JOURNAL OF MEDICAL EDUCATION TECHNOLOGIES

TIME Project Interactive Patient Simulations:

Experiential Learning in the Medical School Classroom

by William G. Harless, Ph.D., Marcia A. Zier, 
Jean E. Smith, Ross Dube, Robert C. Duncan, Ph.D., and William R. Ayers, M.D.

ABSTRACT

The Technological Innovations in Medical Education (TIME) Project has developed, implemented, and evaluated an interactive videodisc patient-simulation model that is voice activated and designed for use by faculty in the classroom. The interactive patients provide faculty with a new method for problem-oriented teaching and performance-based evaluation. They give students the opportunity to experience clinical decision making and problem solving in the classroom, with the added advantage of being able to discuss the patient’s situation with their professor and peers. Seven interactive patient simulations are now being used by faculty at Georgetown University School of Medicine and three other medical schools to teach students at all levels of training.

Background

The trend in contemporary medical education is toward a more practical form of teaching that revolves around the care of patients. This trend was fueled in 1984 by the publication of the Report of the Panel on the General Professional Education of the Physician and College Preparation for Medicine entitled "Physicians for the Twenty-First Century: The GPEP Report" (Association of American Medical Colleges).

The GPEP Report was comprehensive in its review of the existing problems in medical education and provided a conceptual framework for the changes required to prepare medical students to practice medicine in the future. Among its recommendations was replacing a major portion (up to 50 percent) of lecture time with "active learning formats," increasing the clinical relevancy of basic science information, and encouraging faculty to develop new approaches to teaching to enhance their involvement in guiding students’ learning.

Coincidentally, in 1984, the TIME Project, under the auspices of the National Library of Medicine, began to develop a computerized, videodisc-based patient simulation model for faculty use in the medical school classroom (Harless et al. 1986). A voice recognition capability was integrated into the system, allowing the simulated patient to respond to natural, spoken English. The professor could then converse with that patient and create the illusion of a real patient being presented to the class.

The feasibility of faculty using only their voices to control this high-tech simulation was demonstrated at five medical schools; in 1987, a rigorous field test was conducted at three of these schools (Harless et al. 1990). The model has been successfully integrated into the introduction to medicine curriculum at Georgetown University School of Medicine and is being used as a major portion of the final examination for the second-year class. Medical schools at the University of Illinois-Chicago, University of Wisconsin, and University of Oklahoma are actively using the simulations at various levels of medical student training.

The TIME simulated-patient model has proved to be a solution to many of the educational problems identified in the GPEP Report. Specifically, the model does provide faculty with the opportunity to transform the lecture hall into an interactive, contextual format where active, experiential learning can occur. And, because the TIME simulation is free from pedagogy, faculty can define their own objectives and design their own teaching strategies. They can then use any TIME simulated patient to offer clinically relevant instruction at all levels of medical school training just as, in theory, a real patient could be used.

Establishing an "Active Learning Format"

Almost without exception, when medical students are intellectually involved in a learning situation, they will master the material and understand what they have learned. If they are also emotionally involved in the situation, they will personalize the experience, remember what they have learned, and be inclined to apply it to similar real-life situations. If students are both intellectually and emotionally involved in a learning situation and, in addition, they have the opportunity to influence that situation, they will feel responsible for the decisions made during the session. This ultimate sense of responsibility is what creates an "active learning format" and is the basis for the powerful educational method known as "experiential learning" – that is, learning through the participation in, and observation and analysis of, an evolving situation. The TIME simulation model is an experiential learning method.

Using TIME Interactive Patients in the Classroom

Unlike other educational simulations that are intended for the individualized instruction of students in a learning resource center (Henderson 1989; Allan R Lane 1991), the TIME interactive patients are designed for faculty use in a large or small classroom. Like all other educational simulations, the effectiveness of the TIME simulation is dependent on the students becoming involved in the patient’s situation. Therefore, the professor’s challenge, and opportunity, in every classroom presentation is to make the students feel that they are dealing with an actual patient.

The use of voice recognition technology and natural spoken English as the input modality makes it possible for the professor to simulate a clinical encounter with an actual patient in a way that is not possible with other input modalities. It is unlikely that an acceptable level of reality could be achieved in the classroom setting if the professor were forced to communicate with the patient by using a mouse or touch-screen capability to select an item from a visible menu. On the other hand, the process of relevant, uncued questions being spoken by the professor and responded to by the patient creates a convincing illusion of an active conversation that captivates the students.

In order to use a simulation, the professor voice-trains the computer to recognize a comprehensive vocabulary of more than three hundred medical history concepts and technical terms that are relevant to the problems presented in the case. This one-hour session is followed by several "rehearsals" with the simulated patient, when the professor becomes familiar with the content of the simulation and practices interacting with the patient until the conversation between them appears spontaneous and natural. The rehearsals also prepare the professor to control the technical aspects and the pace of the simulation.

In the typical classroom presentation, the professor, in concert with the students, conducts a conversational interview with the simulated patient, who appears on a video screen in front of the class. (The size of the screen may vary to accommodate the size of the class). Diagnostic information is obtained from the physical examination and laboratory tests, and clinical decisions are collectively made concerning the patient’s management.

During the presentation, the professor has access to an intelligent terminal capable of dynamically displaying the various subsets of concepts that the system can understand at any specific point. Each concept is randomly accessible when the user gives the specific command. The fifty or so conversational items that appear on the prompt screen at any one time may be spoken in any order, giving the professor the capability of creating the illusion of a natural conversation with the patient. The intelligent prompt screen is not visible to the audience. In fact, there are no computer menus or cues seen by the students that might influence their decisions.

The professor is the guide in this new approach to teaching. The environment is carefully designed so that the professor and students, jointly in the presence of the patient, can discuss relevant bio-physiological constructs, make clinical decisions about the patient’s condition, explore psychosocial aspects of the case, and raise bioethical issues, paradoxically without being constrained by the presence of the patient.

Teaching Strategies

Professors are free to create their own teaching strategies for a presentation. However, they are encouraged to maximize the students’ involvement in the learning experience and to take advantage of the system’s capability to incorporate the students’ thoughts and ideas into the diagnostic and management decisions for the patient.

In one session with Frank Hall (Harless 1986), a gastroenterology patient with alcoholic tendencies, the professor stopped the simulation after the opening scenes depicting Mr. Hall’s arrival at the hospital and a brief exchange with an emergency room nurse had played. Turning to the class, the professor began his presentation:

Physicians are like detectives. We are always looking for clues that will help us know something about the patient. It may be in the way he looks or walks or dresses, or even the way he ... What clues have you noticed up to now?

Calling on individual students, the professor teased out and discussed the important observations, illustrating the fact that useful information can be obtained by careful observation before the physician actually talks to the patient. From then on, the students were alert and observant, and much of the discussion centered around their perceptions and intuitions about the patient and his condition. Another professor presenting Frank Hall to his class of students repeatedly challenged them:

What do you want to ask him? He’s your patient, he’s looking at you. It’s you and Frank.

When a student posed a question, the professor often took the opportunity to discuss the rationale behind it before repeating the question to Mr. Hall.

The drama of the session is often enhanced as the students observe the patient over time and see the effects of their clinical decisions. Various clinical situations arise, with outcomes that are determined by the model’s probability algorithm. Even the professor cannot predict the course of the case. Good clinical decisions and timely interventions increase the probability of an unremarkable course and successful outcome but, as in reality, assure neither.

A professor of gastroenterology, who was presenting Frank Hall to a second-year class if two hundred students, created a very tense situation by leaving all of the management decisions up to the class when Mr. Hall vomited blood during his hospital stay. As the students, seemingly stunned, sat observing the crisis on the twenty-foot screen in front of them, the professor urged them to react:

What do you do? He’s there on the bed, bleeding... What do you do?

Answers came from the audience: "Clear the airway!" "Call for help!" The professor chose from among the students’ therapeutic suggestions, explaining the rationale and emphasizing important management considerations, moving step-by-step through the crisis until the situation was under control.

In a presentation of a TIME simulation involving a morbidly obese female (Patricia Fletcher) who wished to have a gastroplasty, the professor asked the students to collectively decide and justify the optimal management plan for her condition. Based on the patient’s generally good health and the serious stresses in her life, the class voted for conservative therapy (eating disorders clinic) rather than the surgical procedure (gastroplasty) being sought by the patient as the solution to all of her problems.

In another presentation of this simulation, surgery was the professor’s treatment of choice to emphasize the difficulties it poses for managing the obese patient. When a post-surgical crisis arose and the patient was in obvious distress, the students were left to evaluate and manage the situation themselves.

In every case, regardless of the style of presentation, the classroom was transformed into an "active learning format" where participation and involvement replaced passiveness and apathy. Data were gathered that support this claim.

Evaluating the TIME Experience

Data gathered from more than three hundred fifty students during the field test mentioned previously showed that the TIME model is an effective experiential learning method. Questionnaires were administered immediately following every case presentation to assess 1) the degree of the students’ intellectual participation in the clinical problems presented in the simulations, and 2) the level of the students’ emotional involvement with the interactive patients during the presentations.

Students were asked to respond to questionnaire statements on a five-point Likert-type scale (where 1 = strongly disagree and 5 = strongly agree). In addition to the questionnaires, several of the presentations were videotaped, and post-session interviews of a sample of the participating students at each school were conducted and videotaped. The results from the questionnaires provided clear evidence that the majority of students were intellectually involved (Chart 1), as well as emotionally involved (Chart 2), with the simulated patient.

Chart 1

TIME Interactive Case Study Model -- Intellectual Involvement

Chart 2

TIME Interactive Case Study Model -- Emotional Involvement

A notable finding, clearly shown in the charts, was that in a group setting, a large majority of the students at all three schools had a "sense of reality" about the case(s), had a "sense of concern" for the patient(s), and made their "own decisions carefully" about the patient(s). In essence, they became individually committed to the care and management of both Frank Hall and Patricia Fletcher, and they left the session feeling as though they had encountered an actual person. This finding was validated by the videotapes that were made of several of the classroom presentations. The interest level and high degree of the students’ intellectual and emotional involvement, and their active participation throughout the sessions, are a matter of visual record.

From the post-session interviews, there was confirming evidence that students believed that they were responsible for what was happening to the "patient." For example, one student described his reaction to Patricia Fletcher as follows:

I was scared when she went into distress. I asked myself, "What would I do?" and I didn’t know... You don’t know really how you feel until you’re faced with that issue... and in that crisis situation I didn’t know what to do, and I didn’t like it!

Even though this student was part of a large group presentation, he felt a personal responsibility to act. Another student in the same group stated:

I really felt for the woman... When we started delving into why she was depressed, it was very, very good. You realize that issues are not cut and dried. She was a Ph.D., she had a son that died, a sister that died, and her husband might be homosexual... her whole life is collapsing around her and... the whole issue of sending her to surgery for some quick fix falls in the background.

Such comments demonstrate that the students felt a personal responsibility for managing the simulated patient, as well as intellectual and/or emotional involvement in her medical and psychosocial problems.

Conclusion

The TIME simulation model enhances the elements essential to effective experiential learning – intellectual involvement, emotional involvement, and the opportunity to influence the situation. Using the model in the classroom, medical faculty are able to create believable clinical encounters that compel their students to feel responsible for managing simulated patients’ health problems. This new method of patient-centered, experiential learning responds to at least three recommendations of the GPEP Report by replacing a lecture with an active learning format, presenting basic science information in a clinically relevant setting, and involving faculty with student learning using a new, more dynamic approach.

References

Allan, David M. E., and L. Scott Lane. 1991. Delivering thirty-six IVD programs for nurses and health professionals. Journal of Medical Education Technologies 1(Winter):5-7.

Association of American Medical Colleges. 1984. Physicians for the twenty-first century: The GPEP report, Report of the Panel on the General Professional Education of the Physician and College Preparation for Medicine.

Harless, W. G., M. A. Zier, and R. C. Duncan. 1986. A voice-activated, interactive videodisc case study for use in the medical school classroom. Journal of Medical Education 61:913-915.

Harless, W. G. 1986. An interactive videodisc drama: The case of Frank Hall. Journal of Computer-Based Instruction 13:113-116. ’

Harless, W. G., R, C. Duncan, M. A. Zier, et al. 1990. A field test of the TIME patient simulation mode1. Academic Medicine 65: 327-333.

Henderson, Joseph. 1989. Regimental surgeon: Interactive media for surrogate on-the-job training. Warrenton, VA: Proceedings of the Society for Applied Learning Technology.

About the Authors

William G. Harless, Ph.D., is Director of the TIME Project. Dr. Harless has more than twenty-five years of experience in conceptualizing and implementing innovative educational models to address problems in medical education and evaluation. As a special expert at the Lister Hill Center for Biomedical Communications, National Library of Medicine, Bethesda, Maryland, he developed the TIME model, a voice-activated interactive videodisc patient simulation for medical school classroom teaching. Contact Dr. Harless, The TIME Project, 1616 N. Ft. Meyer Drive, Suite 1430, Arlington, Virginia 22209, (703) 528-3292.

Marcia A. Zier is Internal Program Manager for the TIME Project and coordinates videodisc simulation production and development. Ms. Zier has designed and developed computer-based simulations for education and evaluation since the early 1970s; she has also developed educational software and print materials in the private sector. At the Lister Hill National Center for Biomedical Communications, she was Manager of Design and Production for the TIME Project during the early development of the prototype TIME interactive patient simulation model.

Jean E. Smith is External Program Manager for the TIME Project, coordinating faculty training and use of TIME simulations at medical schools throughout the country. Prior to joining the TIME Project, Ms. Smith was a policy analyst in the Congressional Office of Technology Assessment (OTA), Communication and Information Technologies Program, Washington, DC. She was also program coordinator of the Robert Wood Johnson Clinical Scholars Program when its national office was based at the University of California, San Francisco.

Ross Dube is Systems Analyst for the TIME Project. He is responsible for the design, development, and maintenance of software, and he provides technical support for both the internal and external programs. Mr. Dube’s previous experience includes development of a computer-aided design software package, which is used by architectural firms for building design. He has also designed and developed firmware for digitizing tablets and a flatbed plotter, which are used, respectively, as input and output devices in a graphics environment.

Robert C. Duncan, Ph.D., is a Professor in the Department of Oncology at the University of Miami School of Medicine in Miami, Florida. In addition to many years of experience in the planning and analysis of medical clinical trials, Dr. Duncan has been extensively involved in graduate and medical education. His collaboration with Dr. Harless began with responsibility for the evaluation of several computer-based learning projects, and it currently involves the evaluation and refinement of the TIME model for use in medical education.

William R. Ayers, M.D., is Associate Professor of Internal Medicine and Associate Dean for Undergraduate Medical Education at Georgetown University School of Medicine. He is responsible for medical informatics and coordinates the program on computers in medical education. Dr. Ayers is a Fellow of the American College of Physicians and a member of both the AMA’s Section on Medical Schools and the Liaison Committee on Medical Education’s accreditation teams. He has been using TIME interactive patients in the classroom since 1987.

--JOURNAL OF MEDICAL EDUCATION TECHNOLOGIES, Volume2, Number4, Spring 1992 (p.3-8)