education, also discusses its importance in relation to building relationships of trust. This focus on overcoming perceptions of risk and threat is well suited to the environment of remediation and provides a credible explanation of how intervention strategies can create the contexts that lead to trusting relationships in a remediation programme, in turn leading to a readiness to explore a problem. As a mechanism, psychological safety describes a process in which the opportunities and resources provided by the programme (i.e. the provision of a safe space and the advocacy role) alter the reasoning of individuals, enabling them to take the psychological risk of exploring their performance or behaviour.

1. Newman et al.:81

[Psychological safety is] the degree to which people view the environment as conducive to interpersonally risky behaviours.

2. Edmondson et al.:82

Psychological safety today is seen as especially important for enabling learning and change in contexts characterized by high stakes, complexity, and essential human interactions, such as hospital operating rooms . . . psychological safety plays a vital role in helping people overcome barriers to learning and change in interpersonally challenging work environments.

3. Tsuei et al.:83

PS [psychological safety] appeared to free them to focus on learning in the present moment without considering the consequences for their image in the eyes of others. Feeling safe also seemed to facilitate relationship building with the mentors.

These CMOcs identify the different contexts that can facilitate psychological safety, including a confidential and safe space for discussion and support provided by advocacy. Some papers note that part of this advocacy role may be acknowledging a doctor’s intentions and professional motives. Extracts 8–15 (see Box 2) imply that a more psychologically safe environment can be created with such validation, and this has informed the contexts noted in CMOc 2.

For CMOcs 3 and 4, we assert that the advocacy role may be most effective when this individual has no role in summative judgements about the performance of the doctor, and this point is explicitly noted in the literature (see Box 2, extracts 16–19). Although not developed in the literature, the notion of trust here offers a reasonable explanation (i.e. if someone is feeling judged then it is likely to be more difficult for them to build a trusting relationship and feel that they are being validated). This was also noted in stakeholder group discussions.

If a remediating doctor is ready to explore a potential behavioural or performance concern, and there is enough trust to validate others’ perceptions and judgements, a further aspect of developing insight is to accept the need to improve performance or change behaviour.84 The acceptance comes when a remediating doctor realises that their behaviours or performance do not meet the standards that they consider to be important to the practice of medicine. However, as noted above, the fear and anger that accompany remediation may mean that, rather than accept these standards, a remediating doctor may just reject them or reject the feedback itself. Therefore, the important question is how can feedback be delivered in such a way that a remediating doctor accepts that they must improve their performance or change their behaviour?

Framing feedback 1: juxtaposition

One strategy is to use the remediation process to elicit the remediating doctor’s own ideas about what makes a good doctor, and then compare their own behaviours to these standards. One ethics remediation programme in the USA starts by asking all participants on the first day of the programme to describe the ‘ideal healer’. On the second day of the programme the participants describe their own actions and are presented with their previous ideals.85,86 The authors expressly stated that the purpose of these programmes is to create a ‘professional dissonance’ or ‘ethical discomfort’85 that can be reconciled through an acceptance that the remediating doctor’s behaviour does not conform to the expectations of either the profession or the remediating doctor’s idea of what constitutes a good doctor. As these values come from the remediating doctor, rather than an external source, they may be perceived as being more authentic and less challengeable.

Acceptance may also occur when a remediating doctor’s own perceptions of their performance or behaviours are juxtaposed with the perceptions of trusted colleagues. Samenow et al.79 describe a university hospital’s professional behaviours remediation programme. The first part of the programme for a referred doctor involves collecting feedback on behaviour from different members of staff. Similar to the aforementioned ethics remediation programme, the express purpose is to create a ‘disorientating dilemma’ where the remediating doctor contrasts their own perceptions of their behaviour with that of their colleagues. Another version of the juxtaposition strategy is to use video simulations in which actors act out unprofessional behaviours. The remediating doctor views a number of simulations, the last of which (unbeknown to them) has been specifically devised to show behaviours similar to their own behaviours.65 Again, the purpose here seems to be contrasting their views on professional behaviours with their own behaviours.

Realist analysis of framing feedback using juxtaposition

Dissonance as a mechanism appears quite explicitly in the literature. In Box 4, extracts 2 and 3 discuss this in almost realist terms. In describing the disorientating dilemma or ethics discomfort that is induced through the strategy of juxtaposition, the authors offer a causal explanation of the way in which change can occur. This is based on the premise that the remediating doctor must accept the need to change performance or behaviour, which is noted or implied throughout much of the literature. This was also a continuing theme in stakeholder group discussions. Similarly, in extract 4, a process of guided reflection is used to enable the remediating doctor to compare their own actions with how they were perceived by others.

1. Caldicott and d’Oronzio:85

If the goal of the Program is to revitalize a commitment to the profession and reinforce its ethical values . . . the first two modules provide the experiential basis, ethical discomfort, and emotional energy for the undertaking. The task of the remaining five modules is to . . . create professional dissonance.

2. d’Oronzio:86

The contrast between knowing the idealized virtues and behaving in ways judged to be unprofessional creates a dramatic, dynamic tension.

3. Samenow et al.:79

The survey serves two purposes: to give objective feedback to physicians on how they are perceived, which they can then compare with their self-perceptions, and to monitor progress post intervention for as long as necessary.

4. Rumack et al.:87

We have found that some residents recognize such [unprofessional] behaviors only when they are videotaped and given directed feedback.

Context–mechanism–outcome configuration 5 (Table 4) describes the process whereby strategies that create this juxtaposition invoke change through dissonance. The outcome created here is the acceptance of the discrepancy between their own and acceptable behaviours.

Framing feedback 2: specific data from different sources

Gaining insight from accepting feedback is important at all stages of remediation. Initial presentation of feedback may be part of what Papadakis et al.58 have termed the ‘feedback conversation’.58 As remediation progresses, feedback and reflection on new skills and progress will create more insight and more progress. Trust may then link into feedback, in that the remediating doctor may react more readily to the feedback received if they have a trusting relationship.

A frequently identified barrier to successful remediation is the negative emotions of the remediating doctor throughout the remediation process.49,51 In the remediation environment, therefore, providing feedback does not mean that feedback will automatically effect behaviour change.55 Rejecting feedback may be likely throughout remediation. One of the contexts that may lead to a rejection of feedback is when a doctor perceives that accepting the feedback is a threat to their career or their professional identity.51 There are a number of strategies within the literature that are used to manipulate the context within remediation programmes to mitigate or ameliorate these emotional barriers to engaging in remediation.

A relatively common argument in the literature (and one that also arose in the stakeholder group) is that feedback needs to be as clear and as specific as possible so that it might not be so easily dismissed or discredited. In particular, the feedback should be articulated in ways that are unambiguous and objective.52,63,65,71,88,89

This is particularly challenging when the issue is a behavioural one, as specific and direct feedback is difficult to gather. However, it is also particularly important, as receiving feedback that is non-specific may be interpreted by the remediating doctor as being victimised by the organisation or specific individuals within it. This raises the issue of the role of soft intelligence (i.e. an ear to the ground) within remediation programmes. The literature suggests that soft intelligence may be particularly useful in gaining a better understanding of the nature of performance or behavioural concerns and, therefore, help tailor a remediation programme appropriately.50,52,56,66,67,69,73,79,87,89–93 However, such data may be less useful in invoking behaviour change for the participant, given the importance attached to specific and unambiguous feedback.

In addition to having specific data, remediating doctors are more likely to validate feedback when it is derived from multiple sources.42,46,52,58,79 Although the reason for this is not really described in the literature, it seems logical that feedback is much more difficult to deny if it comes from a range of sources. Patient feedback may be particularly useful, as provision of safe and effective care is central to medical professional identity (see Framing feedback 3: affirmation, normative frameworks and feedback standard discrepancy).

Realist analysis of how specific feedback from different sources can lead to acceptance

Context–mechanism–outcome configuration 6 is derived from a substantial body of literature that suggests that feedback is more effective when it is specific (Table 5). This is clearly supported in Box 5, extracts 1–3, whereas extract 7 notes that feedback in the absence of specific data may be unconvincing. The importance of a variety of sources of feedback is noted in a number of the extracts (see Box 5) and was strongly supported in stakeholder group meetings. Validation describes a simple but important mechanism whereby, once the feedback is recognised as a legitimate and accurate description of behaviour, the remediating doctor will logically recognise the discrepancy between their own perceptions of their performance or behaviour, and others’ perceptions.

1. Mar et al.:71

Specific comments from supervisors may also be quoted, as they provide concrete examples of suboptimal clinical performance or behavior.

2. Rougas et al.:52

This data should be objective and multifaceted – all available avenues for gathering specific, observable behaviors should be sought from other health professionals (nurses, social workers, pharmacists, physicians, etc.).

3. Liu et al.:93

. . . instead of informing a resident that he or she is ‘underperforming,’ the more specific ‘you are reading an average of five CT scans a day while your peers are averaging 10 per day’ clearly conveys the issue . . .

4. Paice:46

The decision that remedial training is necessary should be based on objective evidence from more than one source.

5. Roberts et al.:59

Our research suggests that nearly two-thirds of individuals who have demonstrated patterns of disruptive behavior and who are provided reliable data delivered in a respectful, non-judgmental way will respond.

6. Rumack et al.:87

We have found that some residents recognize such [unprofessional] behaviors only when they are videotaped and given directed feedback.

7. Bhatti et al.:67

. . . more than half of the problematic residents were identified through personal communication between a faculty member and the PD. Because no documentation or objective evaluation is available in many such cases, it may be difficult to convince residents of their deficiencies, and litigation is a possible consequence.

8. Papadakis et al.:58

The core message of this [first] meeting is: ‘We have important feedback that you need to hear. We have data from many sources that we can present to you . . . in a de-identified fashion with specific behavioral issues that you will need to change . . .’

9. Samenow et al.:79

The physician receives standardized objective feedback from colleagues, staff, and administrators that can then be compared with his or her own responses . . .

10. Brown et al.:42

The basic data are obtained from historical information regarding the individual provided by their employer, MSF, including self-assessment, psychometric profiling and the results of a semi-structured interview.

In CMOc 7, this outcome of the gap between the remediating doctor’s perceptions and the actual behaviours or performance becomes the new context that triggers the mechanism of dissonance. CMOc 8 describes what happens when feedback is not framed in this way and therefore becomes perceived as a generalised judgement on the remediating doctor as a person. Denial (discussed in more detail in Realist analysis of framing feedback: affirmation, normative frameworks and feedback standard discrepancy) is the remediating doctor’s way of reconciling the uncomfortable professional dissonance they experience.

Framing feedback 3: affirmation, normative frameworks and feedback standard discrepancy

Having specific feedback from multiple sources may be important for that feedback to be validated, but it may not be enough, given the emotional barriers associated with engaging in remediation. The literature points to a number of other framing strategies. If a doctor gets feedback that does not fit with their professional identity, a state of acute identity dissonance may be alleviated only through denial of the feedback itself.57 Some intervention strategies appear specifically directed at framing feedback in such a way that it protects the professional identity of a doctor being remediated, namely affirming the remediating doctor’s strengths, framing discordant data in accordance with relatable normative frameworks, and presenting feedback in a way in which the gap between actual and perceived standards is neither too big nor too small.

Affirming a remediating doctor’s strengths may increase the chances that any corresponding negative feedback is validated.50,94,95 Moreover, care should be taken to ensure that the feedback is delivered in a way that is not judgemental or demeaning.50 The important point here, as noted above, is that feedback that undermines a doctor’s professional identity may result in rejection of that feedback.

It may also be important to frame feedback according to norms that a doctor can relate to and that are congruent with their professional identity. For example, respect for colleagues may be a difficult normative reference to grasp if the remediating doctor thinks that they are motivated by appropriate norms. However, doctors may find it easier to relate to normative frameworks around patient safety.52 In other words, if possible, it may be useful to ‘honor the intent’ of the remediating doctor, while pointing out that the behaviour poses safety risks.53

In our stakeholder meetings there was discussion on the importance of framing feedback so that it seems manageable. It may be important to deliver feedback in a way that encourages perspective, conveying to the doctor that the issue to be addressed is one part of their practice and not their practice in its entirety. The discussions included suggestions that effective coaching can be used to remind a doctor that they have already achieved a great deal by getting through medical school and have proven themself to be a capable practitioner.

Realist analysis of framing feedback: affirmation, normative frameworks and feedback standard discrepancy

The context for CMOc 9 is the fear that characterises remediation (Table 6). Denial functions here as the mechanism, as it is a way for the remediating doctor to reconcile what we previously described as uncomfortable professional dissonance without accepting the feedback on their performance or behaviour. Citation searching from one of the returned papers96 pointed to some of the wider literature on denial in the medical education literature. This suggests that denial may take different forms (i.e. denial of the feedback or denial of the standards to which they are being held). This fits with the remediation literature, which discusses situations where there may be denial of the veracity of the feedback itself (i.e. ‘I don’t believe this feedback reflects my performance or behaviour’)55 or a denial of the standard, or even the values, of the profession to which those standards pertain.57,96,97 Denial has been posited as a mechanism that explains non-engagement or superficial engagement with remediation. Box 6, extracts 1–4 offer examples of the different types of denial that might take place in remediation.

Context–mechanism–outcome configuration 10 explains that affirmation, which has been identified as a mechanism in other realist research on doctor performance,101 may be a process that helps protect professional identity. This was discussed in the stakeholder group and is supported by Box 6, extracts 5–7, which suggest that some positive feedback, and affirmation of a remediating doctor’s strengths, can be helpful in remediation. This is a tentative CMOc, as the literature does not discuss this in much detail.

Context–mechanism–outcome configuration 11 concerns the degree of variance in feedback. This was informed by stakeholder group discussions on the need to make remediation seem ‘manageable’ and to deliver feedback in a way that did not make the process seem overwhelming. In the wider literature on feedback, the gap between the data provided in the feedback and the accepted standards is known as the ‘feedback-standard discrepancy’.96 Again, the use of coaches and advocacy may be important in conveying this. This links to CMOc 6, where specific feedback is important because it is less likely to be seen as a generalised judgement about a doctor’s performance and more likely to allow for the identification of achievable improvements. If a remediating doctor feels that the remediation is manageable, then the mechanism of dissonance, rather than denial, is more likely to be triggered.

Context–mechanism–outcome configuration 12 develops a causal explanation for how a doctor can be motivated to align their performance or behaviours with those of the profession, when feedback is framed in such a way that aligns with their values. Pawson’s seminal work on realist synthesis draws on a number of broader substantive sociological theories that help understand the role of identity in behaviour change. One of these is Merton and Merton’s reference group theory,102 developed in the 1950s largely in relation to the behaviour of US soldiers who were situated either on the inside or on the margins of the established community of soldiers.102 This theory posits that individuals will change attitudes and behaviours to assimilate with the reference group. More recently, Elder-Vass103,104 has developed the idea of norms circles which, like reference groups, invoke change through people assimilating to the norms of the group. Linking this to behaviour change theory, studies have shown that self-identification with the perceived norms of a reference group can significantly predict behavioural intention.105 This is particularly relevant in the field of medical education, given the importance attached to the communities of practice model of learning.106–109

The causal mechanism can be termed ‘normative enticement’, as it describes a situation in which an individual is enticed to the norms of a particular group. This mechanism could crudely but accurately be described as ‘I identify as being part of a particular group and therefore will accept the need to change my behaviour to retain membership of that group, to reaffirm to myself that I am a legitimate group member’. The normative enticement mechanism may run alongside the dissonance mechanism, feeding into it and facilitating its function at various stages throughout the remediation programme. This mechanism also bridges motivation and insight outcomes. The enticement to follow the norms of the profession may lead to acceptance that change is necessary (i.e. insight), but also intrinsic motivation to embark on that change. Motivation is discussed in the next section.

Context–mechanism–outcome configuration 13 suggests that being identified as needing remediation may, in the absence of efforts to frame feedback in a way that seeks to protect the remediating doctor’s professional identity, trigger a mechanism of identity dissonance. This type of dissonance, rather than invoking acceptance, may lead to a reconstruction or rejection of that identity to relieve this dissonance. Again, this is likely to have demotivating influence. This alluded to in Box 6, extract 4.

Motivation

The processes that occur to induce motivation are not divorced from those processes that lead to insight. However, it is worth distinguishing between insight and motivation, as, although there is a logical overlap, motivation arose as a clear and distinct theme in stakeholder group discussions and is addressed specifically in the literature. Moreover, although remediation is by no means a linear process, insight into a performance problem or behaviour logically precedes motivation in building a programme theory.

Although motivation is mentioned relatively frequently in the literature, it is not often presented as a specific outcome. In some of the literature, a lack of motivation is identified as a barrier to remediation (i.e. if the remediating doctor lacks motivation, remediation will be more difficult).40,45,53,110 Other papers discuss, even if only briefly, some of the strategies in remediation programmes that bring about motivation.7,44,48,50,55,57,91,97,111 There is a basic distinction between intrinsic motivation (i.e. linked in values and beliefs) and extrinsic motivation (i.e. ‘I do it because I want to avoid a punishment or earn a specific reward’) that features in the literature and was part of stakeholder group discussions. The distinction provides a useful starting point to begin to think about the intermediate outcomes related to motivation and its role in remediation.

Through examining the literature and through extensive stakeholder discussions, we were able to identify a number of strategies that facilitate intrinsic motivation in remediation programmes; for example, involving the remediating doctor in remediation programme planning to encourage perceptions of control, correcting causal attribution and effective goal-setting.

These are analysed in turn below to explain how and why they work for remediating doctors, in which contexts and to what extent they produce outcomes related to intrinsic motivation.

Involving the remediating doctor in remediation planning: perceived control

Involving a remediating doctor in planning a remediation programme may have a motivating influence. Facilitating a remediating doctor to maintain a degree of professional autonomy may have positive motivational consequences because it protects professional identity and engenders a sense of agency. Being involved in the remediation planning necessitates that the remediating doctor has an active, rather than a passive, role and they may perceive that they have more control over the outcome. In the literature, this is often framed in terms of developing ‘buy-in’ to the programme, suggesting that involving the doctor in the design of the programme may also give them more of an emotional investment.44,48,50,55,57,91,97,111,112

Having input into remediation planning may also help protect professional identity. Autonomy is a key component of professional identity, especially in those doctors who have passed through training and become a consultant doctor (i.e. an attending physician).12 As noted in Facilitating remediation: developing insight, professional identity is part of how doctors learn, and doctors may be intrinsically motivated by the desire to serve patients.55 As highlighted above, invoking the remediating doctor’s own sense of professionalism and professional values may help create professional dissonance, leading to insight. This is because individuals can be intrinsically motivated to adopt certain behaviours or practices when these behaviours are associated with their own professional values, previously discussed in terms of the mechanism of normative enticement (see Table 6).57 In other words, remediating doctors can be motivated because they want to be, or remain, part of the ‘in group’.

The amount of professional autonomy available in a remediation programme may be limited (remediation, by definition, means that the doctor has no choice but to remediate if they want to continue in their existing role or continue in their training pathway). However, reducing the inevitable loss of professional autonomy may be an important part of a successful remediation programme.

The important point here is that remediation necessitates a loss of autonomy, but if doctors have some input into the design of the programme then their professional identity will be less disrupted and the doctor will feel a greater degree of control over the process. Subsequently, they will have more intrinsic motivation.

Realist analysis for involving the doctor in remediation planning: perceived control

In CMOc 14 (Table 7), the extent to which a remediating doctor has input into the remediation programme functions as context. The input into the remediation programme increases the perception of control. Perceived control has been identified as a behaviour change mechanism in previous realist reviews.113 This is consistent with behavioural change theories that have been used to understand and develop remediation programmes, and were identified through supplementary searching, which emphasise the importance of perceived control in motivation.114 This provides a credible causal explanation for the substantial emphasis placed on creating remediating doctor involvement in remediation planning in a number of remediation programmes worldwide. This is illustrated in several of the extracts in Box 7.

Conversely, in CMOc 15, the loss of professional identity triggers normative rejection, which leads to a lack of motivation. This normative rejection is the converse of the CMOc of normative enticement in CMOc 12. The remediating doctor may be less likely to reject their professional identity if they maintain a degree of autonomy in the remediation programme.

Correcting causal attribution

If a remediating doctor does not believe that they have the ability to improve their performance or their behaviour then they will not be motivated. This is likely to happen when the remediating doctor incorrectly attributes the problem to something they cannot control. Durning et al.116 use the example of a remediating doctor who attributes his poor performance to a lack of ability, which is demotivating because ability is perceived as an innate and immutable characteristic. However, after investigating the causes of the problem and finding that it is related to study techniques – something over which the remediating doctor has control – motivation is increased.116

Similarly, for behaviour, literature points to the importance of doctors identifying particular emotional triggers. If a remediating doctor is able to recognise particular triggers for poor behaviour, they may become more aware of why they are behaving in this way and how they might seek to change their behaviour. In some papers the process of self-identification is part of an initial conversation, in which the remediating doctor is encouraged to explore their own perceptions of a particular problem.90,117

This supports the strategies for post-referral assessment4,7,45,50,52,56,65,71,76,78,79,89–91,116–118 that examine not just the identified performance or behavioural problem, but also the possible causes of that problem. Accurate causal attribution will also enable the development of individualised and targeted remediation plans, a feature of a large number of remediation programmes.7,11,47,73,100,118–123

This points to a further overlap between insight and motivation. Identifying the causes of a performance or behavioural issue is important in developing insight (i.e. the remediating doctor knows what is triggering a particular set of behaviours or performance and therefore has greater insight). This kind of insight leads directly to motivation, as it may enable doctors to overcome problems with causal attribution and therefore improve self-efficacy.

Realist analysis for correcting causal attribution

In CMOc 16 (Table 8), the causal mechanism is again a perception of control. In this case, the perception of control comes from correctly identifying the issue itself and through this process understanding that the problem is manageable. Again, this is supported by the wider literature on motivation and learning that highlights the motivational consequences of someone knowing what they are capable of and what they can control.124

When a remediating doctor can be coached through specific strategies to improve learning then they may also improve their self-efficacy, which in turn leads to more engagement. This process is explained in CMOc 17. Self-efficacy in the setting of remediation is about having the confidence that you can successfully remediate and achieve the goals that have been put in place. There are clear overlaps here with the perceived control mechanism, because having confidence to put knowledge into action implies a belief in control over the application of this knowledge. This is informed in Box 8, extract 3. Extract 4 informs the overall points about the relationship between self-efficacy and motivation.

1. Durning et al.:116

The supervisor wants to improve George’s self-efficacy beliefs, and an important first step is to challenge his attribution that his lack of success is due to his innate ability instead of his approach to learning.

2. Samenow et al.:79

Dr. Jones recognizes that his perfectionist drive and extreme work ethic have cost him dearly. He tells the group how, in his efforts to be self-sufficient and the ‘best’, he has pushed people away, including colleagues, family, and loved ones. He identifies this loneliness as a major contributor.

3. Liu et al.:93

It is also important to assess residents’ confidence in their knowledge, because actionable knowledge is only the knowledge the person is confident enough to use.

4. Prescott-Clements et al.:7

Similarly, a lack of self-efficacy or confidence can be encountered by individuals who are performing poorly (suggesting a degree of insight), which can also reduce intrinsic motivation to engage with remediation through negative feelings about ability, unless addressed quickly through support.

5. Swiggart et al.:89

Detailed inquiry of her social history focused upon her family, using a family genogram to depict individual family members, addiction issues, and relationships.

The role of self-efficacy in enhancing motivation is noted in the wider literature on behaviour change that was uncovered through supplementary searching on the topic of self-efficacy and behaviour change (see Box 8, extract 6).114,125 CMOc 18 explains how strategies that are related to identifying emotional triggers for behaviour work. Self-awareness here describes the process of identifying those triggers and becoming aware of when and how they operate to change behaviour.

Goal-setting

There are other features of interventions that may induce self-efficacy. One of these may be goal-setting. A number of interventions note the importance of having clear and achievable goals set out in the remediation programme.52,55,71,87,91,118,126,127 There are various ways that this goal-setting is framed, but the basic premise is that goals should be both clear and achievable. Telling a doctor to ‘improve background radiology knowledge’ is not as helpful as ‘improve knowledge of chest radiography, anatomy and pathology’.71 The clarity is not just about the goals themselves, but ensuring that the doctor understands how these goals are being assessed.118

Although developing self-efficacy (i.e. the sense that the goals can be achieved) may be an important component of a remediation programme, it does not necessarily mean that very high self-efficacy will always facilitate effective remediation. Indeed, there may at times be an inverse relationship between levels of self-efficacy and successful remediation if the doctor’s self-efficacy is not based on insight. Studies on doctor performance, and professional performance more generally, have shown that poor performers tend to have the extra burden of being less likely than higher performers to judge their own performance level accurately.128–132 If a remediating doctor has very high self-efficacy beliefs going into a remediation programme then this may suggest a lack of insight and may impede remediation.93 The corollary to this is that low self-efficacy may be a sign of insight, as the remediating doctor is aware of their deficiencies, although this may also induce low intrinsic motivation.7 Therefore, it seems that there is a fine line in remediation programmes. Feedback has to be direct enough to ensure that an individual is faced with credible evidence of the need to change (we called this juxtaposition; see Framing feedback 1: juxtaposition). However, careful goal-setting and support may be important in inducing the self-efficacy required to believe that goals are achievable. This relates to what was discussed in Facilitating remediation: developing insight, about framing feedback in manageable way.

Realist analysis of goal-setting

In CMOc 19 (Table 9), the extent to which the goal is perceived to be achievable functions as context and can be changed with specific, measurable, achievable, relevant, time-bound (SMART) goal-setting strategies and goals that are tailored to the individual being remediated. Again, the focus on self-efficacy as a mechanism fits with behaviour change models that underpinned some of the literature on remediation that were found through citation searching. This was also a point of discussion in the stakeholder group. Box 9 sets out the data that we have drawn on to develop this CMOc.

1. Katz et al.:91

By setting concrete and achievable goals, and ensuring that residents understand and accept them, program directors define performance targets. Success or failure will be defined by whether or not these targets are eventually met.

2. Katz et al.:118

By making the criteria transparent, the resident was given every opportunity to target the goals and accomplish them in a reasonable timeframe. The time allowed was appropriate for the resident to succeed or fail on his own merits.

3. Mar et al.:71

A clear and comprehensive set of remedial objectives is then created directly from the issues identified in the background assessment. An example of . . . [a performance goal] ‘to improve the knowledge base of ultrasound physics and anatomy’, while a behavioral task may be ‘to improve attendance and punctuality’.

4. Rougas et al.:52

Several practical approaches have been utilized by educators across disciplines including frameworks such as the ‘SMART’ goals: specific, measurable, achievable, relevant, and timely (Day & Tosey 2011). This type of framework helps ensure that individual learners know what is expected of them in a defined time period with established consequences.

5. Rumack et al.:87

Regardless of how many deficits a resident may have, it is wise to choose one or two skills per competency to remediate first (6,8). Choosing too many may confuse the resident and weaken their response.

6. Sargeant et al.:55

The coaching conversation then moves to guiding the physician to set specific and realist goal(s) and to developing a plan for attaining the goals. Planning involves paying special consideration to the factors which may both impede and enable progress and success, setting timelines, and considering how success will be measured.

Destigmatising remediation

If professional identity is important for intrinsic motivation, then all efforts to mitigate the negative impact of remediation on a remediating doctor’s professional identity are potentially important. In this respect, a number of papers note that the stigma attached to remediation acts a barrier to engagement.8,41,133 If a doctor feels stigmatised, then they will feel isolated from their colleagues and this can have negative implications for identity development. The importance of professional identity in remediation therefore brings into the focus some of the broader contextual issues around how remediation is conceptualised within the medical profession and the extent to which it stigmatises remediation.12

A study in Canada considered the perception of remediating doctors from various stakeholders within the medical community. It found that those involved in remediation programmes often held ambivalent perceptions, describing remediating doctors as being both ‘one of us’ and ‘not like us’.134 This is important as it suggests that negative perceptions exist and also that it may be possible to nurture more positive perceptions of remediation. This finding emphasises the relevance of in- and out-group dynamics that may impact professional identity for remediating doctors. Similarly, in the UK, it has been noted that policy developments have simultaneously sought to normalise remediation, while also creating a contradictory and unhelpful climate of fear around doctor performance.8

One approach to mitigate this barrier is to seek to remove the stigma attached to remediation. This is noted as important in a number of papers.8,40,84,91,117,135 However, the literature does not elaborate on what this looks like in practice beyond renaming the process itself,91 ensuring that the process is not framed as punishment84 and seeking to conduct remediation in such a way so that the remediating doctor is not easily identified as such by peers.41

A further approach to lessen the sense of isolation and the stigma attached to remediation is to bring remediating doctors together. Group remediation is advocated in a number of papers.4,51,73,74,85,89,95,120,121,136–138 The process of working in a group may help doctors feel less alone and the process of seeing colleagues progress through and complete a remediation programme may be motivating.85 However, as noted in stakeholder group discussions, group work may in many situations be impractical. Other ways of networking with doctors undergoing remediation may be more pragmatic.

Realist analysis for destigmatising remediation

Context–mechanism–outcome configuration 20 provides a causal explanation for the emphasis placed on seeking to remove the stigma of remediation, which is discussed in a number of papers (Table 10 and Box 10, extracts 4–6).41,84,135 Again, this is a plausible explanation because it is congruent with broader notions of the importance of professional identity and negative impacts of isolation on motivation. This was derived, in part, from the discussions with the stakeholders, including those undergoing remediation, which noted how doctors would often not disclose their remediation to friends or family, let alone other doctors, and the ensuing sense of loneliness experienced.

1. Krzyzaniak et al.:133

Major themes identified as barriers to remediation were: (1) faculty concern about the premature labelling of residents as ‘struggling’ and the stigma that comes with remediation.

2. Price and Archer:8

This culture of fear is in part fuelled by the negative associations with remedial interventions; as recognised by the Steering Group on remediation, the term remediation has ‘negative connotations’.

3. Harthun et al.:117

Significant effort is expended to remove any hint of punishment in any aspect of the program. Also, we actively minimize stigma associated with poor scores and discuss this aspect of performance with the residents.

4. Gunderman and Rackson:135

One strategy for encouraging the prompt reporting of learning difficulties is to diminish the generally pejorative connotation of remediation. If there is any profession in which remediation should be widely understood, it is medicine.

5. Kalet et al.:84

Remediation should be reframed from a matter of punishment and stigma to a form of training that many, if not most, will need and benefit from at some point.

6. Shapiro et al.:41

The less the resident’s normal schedule is disrupted, the less he or she is distinguished and differentiated from peers and the less likely he or she will see the remediation as a punishment.

7. Caldicott and d’Oronzio:85

In a setting that features face-to-face peer communication and feedback, participants lose their feeling of isolation. Exchanging with other accomplished and dedicated healthcare professionals who have also acted unprofessionally . . . participants realize that committing a transgression and experiencing the consequences do not nullify their own professional accomplishments and dedication.

Context–mechanism–outcome configurations 21 and 22 offer a causal explanation of what happens when the framing of remediation creates isolation. In CMOc 21, the sense of alienation – broadly speaking, the process of detachment of things that belong together – can be understood as a causal mechanism, leading to isolation. Extract 7 (see Box 10) advocates the use of group remediation to lessen isolation. In CMOc 22, the corresponding sense of isolation invokes the mechanism of normative rejection, as discussed above. In this case, it is the sense of isolation that leads the remediating doctor to dissociate with the community of practice in which they work. These are not well substantiated in the literature and would benefit from further testing.

Although intrinsic motivation may be the ultimate goal of remediation, extrinsic motivation is widely used and deemed effective. One paper reports a survey of residency programme directors (n = 253) in the USA. The threat of termination or not progressing were reported in the survey as the most effective strategies of ensuring engagement with remediation programmes.139 A number of papers noted the importance of having clear consequences for not improving performance or behaviour.58,71,76,78,122,140 In the stakeholder group discussions, it was also suggested that understanding the consequences of engaging (or not) with a remediation programme may also be considered an aspect of insight (i.e. insight into the consequences of future actions).

The way in which extrinsic motivation works and interacts with other mechanisms within a remediation programme is not clearly expressed in the remediation literature. In a sense, it is presented as common sense logic (i.e. if a remediating doctor understands that non-engagement will be bad for their career, then they are more likely to engage). However, as our earlier findings and CMOcs have demonstrated, it is difficult to explain how extrinsic motivation alone could induce sustained changes to performance or behaviour. Even if a remediating doctor is sufficiently motivated by the potential consequences of not engaging, the engagement could be superficial or transitory if they are not intrinsically motivated to engage.

Clear consequences

However, outlining clear consequences in remediation may play a role in maintaining engagement throughout a remediation programme141 and, once engaged in a remediation process, this may provide extra motivation to keep a remediating doctor on board. The consequences can escalate if there is continued lack of improvement at specified intervals.53 Inducing extrinsic motivation through clearly articulating the consequences of actions is not advocated as a standalone strategy, rather it works to support strategies aimed at stimulating a remediating doctor’s intrinsic motivation. Hanna et al.95 report that mandatory engagement with a remediation programme encouraged attendance, but not necessarily engagement.95 If a remediating doctor is not intrinsically motivated to change, then they may also re-evaluate their goal. In stakeholder group discussions it was suggested that older doctors, who are nearing retirement, may choose to accept restrictions on the scope of their practice.

The potential for external motivation may be contingent on the perceptions of self-efficacy and perceived control. If a remediating doctor understands that there will be consequences if they do not improve, but does not believe that the improvement is possible, they may change their goals or their commitment. A doctor may adapt their goals for other reasons, such as being near retirement, and, from a patient safety perspective, this may not be a bad thing. Doctors still in training may use the opportunity to reflect on their career and choose a different career path. A number of papers highlighted the importance of systems for ‘coaching out’ of a particular programme of study.40,50,66,69,91,142

Realist analysis for clear consequences

Context–mechanism–outcome configuration 23 (Table 11) posits that a process of evaluating costs and benefits will be a causal factor in determining the level of motivation. However, the outcomes of this evaluation will be contingent on a number of factors. The processes overlap, so whether a remediating doctor engages in remediation or decides to change their goal may depend on processes present in some of the other CMOcs presented earlier (Box 11). For example, whether or not they experience affirmation and feel psychologically safe, perceive that they have some control over the outcome, and feel part of the medical profession with which they identify and wish to retain in-group status.

1. Sparks et al.:62

A formal performance review and written contract, with specifically identified problems and general themes, recommendations for time-based successful behaviors, and clearly identified consequences for unsuccessful behaviors, was initiated.

2. Krzyzaniak et al.:40

A resident cannot be indefinitely remediated . . . Thus, we must establish a time frame for the remediation goals to be accomplished.

3. Domen:69

The remediation plan should only target and address the specific issue(s) identified, should have clearly defined goals with defined measures/assessments . . . and clearly delineated consequences or next steps if the resident fails the remediation plan.

4. Kalet et al.:54

. . . learners should be required to participate in remediation and be held accountable to engage actively with the remediation team in developing an individualized remediation plan, to initiate and complete the remediation activities, and to successfully undertake whatever assessment is deemed appropriate.

Facilitating practice change/improving performance or behaviour

This section outlines our findings related to the way in which remediation works to improve performance or behaviour. Again, it is worth re-emphasising that we do not view remediation as a linear process, and insight and motivation would logically increase as performance or behaviour improves, with these processes working alongside each other. This section specifically focuses on the processes that occur to embed changes in behaviour or performance at the point at which someone has some insight into the problem and some degree of motivation to change.

Qualified doctors are high achievers who have proven to be competent learners. The way in which remediation programmes may work to improve performance and the specific strategies used will, in many ways, be similar to those used in medical education more broadly. We do not want to repeat tried and tested theories of education here. Rather, we seek to focus on the aspects of learning or behaviour change within a remediation programme that are specific to remediation and the particular challenges, such as the fear of engaging in remediation and its threat to professional identity development, that are associated with undergoing remediation.

Practising new behaviours/skills

Remediation may be most effective at changing behaviour or performance when remediating doctors get the opportunity to practise and master new behaviours or skills. Two papers70,143 in the returned literature explicitly discuss learning in a remediation programme in terms of an experiential learning cycle. Experiential learning theories posit that learning occurs most effectively when the learner gets the opportunity to move from abstract reasoning to concrete experience, and then reflect on that experience to develop learning.70,143 Processes of guided reflection (discussed below) are likely to be integral to that process if the remediating doctor is practising new skills in the workplace.

Interestingly, some studies also noted the importance of practising behaviours in a professionalism remediation programme.51,65 One paper51 notes that, during a course on boundary violations, it is not a lack of understanding of ethics issues that besets the referred doctors, but a lack of understanding of how these apply to clinical situations. Therefore, practising behaviours in a simulated environment is an important component of behavioural change.51 Coaching may also be used to facilitate practising new behaviours. This links to reflection (see Table 2, CMOcs 10, 11, 25 and 26), in that the coach can discuss progress with a remediating doctor, help them rehearse correct behaviours59 and set proper boundaries.76

Realist analysis for practising new behaviours and skills

Context–mechanism–outcome configuration 24 (Table 12) describes the repetition of performance or behaviours as contexts, as noted in Box 12, extracts 1 and 2. This seems fairly obvious, but practising new skills and behaviours is not always easy in remediation programmes. When simulation is not available, guided reflection on workplace interactions may be even more important. CMOc 25 links the repetition of skills/behaviours to guided reflection (see Box 12, extract 3). The process of guided reflection facilitates the remediating doctor to integrate the new knowledge and experience into their learning. Integration of new knowledge and experience is a mechanism that underpins learning theories such as the Kolb learning cycle, and so it is no surprise that it functions here. What is important is that, in remediation, the circumstances under which the intervention occurs can make this a more complicated process, especially if the problem is a behavioural one.

1. Parran et al.:51

Boundary maintenance skills require practice. Therefore, each participant practices from both the physician and patient perspective the clinical skill of “Getting to ‘no.”’ The specific scenario that is focused on in this skill practice exercise involves practicing saying “no” when asked to prescribe controlled drugs . . .”

2. Roberts et al.:59

. . . a ‘coach’ – either a chief resident or junior faculty member, who could meet with the intern frequently (weekly to biweekly) to discuss specifics of the incidents that are troublesome, get him to ‘own’ the behaviors that are incorrect and work to rehearse and then reinforce the proper behaviors.

3. Egener:76

A series of visits [from a coach], 1–2 hours in duration, spread 2–8 weeks apart . . . Most often as they practice acknowledging others’ perspectives, their own view of what constitutes ‘the truth’ becomes less black and white.

4. Rumack et al.:87

When professionalism lapses present as poor interpersonal and communication skills – yelling, berating, being disrespectful, or poor team communication – training in a simulated environment can be useful. We have found that some residents recognize such behaviors only when they are videotaped and given directed feedback.

Guided reflection

As noted above, the literature suggests that the way in which feedback is delivered in a remediation programme is important in determining whether or not that feedback is accepted. In particular, we noted that feedback that is both specific and from multiple sources is more likely to induce dissonance rather than denial. Building on this, feedback works to enhance performance when a practitioner reflects on what that feedback means,56,65,87,115,126,144 and this process works best when it is a guided process (i.e. reflection will not automatically follow feedback, but with guidance from a coach or mentor, reflective practice can aid behaviour change). This is noted for both clinical skills and knowledge issues,63,87,126 as well as professionalism.56,87 It seems that a key role of coaches or mentors is to guide the remediating doctor in reflection.

Guided reflection may be easier to facilitate with face-to-face meetings with a coach, mentor or supervisor. Our stakeholders suggested that in remediation coaching situations this guidance can start with an open question, inviting the remediating doctor to reflect in general on their experience throughout the remediation programme. If the meeting is face to face, then body language may be important in determining whether reflection is genuine or if they are just going through the motions. These meetings are also an opportunity to invite the remediating doctor to consider what they think the consequences, either positive or negative, are likely to be as a result of their efforts and progress so far.

Meetings with someone external to the organisation may be more effective than meetings with someone from within the organisation. This is because guided reflection will seem less threatening. For example, if someone is asked ‘what do you think the consequences might be if you do not engage in this programme?’ this may seem like an implied threat if it comes from someone senior within the organisation. If it comes from someone outside the organisation, or someone that is deemed impartial, it is much more likely to be perceived as a genuine question for reflection. This links back to the earlier points of receiving feedback in an environment that is perceived as being safe by the doctors involved (see Table 2, CMOcs 1 and 3).

In addition to face-to-face meetings, written reflective logs of experience and learning will help a remediating doctor to learn from their experiences.49,51 The purpose of these logs is that they are meaningful to the remediating doctor and, therefore, should not form part of the assessment process. However, it may encourage engagement if a supervisor checks to ensure that reflections have been completed.71 The idea that assessment of written reflections could diminish their effectiveness was also raised by stakeholders. Stakeholders were broadly agreed that such written reflections should not be summatively assessed, but noted that they might form part of a process of formative assessment that would enable remediating doctors to progress through the remediation process.

In general, the process of reflection may be most effective when divorced from the assessment of progress entirely. Like feedback presented initially, feedback on progress should also be as specific and objective as possible (see Table 2, CMOc 7). In our stakeholder group, it was suggested that multisource feedback, including feedback from patients, may be the most effective in engendering change, although clearly the optimum type of feedback would be dependent on the issues being remediated.

Realist analysis for guided reflection

Context–mechanism–outcome configuration 26 (Table 13) is derived from the importance attached to guided reflection through regular meetings, clearly evidenced in Box 13, extracts 1–6. This process of guidance functions as new context in a remediation programme. A common theme in these extracts is the notion that the process of regular guided reflection is used to ensure that remediating doctors are engaging with feedback and making sense of it, hence ‘making sense’ is a causal mechanism that helps change practice. The intervention strategies of face-to-face meetings and open questioning came from stakeholder group discussions. CMOc 27 illustrates how the mechanism of dissonance operates throughout a remediation programme, with the process of guided reflection facilitating the function of the dissonance mechanism at various points in the remediation journey.

1. Decoteau et al.:63

. . . one-on-one sessions [were used] to assess the learner in their abilities to review, analyze and learn . . . Based on the assessment, the teacher identified any particular deficits that the learner possessed and provided structured feedback to the learner to allow for a more refined approach in studying and completion of assigned SCORE.

2. Rumack et al.:87

Second, frequent feedback on the resident’s progress is essential for coaching to be effective. Third, the resident needs to take time to reflect on their actions with their mentor or coach to understand what progress or barriers are occurring.

3. Guerrasio et al.:65

After each scenario, the individual watched the video with the actors and remediation team, and was coached through self-reflection, receiving feedback and suggestions for change in an iterative fashion with goal setting for subsequent scenarios.

4. Warburton and Mahan:56

Coaching should focus on ensuring that the learner incorporates proper knowledge (virtues), attitudes (professional identify), and skills (professional behaviors) into practice through discussion, reflection, problem-solving scenarios, and ultimately, actions in real clinical settings.

5. Stirling et al.:145

Trainee receives feedback on their performance (30 minutes). On completing their self-assessment the trainee meets with the assessment panel. This constructive feedback focuses solely on the trainee’s performance during the PgWSE [Postgraduate Ward Simulation Exercise], and is structured using stages 1–6 of Pendleton’s guidelines, as follows.

6. Whiteman and Jamieson:49

We find that [written reflective logs] directly challenges the commonly encountered passivity of practitioners in this situation, which can originate from a practitioner’s perception of continuing education as a transfer of information from experts to passive learners.

7. Parran et al.:51

[Written reflective logs] provide physicians an opportunity to consider their strengths and weaknesses and help them to successfully integrate new learning into existing knowledge.

8. Mar et al.:71

These [reflective log] entries are for the resident’s use only, but their regular completion should be demonstrable.

Context–mechanism–outcome configuration 28 posits that the position of the individual providing the feedback is a context that determines how threatening the feedback is perceived to be. This links back to psychological safety, in that engagement and insight are more likely to follow when a remediating doctors feels safer. The idea that perceptions of safety will be enhanced if the person guiding the reflection is external to the employing organisation was derived from stakeholder group discussions.

Context–mechanism–outcome configuration 29 is based on the evidence (see extracts 6–8 in Box 13) that producing written reflections serves as contexts that trigger engagement. For example, extract 7 (see Box 13) describes this process as challenging ‘passivity’. This engagement allows the remediating doctor to integrate new knowledge and experiences, which is a mechanism that provides new insights and therefore more authentic learning and changes to performance or behaviour.

Programme theory

The complete programme theory of how remediation of doctors produces its effects is presented in Figure 3.

FIGURE 3 Programme theory of remediation. Reproduced with permission from Price et al.1 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure. (Figure not reproduced here; see the original article.)