1.1 Foundations of clinical practice

This chapter aims to describe key knowledge that, while not directly involved in clinical work, is essential for expert clinical practice and therefore should be regarded as equivalent to basic sciences. Like the basic sciences, this knowledge underpins our clinical practice and guides us when faced with new situations.

Surgery and medicine rely on research and innovation to drive improvements in clinical practice. As an expert surgeon, you must know how to assess these innovations and be able to judge objectively whether or not you should incorporate any apparent improvements into your clinical practice.

A common public assumption is that all medical treatment is beneficial. This may be a marker of the success of modern medicine, but some of the major advances of the last 50 years have been the acknowledgement that medical treatments cause harm as well as benefit, and the development of ways of quantifying the relative risks and benefits of treatments. The most recent manifestation of this trend is the current focus on patient safety—preventing harm to patients within the healthcare system.

Another recent trend is an increased focus on the patient as a consumer. There has been a shift towards healthcare provision structured around patient needs rather than simply providing the healthcare that healthcare workers think is best. There are costs and frustrations associated with this approach, but advantages in patient satisfaction and flexibility of healthcare provision. Healthcare evolves to solve patients’ problems rather than the other way round.

For patients to be actively involved in their healthcare, they need to be informed consumers. Therefore part of the healthcare process is to educate the patient about the options available. Patients now have ready access to plentiful information, most of which is not peer reviewed, and may be covert advertising. This can result in health ‘wants’ being confused with health ‘needs’. Surgical practice needs to evolve to meet these challenges.

The problem with structuring healthcare around patient demands is that healthcare, unlike buying a meal or holiday, does not usually result in an immediate objectively good or bad outcome. If you asked a doctor what treatment they would want from a surgeon, they would want a safe, experienced surgeon and anaesthetist operating in a good environment with motivated conscientious nursing staff. They would be in a fairly good position to judge this.

Patients have a limited number of surrogate measures on which to base their assessment of quality (current measures proposed to define ‘good’ healthcare in the United Kingdom (UK) include time on waiting list, hospital cleanliness, and ‘nurse empathy’. Tables of operative results enable comparison of the quantity and timeliness of operations, but not necessarily the outcomes for the patients). Government ratings may have little or no relationship to the outcomes that actually matter most to the patient or society, e.g. freedom from pain, employment, and good quality of life.

In trying to measure healthcare quality, there is a technical aspect (expertise, outcome), an interpersonal aspect (attitude, behaviour), and a service aspect (accessibility, environment). As patients usually cannot objectively judge the technical aspects, they tend to base their decisions on the interpersonal and service aspects, which are more amenable to measurement.

The technical aspects such as safety and competence tend to be taken for granted, which is unfortunate because the technical aspects are usually the major determinant of long-term outcome. Moreover, the really important outcomes are difficult to measure. For example, there is a long time lag between a poorly sited or poorly designed joint replacement and its ultimate failure. For these reasons, poor quality work can appear cost-effective if not evaluated carefully, and in a way that considers all costs.

In summary, healthcare quality measurement is vitally important, but often heavily flawed because of the temptation to use surrogate measures that are politically expedient or easy to measure, rather than concentrating on the outcomes that actually matter.

Performing an intervention such as an operation will have different results in different people. Humans are not linear predictable systems, which is probably why they have survived so long.

There are many different factors that contribute to the operative outcome: type of operation, preoperative function, risks from the operation, surgical skill, and risks from the anaesthetic. The contribution of these factors can be estimated from analysis of different populations of patients who all undergo the treatment, but this would ignore factors unique to that patient and that surgeon.

Imagine that you are the patient. You don’t really care about all this. You want to know: ‘Will this intervention, performed by this person, benefit or harm me, and by how much?’ (Box 1.1.1).

To answer this, and be legitimately reassuring, it is necessary to demonstrate that:

Only when these conditions are fulfilled is the risk of harm to the patient minimized. Harm may occur in different forms: medical, financial, and social.

Medical harm is the harm directly resulting from the treatment, and may be physical or psychological.

The notion of medical harm has been crystallized in the minds of the UK public by cases such as Harold Shipman and paediatric cardiac surgery in Bristol. Harold Shipman was a general practitioner who murdered approximately 200 patients by giving lethal doses of opiates. Some cardiac surgeons in Bristol in the 1990s continued to perform complex paediatric operations despite overwhelming evidence of very poor results, heroically documented by an anaesthetist, Steven Bolsin, at great personal and professional cost.

Although these differ in being acts of active harm (Shipman) and passive harm (Bristol) in that Shipman set out to harm patients, but in Bristol the harm was not intended, both sets of events have been seen by the establishment as indicators of inadequate self-regulation by the medical profession. Steps such as revalidation and competence-based assessments have been introduced.

Revalidation is designed to identify poorly performing doctors and ensure that they are retrained. Revalidation is very expensive to perform if such assessments are to be both valid and reliable. Revalidation is likely to evolve to a system of re-taking one’s exit exam every 5 years, as in the United States of America (USA).

Continuing Medical Education is an important part of ensuring continuing competence, but the providers of such education often have ulterior motives, and the end product (any benefit to patients) is not measured, just the process.

In addition to compulsory revalidation, league tables of mortality offer a superficial view of competence, and are being implemented for some surgical procedures in the UK, and have been available in the USA for some time.

As doctors, we need to recognize both the ageing of our knowledge, and the changing nature of our jobs. Specialties such as cardiothoracic and vascular surgery have changed completely in 10 years, and technology such as robotics will have an increasing role in trauma and orthopaedic surgery. It is likely that periodic retraining will be a normal expectation and will be written into job plans in the future.

By medicalizing a normal human condition, it is possible to create ‘diseases’ where none previously existed. This is obviously good news if you have a product that cures the ‘disease’. The widespread usage of antidepressants for a rather hazy definition of ‘depression’ medicalized many thousands of people who may have had no more than transient unhappiness. Direct-to-consumer advertising, allowed in some countries, encourages consumers to self-diagnose and approach doctors for specific branded products. These behaviours encourage expenditure on health ‘wants’ rather than health ‘needs’ which disadvantage the poor and less articulate and result in financial harm (see following section).

Back pain is an enormous social burden. In the UK, 50 million workdays are lost each year and 500 000 people are on long-term incapacity benefits for back pain. The advent of magnetic resonance (MR) scanning has enabled everyone to have a diagnosis, but it is likely that this validation of non-specific ‘abnormalities’ has not benefited patients or society.

There is direct financial harm to the patient and society in every illness through lost earnings and inability to participate in social activities. However, in a single healthcare provider system like the UK National Health Service (NHS), there is also an opportunity cost to every treatment.

For example, an opportunity cost occurs if one spends X pounds on drugs for Alzheimer’s disease patients, then one cannot spend X pounds on hip prostheses. The decision about which treatment is the best option is laden with value judgements. Central questions are:

Performing these calculations necessitates value judgements on quality of life, which inevitably have a subjective element.

Engaging in futile or damaging treatment not only causes direct damage and waste, but also deprives another patient of an effective treatment. This makes the continuing presence of homeopathic hospitals funded by the NHS highly illogical, as anything that cannot prove superiority over placebo is just money diverted away from effective healthcare (Box 1.1.3).

The placebo (Latin for ‘I will please’) effect is more than just what one attributes the apparent success of the ‘snake oil’ peddled by the quack. It is part of the psycho-social interaction that is present in any therapeutic interaction. It is heterogeneous, and has an interesting history.

The placebo effect appears governed by several factors, which appear additive:

It is sometimes argued that it does not matter what method is used, as long as the patient gets better. This sounds plausible, but if one is spending public funds on healthcare, one has a duty to ensure best value. Therefore treatments that cannot prove their effectiveness over a placebo should not be publicly funded.

By understanding and enhancing the placebo value in medicine, one can maximize patient satisfaction and the chances of patient compliance, both of which are likely to benefit your patient.

In the 1960s, prior to gastric acid-suppression drugs, gastric free zing was commonly performed. By freezing the gastric mucosa, acid-producing cells were damaged and acid formation was suppressed allowing the ulcer to heal and the patient was cured.

In an era where acid suppression drugs are cheap and effective, it is difficult to understand the disease burden of peptic ulceration to patients and society in general. It was found that reducing the temperature of the gastric mucosa to 5–10°C markedly reduced acid secretion. The exciting new technique of freezing the gastric mucosa was, in effect, a miracle cure for many patients.

In 1962, a trial of 24 patients by a past president of the American College of Surgeons showed that all had relief of symptoms of duodenal ulcers 6 weeks after having had a gastric freezing operation. More than 2500 gastric freezing machines were sold, and hundreds of thousands of freezings were carried out. Patient satisfaction was high, and a review of the procedure commented that ‘the method is easy and the clinical results are astoundingly good’.

Although more than 100 articles about gastric freezing were produced in the following years, it was not until 1969 that a placebo-controlled trial of gastric freezing was published. This study, with approximately 75 patients in each arm had sufficient power to detect a 20% difference between the treatments, if one existed¹. There was no difference between the treatment and control groups.

In short, a lot of time and energy and money had been expended on a placebo treatment. In addition active harm was done to patients—fatalities ocsurred in patients under going the treatment—and the financial harm was immense. Without the insight that properly conducted trials provide, doctors are little more than purveyors of snake-oil (Figure 1.1.2).

To prove if one treatment is better than another. What sort of trial you do depends on what you are trying to prove:

Trial design is more important than trial analysis: if the design of a trial is flawed, no amount of fancy analysis can remedy this. It is vital that you do some background reading and talk to someone with experience of trial design before you start collecting data. Any trial involving human subjects also needs ethical review before it can start.

A literature search should be performed to see whether anyone has already done the trial, but also to generate ideas about how best to perform the trial.

When thinking about how many patients you will need for your trial, you need to consider incidence and prevalence. Incidence is the number of new cases of a condition, e.g. fractured neck of femur. Prevalence is the number of people with a condition at a particular point in time, e.g. back pain. If your disease is uncommon, recruiting a sufficient number of patients to be able to adequately power the study will be difficult.

Any trial will need ethical approval. All NHS Trusts and universities have a structure of ethics committee(s) for approving and monitoring clinical research. These committees are charged with ensuring that all research is of good quality and safeguards the rights of the patients and the reputation of the institution.

Informed patient consent is, excepting a few defined hyperacute situations, an essential component of any research, and the patient consent forms and accompanying information will be carefully scrutinized by the ethics committee.

Bias occurs when you fail to measure what you intend to/should measure. This may happen for a variety of different reasons—an example might be if you did a satisfaction survey of patients you saw in outpatients following hip surgery. You might achieve a very good satisfaction score, but by limiting your sample to those who are able to come to your outpatients, this may not be an accurate reflection of the true picture. Bias is therefore the inclusion of an error that distorts the trial in a non-random way (Box 1.1.6).

Confounding occurs when you fail to measure or spot an extraneous factor that accounts for your findings. Say you had a group of patients in whom you measured premature failure of hip prostheses and doughnut consumption, and found a strong relationship. Does this mean that doughnuts cause prosthesis failure? Or might it be that there is another variable that you have not measured that would explain this—a confounding factor?

Bias can be minimized by blinding, randomization, and concealment, but it is not always possible to randomize and blind in surgery.

The Hawthorne effect is named after experiments done at the Hawthorne factory in the 1920s when factory workers were assembling relays for telecommunications equipment (Figure 1.1.4). Researchers looking for the optimal conditions for factory work increased the ambient light and found that production rates improved. They then reduced the ambient light and found that production rates improved again. Eventually the researchers deduced that the presence of the researchers was the over-riding influence on the workers’ performance. Therefore the Hawthorne effect is that just by measuring something, one changes it. Having a control group ensures that the Hawthorne effect does not affect the results.

The consumer of knowledge can never know what a dicky thing knowledge is until he has tried to produce it.

F.J. Roethlisberger, investigator at Hawthorne.

Objective or ‘hard’ outcomes, such as death, are preferred in research as there is little debate about their significance. Subjective or ‘soft’ outcomes, such as a change in a pain score, may not be valid or reliable across different populations of patients. If a validated measure of outcome is available, this makes design easier. If any of the measures used are in any way subjective, e.g. fracture angulation or code, it is best to have more than one independent observer and measure agreement between these observers.

If the ‘hard’ measure is rare or difficult to measure, then surrogate measures may be used. Death due to pulmonary embolism following joint replacement is quite rare, but using deep vein thrombosis (DVT) as a surrogate is fraught with difficulty as DVT is very common, but when limited to below the knee, does not seem to be associated with adverse outcomes.

The gold standard for pharmaceutical research is the double-blind placebo-controlled randomized controlled trial (RCT). RCTs are also the gold standard in surgery, and provide a definitive answer, e.g. for arthroscopic surgery in osteoarthritis. However, RCTs of any sort are uncommon in surgery. This is partly because in order to ensure the full placebo effect in the control group, one must perform a sham operation, and this is ethically and practically difficult. A pragmatic alternative is to randomize based on expertise—if there are two groups of surgeons available with expertise in variants of a surgical procedure, one can randomize patients between these two groups.

Another technique that is used to deal with the difficulty in conducting large RCTs is to use meta-analysis to combine the results of similar small trials.

In an ideal world, all medical and surgical practice would be based on evidence from well-conducted, multicentre, double-blinded, randomized placebo-controlled trials. Unfortunately such trials are very expensive to organize, and therefore are only performed to answer the most large-scale health questions, e.g. the CRASH (Corticosteroid Randomisation After Significant Head injury) and CRASH2 trials relating to the role of tranexamic acid in patients with major haemorrahage from trauma.

Meta-analysis is a structured review of all the trials relating to a particular research question. It is used when there are many small trials that relate to a research question, but none of them have the power to resolve the question on their own. The quality of evidence from each trial is weighted in a structured and predetermined way to allow a definitive assessment to be made. Meta-analysis may have the power to quantify benefit or harm that smaller trials may not be able to do, particularly if these effects are uncommon.

Careful judgement is necessary to combine the results of several trials in a way that does not prejudice the final result. The Cochrane Collaboration (Box 1.1.7) is an independent organization that organizes and validates meta-analyses of clinical questions.

In the 1990s, intravenous colloid fluids, e.g. Haemaccel, Gelofusine were widely used for resuscitation as they were effective at improving a patient’s blood pressure quickly, and the large protein molecules did not leak out of damaged capillaries as quickly as crystalloid. The downside of this is that once they did leak out, they tended to cause resistant tissue oedema, which is bad if the tissue involved is the brain.

In 1998 a meta-analysis performed by the Cochrane Collaboration suggested that colloids gave no benefit over crystalloid fluids, e.g. saline, and appeared to cause harm (up to 5% increase in mortality) in critically ill patients. This, together with their high cost (20 times that of crystalloid), are why colloids are not now used in resuscitation.

The volume of medical literature is vast—there are more than 40 000 biomedical journals, and this number doubles every 20 years. The quality does not. How can you filter this soup of raw data? How can you work out whether something is true or not? If true, can you apply this new knowledge to your daily practice?

The Bulstrode criteria for medical literature are short and sharp:

While the first and last of these depend on subjective judgements, the key skill is the ability to read a paper, spot flaws, and be able to judge whether it should change your clinical practice.

Professor David Sackett, one of the originators of evidence-based medicine, developed a set of criteria that expands on the Bulstrode criteria to help you evaluate whether a trial is valid and applicable to your practice (Box 1.1.8).

The conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients.

The term ‘evidence-based medicine’ was coined by a group of physicians at McMaster University, Ontario, in the 1980s. The aim was to encapsulate the information search and analysis techniques that had been learned from epidemiology in a form that could be used at the bedside. Although this may seem like second nature now, this was in an era before the Internet, PubMed, Google Scholar, and online journals.

One of the problems in traditional medical practice is that doctors tend to carry on doing the things they were taught in medical school. One of the arguments against old-style medical school curricula was that it did not teach the skills for life-long learning. The integration of the science behind ‘evidence-based’ practice into the medical school curriculum has ensured that future generations of doctors will be able to identify and challenge outdated or dangerous practice from an objective viewpoint.

With the many claims and counter claims of patients, healthcare providers, researchers, pharmaceutical and device manufacturers, one needs a league table to be able to work out which evidence trumps others (Table 1.1.1).

It is all very well having found and assessed the best evidence, but you have been wasting your time if you cannot put it into practice. Approaching people and telling them that their practice is out of date and/or dangerous does not always result in the desired outcome (see Bristol, discussed under Medical Harm earlier).

Unless there is a pressing urgency, e.g. ongoing risk of significant harm to patients, it is necessary to prove that there is a problem with the current system. The way to do this is through audit. Audit is different from research in that audit is ensuring optimal usage of current knowledge, whereas research is about creating new knowledge.

An audit is performed to measure the organization’s performance against agreed standard measures of performance. These might be international or national guidelines (e.g. National Institute for Health and Clinical Excellence, NICE), ‘best practice’, or College standards.

Guidelines are advisory summaries of evidence that are used to guide treatment. They are not supposed to be rigidly interpreted, but if one is going outside their advice, one should have good reasons. Protocols are designed to be rigid: they are standardized routines that can be used to streamline and minimize defects in a process. Both guidelines and protocols have inherent dangers in that they tend to stop people thinking about what they are doing, and may stifle innovation.

Any guideline or protocol must be both exhaustive and exclusive. It must cover all eventualities that will occur, and must reliably exclude other conditions that may look similar. Many protocols, elegantly simple in their first draft, can become so complex as to be unworkable when revised after testing. This is a measure of the inherent complexity within healthcare, which is why we need doctors who have the broad range of knowledge and skills to manage these problems, rather than just technicians.

Make everything as simple as possible, but no simpler.

Albert Einstein.

Audit is the process by which we ensure that we are meeting the standards that we set ourselves. Audit differs from research in that research is looking for new knowledge. Audit is about ensuring that present knowledge is being used in the most productive fashion. Audit may be used to generate and refine research questions, but is usually used in the context of a quality improvement framework.

Audit can be prospective, but is usually retrospective—at least in the first instance. For example, to audit time to theatre for open fractures, an easy way would be to collect all patients with a discharge diagnosis of open fracture. However, this would miss patients who had died or been transferred. A better way would be to prospectively collect the data on patients who come in through the Emergency Department with open fractures.

Types of audit:

Audit is an important part of Clinical Governance, being the way that clinicians measure their own performance against their peers, and also a structured way of demonstrating necessary service development. Audit data can be compared against agreed standards/national guidelines e.g. NICE/Scottish Intercollegiate Guidelines Network (SIGN)/‘best practice’ from the literature or College.

The most difficult part of the audit loop is usually making the change, and this is why audit usually takes place within the framework of Clinical Governance. This provides the oversight and a mechanism for pushing through the process changes necessary. If the audit requires a questionnaire, time spent researching the format and construction of questionnaires will substantially increase the quality of the information garnered.

This rather nebulous term is derived from the notion of Corporate Governance, used in business. Corporate Governance is the process of ensuring that the people who run the company (senior management) act in accordance with the interests of the people that own the company (the shareholders, represented by the Board).

Clinical Governance on the other hand means the process of showing that the work you do is as close to best practice as is feasible, and therefore is a quality assurance process. It was defined as:

A framework through which NHS organisations are accountable for continuously improving the quality of their services and safeguarding high standards of care by creating an environment in which excellence in clinical care will flourish.

Scally and Donaldson (1998).

From a clinician’s viewpoint, Clinical Governance is important, as apart from clinical adverse incidents it is the only forum in which to demonstrate and resolve problems with clinical service delivery. Clinical Governance incorporates what were Morbidity and Mortality meetings, but should be proactive, identifying problems before they result in significant patient harm, incorporating best practice and a framework for service development/quality assurance.

Your decision as to whether to offer a surgical solution to a patient’s problem is based on clinical diagnosis and investigations. An expert should have insight into their own process of assessment, to be able to teach others but also to understand how and why things go wrong.

Much time, ink, and paper has been expended in pursuit of understanding the process of diagnosis. The short answer is ‘we don’t know’. However, some basic principles seem to be that:

There is no such thing as clinical reasoning; there is no one best way through a problem. The more one studies the clinical expert, the more one marvels at the complex and multidimensional components of knowledge and skill that she or he brings to bear on the problem, and the amazing adaptability she must possess to achieve the goal of effective care.

Norman (2005).

Research into clinical reasoning helps us understand some of the processes that underlie clinical decision-making.

If a trainee has persistent problems with poor clinical judgement, then this is usually handled through their training programme. If a colleague shows poor clinical judgement, this should be taken up through the Director/Divisional Director/Medical Director. If these avenues do not resolve the problem, in the UK there is the National Clinical Assessment Service, part of the National Patient Safety Agency (http://www.npsa.nhs.uk).

As investigations have become more complex and are used earlier in the disease process, their flaws have become more apparent and caution is needed in interpreting the results. There are risks and benefits with all investigations but as the treating doctor, there is a duty of care to use them appropriately. Some tests may give misleading results if used in unsuitable groups of patients.

A patient with longstanding mechanical back pain has an MR scan of their back. An abnormality is noted but is perceived by the radiologist to be the result of normal wear and tear, but is mentioned in the report. The patient and their family may put pressure on the doctors because ‘something must be done’. An operation may be performed that is both unsuccessful and/or has complications in the surgery, anaesthesia, or recovery period. Significant medical and financial harm has occurred to both the patient and society.

In this situation, the scan could be said to be a false positive—the scan implies that there is a significant abnormality that is responsible for the patient’s pain, whereas in fact this is not the case. It could reasonably be argued that it was not the scan that was a false positive, but the diagnostic process as a whole. The MR scan may be very accurate at finding abnormalities. However, what is required is a scan that can:

This is why we need clinicians to interpret these scans in the light of clinical judgement. The ability to weigh all this grey information factors accurately to produce an appropriate treatment plan is one of the hallmarks of an expert clinician.

Both society and individuals seem far more tolerant of false positives than false negatives. It is seen as perfectly acceptable to provide too much treatment rather than not enough, whereas in reality both may result in similar rates of harm in the widest sense. A good, if politically charged, example of this would be screening for breast cancer:

For every 2000 women invited for screening throughout 10 years, one will have her life prolonged. In addition, 10 healthy women, who would not have been diagnosed if there had not been screening, will be diagnosed as breast cancer patients and will be treated unnecessarily.

Gotzsche and Nielsen (2006).

Such information is difficult to communicate to the public. By way of contrast, one does not have to open too many newspapers to find sensational reports of a catastrophic ‘missed diagnosis’: these are the opposite: the false negative.

These two situations, the false positive and false negative, can be put into a table together with the true positive and true negative (Table 1.1.2).

Apart from the cost, anxiety, and unnecessary investigation that a false positive test entails, there is another form of damage that of which an orthopaedic practitioner should be particularly aware. The radiation dose used in diagnostic testing, particularly in computed tomography (CT), is very large (Table 1.1.3).

The use of the ‘scanogram’ (CT head/neck/chest/abdomen/pelvis) in blunt trauma is associated with significant risks of lethal malignancy—approximately 1:500 for children and 1:3000 in adults. Patients have a right to understand the risks and benefits of investigations that are being performed, and the clinician has a duty to ensure they have access to information to make those choices.

Consent is the legal framework used to cover the agreement between a patient and a doctor to undertake a procedure. There are local rules for how consent is obtained and variation between different countries’ legal frameworks; therefore general principles will be discussed.

For consent to be informed a patient must have capacity to give this consent. The rules governing the assessment of mental capacity have been clarified in English law. The key components of this are:

A patient must be assumed to have capacity, irrespective of age/disability/behaviour/beliefs/illness (including mental illness) or the fact that you may disagree with their decision. Each decision as to whether a patient has capacity must be a new assessment, and it is possible that a patient can have capacity in one situation but not another. All efforts must be made to give the patient the maximum chance at establishing capacity.

The Mental Capacity Act (2005) uses four tests to establish capacity. The patient must be able to:

If the doctor’s opinion is that the patient lacks capacity to consent, and it is not an emergency, then medicolegal advice should be sought, through the hospital’s usual channels.

In an emergency, one may do what is necessary to preserve life and limb, but no more. Verbal consent should be sought, but is not essential.

Advance directives may still over-ride life-saving treatments, but only if they specifically include life-threatening situations.

Advance directives are now legally binding in parts of the UK. In an emergency, an advance directive may be over-ridden if the patient’s condition is life threatening, unless the directive specifically refers to life-threatening condition being included, e.g. blood transfusion by Jehovah’s Witnesses.

If there is doubt about a local situation, medicolegal advice should be sought.

Informed consent does not mean: ‘sign here and I will do the operation’. Consent is a two-way process and may need to include negotiation about what will and will not be performed, and a frank discussion of the likely outcomes of these actions. Consent should include an understanding of what will occur if the operation does not take place.

Part of the process of negotiation is to manage the patient’s expectations. If a patient has unrealistic expectations of surgery, this is the time to address this. Failure to understand the risks and likely outcomes is far more likely to lead to dissatisfaction and adverse outcomes for all concerned.

The notion of adequate risk disclosure has changed in the last few years from that of what a doctor deems necessary to what the patient deems necessary. The old legal definition of ‘acceptable medical practice’ (the Bolam test) being defined as acting ‘in accordance with practice accepted by a responsible body of medical opinion’ has been superseded by a patient-focused test—‘what a reasonable patient would want to know’ (Box 1.1.9).

Risks should be disclosed if the risk is more common than 1:1000, or is very serious—death or serious disability. The advice about risks must be recorded on the consent form. Risk should be numerically quantified and documented whenever possible. If a patient enquires about a specific risk they must be advised of the likelihood of this occurring.

Guidance on consent was published by the General Medical Council in 2008, which takes account of these changes and is freely available through their website (http://www.gmc-uk.org).

There are many examples of the public failing to understand medical risk or misinterpreting information, sometimes due to sensational or misreporting of the media, e.g. risks from MMR immunization.

Expression of risk in percentages is a good start, but may not be readily interpreted by patients with linguistic or educational barriers to understanding. Other approaches to help patients understand risk include the use of:

As an expert clinician, you have a role and a duty to educate. You did not become an expert without the educational input of a large number of teachers. From Figure 1.1.9, as a conscious competent at the end of postgraduate training, you will be at the peak of your educational ability as a ‘conscious competent’.

The danger is the ease of slipping into being an unconscious incompetent—you can do the job, but don’t know why you are doing what you are doing. Regular teaching helps ensure that knowledge and skills are kept up to date. We can all remember inspirational teachers who appeared to effortlessly communicate complex ideas in a fun way. Good teaching does not just happen—it is the result of careful preparation of content, structure, and presentation, and the following section aims to help with this.

The model of learning developed by Kolb is helpful in considering medical teaching. Trainees are generally not short of concrete, i.e. real-life, experiences, but the trick is to teach in such a way as to build on these experiences and incorporate the theory and facts from textbooks into deeper knowledge.

When planning teaching, there should have some overall aim stated—what should the student be able to do/understand at the end? This aim can be broken down into component goals—specific chunks of learning that build together to achieve the aim. Telling students these aims and goals at the beginning of teaching helps them understand how the information is going to fit together into something that will be useful. This increases the chance that teaching becomes learning.

The most common teaching an orthopaedic surgeon will do is teaching practical skills, and for this it is helpful to have a minimal structure that helps organize how you transmit the information in a way in that is likely to be understood and retained.

By explaining what you are going to do, and breaking it into stages, you allow the learner to look for the different stages when you put them together as a demonstration. The learner then imitates what you have demonstrated, and then practises this to a level of competence.

When you are teaching practical skills it is important to be able to help the person you are teaching to improve. If you think back through your training, you will be able to think of times when this has not been done particularly well. A good acronym for this is PQRS:

While lectures are superficially efficient at teaching, the amount of learning is highly variable, and difficult for the lecturer to gauge. Dividing large groups into small groups for teaching has many advantages:

Small-group teaching is initially quite challenging for many teachers as there is an implicit loss of control of the whole process, and there is a danger the whole thing goes wrong (very rare, providing the task is well structured).

Specific small-group techniques that may be useful are:

Bedside teaching is a particularly difficult skill to do well, yet is the one that has the most chance of inspiring students. A common complaint from students is that they rarely have the opportunity to examine patients with a senior doctor to guide them and give feedback.

Bedside teaching gives an opportunity for the expert to articulate their thought processes so that students may learn how an expert weighs different factors to make decisions. It provides an excellent opportunity to involve the patient both in decisions regarding their management and in the education process, and the evidence is that patients appreciate and enjoy this.

The ubiquity of electronic presentation using data projectors has spawned a number of habits that can detract from learning. There are a number of things that you can do to lessen the pain for the audience:

Assessment is the term given to establishing whether the students have learned anything or not. Although this may sound obvious, care needs to be given to establishing the purpose of the assessment. Is the purpose:

Training has generally moved away from the former towards the latter. Assessment should therefore be tied very closely to the curriculum: students should have a very clear idea of what is to be tested and the pass rate should be very high, as only the essentials—the core curriculum—must be tested, but they must be passed.

Teachers are always worried about examination materials, particularly practical (OSCE—objective structured clinical examination) stations getting out into the students’ domain. This is going to happen anyway. The consequences of this can be minimized, and the effect harnessed by giving open access to certain assessment materials. For example, by publishing model OSCE stations with their marking schema, one ensures that the overall performance is much improved, this is: ‘using the tail to wag the dog’.

A simple way of evaluating education is to use an anonymous questionnaire. Questionnaires need careful construction to ensure that you receive valid and reliable responses.

Medicine is a science of uncertainty and an art of probability.

Sir William Osler (1904).

This chapter has provided an overview of some of the key supporting skills for expert clinical practice. There is a limited amount of material that can be included in such a textbook as this, and therefore this chapter is little more than a dégustation menu.

The hope is that reading this chapter will equip the reader to go out and negotiate some of the uncertainties that will face them in everyday practice, and also to act as a roadmap for further reading for those who are looking for more information.