Educational Applications of Computers in Medical

Peter G. Coggan, MD, Marilyn Hoppe, PhD, and Ralph Hadac, PhD Seattle, Washington

Computers have been used in various aspects of data provoked interest in computer applications medical education since the 1960s. Recent techno­ for the management of educational activities and logical developments have resulted in an explosion student assessment. Further interest was gener­ of more powerful and economical computer hard­ ated by the computer’s potential role in determin­ ware, which is now as easily available as television ing the diagnosis and prognosis of disease and in or stereo. In this article the use of computers in developing therapeutic plans in patient care. Pre­ medical education will be explored by looking dictions were made that the available technology at past and present practices and problems. How would save time, improve efficiency and skills, computers will be used in medical education and and provide unique learning opportunities. the implication of their future use for medical stu­ Early efforts to apply computer science to med­ dents, residents, faculty, and practicing ical education resulted in a number of interesting will be examined. but limited projects. The computer of the 1960s was a large mainframe machine filled with vacuum tubes and miles of wire. Because of the great ex­ pense involved and the need for technical mainte­ nance expertise, computers were located only in large business and governmental or educational The First 20 Years centers. Computer applications during this time Computers were recognized at an early stage in were largely industrial or military, although at their development as a new technology to be ex­ least one large computer-based instruction project ploited in —particularly in medical edu­ for postsecondary education, PLATO (Program cation. The ability to manipulate large amounts of Logic for Automatic Teaching Operations), was developed at the of Illinois. PLATO is a large and expensive centralized computer sys­ Dr. Coggan is Assistant Professor of and tem capable of storing great amounts of data and Dr. Hoppe is Resources Coordinator, both executing complex instructional strategies that are in the Department of Family Medicine, and Dr. Hadac is Research Associate in the Division of Research in Medical teacher centered and controlled. PLATO marked Education, University of Washington, Seattle, Washington. a successful application of computers to educa-

1984 Appleton-Century-Crofts

66 THE JOURNAL OF FAMILY PRACTICE, VOL. 19, NO. 1: 66-71, 1984 EDUCATIONAL APPLICATIONS OF COMPUTERS

tion, but was not widely adopted because of its Table 1 summarizes some of the problems iden­ expense.1 tified from various sources in each of these six The next stage in this evolutionary process was categories. As can be seen, developers were faced the introduction of the minicomputer. Still rela­ with many difficult problems. tively expensive, the minicomputer relied heavily on transistors for its internal operation, making it a much more compact unit than its predecessors. A large-scale computer-based instruction project, TICCIT (Time Shared Interactive Computer- Controlled Information Television), was devel­ Advent of the Microcomputer oped using this new technology. TICCIT was to­ Although educational problems with the first tally learner centered and organized, using the computer systems and computer-based instruction computer and instructors as resources. Included in models could have been overcome with time, eco­ this instruction were some of the first computer­ nomic and technical difficulties presented the big­ ized medical for instruction.1 gest challenge. Fortunately, a breakthrough in During the same period that TICCIT was being technology, the development of the microchip,10 launched, a joint project was beginning to take helped to reduce greatly the number of technical shape between the National Board of Medical difficulties, and in time, mass production tech­ Examiners and the American Board of Internal niques helped to dissolve economic barriers. Medicine. This project, the Computer-Based Ex­ This single technological development, the mi­ amination Project,2 sought to develop a computer- crochip, resulted in the production of much based system for evaluating clinical competence cheaper and more reliable computers of manage­ through the use of computer programs that simu­ able size, which rapidly became available in lated patient problems. Although the project has schools, businesses, and homes across the coun­ not yet been completed and implemented, it marks try. Although several substantial problems in thg a major step forward in the use of computers in use of computers in education were solved by this medical education. breakthrough, many still remained as major im­ pediments to large-scale use. A number of the problems enumerated in Table 1 persist. Production and distribution remain prob­ lems, as there are few tangible rewards for those Problems who write computer programs. The development Although all three of the projects mentioned are of new materials is not viewed with the same favor landmarks in the evolution of computers for use as research or traditional teaching activities. The in instruction, widespread adoption of these considerable effort required to produce software computer-based instruction models has not oc­ materials is therefore confined to a few individuals curred. According to Anastasio and Morgan3 in a who are able to see a potential profit or who have a report written as early as 1972, a number of diffi­ special interest. As a result, the quality and quan­ culties have presented roadblocks to wider use of tity of materials are lacking. While fun to com­ computers in instruction: plete, most existing computer-assisted instruction programs or patient management problems are not challenging or flexible enough for most physicians. The effectiveness of computer-assisted instruc­ The complex factors inhibiting widespread use of com­ tion has yet to be convincingly demonstrated, par­ puters in instruction have three principal dimensions— educational, economic, and technical. These three di­ ticularly in medical education. To add to the woes mensions are reflected in the six categories . . . a) pro­ of computer-assisted instruction, the underlying duction and distribution of instructional materials; b) instructional theory necessary to develop valid demonstration of the effectiveness of CAI (computer- strategies is still poorly conceived. It is not diffi­ assisted instruction); c) theory of instruction; d) educa­ cult to see the role of the computer in the storage, tional system and the teacher; e) cost; and f) technical retrieval, and manipulations of data that may be research and development. presented in a form assisting in the diagnostic,

THE JOURNAL OF FAMILY PRACTICE, VOL. 19, NO. 1, 1984 67 EDUCA TIONAL APPLICA TIONS OF COMPUTERS

Table 1. Problems of Educational Applications of Computers

Categories3 Problems (1955-1975)

Theory of instruction Poor understanding of the computer's capabilities in education Information processing potential not reached Education system Poor understanding of the role of the and the teacher student and the teacher Perception of threat to job security Not well integrated into medical curriculum4,5 Fear of new technology4 Demonstration of Only small positive effects realized effectiveness for attitude and achievement6,7 Few good examples available

Technical research Reliability of equipment and development Difficult mode of input Languages difficult to use8 Monosensory9 Production and distribution Lack of faculty interest Dispersion of responsibility Lack of incentives Lack of good materials Lack of production methods

Cost High cost of hardware and software5

prognostic, and therapeutic decision-making proc­ familiarity with new languages may also dull fac­ ess. It is exceedingly difficult, however, to see ulty interest. how a new technology, which is perceived as a The picture is not all bleak, however. Some ini­ threat and is not well understood, can be of help tial problems have been wholly or partially solved. in educational programs in which there is little The microcomputer has made hardware cheaper, teacher expertise. The potential offered by com­ smaller, more reliable, and more widely distrib­ puters and their application to specific and appro­ uted. Microcomputers have the advantage that priate areas of teaching is not well developed, thus programs on “ diskette” may be readily distributed limiting their usefulness. and exchanged, but they are limited in that, gen­ Teachers in the public school system are rapidly erally speaking, programs developed on one com­ adding computerized instruction and courses in puter may not be easily used on others. The mem­ computer literacy to the . Medical ory capacity of a modern microcomputer may also schools have been slower to adopt these new limit its use, although the development of hard methods; consequently, entering students fre­ disks and the availability of increasing memory quently are more familiar with computers and their capacity are changing this situation rapidly. uses than the faculty. The difficulty of gaining The sophistication of available programs may

68 THE JOURNAL OF FAMILY PRACTICE, VOL. 19, NO. 1, 1984 EDUCATIONAL APPLICATIONS OF COMPUTERS not be great, but the number and variety of pro­ opment for continuing medical education, where grams are increasing. Experience is accumulating they hold some promise as a combined teaching and an improved product can be anticipated. and self-assessment tool. These programs, however, do not currently offer flexibility for the learner by responding to his or her needs or learning style.

Evaluation and Assessment A number of programs developed around pa­ tient management problem models score the learner as the problem proceeds as a means of self- Current Educational Applications assessment. Cumulative scores may be registered of Computers on the diskette, and then may be available to both Of the many computer uses currently being ex­ teacher and learner. Again, video applications are plored, the following categories summarize the being developed, but these programs are not cur­ major uses in medical education. rently being used for evaluation or assessment. One computer application that is available and does work well is the storage and retrieval of multiple-choice questions. Large numbers of multiple-choice questions may be stored easily on Learning diskette and retrieved for an examination in a ran­ Many programs are now available on diskette. domized fashion or as blocks of questions on spe­ Most have been developed for Apple II or Apple cific topics. lie computers, which seem to be by far the most commonly found. The programs are often interest­ ing and stimulating, but require the learner to fol­ low a fixed tutorial direction without offering much flexibility. Several centers are working on Information Management the integration of videotape or videodisc materi­ A successful use of computer technology in als, but these developments are still in their education is in conducting literature searches. infancy. Catalogues of published articles have been acces­ A number of clinical simulations are available sible by computer for several years. Two of the for use specifically on microcomputers. The simu­ most familiar are MEDLINE and MEDLARS, lations follow a format similar to that for programs both of which offer key word searches. MED­ developed for PLATO but remain relatively unre­ LARS also provides an annotated printout. sponsive to the needs of the individual student. Computers may also be used in the organization Although capable of teaching and testing knowl­ and storage of large data bases for medical use. edge and clinical reasoning and demonstrating Programs such as AMA/NET (or Telenet) are skills, such programs cannot test skills, perform­ being developed under which the vast volume of ance, or attitude. Most of these programs use a current medical information can be stored and re­ linear format in which information is presented to trieved for use in solving clinical problems. the learner, and an understanding of the new in­ formation is immediately tested. Further help is offered where learning is deficient, and a summa- tive test is offered at the end of the program. Most Documentation of Learning Experiences of the patient management problems follow the familiar linear design. Mainframe computer pro­ Many programs document the clinical grams are able to offer more complex branched experiences of their residents for educational pur­ programs, but these are not yet available for use poses and to provide evidence for privi­ on microcomputers. leges. Most such computer programs currently in Clinical simulations are in the process of devel­ use rely on information obtained from patient bill-

THE JOURNAL OF FAMILY PRACTICE, VOL. 19, NO. 1, 1984 69 EDUCATIONAL APPLICATIONS OF COMPUTERS

ing data. There is little consistency among pro­ ized data base from a central source. Advances in grams and, as yet, no national agreement exists on data storage and retrieval, coupled with a greater what should be recorded or what constitutes the ease in assessment, will lead to the development of essential experiences that should occur during a more individualized curriculum that will allow each resident’s training. students to progress quickly through areas already mastered or those found to be relatively easy. More time may be assigned to areas of weakness or special interest. More interactive computer programs will allow flexibility not only for learning but also for testing The Future and prescriptive feedback. Areas of weakness will Given the rate of change in computer technol­ be more easily identified, and specific remediation ogy in recent years, it is probable that many or will be available with retesting until competence most of the problems enumerated in Table 1 will is achieved. Thus the curriculum is likely to be be overcome. It seems likely, however, that prog­ driven almost entirely by the needs of the individ­ ress will be slow as software is perfected, new ual, and the result may well be a profound change skills are learned and disseminated, and educa­ in the role of the faculty. tional theories are developed. Once these prob­ lems are surmounted, the impact of computers on medical education could bring about profound changes in the learning process for medical stu­ dents, residents, and practicing physicians. Equal­ Impact on the Teacher and ly far-reaching are the effects that can be antici­ Teaching Program pated in the role of faculty and the institutions they The computer will be able to assemble and serve. present information in a form designed to fit the needs of the individual learner, offering greater flexibility than current instructional methods. The need to impart large numbers of facts through a lecture or seminar is likely to diminish greatly. The Impact on the Learner teacher’s role will therefore become that of a men­ Perhaps one of the most obvious applications of tor, advising and guiding the student in the choice computers in medical education—and one that is of a wide range of learning opportunities. It may acutely needed—is data retrieval. The increase in be possible to teach skills information available to physicians has been ex­ through the use of computerized holograms or ponential as research has advanced medical models, but it is unlikely that the computer will knowledge. Maintaining knowledge at a current supplant the teacher as a clinical role model or level is a problem for all physicians and an even personal advisor. greater challenge to medical students, who have The development of more individualized learn­ not yet developed a cognitive framework in which ing will change curricular and insti­ to integrate new information. Computers offer the tutional structures. The admissions process will potential for storage and retrieval of information select students who are more self-directed. Cur­ that may be accumulated over time. Information ricula will become self-paced and competency may be stored in a way most compatible for the based. Testing of carefully selected competencies individual’s learning and problem-solving style. will become a major preoccupation of the faculty, Rote memorization of large quantities of facts will who will require demonstration of a comprehen­ no longer be necessary, freeing physicians from sive range of knowledge and skills to a required this discipline and allowing more opportunities to standard prior to graduation. The development of focus on clinical reasoning and interpersonal appropriate instructional assessment software will skills. become a valued faculty activity and will require As computer networking develops, it will be­ the nurturing and development of resource teams come possible to revise and update a computer- to aid in programming and implementation.

70 THE JOURNAL OF FAMILY PRACTICE, VOL. 19, NO. 1, 1984 EDUCATIONAL APPLICATIONS OF COMPUTERS

The student academic record is likely to be­ Conclusions come computerized and will be much more com­ prehensive than at present. An exposure to The pace of technological change in the com­ specific clinical situations and procedures will be puter field over the past two decades has been required and monitored closely. Documentation breathtaking. It is tempting to predict that the of experiences in medical school will be linked to computer promises to be a panacea for all educa­ residency training, where such documentation is tional ills. If the rate of progress in its development already assuming great importance. continues, it is unlikely problems that currently Departments of medical education are likely to are impediments to the integration of computers in develop more expertise in individual learning medical education will remain so. Technological styles, the teaching and assessment of clinical rea­ advance, however, is rarely a complete answer soning skills, and the identification, teaching, and in and of itself, since it demands an adaptation assessment of fundamental history taking, physi­ by affected individuals. Ultimately the success of cal examination, and interpersonal and procedural computers in medical education will depend on skills. The primary teaching and assessment tool their acceptance and integration by both learners will, of course, be the computer. and teachers. It is hard to imagine a complete re­ jection of this new and exciting educational tool, but it is equally difficult to predict to what extent it will receive acceptance and how the present educational system will adapt to its many uses. Impact on Medical Practice Clearly the personal relationship between teacher The manipulation of huge numbers of medical and student will remain a central element in facts will result in a greatly increased understand­ the educational process, but that relationship is ing of disease and health. The data base developed bound to change. It will be intriguing to watch its on computers in medical school will be refined, evolution. revised, and extended in practice, contributing to a higher standard of medical care, stimulating new strategies for problem solving, and allowing more time for a more personal approach to patients. Licensing examinations will likely become com­ petency based, dependent not on a specific period of training but on demonstrable performance at a specific level. Testing will continue until a satis­ References factory performance is obtained. Thus, highly 1. O'Neil HF Jr: Computer-Based Instruction: A State- competent students whose level of test perform­ of-the-Art Assessment. New York, Academic Press, 1981 2. Senoir JR: Toward the Measurement of Compe­ ance is outstanding may receive a relatively short tence in Medicine. Philadelphia, John R. Senoir, 1976 examination, whereas those whose performance is 3. Anastasio EJ, Morgan JS: Study of Factors That Have Inhibited a More Widespread Use of Computers in the marginal could be tested at length to ensure that Instructional Process. Washington, DC, EDUCOM, 1972 minimal competency levels have been reached. 4. Schwartz MW, Hanson CW: Microcomputer and computer-based instruction. J Med Educ 57:303, 1982 Continuing medical education is likely to be 5. Swanson DB, Friedman CP: Microcomputer applica­ conducted through computer terminals in the tion in undergraduate clinical training. In Research in Medi­ cal Education: 1982. Proceedings of the Twenty-first An­ office or home. Programs will be increasingly so­ nual Conference. Washington, DC, Association of American phisticated and more interactive and will provide Medical Colleges, 1982 6. Kulik JA, Kulik CC, Cohen PA: Effectiveness of constructive, prescriptive feedback, raising the computer-based college teaching: A meta-analysis of find­ possibility of continuing assessment of compe­ ings. Rev Educ Res 50:525, 1980 7. Marion R, Niebuhr BR, Petrusa ER, Weinholtz D: tency in practice through computerized testing. Computer-based instruction in basic medical science edu­ Certainly relicensure is likely to be conducted in cation. J Med Educ 57:521, 1982 8. Burson J: Education at your fingertips. Am J Med this manner. Documentation of continuing compe­ Tech 48:573, 1982 tency will be important for renewing hospital 9. Abdulla AM, Henke JS, Weitz FI: Microcomputer- based medical education. South Med J 74:1251, 1981 privileges and perhaps for obtaining malpractice 10. Boraiko AA: The chip. National Geographic 162:421, insurance. 1982

THE JOURNAL OF FAMILY PRACTICE, VOL. 19, NO. 1, 1984 71