Trust is central to all healthcare. A patient trusts the doctor’s diagnosis and recommendations. The doctor relies on a radiologist to read the image correctly. When the doctor refers a patient to a therapist, she trusts the therapist to assess for and provide appropriate intervention. The therapist, in return, relies on the doctor’s medical diagnosis and prognosis. And when a doctor tells a nurse a patient’s prognosis, the nurse accepts the information is accurate and will discuss the prognosis with the patient. In training, a senior nurse will trust a student nurse to make observations accurately and, later in training, the same nurse will trust the same student to give out medication. At a higher level, a surgeon will trust a trainee to diagnose a patient with acute abdominal pain and to operate to remove the inflamed appendix. The process of trusting another professional to undertake an activity is entrustment. Entrustment has happened since professions arose. A characteristic of a professional is the passing on of knowledge and skill, and inevitably and eventually, the teacher has to trust the student. But how does anyone decide? That is the topic of this blog post. This blog post complements an earlier piece on entrustability here; you should also read it.
Context – assessing ability and capability.
This discussion concerns the entrustment of a professional healthcare worker to undertake a complex activity, an activity where the process and outcome are not entirely predictable and unexpected events may occur, requiring flexibility and adaptation.
Historically teachers made judgments about the ability of their pupils informally but soon identified factual knowledge as something that they could test without using judgement. However, pupils had to write essays or show their workings even in the early examinations and marking these required some judgment.
For more practical skills, such as chemistry, mechanics and biology, examinations could test knowledge and simple skills. Moving into healthcare, these came to be known as competencies. Competencies are not straightforward. A carer of a specific patient can become competent in quite skilled tasks – managing a tracheostomy of a ventilator, for example. They would not be adept with any other patient. A healthcare professional might be more generally competent at managing patients with tracheostomies and ventilators in an intensive care unit but would not necessarily be qualified to undertake the same tasks in a care home.
Competencies emerged as the assessment term used in healthcare. Various systems of classifying competency arose. One framework was called Miller’s pyramid. (here) This organised performance of activity thus:
- Knows. Having the required knowledge. Is a necessary condition but not sufficient. Tests of knowledge do not predict competence, although a lack of knowledge might cause failure or incompetence.
- Knows how. The person can use their knowledge in practice. They develop competence. This can also be tested. It shows someone could undertake a task, not that they do.
- Shows how. The person can demonstrate how to undertake the activity in practice, only not in an actual clinical situation.
- Does. Performance in practice. The person successfully undertakes the activity. The person collects evidence of successful performance.
The problem with this classification, and the problem with most of the other competence assessments, is that it takes no account of context. Consider assessing the mental capacity to use money wisely of a patient with poor executive functioning after a traumatic brain injury, someone who has many superficial ‘friends’. In a quiet consulting room, she will give a credible, rational account of what she would do and why, and of what she would not do and why. However, in a busy, noisy pub, she will readily give a large sum to an acquaintance, contrary to her stated intent.
Similarly, a doctor may perform a complex task, such as making a diagnosis and undertaking a procedure, very successfully when there are no other demands, and knowledgeable nurses can assist with the process. However, halfway through an eight-hour acute take, with only newly qualified nurses to help and three more sick patients waiting, the doctor might not perform well, especially if the patient fainted as the doctor started the procedure. Before starting the shift, she may have been competent, but she was probably not entrustable.
Olle ten Cate, the prime mover behind much entrustment research, suggests that the current top level in most competency scales concerns observations of past performance. He indicates that entrustment concerns an expected ability to perform in the future in any circumstances.
The question is not whether the professional will manage, without help or risk, every potential problem that might arise. No person should believe that they can handle, without any support, every possible situation they face. And assessors should not expect that. The question is whether the assessor can trust the trainee to resolve most problems while also acknowledging when further advice or help is needed.
Olle ten Cate has added a fifth level to the Miller’s Pyramid – “trusted with future care”. (here)
To summarise the first point. Professionals need to learn to undertake complex activities safely and independently. Until recently, an assessor verified that someone could perform tasks safely and independently by observing their performance, usually on many occasions. If they did so, they were considered competent at that task. The development required an additional step. Somehow, to be discussed later, one needed to decide whether someone would perform the job safely and independently in the future. If the assessment process confirms that the professional would be safe and independent in future, then they would be entrusted with that task. A competent person can undertake an activity; an entrusted person is capable of undertaking the activity.
Context – complexity of healthcare
A second factor that has required entrustment is the complexity of many healthcare activities, especially the activities undertaken by the more experienced members of any profession, be it medicine, nursing, therapy, or any other profession.
Assessing competency is only possible for constrained, relatively short activities where the steps are reasonably clear-cut and constrained and predictable. Activities well suited to competency include, for example, taking blood pressure, inserting a chest drain, passing an endotracheal tube, taking a complete history, undertaking an ASIA assessment of a person with an acute spinal cord injury, and undertaking cognitive behavioural therapy. Although many of these activities require considerable experience, they are nonetheless easily defined and measured.
The General Medical Council used competencies to certify that a doctor was safe to be an independent practitioner. Competencies are still used by many professional organisations to authenticate professional safety and to give a licence to practice independently. This approach has an undeniable problem. To be a consultant (a senior doctor) in any medical speciality will require innumerable competencies, well above 100. Further, in practice, a doctor will only need good competence in a small subset, but no one knows what that subset will be at the time of certification.
In addition, being competent at a series of specified activities did not show that the doctor was safe and effective. It overlooked several vital aspects of clinical professional practice. Can the professional prioritise when under pressure, can the professional organise their work to complete a series of interrelated tasks, and can the professional adapt to patient needed? For example, one may be competent to insert a chest drain, but one needs to decide if it will be effective, whether it is the correct action, and whether it has a greater priority than some other task like stopping bleeding.
These characteristics are vital to successful, safe clinical practice. These characteristics are not competencies. They depend upon much higher-level knowledge and skills, a person’s attitude, communication ability, ethical awareness etc.
This realisation led the General Medical Council in the UK and similar organisations in other countries to move towards a small set of high-level educational outcomes. In many countries, they are referred to as Entrustable Professional Activities. In the UK, they are referred to as capabilities. The Rehabilitation Medicine curriculum has six generic capabilities, identical for all doctors, with an additional eight rehabilitation-specific capabilities. Other specialities have different numbers of speciality-specific capabilities. Moreover, other bodies responsible for medical education have other names for the high-level outcomes. For example, the Royal College of Psychiatrists has 19 “intended learning outcomes” for General Psychiatry.
The essential feature of the higher-level outcomes is that they are all complex. Therefore they contain much that is unpredictable, and the outcome of the activity cannot be judged individually. The senior professional needs flexibility and adaptability must know when to ask for help and when to offer support to another, they must have good interpersonal skills, etc. Any simple test cannot capture these features.
To summarise this part, having the large number of competencies needed to be a senior member of a profession was an inadequate way of assessing a person’s future ability to practice safely and effectively. Moreover, it was inefficient to evaluate every one of the 50-100 possible competencies. This problem led the bodies responsible for certifying doctors in several countries to focus on fewer complex high-level training outcomes, capabilities that the professional could undertake activities safely in future. In the UK, in medical specialities, these are referred to as capabilities in practice. (here)
Model of entrustment
A process model helps in understanding how to improve the process systematically. I will discuss a model developed for clinical entrustment decision-making, shown in this figure here. The model is taken from two papers, here and here, and it has four sets of factors that may influence a decision.
The first set relates to the trainee being considered for a task or activity. Research suggests that five different characteristics influence the decision. An obvious one is an ability to undertake the activity. Does he know what to do? Has he the skills to do it? Has he done it before? How well?
There are, however, four other groups of characteristics that have a significant influence on any decision. They include:
- the trainee’s integrity, truthfulness and benevolence;
- the reliability of the trainee, the extent to which he can be depended upon and takes responsibility for his actions;
- his humility, his ability to admit uncertainty and ask for help and accept feedback; and
- his agency, the ability to take control actively for learning, cooperating and considering risks that might occur.
Trainers differ in their ability and willingness to trust someone to undertake an activity for which they are ultimately responsible. This characteristic could well be the most significant factor, one that will only change with experience. The second factor of importance is the relationship between the trainer and trainee. Is it a long-standing, close relationship or a recently started relationship, sharing little in common? Thirdly, several factors may influence the decision: how experienced is the trainer in the activity to be entrusted, how successful or stressful have previous decisions been, and is the trainer familiar with the context? These and other factors will have an impact on the decision.
The third matter to consider is the activity. Is it straightforward, with little risk and a predictable course and outcome, or is it very complex, with many unknowns and uncertainties? Is there time pressure, such that the activity needs to be done by a deadline, or will the trainee have time to think carefully? To what extent does the activity also depend upon others from different professions or different specialities.?
The last area of influence is the context for the activity. Many factors comprise the context, such a the staff available – experienced or inexperienced, good or insufficient – and their attitude towards trainees being entrusted with an activity. Second, is the activity the only matter the trainee has to concentrate upon, or are there many other activities occurring? The culture of the organisation will also have an impact. Is it a teaching centre, well used to trainees taking responsibility for the first time, or not? Even a simple matter like the time of day may influence the decision. Last, how great is the entrustment being considered? Total, with the trainer being remote, or will there be a degree of monitoring and supervision?
Use of this information.
This model should not be considered a formulaic approach, measuring each variable to decide. It helps by drawing attention to the factors that most people expect to be relevant – when or if they considered them. It helps to make explicit what is otherwise implicit and often overlooked. It also allows some planned gradation in progress, with increasing case complexity or contextual complexity occurring separately rather than simultaneously. Last, the model may facilitate an analysis of an incorrect judgment.
Making an entrustment decision
in UK Rehabilitation Medicine
This section of the blog post is focused on making judgments on high-level outcomes. Most of the research has considered entrustment at the level of quite complex activities that exceed competencies but do not extend as broadly as the high-level outcomes used in the UK. Thus, for example, papers reviewing entrustment scales (here) are of little value in the UK context. Moreover, the research focuses on Entrustable Professional Activities (EPAs), whereas the six generic capabilities required in all training programmes are general skills that apply across all professional activities.
Rehabilitation Medicine is also a speciality where doctors, like all other professions, work within a multiprofessional team, and all their expertise and skills must be considered in that context. In contrast, surgeons and many physicians undertake specific professional activities such as operating, undertaking interventional cardiology, managing an acute illness etc. It is easier to evaluate a professional’s performance of these activities.
Rehabilitation Medicine is much more like Psychiatry, or Palliative Medicine, or Geriatric Medicine. The expertise is in using medical knowledge and skills (a) within a multi-professional team and (b) within a long-term relationship with a patient. Some rehabilitation doctors will have procedural and specific skills, but they are all a small part of a greater whole.
Consequently, the high-level outcomes, the Capabilities in Practice (here), are not easily measured by direct reference to a scale.
The solution lies within the curriculum. The basis of entrustment is that the person is trusted to perform an activity under any circumstances. The curriculum explicitly requires trainees to gain experience in more or less any situation that may face in future. Thus a trainee should perform in all geographic settings – intensive care, social service day centres, patient homes, etc. The curriculum requires that a trainee gains experience in all locations. Equally, a trainee should assess the rehabilitation needs of patients of all ages and with any disorder or condition. The curriculum requires evidence of this.
The curriculum also requires work-based assessments and encourages trainees to obtain this over various conditions from various assessors. The entrustment literature stresses the importance of constructive comments and feedback and the relative unimportance of a score or rating. (here) The Rough Guide (to the curriculum) encourages trainees to ensure that they get constructive, relatively detailed comments. These provide helpful evidence.
The curriculum and the portfolio provided to collect evidence do not mention the essential trainee characteristics of integrity, agency, reliability and humility. Trainers can mention these features in their comments and feedback. The consultants providing Multiple Consultant Reports reports could be reformatted to consider these characteristics. Trainees could also record examples illustrating how they demonstrate these attitudes in their reflective entries on the portfolio.
The curriculum does not mention competencies in specific activities. The syllabus does give 39 areas where a trainee can achieve competency. It is unlikely that anyone will acquire all 39! However, the syllabus guides the trainee and trainer to develop the vital characteristic needed for entrustment in clinical activities and capability at the task. These are not recorded directly but will be assessed in case-based discussions and other work-based assessments.
Last, the research papers suggest that the trainee and trainers should build a body of evidence showing increasing entrustment and that a group should make that final decision.
To conclude, the current training programmes provide trainees with the opportunities to increase their entrustability and with the opportunities to collect evidence to show that. Trainers are responsible for providing helpful, constructive feedback and simple comments every time they undertake a work-based assessment or structured learning event. Trainees are accountable for seeking evaluations from various people, covering a variety of different clinical situations, and they should ask assessors to make comments and give constructive feedback. The panel at the Annual Review of Competency Progression should make the final decision of entrustment.