Patient-based not problem-based learning: An Oslerian approach to clinical skills, looking back to move forwardAD Franklyn-Miller, EC Falvey, PR McCrory
Department of Exercise and Sports Medicine, Centre for Health, Faculty of Medicine and Dentistry, University of Melbourne, Australia
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0022-3859.57404
Source of Support: None, Conflict of Interest: None
Context : There have been significant changes in the past decade in both the curriculum and its delivery, in undergraduate medical education. Many of these changes have been made simultaneously, preventing clear assessment of outcome measures. The move away from a pre-clinical science grounding, to an integrated 'problem-based learning (PBL) approach' has been widespread in many countries across the world. Purpose : One effect of these changes has been the way in which clinical skills, in particular history and examination are taught. By integrating clinical scenarios earlier in the undergraduate course, clinical skills are increasingly taught in tutorials. This approach, when used in the pre-clinical setting may have shortcomings in the development of the ability to construct a differential diagnosis. There has been little evidence that PBL improves problem-solving ability and this is critical to the differential diagnostic process. The concurrent decline in anatomical teaching and understanding contributes to this difficulty. Discussion : The authors outline a model which clinicians can re-emphasize to students and juniors based on the fundamentals of clinical practice. The apprenticeship is more important than ever in the days of small group learning. The relinquishing of the traditional model of undergraduate medicine is of concern. The effects of educational reform should be examined by further research into the competencies of graduates entering higher professional training, before it is accepted that this change has been for the better.
Keywords: Clinical education, continuing medical education, curriculum development, problem-based learning
Every physician needs to be able to make a correct diagnosis, an essential first step to the next one; treatment and prognosis'.  How the medical student formulates this diagnosis is being altered by developments in medical education. Over the last decade a paradigm shift in the way undergraduate medicine is taught has occurred. Part of this change is the move away from traditional didactic teaching to problem-based learning (PBL) models.  This has already occurred in many countries (e.g. UK,  USA  and Australia  ) and on a smaller scale trials are occurring in India and West Africa following the international trend. , PBL has progressed from the original McMaster concept of postgraduate application, to its introduction into the first semester of undergraduate education. Part of the problem in identifying whether these changes are successful is the mutation of McMaster PBL into different degrees and methods of application.  Dolmans et al. identify three common threads of patient-based problems as stimuli, tutors as facilitators and small group work.  These features are no different than the traditional methods of ward round teaching in the clinical phase of medical school, so educationalists need not have looked beyond this or made such widespread changes to examine the effects.
The total transformation in teaching strategy at many Universities  has led to a fundamental deviation from the absolute requirement for students to build a systematic knowledge base of anatomy, pathology and physiology. These were traditionally the means to construct a diagnosis, and have been replaced in recent graduates, with a common problem-solving framework and 'composite' learning experience. 
The motivation for this change is due to the perception that much of the undergraduate curriculum was overcrowded with factual information of little practical use and inhibited students from developing into creative thinkers  accompanied by the drive for educational reform.  These were good reasons for review; but the changes that have followed appear more widespread than their architect's plans. Changing trends in healthcare have pushed a different set of demands on doctors of the future, the General Medical Council acknowledge that the newly qualified doctor must be able to take advantage of postgraduate training  and indeed laid out a progressive curriculum in 'Tomorrow's doctors'.  By removing much of the factual load from the undergraduate curriculum it can be filled with 'essentials' such as communication skills, teamwork, ethics, information and technology and audit and management.  But a doctor must still primarily be able to construct a diagnosis and centuries of teaching medicine cannot be ignored.
Problem-based learning (PBL) evolved from the educational theory of how we actually learn during our clinical practice. The patient poses a scenario in which we may recognize familiar patterns. From this we construct a differential diagnosis. By targeted questioning and examination we can narrow this differential and ultimately arrive at the diagnosis with or without further investigation or tests. The PBL approach acknowledges this by teaching the 'flow chart' for common scenarios in isolation from the bigger picture. Upon reaching the diagnosis, we direct the student to reference sources for more information on how to deal with the problem, following the best intentions of evidence-based medicine (EBM). Ideally, this method allows the student to identify current knowledge and gaps in this knowledge that will stimulate wider reading outside of the tutorials. The knowledge and skills we acquire allow us to inform our practice from the repeated application of these principles. It is our patients who thus provide the stimulus for lifelong learning.
A scenario built on common findings, modeled by educationalists, to teach the presentation of myocardial infarction or left cerebellar infarction, will direct the delivery of this module to the intended diagnosis. Practicing medicine in real life is seldom as convenient. It is acknowledged that experienced actors can perform a convincing role  but it is left to the student to fill in the missing parts of the differential in which they may not have the luxury of a tutorial. Can a doctor accurately assess what he or she requires to construct a differential diagnosis without the fundamental basic science knowledge base behind this process?
Throughout the world in institutions where PBL was introduced, numerous other changes (e.g. curriculum changes, admission criteria etc) occurred at the same time, making assessment of any putative benefits of these new courses difficult. The opportunity to validate outcome measures, based on these changes, has been lost. Educationalists widely assume that these changes were for the better,  although specialist colleges and other professional groups are now more skeptical. , Maudsley suggests that PBL is borrowed by universities for 'prestige' and 'subversion.' 
The evidence supporting PBL as an effective educational tool at undergraduate level is limited. McMaster originally intended it as a postgraduate tool where existing educational theory was very similar with the PhD students being guided to think for themselves based on core knowledge from their undergraduate education. Reviews to date have highlighted that although successful at improving information retention, PBL has not improved overall knowledge and competence.  Students certainly like the model when compared to traditional education, , but extrinsic factors to small group learning could easily influence attendance and attitudes. Some reviews have shown marginal outcome improvements but not in proportion to the radical shift and the costs of PBL implementation.  Disappointingly, there is little evidence that PBL results in improvement in problem-solving skills. , A study of psychiatry trainees (some may say the perfectionists of history-taking) indicated that although the PBL curriculum was more successful than the traditional learning method in terms of examination success, the improvement was likely to be due to students learning the volume required more effectively rather than through enhancing their diagnostic or problem-solving ability.  Yalcin et al.  have demonstrated short-term gains in problem-solving skills and scientific thinking, but these were confined to first-year education. The applicability of this enormous change should have been examined prior to its implementation, by assessing the end product, at higher professional examinations.
There is no evidence to suggest that the skills of history and examination are better served by the PBL route. Indeed the personal experience of the authors teaching medical undergraduates from a 'modern PBL style curriculum' suggests that the problem-based approach limits the development of questioning faculties and applying a 'cerebral sieve' to approaching diagnosis. The need to fit a scenario into a pattern leads to frustration when this is not possible. Students do still perceive the need to confront real patients to acquire real skills  but much of the evidence in support of the new PBL curriculum is qualitative or opinion-driven  rather than objectively assessed. At the present time, there is insufficient data to comment on the overall quality of future qualified physicians.  But this is the area where research must be directed.
Those responsible for medical education and curriculum design have traditionally included teaching basic clinical skills (such as history-taking and physical examination) in their remit. The days of the hospital 'teaching round', however, where an experienced physician demonstrates the techniques and skills in elucidating facts from the patient appear lost.  Where PBL-trained graduates attend such rounds, the personal experience of the authors is that these graduates lack the systematic body of medical knowledge to take full advantage of the learning opportunity. Although provision can be made for rapid internet searches, these tend to be more useful for specific knowledge points (e.g. drug doses, side-effects etc), rather than the clinical pattern recognition, often with limited or incomplete data, that forms the basis of medical practice.
Although 'communication skills' is the current vogue terminology within PBL curricula, the ability to apply history-taking skills in a real life patient is limited. An actor (in place of a real patient) may be prompted to relate 'no headaches, dizziness, weakness or loss of sensation' but how often do our real-life patients present which such direct and specific exclusions? This reproducible, sanitized, highly structured learning developed by educationalists leaves no room for development of the art of clinical questioning and must be a concern for future generations of doctors. How can this compare to real life, where the skills of the clinician are required to enable the rapid recognition of a diagnosis in the absence of a classical disease history, quite often under the influence of temporal and emotional stressors? The recognition of such complicated presentation patterns requires an expert knowledge base from which to select the differential diagnosis, not merely the skills to fit the situation to manufactured learned cases, or find the answer via a literature search.
It is unlikely that we will recognize any true decline in the clinical skills and diagnostic power of the next generation of doctors until too late. It takes over 15 years to fully train in a speciality and this will result in a hangover effect where changes in learning methodology will not be apparent until these PBL-trained classes of doctors complete their higher training. Here they will face the more traditional model of learning and already colleges are noting the lack of preparation of entrants for these first professional examinations. 
Many authors agree that PBL graduates consistently demonstrate knowledge of basic science inferior to those students taught on traditional courses. ,,, The effects of this on diagnostic skills may not be tested until Registrar or Fellowship level, when the increased burden on existing senior consultant staff may become apparent.
The experience and attributes of the skilled clinician are long recognized.  The rise of EBM means we may have left some of these skills behind in the pursuit of the skills of problem definition, searching, evaluating, and applying original medical literature. Beyond the growing concerns of both clinicians and those involved in specialist training, studies have highlighted the importance of clinical proficiency. Reilly demonstrates that physical examination changed the diagnosis in a quarter of patients admitted under internal medicine units and as such, specific training and feedback of trainees is essential.  Johnson et al. further demonstrated that the clinical examination skills correlated directly with the time spent on bedside teaching by senior physicians.  Tamblyn has also demonstrated a causal link between poor history-taking skills and litigious complaints against those doctors. 
Bedside teaching, the mainstay of many generations of clinical education has been lost in favor of simulated patients and situations to best fit the curriculum teaching modules and the examination standardized process. Sadly, clinical scenarios rarely 'fit in' to individual categories, an unpalatable truth which poses a fundamental problem for any PBL curriculum. There are those who suggest that these basic skills are best left behind to further more modern diagnostic techniques centered on imaging and laboratory tests. They argue that the diagnostic power of history and examination is poor, and that no single symptom or sign is sufficiently discriminatory. 
The judicious use of such advances in investigations has improved understanding and diagnostic yield in some situations, but also driven sharp increases in their use diagnostically, economically, and medico-legally as physicians seek to protect themselves from the threat of litigation. Iglehart looked at the use of imaging services in the US  and found a spiraling use of magnetic resonance imaging (MRI) which did not improve diagnostic yield.
By focusing on a thorough clinical history and examination using the traditional teaching model, not only do we have the potential to reduce the burden of unnecessary investigation, but also to target investigation more accurately and appropriately.
Sir William Osler (1849-1919) insisted that his students learned from seeing and talking to patients and this approach was instrumental in the establishment of medical training programs within hospitals. He himself liked to say, "He who studies medicine without books sails an uncharted sea, whereas he who studies medicine without patients does not go to sea at all." Osler believed that the best textbook was the patient himself, and the application of anatomy, physiology, and pathology were crucial in understanding the patient's condition. The clinician draws on his knowledge base to refine a possible diagnosis from 'triggers', narrowing down the differential diagnosis to form the basis of further investigations. Many esteemed colleagues agree , but with didactic anatomy and basic science teaching being out of 'vogue' it is in dangerous decline.
Pattern recognition forms a critical part of the diagnostic process. This allows the discrimination between a number of diagnostic possibilities and subsequently focuses the physical examination and future investigation. 
In order to refocus attention on such a vital area we outline a four-step compartmentalization of the diagnostic process, [Figure 1] which emphasizes history and examination in an effort to limit investigation to the final step in the process. This is little different from conventional thinking but dividing it into a four-step approach is a means by which the increased emphasis on traditional skills can be focused by the clinician, teacher or student.
Step 1: Define and align
We suggest the clinician defines the area of the patient, which is the likely source of the pathology as the initial step. For the patient presenting with a laceration to the anterior aspect of the wrist in the Emergency room, the obvious 'define and align' is the wrist, but the clinician should also include those structures potentially damaged in the area, such as median nerve, flexor digitorum and radial and ulnar artery. It can be seen that this will align the questioning and subsequent examination to focus on particular structures that may have been injured far more than just thinking 'wrist'.
Step 2: Listen and localize
Following on from Step 1, an inability to move fingers or a change in sensation in part of the hand requires specific questioning to narrow down the examination. It is as important to pose the question: What it is not, as well as what it is? The patient with a laceration may well be suffering from preexisting conditions, which prevent full hand function, such as a more proximal neuropathy; so important negatives require attention and training in application. To exclude other sources of pathology, questioning should be targeted at relevant systems or anatomical areas. The consultation must be approached, following Step 1 with a list of conditions that must be excluded. This does require knowledge of the underlying anatomy and pathology. Unfortunately, one of the striking problems of the shift to PBL style teaching in medical schools is the de-emphasis of teaching anatomy and physiology in a structured format. This has the potential to limit the diagnostic acumen of the PBL-trained clinician.
Step 3: Palpate and re-create
To examine with specific purpose a movement or test, not to just do such a test for completeness, but to identify specific pathology. Too often as students we attempted to feel a praecordial 'thrill' on examining the cardiovascular system, or elicit Murphy's sign without thinking why we were attempting this skill. In the wrist laceration it is important to tie the test with the potentially damaged structure, holding the other fingers in extension whilst testing flexor digitorum superficialis, or by looking for flexor digitorum profundus injury by isolating the distal interphalangeal joint. Without this basic understanding, an important injury may be missed.
Step 4: Alleviate and investigate
In the example of the wrist laceration, it is likely no further investigation would be required given an in-depth history and examination. In the case of ambiguity, an MRI request form with 'Laceration of Flexor digitorum profundus' is of far more value to the patient and radiologist than 'tendon damage'. Some conditions can be alleviated by intervention, or maneuvers and these should be attempted where appropriate. These are diagnostic tests in themselves as their effect is guiding the investigation further.
The evolution of PBL models of medical education has driven the decline of bedside teaching. The lack of understanding and practice of history and examination skills in real patients risks a parallel decline in the diagnostic skills of the clinician of tomorrow. Without an understanding of the anatomy of an area, the clinician has little chance in examining it correctly or thoroughly. In many cases, investigations and special tests are carried out as function of habit and blind searching rather than specific targeted testing based on the application of traditional Oslerian diagnostic skills.
The authors recognize that there are benefits from integrating a PBL approach, as demonstrated in the meta-analysis by Dochy et al.,  but these are restricted to knowledge application and student satisfaction. These benefits alone are not sufficient on which to build a generation of future clinicians. The fundamentals of medical education require a didactic approach. This more recently has been recognized by some who initially were driving the PBL approach, who are now seeking to slow down the race, if not wind back the clock. 
We, as clinicians have the opportunity to redefine PBL teaching and the authors believe that including anatomy and physiology at the core of this process is essential. The true PBL method requires that the 'knowledge facilitator' does not provide content knowledge but guides the students in the search strategies. We would argue that this approach is insufficient for medical undergraduate teaching. Where the PBL approach includes a wider and more integrative tutor base, where a biochemist and an anatomist interact alongside a clinician and a hematologist one can see the positive benefits. This is where the true wider understanding and stimulation of the student will come from. An expert explaining firsthand in small groups where the precise detail fits into the clinical picture, will inspire the next generation to explore the basic sciences underpinning our specialty. They have to be allowed to use their specialist knowledge, unlike the McMaster facilitators.
The authors recognize the limitations in the requirement for appropriate levels of funding and staffing. Academic departments already struggle to recruit and maintain facilitators for PBL. , It is up to practicing clinicians to volunteer their time for the benefit of teaching of undergraduates so that these skills are not lost, or indeed for hospitals to set aside time in all doctors' schedules for teaching of students. Rather than isolate the student back to the pre-clinical lecture-based learning, let us combine the skills and knowledge base of existing clinicians to inspire the next generation. In order to do this we need to combine the benefits of both schools of thought, not lose one in favor of the latest trend. Particularly, in the teaching of anatomy and pathology, which are the cornerstones of clinical practice as defined by Sir William Osler. 
Existing research is limited by the heterogeneity of PBL and its implementation.  It is also usually targeted at short-term rather than long-term effects and challenges the 'whole learning' outcome rather than areas where it has been shown to be weak. Sanson-Fisher suggests that given the lack of substantive evidence behind PBL it is reasonable to question why medical schools have been so quick to join, and postulates it is the desire to innovate and not lag behind their competitors.  The authors agree with the summation of Dolmans et al.  who conclude that PBL research focused on one outcome measure will always give mixed outcome measures. Maudsley et al.  in a well-designed study highlight that a hitherto unmentioned problem with PBL is that poorly directed sessions by the tutor or poorly prepared sessions by the student deliver nothing in terms of education but does support the use of expert tutors imparting their knowledge, perhaps a small step back towards traditionalism. Koh et al.  refreshingly have begun the process of evaluating effects on physician competency and do show some encouraging results, they are at present focused on more social dimensions but remain positive. Further work has begun looking at examination success at a higher level,  again this demonstrated no statistical difference between PBL and traditional schools, although as previously cited, the difficulty is in the defining of what remains a traditional or PBL school. It is recognized that the systematic reviews in existence have strict inclusion criteria which limit their analysis to end goals  but one can analyze the method of learning.
We must utilize the benefits of small group teaching, but with expert tutors who understand the clinical application of basic science and vice versa. Rather than sending the student off to utilize Google and find erroneous links and eponymous names we should seek to impart knowledge gained over the years but via a more open structured method.
The curriculum for the education of doctors has been refined over centuries of development, and the resultant standards and professional abilities of generations have been maintained. In the last decade, rapid changes both in the content of curriculum and in the method of its delivery have been made with little evidence to support such changes. The Oslerian approach has maintained the integrity and standards of our profession for over 100 years; before it is cast aside perhaps clinicians should reflect upon their curriculum and standards rather than be advised by 'educational experts'. In adopting these curriculum and educational changes so readily, have we have forgotten the meaning of what is a doctor? The ability to construct, from a detailed history and clinical examination, a differential diagnosis for the patient.
The authors believe that in a rush to follow the trend medical schools have made wholesale changes without putting in place safeguards to effectively appraise, in a robust scientific manner the impact that this paradigm shift will and has had on postgraduate levels of diagnostic and scientific knowledge. The Oslerian principles and basic science knowledge have been lost with the decline of traditional teaching in favor of ethics, IT, teamwork, communication skills and management. Parallels can be drawn with the modernization of the UK postgraduate training with the development of rapid changes without thought as to the end product. Concerns are being voiced over this process due to its accelerated course in a similar vein. [ 53] We as a profession have abdicated responsibility for the organization of medical education to non-clinicians; let us not abdicate our responsibilities to our profession itself. We should focus research on the successful attributes of experienced working clinicians, listen to their advice and apply it to educational frameworks rather than the reverse, as evidenced by the spread of PBL. It is not the fault of PBL, just its speed and method of introduction that has led to this dramatic shift. PBL many benefits can be integrated within a traditional curriculum but can never wholly replace the building blocks of a physician's training. It is up to today's clinicians and the Royal Colleges to reaffirm what underpins a medical education and to develop a new model appropriate to educate tomorrow's doctors.