clinical skills training

A model of clinical problem-based learning for clinical attachments in medicine

Derek C Macallan,1 Andrew Kent,2 Sandra C Holmes,3 Elizabeth A Farmer4 & Peter McCrorie3

CONTEXT Problem-based learning (PBL) has been widely adopted in medical curricula for early-years training, but its use during clinical attachments has not been extensively explored. OBJECTIVES This study aimed to develop and evaluate a new model, clinical problem-based learning (CPBL), to promote learning skills, attitudes and knowledge during clinical attachments. METHODS The CPBL model takes the principles of PBL and applies them to learning during clinical attachments. Real patient encounters are guided by a list of broadly dened case types to ensure curriculum coverage. By discussing history taking and examination in the context of differential diagnosis and problem listing, students generate learning objectives relating to clinical skills, disease mechanisms and clinical management. These are explored through self-directed learning before the second tutorial, in which the tutor takes the role of expert, demonstrating how

learned material translates into clinical practice. We evaluated which components contributed most to the success of the model using semi-structured questionnaires, focus groups and a consensus (Delphi process) method. RESULTS Students found CPBL a positive learning experience. Identication of suitable cases for discussion was readily achieved, although follow-up was sometimes difcult. The tutors level of expertise and a nonthreatening learning environment, conducive to student questioning, were highly rated contributors to successful CPBL. Comments reinforced the view that CPBL is a parallel teaching approach that helps structure the teaching week, but does not replace traditional bedside teaching. CONCLUSIONS Clinical problem-based learning was well received in clinical placements. Key elements were the learning interval, the involvement of expert tutors and a non-threatening learning environment.

Medical Education 2009: 43: 799807

doi:10.1111/j.1365-2923.2009.03406.x
1 Division of Cellular and Molecular Medicine, St Georges, University of London, London, UK 2 Division of Mental Health, St Georges, University of London, London, UK 3 Centre for Medical and Healthcare Education, St Georges, University of London, London, UK 4 Graduate School of Medicine, University of Wollongong, Wollongong, New South Wales, Australia

Correspondence: Prof. Derek Macallan, Professor of Infectious Diseases and Medicine, Centre for Infection, Cellular and Molecular Medicine, St Georges, University of London, Cranmer Terrace, London SW17 0RE, UK. Tel: 00 44 20 8725 0283; Fax: 00 44 20 8725 3487; E-mail: macallan@sgul.ac.uk

Blackwell Publishing Ltd 2009. MEDICAL EDUCATION 2009; 43: 799807

799

D C Macallan et al
When faced with the task of designing a clinical attachment curriculum to follow 2 years of PBL in a new graduate programme in medicine, we considered these objections outweighed by the benets of maintaining a consistent educational approach throughout the course. We felt that clinical reasoning could not be considered to be fully developed after such a short period and believed that students would benet from continued structured teaching. We thought that clinical teachers could be positively engaged in novel learning processes and were conscious of the wealth of clinical teaching material available which we wished to exploit, whilst retaining some control over curriculum coverage. We therefore developed, piloted and evaluated a model of clinical PBL (CPBL) based around real patient encounters in real clinical settings. Although it shares the same theoretical basis, this model of CPBL differs in several major respects from traditional PBL. Our evaluation focused not just on whether or not CPBL was successful, but also on which components contributed most to successful implementation.

INTRODUCTION

Problem-based learning (PBL) has been widely adopted by many medical schools, in which it frequently constitutes the primary teaching tool in the early years of medical training.1,2 Problem-based learning was initially developed for pre-clinical teaching3 and this remains its primary setting, albeit in many guises.4,5 As a common goal of most PBL programmes is to develop clinical reasoning skills, it is perhaps surprising that there has been little exploration of PBL as a teaching tool during clinical attachments. Attempts to extend PBL into the clinical years have included continuing paper case-based tutorials,6,7 paper-based patient simulations offered in addition to clinical experience,8 task-based learning,9 interdisciplinary clerkships10 and the use of real patients in general practice PBL tutorials.11 There remains, however, in most curricula, a remarkable lack of crossover between PBL and clinical teaching.12 This lack of crossover is even more surprising when one considers that the theoretical basis of PBL has much in common with the way students learn on clinical attachment. Whether considered from the viewpoint of information-processing theory, selfdetermination theory or control theory (reviewed by Albanese2), clinical learning frequently parallels the PBL process. Furthermore, the group dynamics fundamental to cooperative learning in PBL2 are commonly found in clinical attachments. Problembased learning generally consists of a problem trigger, clinical reasoning to address the problem, interactive group discussion, identication of learning needs, self-study and a chance to revisit the problem with the benet of further knowledge.2,3 These steps mirror the clinical process and thus can readily be reiterated in the clinical environment. Furthermore, the three so-called essential characteristics of PBL problems as a stimulus for learning, tutors as facilitators and group work as stimulus for interaction13 are also readily available in this setting. Why then has there not been further development of PBL in the clinical years? Three hindrances merit consideration. Firstly, it has been argued that as students completing a PBL course enter the clinical years as fully edged problem-based learners there is no need for a structured PBL tutorial process in the clinical years.14 Secondly, resistance from clinical teachers and students may inhibit implementation.12,15 Finally, tutors wishing to pursue such an approach nd a dearth of models in the literature on which to base their own practice.1417

METHODS

Development of the CPBL model Context The development of this model occurred within the context of a new 4-year, graduate-entry programme. In the rst 2 years, progressive-release PBL represented the main learning method18 and tutorials were facilitated by non-experts, with curriculum content based on six modules.19 Students then spent almost all of Year 3 on clinical attachments, including two 12week general clinical attachments (GCAs) in wardbased medicine and surgery departments, which represented the primary setting for the development of the CPBL model, and three 5-week blocks in, respectively, psychiatry, obstetrics and gynaecology (O&G) and paediatrics. Following a literature review, panel discussions among teaching staff identied critical elements of the PBL process and generated options for incorporating such elements into the clinical setting. A structure was developed and distributed to relevant stakeholders (clinician-teachers and medical educationalists), and then modied in response to comments. The resulting structure is detailed in Table 1; a quick-reference summary is given in Figure S1,

800

Blackwell Publishing Ltd 2009. MEDICAL EDUCATION 2009; 43: 799807

Clinical problem-based learning

Table 1 Outline of clinical problem-based learning structure for medical and surgical attachments

Session Beforehand

Tutor Allocate case and designate lead student Allocate scribe to annotate discussion (whiteboard ip-chart)

Student group Lead student clerks patient Lead student brings appropriate investigations (bloods, X-rays, electrocardiogram, etc.) to session

Key questions issues

Tutorial 1

Tutor acts primarily as facilitator Presentation Lead student: one-line summary of presentation, then stops Hypothesis generation Group brainstorm to develop hypotheses to explain presentation Develop the focus of Group discussion to relate history acquisition hypotheses to history

What might be going on here? How do these different conditions present? What are the important things we need to know about the patient?

Detailed history

Identify learning issues

Focus on clinical signs

Review hypotheses Focus on mechanisms Finalise learning objectives Review at bedside

Either, lead student plays the role of the patient, other students take history from him or her, or lead student presents history in the third person Students collectively agree on what they need to research before the next tutorial Relate examination to hypotheses and history Lead student presents examination ndings by answering questions raised by the other students Group discussion Group discusses likely pathology and underlying aetiology Group agreement The whole group visits the patient on the ward Review history Demonstrate clinical signs Students use textbook, online and staff resources to answer learning objectives

What do we need to understand to approach this case effectively? What should we look for on examination?

What do examination ndings tell us? What disease states or processes are involved? All objectives are for all students

Learning interval Tutorial 2

Lead student follows case daily, referring to notes and further investigations Patient-focused: what have you learned that specically applies to this case? What management decisions were made and why? Expert discussion of clinical management and reasoning

Tutor switches from facilitator to expert role Summarise case; set Group members review what they agenda have found out about their learning objectives Case progress; explain Review investigations, management, decision making progress Revisit patient at Demonstrate key issues discussed bedside Summarise learning Tutor leads discussion and ensures Identify outstanding issues students have covered all appropriate and any further learning aspects objectives Case report Lead student produces case report Report is distributed to whole group

Write-up

Part of summative assessment

Blackwell Publishing Ltd 2009. MEDICAL EDUCATION 2009; 43: 799807

801

D C Macallan et al
published online as Supporting information. It is derived from the PBL approach used earlier in the course, consisting of paired teaching sessions separated by a self-directed learning interval, but is based on learning through real patients in the clinical setting (Figure 1). In order to ensure that material covered curricular content, rather than arising solely from opportunistic teaching, we generated a list of indicative case types to direct CPBL sessions by blueprinting our learning objectives database (Table S1a). Balance was achieved by maintaining the six-module structure.19 Case denitions were broad enough to allow patient material to be easily identied within the usual in-patient complement, but narrow enough to ensure curriculum coverage. Although initially developed for GCAs, CPBL was subsequently rolled out in psychiatry, O&G and paediatrics, each of which had its own set of indicative case types (Table S1bd). Tutor training consisted of a series of half-day workshops. Format Prior to the CPBL tutorial, a designated (lead) student, chosen on a rotating basis, takes a history from and examines the selected patient. In the rst group session, students are led in a discussion of how history taking and examination contribute to differential diagnosis and problem listing. During this, the tutor acts as facilitator. One student (the scribe) annotates comments from which learning objectives, relating to both clinical skills and mechanisms of disease, are identied. If possible, the case is reviewed at the bedside. In the second tutorial, conducted a few days later, the tutor adopts a more expert role, demonstrating how learnt material translates into investigation, diagnosis and management. If the lead student completes a written case report after the tutorial, it can contribute to his or her summative assessment. Pilot tutorials found that students appreciated clinically grounded learning and felt that the dissection of the history-taking process made them think more about how they asked questions. It rapidly became apparent that case selection was critical and that a poorly chosen case limited the scope of the tutorial. Students appreciated time spent at the bedside. A group size of 10 was too large, and groups of four to six were more successful. Further modications were made in the light of questionnaire evaluation, staff feedback and focus groups to produce the working model proposed. Evaluation of the CPBL model Students participating in the evaluation were drawn from two cohorts of Year 3 students in sequential years of the programme, based at three different hospitals. At all sites, weekly, paired CPBL sessions were delivered by trained clinicians; typically about 1.5 hours was spent on the rst and 1 hour on the second tutorial. Evaluation consisted of questionnaires, focus groups and a Delphi assessment. Questionnaire evaluation Students were asked to grade their agreement with 11 statements related to the structure and delivery of CPBL on a 5-point Likert scale; both positive and negative statements were included to avoid bias. Participants were also asked whether tutorials followed the two-session structure and to estimate the amount of time spent, rstly, at the bedside, secondly, in reviewing learning objectives in the second tutorial, and, thirdly, writing the case report. Finally, students were encouraged to submit free text comments indicating helpful and unhelpful aspects of CPBL and ideas for modication. Focus groups All students in the rst cohort took part in focus groups. Students were invited to discuss positive and negative aspects of CPBL and to give suggestions for its development. Delphi consensus approach In order to provide a robust evaluation of specic components contributing to the success of CPBL tutorials, a consensus method was used, consisting of a two-panel, three-round, adapted Delphi technique. A well-executed Delphi technique provides an effective method of group communication.2023 By using an iterative process with controlled feedback, it enables consensus about a given issue to be measured whilst maintaining inter-participant anonymity.24,25 Both students and teachers were asked: What factors contribute to the success of CPBL? Statements were collated and reviewed by two of the authors; those with identical or very similar wording were combined to a single entry to produce a non-hierarchical list of 83 different statements. This list was then circulated to the students, who rated the importance of each factor using a 5-point scale on which 1 = essential and 5 = undesirable. Once the ratings

802

Blackwell Publishing Ltd 2009. MEDICAL EDUCATION 2009; 43: 799807

Clinical problem-based learning

had been collated and analysed, the list was re-circulated with, for each factor, the group median score and the students own score. Each student was asked to re-rate his or her scores in the light of the median; in instances when students wished to maintain a difference > 1 score point between their own and the median score, they were invited to give specic comments. Students comments on evaluation forms demonstrated that they particularly appreciated the tutor taking on the role of expert in the second tutorial. Negative comments emphasised the need for complementary clinical skills teaching and the importance of continuing bedside teaching. Inevitably, the constraints of a more structured learning week raised the problem of timetable clashes as CBPL operated alongside other ward-based teaching. Focus group feedback In focus groups, the dominant theme to emerge referred to the value attached to contact with expert tutors, especially in the second tutorial, when the tutor moved from a facilitative to a more expert role. Students found this helpful because it contextualised their textbook learning in the clinical environment. The importance of seeing real patients and spending a substantial part of the tutorial time at the bedside was also emphasised. Students raised three areas of concern. These referred, rstly, to the loss of exibility introduced by structured learning; secondly, to potential conicts between structured tutorials and more informal bedside teaching, and, thirdly, to the

RESULTS

Questionnaire evaluation When questionnaire responses were ranked in order of average scores (Table 2), it was noted that most students followed the case list and that learning objectives were clearly dened. Most students found the sessions a positive experience. Importantly, most had no difculty in nding suitable cases. Although the two-session structure was not always followed, it did apply to over 80% of tutorials. On average, over 10% of tutorial time was spent at the bedside and about half of the second tutorial was spent reviewing learning objectives. The case report typically took about 5 hours to complete (range 116 hours).

Table 2 Evaluation of clinical problem-based learning. Scoring for 11 key statements, ranked in order of agreement*

Cohort Session Question 3 6 9 1 11 8 5 10 4 7 2 Response rate, n (%) We followed the indicative case list closely Learning objectives were clearly dened Doing my write-up was helpful I enjoyed the clinical PBL sessions Reading other peoples write-ups was helpful Follow-up of the case was often impossible The tutor rarely followed the proposed structure I always read other peoples case write-ups Usually the case was one of our own patients We usually had different tutors for the two tutorials Finding a case for CPBL was difcult

1 GCA1 34 (97) 4.2 3.5 3.6 3.1 3.6 2.6 2.8 3.1 2.8 2.2 2.6

1 GCA2 20 (57) 4.1 3.4 2.9 3.6 3.1 3.3 2.7 2.5 2.9 3.2 2.0

2 GCA1 32 (91) 4.3 3.4 3.8 3.4 3.3 2.7 2.7 2.6 2.3 2.5 2.3

2 GCA2 11 (31) 4.3 3.9 3.1 3.1 2.3 3.3 2.4 1.6 2.9 2.5 1.7

All students Mean 97 4.2 3.5 3.5 3.3 3.3 2.9 2.7 2.6 2.6 2.5 2.3 0.9 1.0 1.1 1.0 1.1 1.0 1.0 1.3 1.2 1.7 1.0 SD

* Scores represent the level of agreement (1 = strongly disagree, 5 = strongly agree) GCA = general clinical attachment (GCA1 and GCA2 refer to the rst and second GCAs undertaken); SD = standard deviation; PBL= problem-based learning

Blackwell Publishing Ltd 2009. MEDICAL EDUCATION 2009; 43: 799807

803

D C Macallan et al
relationship between CPBL and the overall learning objectives of the course. Conicting views were expressed about the desirability of having the same tutor for both sessions. At one participating hospital the rst tutorial was taken by a junior teaching fellow, who was joined by a consultant for the second session. Although this practice differed from our original model (in which both sessions were led by the same tutor), students felt this alternative worked well; junior fellows helped students identify key learning objectives and focused the tutorial appropriately. Delphi evaluation Fourteen students and six staff members responded with a total of 110 statements in response to the question: What factors contribute to the success of CPBL? After conation of duplicates, 83 statements were included in the Delphi exercise. Statements were classied by topic by two experienced tutors. Most statements referred to the CPBL process and structure (31% and 19%, respectively); a further 15 referred to what we termed the learning environment (Fig. 1). All students were invited to grade the importance of the statements; a response rate of 71% (24 35) was achieved. The importance ascribed after grading and re-grading is shown in Table 3. For the most important attributes, there was very good agreement between rankings; no scores were > 1 point distant from the median for the rst four statements (Table 3). There was somewhat less agreement over the least important factors, although in all but one, more than 90% of replies were within 1 point of the median. Issues upon which there was least consensus (at least three responses > 1 point distant from median) included: covering the individual case quickly to enable discussion of more general issues of disease; second session feedback; discussion in the second session directed by the facilitator reecting the learning objectives; time limit to sessions; discussing difcult questions not in textbooks, and the non-provision of didactic teaching by tutors. Comments made to explain discordant responses tended to be idiosyncratic and did not reveal any consistent themes or consensus.

DISCUSSION

Curriculum 1% Clinical experience 4% Case attribute 9%

Teacher attribute 10%

CPBL process 31%

We believe that the model of CPBL proposed in this paper (Table 1; Fig. S1) represents a useful addition to the clinical teachers tool kit. The majority of our students considered it a positive learning experience. It was achievable within the context of a busy clinical environment and most students were able to identify suitable cases and develop clear learning objectives from the process. Importantly, it allowed students to follow a curriculum structure, rather than using cases opportunistically, whilst still utilising real patients as learning triggers. In our original model, we anticipated that case follow-up would be an important component of the learning week, but, in practice, this remained largely unrealised as a result of rapid patient turnover. In specialties where the average patient stay is very short, such as in maternity units, gynaecological surgical wards and paediatrics, the second tutorial must often be performed without direct involvement of the initial patient; in these settings the CPBL approach was modied so that all patient contact was concentrated within the rst tutorial. When we evaluated students experiences, a common theme to emerge concerned the value ascribed to expert input, especially in the second tutorial. Although some authors describe a model of PBL in which the tutor remains predominantly facilitative,2630 our students greatly appreciated it when expert clinicians demonstrated how clinical reasoning applied to the case. Such modelling helped students identify and prioritise key issues and reach realistic, contextualised management solutions to real clinical problems, particularly when such

Learning environment 15% CPBL structure 19% CPBL content 11%

Figure 1 Distribution of factors contributing to the success of clinical problem-based learning (CPBL) according to the classication of statements generated in the rst round of the Delphi exercise (after the conation of duplicates). Values represent the percentages of the 83 statements that referred to each topic area

804

Blackwell Publishing Ltd 2009. MEDICAL EDUCATION 2009; 43: 799807

Clinical problem-based learning

Table 3 Most and least important attributes for success of clinical problem-based learning

Rank Most important attributes Being able to be wrong and be corrected Being able to ask simple questions The chance to ask any questions of an expert in the eld regarding management and other things related to that case Tutors willing to teach and listen Clinicians with relevant knowledge leading the PBL Expert for second session Clinically relevant material discussed Going through systematic approaches to differentials Taking a full history, examining and following up your allocated patient Tutor responds to student needs Two tutorials per case Two sessions: introduction and expert session Allowing times for questions Enthusiastic facilitators A facilitator who knows the process and the topic very well An expert forum second session Interactive teaching during second session Supportive environment A topic with sufcient meat and relevance to be discussed by the group Recap of basics where necessary Least important attributes Seeing others strengths and weaknesses Case write-up Having groups with mixtures of backgrounds Second PBL feedback Freedom to choose a case not in the suggested list 79 80 81 82 83 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 1 2 3

Median

Deviation

Mean

1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1.3 1.3 1.3 1.3 1.5 1.5 1.5 1.5 1.6 1.6 1.6 1.7 1.7 1.7 1.7 1.7 1.7 1.7 1.8 1.8

3 3 3 4 4

2 2 2 4 0

3.0 3.0 3.3 3.5 3.7

The table shows the 20 most important (lowest scores) and ve least important (highest scores) statements from the total list of 83. Scoring: 1 = essential; 2 = very important; 3 = neither important nor unimportant; 4 = unimportant; 5 = undesirable. The deviation indicates the number of responses > 1 point distant from the median PBL = problem-based learning

problems did not tally with textbook denitions. When senior clinician teaching time is limited, using juniors for the rst tutorial allows senior staff to concentrate on the second, expert session. Delphi analysis reiterated the value of clinical experts, but also highlighted the relational aspect of the tutorial. The value ascribed to an environment where they could ask simple questions and make mistakes tallies with the assessment by Dornan et al. of the importance of the social dimension of professional

learning.12 Clinical PBL creates a non-judgemental environment in the rst tutorial by being predominantly facilitative; participants are not expected to demonstrate an extensive prior knowledge base in the subject area being discussed. This represents an advantage of CPBL over single-session, opportunistic bedside teaching, in which students are often unprepared and teachers easily overestimate students prior knowledge, with resulting mutual discouragement. In CPBL, by contrast, students are initially expected only to dene problems clearly and identify the cognitive

Blackwell Publishing Ltd 2009. MEDICAL EDUCATION 2009; 43: 799807

805

D C Macallan et al
dissonance between what they know and what they need to know. In the second tutorial, students are prepared and can utilise the expertise of senior clinicians much more effectively and condently. In terms of important components, our ndings suggest that, as for PBL,12,14,31 the quality of individual tutors remains an important factor. Relatively little comment was made about the structure of the tutorials, except that timetabling was sometimes problematic. We cannot comment on whether structures were of secondary importance or were considered at least satisfactory and thus taken for granted. Case report writing received mixed responses. Overall it did not feature strongly as a positive contributor to learning, although reading other students case reports was rated as helpful by many, but it did constitute a formally assessable piece of work. Several limitations in this evaluation warrant consideration. Firstly, our sample size was small, although it represented more than two-thirds of the student year group. Care must be taken in generalising to other courses or schools where PBL is not as deeply embedded in the curriculum or where different styles of PBL are used. In addition, the enthusiasm among students involved in a new course may have biased results favourably. There may be other aspects of the learning environment, such as the availability of resources (library facilities, computers, junior medical staff, etc.) that are very important but did not emerge in our evaluation because they were taken as givens. We tried to avoid bias in reporting a large volume of qualitative feedback by using the Delphi approach, which has been well validated as a tool to provide consensus views and has been used extensively in health sciences research.24,25 This assessment has considered CPBL from the students viewpoint. Of course, students may not correctly perceive those factors that truly enhance their learning. This is a common limitation seen in analyses of PBL approaches, but more objective measures are difcult to apply in this context.2,32,33 However, nal results from the rst years of the course were very favourable and we propose to continue to develop and implement this model as the curriculum is developed further. Specically, we intend to assess whether it will be as successful with school-leaver entrants as with graduates, whether it might be applied elsewhere in the programme, and whether simpler constructs may be as effective. In addition, we intend to investigate optimal training methods and support for clinical teachers participating in CPBL. In conclusion, we have shown that it is possible to introduce the principles of PBL into a busy teaching hospital environment by using a tool such as CPBL. It allows structured learning in a clinical environment and complements, but does not replace, other forms of clinical teaching. As well as being popular with students, this guided process of learning allows students to approach real patient problems systematically with condence, despite their limited background knowledge. Generating appropriate learning objectives and providing opportunities for selfdirected learning before expert tutorials combines the best of facilitative tutoring with opportunities to think through a problem with a clinical expert and role-model.

Contributors: DCM, AK, LF and PM formulated the approach

and developed its application with assistance from SCH. DCM and AK performed the analysis and wrote the rst draft of the manuscript. SCH collected and analysed the data from questionnaires and focus groups. All authors contributed to the revision of the manuscript and approved the nal version for publication. Acknowledgements: we would like to acknowledge the contributions to the development of the clinical problembased learning model of Emma Baker and Penny Neild in general internal medicine, Jonathan Round in paediatrics, and Kevin Hayes in obstetrics and gynaecology. Funding: DCM received funding from the UK Medical Research Council during the course of this work. Conicts of interest: none. Ethical approval: not required.

REFERENCES 1 Jolly B. Problem-based learning. Med Educ 2006;40 (6):4945. 2 Albanese MA. Problem-based Learning. Edinburgh: Association for the Study of Medical Education 2007. 3 Barrows HS. How to Design a Problem-based Curriculum for the Preclinical Years. New York, NY: Springer 1985. 4 de Leng BA, Dolmans DHJM, Muijtjens AMM, van der Vleuten CPM. Student perceptions of a virtual learning environment for a problem-based learning undergraduate medical curriculum. Med Educ 2006;40 (6):56875. 5 Mamede S, Schmidt HG, Norman GR. Innovations in problem-based learning: what can we learn from recent studies? Adv Health Sci Educ Theory Pract 2006;11 (4):40322. 6 ONeill P, Duplock A, Willis S. Using clinical experience in discussion within problem-based learning groups. Adv Health Sci Educ Theory Pract 2006;11 (4):34963.

806

Blackwell Publishing Ltd 2009. MEDICAL EDUCATION 2009; 43: 799807

Clinical problem-based learning

7 ONeill PA, Morris J, Baxter CM. Evaluation of an integrated curriculum using problem-based learning in a clinical environment: the Manchester experience. Med Educ 2000;34 (3):22230. 8 Nash PP, Schwartz RW, Middleton JL, Witte FM, Young B. A student-centred, problem-based surgery clerkship. Acad Med 1991;66 (7):4157. 9 Harden R, Crosby J, Davis MH, Howie PW, Struthers AD. Task-based learning: the answer to integration and problem-based learning in the clinical years. Med Educ 2000;34 (5):3917. 10 Speer AJ, Stagnaro-Green A, Elnicki DM. Interdisciplinary clerkships: educational models of the future? Am J Med 1995;99 (5):4513. 11 Dammers J, Spencer J, Thomas M. Using real patients in problem-based learning: students comments on the value of using real, as opposed to paper, cases in a problem-based learning module in general practice. Med Educ 2001;35 (1):2734. 12 Dornan T, Scherpbier A, King N, Boshuizen H. Clinical teachers and problem-based learning: a phenomenological study. Med Educ 2005;39 (2):16370. 13 Dolmans DHJM, De Grave W, Wolfhagen IHAP, van der Vleuten CPM. Problem-based learning: future challenges for educational practice and research. Med Educ 2005;39 (7):73241. 14 Taylor D, Miflin B. Problem-based learning: where are we now? Med Teach 2008;30 (8):74263. 15 Farmer E. Not more PBL! In: Schwartz P, Mennin S, Webb G, eds. PBL Case Studies. London: Kogan Page 2001;7882. 16 Barrington D, Wing L, Alpers J, Latimer K, Prideaux D. Evaluation of a change from traditional case studies to patient-based, problem-based learning: a case study. Med Teach 1997;19:1047. 17 Finucane P, Crotty M, Henschke P. Clinical problem solving (CPS) as a teaching strategy: a dual teacher approach. Med Teach 2001;23 (6):5725. 18 Finucane P, Nicholas T, Prideaux D. The new medical curriculum at Flinders University, South Australia: from concept to reality. Med Teach 2001;23 (1):769. 19 McCrorie P. Tales from Tooting: reflections on the first year of the MBBS graduate entry programme at St Georges Hospital Medical School. Med Educ 2001;35 (12):11449. 20 Couper MR. The Delphi technique: characteristics and sequence model. ANS Adv Nurs Sci 1984;7 (1):72 7. 21 Delbecq AL, van de Wen A, Gustafson D. The Delphi technique. In: Delbecq AL, ed. Group Processes for Program Planning. Glenview, IL: Scott Foresman Publishing 1975;83107. 22 Jones J, Hunter D. Consensus methods for medical and health services research. BMJ 1995;311 (7001):376 80. 23 Linstone HA, Turoff M. The Delphi Method: Techniques and Application. Reading, MA: Addison-Wesley 1975. 24 Fiander M, Burns T. Essential components of schizophrenia care: a Delphi approach. Acta Psychiatr Scand 1998;98 (5):4005. 25 Fiander M, Burns T. A Delphi approach to describing service models of community mental health practice. Psychiatr Serv 2000;51 (5):6568. 26 Bochner D, Badovinac RL, Howell TH, Karimbux NY. Tutoring in a problem-based curriculum: expert versus non-expert. J Dent Educ 2002;66 (11):124651. 27 Camp MG, Anderson AS. Expert tutors versus non-expert tutors in PBL. Acad Med 1993;68 (5):3535. 28 Eagle CJ, Harasym PH, Mandin H. Effects of tutors with case expertise on problem-based learning issues. Acad Med 1992;67 (7):4659. 29 Gilkison A. Techniques used by expert and non-expert tutors to facilitate problem-based learning tutorials in an undergraduate medical curriculum. Med Educ 2003;37 (1):614. 30 Hay PJ, Katsikitis M. The expert in problem-based and case-based learning: necessary or not? Med Educ 2001;35 (1):226. 31 Dolmans DHJM, Gijselaers WH, Moust JHC, De Grave WS, Wolfhagen IHAP, van der Vleuten CPM. Trends in research on the tutor in problem-based learning: conclusions and implications for educational practice and research. Med Teach 2002;24 (2):17380. 32 Colliver JA, Markwell SJ. Research on problem-based learning: the need for critical analysis of methods and findings. Med Educ 2007;41 (6):5335. 33 Hamdy H. The fuzzy world of problem-based learning. Med Teach 2008;30 (8):73941.

SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Table S1. Examples of indicative cases for clinical problembased learning sessions. Figure S1. Schematic of CPBL Structure. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than for missing material) should be directed to the corresponding author for the article. Received 20 October 2008; editorial comments to authors 16 January 2009; accepted for publication 7 April 2009

Blackwell Publishing Ltd 2009. MEDICAL EDUCATION 2009; 43: 799807

807