AUC Academic Conference 'From Virtual to Reality' The University of
Queensland 1996AUC Academic Conference 'From Virtual to Reality' The University of
Queensland 1996
Back to Home Page ...... by Presenter ...... by Paper Title ...... by Keyword
Paper Title:
WANDA HOLMES:
DEVELOPING DIAGNOSTIC REASONING SKILLS THROUGH A COMPUTER-BASED CASE SERIES
Presenter:
Dr Nicholas J.C. King, The University of Sydney
Authors:
Dr Nicholas J.C. King, Deborah A. Bryce, Celia F. Graebner,
The University of Sydney.Dr Maria Evans, The University of Missouri
Assoc. Prof. J. Hurley Myers, The Southern Illinios University
Keywords: Clinical Reasoning, Pathology
Faculty area: Medicine
Abstract
The interaction of skills and knowledge involved in evidence-based professional practice makes learning in this area a complex process, requiring critical reflection not readily captured in individual tutorials or case studies. This abstract presents a prototype case from a working environment for medical students which addresses these issues by providing tools for integrating the results of simulated tests and examinations into a developing diagnostic hypothesis.
Salient features of the environment include an audit trail of enquires, decisions and current hypotheses for each individual student to be used in self-assessment or group debriefing; interface features constructed around relevant taxonomies of activities and procedures for each phase of clinical investigation; on-line reference resources providing the user with consistent contextualisation of the current case material, for example, sets of normal result ranges for lab tests. The common base of reference materials and common investigative tools offer economies of scale in the development of a case series, and economies of effort for case designers and students.
INTRODUCTION
This paper will describe the history behind the development of a prototype, salient features of the program that support learning in medical school (both for student and faculty), and potential applications of this program to address key problem areas in education.
Medical students face the problem of factual overload (1), difficulty in integrating basic and clinical science (2) and are often ill-prepared for the clinical reasoning process (2-4). Although students in a traditional course are educated in basic sciences they are often unable to access that knowledge in a clinical setting (3). Research shows that learning in a relevant context increases the accessibility and practical application of that knowledge (3,5,6). Research also indicates that presenting a problem first and allowing students to struggle and resolve this problem is more effective for learning than presenting relevant problems after teaching has occurred (6). For problem-first learning to be most effective however, useful feedback on the problem must be provided (6). The Wanda Holmes case addresses these key issues in medical education by presenting a relevant context for learning, providing means of practising the clinical reasoning process and providing feedback on that process. (references)
Evolution of the prototype
The Wanda Holmes prototype is constructed of two independent, but related, shells: a diagnostic reasoning (DxR) shell and a Pathology shell. The DxR shell, developed at Southern Illinios University, was modelled on paper-based modules used by students in the PBL-stream. The paper-based modules allowed students to work through a patient case, but because most learning occurred in small groups, it was difficult to determine just how much basic knowledge and clinical reasoning skills an individual student had in a particular area. An evaluation tool was needed that would allow students to work through the clinical reasoning problem solving process individually. In response to this need, the hypercard-based computerised DxR shell was developed. Its advantages lay not just in the provision of a wider range of diagnostic tools than the PBLM's, but in the provision of a tool to analyse the problem solving process of individuals, and provide faculty the facility to monitor the performance, misconceptions and gaps in knowledge of both individuals and groups.
A student can diagnose and treat a patient using the DxR program, however, there is no option to pursue, and confirm, the diagnosed pathophysiology. Incorporating a pathology lab was seen as a way of providing a clear context for learning pathology. Basic knowledge and skills in pathology, could be integrated in the setting of a patient case that was familiar, and had personal input from the student. The pathology lab would allow students to apply their pathological knowledge, test the validity of their diagnosis and complete the diagnostic cycle.
It is important to note that although history, examinations, tests and analyses of pathological specimens are simulated in this program, to a varying degree of realism, the value of the program lies not in simulations, but the design aspects which provide structured support for the clinical reasoning process. The environment gives the learner flexibility, a relevant context and a safe environment to practise hypothesising and integrating basic and clinical science.
A WALK THROUGH THE WANDA HOLMES CASE
What do the students do when they work through the Wanda Holmes case? Essentially the program is a simulated patient case that allows students to ask questions (from a question bank), to perform a physical examination using clinician's tools, and to order tests to track down the nature of the patient's problem, then confirm the pathology.
Patient Presentation and Interview
The student is presented with a photo, the patient's age, and a brief description of the presenting complaint, usually two or three lines. The student has a choice to progress to either interview [Figure 1], physical examination or laboratory tests. Students select their questions from an extensive question bank, which is organised into categories covering presenting illness, medical history and individual body systems. At transitional points in the investigation, the student is requested to formulate the patient's problem and confirm or restate their provisional diagnoses.
Physical Examination
Examination tools include looking, feeling, moving, ophthalmoscope, thermometer, stethoscope and stop watch (eg. pulse rate). Four views of the of the patient's body can be selected. For example, a chest view would be selected to listen to the patients heart; the stethoscope tool must be placed over the appropriate parts to access heart sounds.
Students are required to interpret findings, such as heart and respiratory sounds and fundus of the eye images. A "consult" tool, will allow comparison with an expert's interpretation, but is available only after the student has committed to his or her interpretation. If a student is not ready to interpret, s/he may continue the investigation, however all findings must be interpreted before the end of the case. Alternatively, the student can exit the program and investigate, for example the fundus of the eye, and return when more confident.
To support learning, faculty can provide access to a "library" tool. This incorporates information pre-determined by faculty as important to learning, and could include, a learning document, hypercard stack, short video or access to medline. In this situation students do not have to quit the case to access resources
Laboratory Tests
A broad range of lab tests are available, the more commonly ordered tests being separate from the comprehensive list. Prior to ordering, students may access information about the test and the circumstances in which it is used, as well as cost. Normal values for results can be viewed and compared to the patient's test results. In some situations, the student will be asked to justify the ordering of a test. The computer does not interpret this written input but simply records it for faculty assessment.
Figure 1. Presenting Illness Screen of Patient Interview
Differential Diagnosis
As their investigation unfolds, students are able to add to, delete, or rearrange their hypotheses. When ready, and before moving on to treatment, students must commit to a final diagnosis, usually the hypothesis they have moved to the top of their differential diagnosis list. Not only must students commit to a diagnosis, for example, aortic valvular stenosis, they must provide a brief but concise pathophysiological description of the diagnosis. The program will recognise the differential diagnosis (recorded in its evaluation paradigm), however faculty interprets students' description of the diagnosis (usually a few lines of text).
Treatment
This section is complex and may be beyond the scope of students at an early stage of their course. Students may opt not to treat the patient, but are encouraged to make an attempt at this section as a learning experience. Examples of available choices include monitoring the patient, administration of drugs, and referral to a specialist.
Pathology Laboratory
Wanda Holmes is a case of appendicitis. The appendix is surgically removed (if students fail to reach the appropriate diagnosis and treatment, the patient is rushed to emergency). The appendix is sent to Pathology for analysis. Students use virtual tools to examine the whole appendix first (weigh, measure, feel, observe) and may order laboratory tests (Figure 2). They may also cut the specimen (once cut it cannot be 'uncut') and send it away for a range of analyses, including bacteria, virus, drugs and DNA testing. Histological slides may be made from any or all of the cut parts of the appendix and those selected may be viewed in the histopathology window. Each slide can be seen from three fields of view, low, medium and high power, and is treated like a virtual microscope slide, moved around the stage, with each view magnifiable one step further. More slides can be made of the remaining cut tissue, if desired. Three predetermined slides are preset to elicit descriptions from the student prior to exiting the Pathology Lab. These are triggered only when the linked slide has been made and the views associated with them have been seen by the student. The description(s) may be postponed until just prior to exiting Pathology. After the slides taken have been viewed, the students may collect the results of other tests run on the specimen from the laboratory and exit Pathology. The students are asked if the pathological evidence supports their final diagnosis. At this point they are able to change their final diagnosis. If the students have not made slides from all the cut parts of the appendix that are triggered to require description, the remaining pictures requiring description will appear at this time. The students are then asked if the new information supports their final diagnosis and again they are able to change their final diagnosis. They then enter the evaluative feedback section of the program.
Feedback
On completion of the investigation, a student's record is passed to a student activity record (SAR), where it is processed according to a preset evaluation paradigm. The feedback categories are: diagnosis; justification of diagnosis; consideration of alternative hypotheses; thoroughness of the work up; efficiency; and cost (of lab tests). Faculty sets the relevant criteria for each of these categories and provides students with feedback on the questions /examinations /tests an expert considered important for each category and why.
Feedback on the histological slides from the pathology lab is also given to students in the form of a comparison with an expert's description. Students rate the level of agreement between the expert and their own description. The self evaluation feedback concludes the Wanda Holmes case.
Figure 2. Pathology Laboratory Screen
PROGRAM FEATURES THAT SUPPORT STUDENT AND FACULTY
As described earlier, the program arose out of a need to assess clinical reasoning of individual students and to place pathology in a more realistic setting. The program needed to support learning by providing an interface whereby students could practise integrating basic and clinical science in a safe, non-threatening (no harm to a patient) environment. In addition it needed to provide information to faculty on how individuals and the student body were performing.
Features Supporting Student Learning
The first feature that supports student learning is the presentation of a relevant , virtual context, in which students can practise applying and integrating their knowledge. Students begin with a patient on whom they have minimal information, as in a clinical setting (4). (references) They then place themselves in the role of an examining clinician and selectively seek out information pertinent to the case. Students are not given key features of the case, they must determine which information is relevant to request, and prise out the key features for themselves. This requires the generation of working hypotheses, and an understanding of tests and physical examinations, their purpose and application. The tools and tests represent the range and variety that would be available to a practising clinician.
A second group of features supporting learning are the aspects of the program that require students to be accountable for their actions and thought processes. Students must identify and state the patient's problem, generate hypotheses, and if required by faculty, identify which hypothesis is being followed when requesting a particular test/exam. They must justify tests ordered and interpret findings, for example heart sounds, x-rays, ECG's. For a comprehensive examination, and to reach the highest level of the evaluation criteria, they must justify their diagnosis, consider competing hypotheses and perform a thorough workup, and do so within the efficiency limits set by faculty.
A third feature of the program supporting learning is the degree of student control. Students can proceed at their own pace, exit and re-enter, and proceed forward or backwards between interview, exam and tests.
Fourth, a notes section acts as an external memory for students, leaving them cognitively free to interact with the content of the information. Results of history/tests/exams can be copied and pasted to notes, edited, and accessed if students do not recall a particular response. Rather than using short term memory to remember data, students are able to access information from the notes and use their short term memory for the more important task of reasoning, hypothesising, questioning, sorting and decision making (7). (references)
The fifth feature of the program that is important for supporting learning is the evaluation of performance. Delayed feedback is appropriate when higher level learning is to be encouraged (8). This is found in the evaluative feedback at the end of the program. Students compare their performance with how an expert would justify, consider competing hypotheses, and perform a thorough workup. They can compare their description of histological slides to that of an expert. Immediate feedback is also given, in the form of a consultant's report, following student interpretation. Through the feedback process students have a basis upon which to identify gaps in understanding, and determine modes of action to correct these gaps (9). (references)
Features Supporting Faculty Learning
A starting point for teachers and faculty in improving learning is discovering what the students know - their understanding and misconceptions (9). (references). Faculty needs to know the outcome of learning for both individuals and the student body. The feedback from individual assessment tells tutors how a student is understanding and interpreting a particular patient case. Group data tells faculty where the successful, and problem, areas are with learning for that particular case. For example, the majority of the students may be overlooking a critical part of the investigation. Feedback on the group thus enables faculty to identify and address problem areas of the curriculum.
The program provides access to individual records; full audit trails of inquiry are recorded as a student activity record (SAR). This is available in summary form and includes, inter alia, the diagnosis, pathophysiological description of the problem, and access to notes, learning issues, missing items (items not requested, but considered essential by faculty) and more. Faculty may view the summary only, or probe deeper to learn about the student's reasoning, for example, the hypotheses considered when ordering a particular test.
Alternatively faculty may wish to access a group performance record. Here, records are grouped according to an evaluation paradigm. Faculty can select a particular category (eg. impressive, acceptable, shotgun), identify students within that category, their diagnoses, time to completion, efficiency levels and missing items. This latter "missing items" record may be used by faculty to target areas that are not being addressed adequately in the teaching & learning process.
If, for example, in a problem based learning curriculum, faculty wants to learn how a student is thinking about a case, prior to and after a study break, preferences can be set that request a student to record a) learning issues (prior to study session), b) external resources accessed and c) hypotheses, on resumption of the case. Learning issues refers to the areas of study a student has decided he/she needs to investigate to understand, and start to resolve, the patient problem.
The tools available to faculty are powerful, by virtue of the degree of control and flexibility that can be brought to a case. Faculty can assemble cases, modify them at any stage (thus obviating the "not invented here syndrome" (10). Evaluation parameters can be set, and if required over-ridden, allowing a selected subgroup to be re-evaluated. Within a case, faculty can alter the degree of stringency of evaluation parameters, thus making a single case adaptable to either an early or more advanced stage of a course. Thus Wanda Holmes could be seen both in the first and last years of a clinical course, with the stringency levels and efficiency limits altered accordingly. (references)
PROJECTING INTO THE FUTURE:
EXAMPLES OF POTENTIAL USESCurrently the University of Sydney is undergoing a change from a traditional to a problem based curriculum. Methods are being investigated to assess students' progress in a PBL curriculum, to monitor learning and, in response to outcomes, investigate strategies to support learning.
Both the DxR and pathology shells are based on a normal shell, into which abnormal findings are entered. This affords economies of scale for case design, making case construction a relatively simple and quick process. Cases may be designed by a collaborative team, ideally producing a product which is more broadly disseminable, or alternatively, an existing case may be altered using case management tools, to suit the needs of a particular institution.
Cases can be available to students to use in their own time, at any potential location on a network. If student records (SAR's) are stored and retrieved from an single site on a network, students can work at their own pace, leave the program to investigate learning issues, and re-enter and complete the program at a distant, but linked, site enabling great flexibility of learning.
Potential applications of the DxR-pathology shell.
A number of possibilities are entertained as to the DxR-Pathology lab's potential uses:
1) In an undergraduate program learning may be stimulated by presenting a patient problem first, prior to a tutorial. Students therefore go to a tutorial with clearly identified problem areas. The tutor also goes to the tutorial informed, having viewed student records and identified areas of misunderstandings or omission. The tutorial can then be effectively used to clarify areas of misunderstanding. Three important principles of active learning are incorporated; that of starting with a problem to stimulate learning (6) ; providing feedback to enhance learning (6,8) ; and identifying what students know and using it to inform teaching and address student misconceptions ( 2,3,9.) (references)
2) Computerised cases could replace paper-based cases in a PBL curriculum. Students could problem solve in groups (effective for the elaboration of knowledge (5)). Subsequent tutorials could be used to clarify problem areas. Faculty in this circumstance would have a record of the group's progress, their concept of the problem, differential diagnoses, and resources utilised to resolve the problem. Records could be used for a number of purposes, for example, as a basis for action research into teaching or to compare student outcomes in longitudinal studies.
3) Students could work through a problem that is a variation on a completed 'problem of the week' in a PBL curriculum, thus allowing scope for elaboration of that problem area. An example would be an unusual presentation of the condition seen in the problem of the week.
4) A patient could be revisited throughout the curriculum to introduce complexity into patient diagnosis and management. For example, Wanda Holmes (1) could present early in the curriculum with a relatively simple problem. In year 2, Wanda Holmes (2), may present with an additional problem. Later still, Wanda Holmes (3) may present with a more complex combination of conditions, (eg. related to age and/or previous medical conditions). This graduated approach could provide one avenue of addressing the problem of dealing with complexity in medical teaching.
5) Cases could be designed for a particular speciality of medicine or surgery, for example, the most commonly encountered or most serious, life-threatening conditions. This may help counteract a common problem for students of inequity of clinical exposure during the short time spent in clinical placements (12).(references)
6) Satisfactory completion of case problems by graduates could contribute to continuing medical education accreditation.
In conclusion, many possibilities arise out of this dual shell for offering economies of scale in the development of a case series and optimising the effectiveness of teaching and learning. It provides opportunities for students to practise integrating their knowledge in a relevant, supportive and safe environment for learning. It also informs teachers of what students know. Having access to how students are thinking about a case provides opportunities to address key difficulties in learning and thus provide a more effective teaching and learning environment for both student and teacher.
Back to Home Page ...... by Presenter ...... by Paper Title ...... by Keyword
1. C. R. Coles, "Helping students with learning difficulties in medical and health care education", Medical Education, vol. 24, pp 300-312, 1990.
2. J. I. Balla, "Changing concepts in clinical education: the case for a theory", In: J. I. Balla, M. Gibson and A. M. Chang (eds) Learning in medical school: a model for the clinical professions, Hong Kong University Press, Hong Kong, pp 3-17, 1989.
3. J. Grant, "Clinical decision making: rational principles, clinical intuition or clinical thinking?", In: J. I. Balla, M. Gibson and A. M. Chang (eds) Learning in medical school: a model for the clinical professions, Hong Kong University Press, Hong Kong, pp 81-100, 1989.
4. H. S. Barrows and P. J. Feltovich, "The clinical reasoning process", Medical Education, vol 21, pp 86-91, 1987.
5. H. G, Schmidt, "Foundations of problem-based learning: some explanatory notes", Medical Education, vol. 27, pp 422-432, 1993.
6. G. R. Norman and H. G. Schmidt, "The psychological basis of problem-based learning: a review of the evidence", Academic Medicine, vol. 67, no. 9, pp 53-61, 1992.
7. R. B. Kozma, "The implications of cognitive psychology for computer-based larning tools", Educational Technology, November, pp 20-25, 1987.
8. D. H. Jonassen and W. H. Hannum, "Research principles for designing computer software', Educational Technology, December, pp 7-14, 1987.
9. D. Laurillard, "Rethinking university teaching: a framework for the effective use of educational technology", Routledge, London, 1993.
10. Costs and Quality in Resource Based Learning On and Off Campus. National Board of Employment, Education and Training Australian Government Publishing Service, Canberra, October 1994.
11. C. R. Coles, "The role of context in elaborating learning", In: J. I. Balla, M. Gibson and A. M. Chang (eds) Learning in medical school: a model for the clinical professions, Hong Kong University Press, Hong Kong, pp 41-56, 1989.
12. A. W. Sajid, G. A. Ewy, J. M. Felner, I. Gessner, M. S. Gordon, J. W. Mayer, C. Shub and R. A. Waugh, "Cardiology patient simulator and computer-assisted instruction technologies in bedside teaching", Medical Education, 24: 512-517, 1990.
Dr Nicholas J.C. King*, Deborah A. Bryce*,
Celia F. GraebnerÝ,
Dr Maria Evans§
and Assoc. Prof. J. Hurley Myers¶*The Department of Pathology and ÝCentre for Teaching and Learning, The University of Sydney, N.S.W. 2006 Australia, §Department of Pathology, University of Missouri, Columbia, MO 65202, USA, ¶Department of Physiology, School of Medicine, Southern Illinios University at Carbondale, Carbondale, IL 62901, USA
Dr Nicholas J.C. King,
The Department of Pathology,
The University of Sydney,
N.S.W. 2006 AustraliaPhone: 61 2 351 4553
FAX: 61 2 351 3429email:
nickk@med.su.oz.au
debryc@anatomy.su.oz.au
Back to Home Page ...... by Presenter ...... by Paper Title ...... by Keyword
