Preamble

The field of medical informatics deals with the use of computing and information technology in medical research, teaching, patient care, and hospital administration. Major areas of focus in medical informatics include: biomedical-systems technology, health-information systems, medical-decision-support systems, image and signal analysis, statistical techniques and modelling, education and training, health-care management, and the human-machine interface.

In December 1994, Faculty Council and Dean Ruedy recognized the need for the development of medical informatics in the Faculty of Medicine, and the Medical Informatics Review Committee was struck to make recommendations regarding the development of medical informatics at Dalhousie over the next five years.

We met regularly and interviewed interested parties in the Faculty of Medicine, including both faculty and students. In addition, the Chair had discussions with the Medical Society of Nova Scotia and the provincial department of health regarding medical-informatics topics.

This document is a result of our deliberations. The specific recommendations are consistent with the long-term aspirations of the Faculty of Medicine as they are formulated in "Strategies for the 1990s: Faculty of Medicine of Dalhousie University" (please see Appendix 1).

1. Executive Summary

Information is the currency of universities; its organization and retrieval is essential for the development of new knowledge. It is therefore desirable that universities implement the most effective techniques for information retrieval, storage, and display. The quality of their computing and information-technology environment will profoundly influence universities' ability to compete in education and research. The quality of the Medical Schools informatics infrastructure will also influence the opportunity to collaborate with other health organizations including hospitals, Departments of Health, Medical Societies and clinicians in remote areas.

The Faculty of Medicine at Dalhousie University has many of the components that are essential to a successful medical-informatics program. In addition to faculty members who are knowledgeable in key areas of medical informatics (such as image and signal analysis, statistical techniques and modelling, decision-support systems, health-care management, and education), there is a vast intellectual resource of medical experts who, because of their knowledge and experience, should be able to contribute significantly in various areas. Two examples are decision-support systems and computer-aided learning.

But in spite of a considerable level of local expertise and a substantial investment in hardware and software, the Faculty of Medicine currently has no overall informatics policy or program. In our opinion, a comprehensive medical-informatics program must be urgently developed, based on the power of high-performance computation, high-speed communications, and strengthened expertise in key areas of informatics. Much of the emphasis should be on projects that build on partnerships and collaborations with other scientific and technical disciplines, with other universities, with industry, with hospitals and regional health boards, the Nova Scotia Medical Society, and the Department of Health.

The first step towards developing a comprehensive medical-informatics program in the Faculty of Medicine is the realization that the field of medical informatics spans a broad spectrum, including research and development, teaching, and clinical care. To coordinate and integrate activities in this broad field, it is necessary to first establish an appropriate governing body with representation from all constituent components of medical informatics. Therefore, we propose that a Medical Informatics Steering Committee be struck, as the first step in the development of a comprehensive and integrated medical-informatics program.

Recommendation 1:

That the Faculty of Medicine immediately strike a Medical Informatics Steering Committee

We recommend that the Dean of the Faculty of Medicine choose members of the Medical Informatics Steering Committee on the basis of their educational background, training, and experience (appendix 1 has a suggested terms of reference). To reflect the diversity of activities in medical informatics, the disciplines represented by this committee should include biomedical engineering and computer science, epidemiology and biostatistics, surgical specialties, internal medicine and family medicine, as well as the highly technical hospital-based specialities such as radiology, laboratory medicine, and anaesthesiology. We are not suggesting a large committee; rather a committee where each member has the responsibility for representing several constituencies. All constituencies must have the opportunity to contribute, directly or indirectly, to the process of developing a medical-informatics program. Ideally the Medical Informatics Steering Committee should comprise no more than seven members, but it would be able to recruit working groups to deal with specific problems. The director of Medical Computing and Media Services should report to and be responsible to the committee. The committee should report directly to the Dean of Medicine.

Immediately after it is appointed, the committee should begin a systematic analytical and design phase; this would include examining the existing organizational structure and infrastructure, and making recommendations for the future development of chosen areas in research, education, and patient care. We would expect the Medical Informatics Steering Committee to complete within the first year of its mandate the documentation and analysis of the existing infrastructure and human resources in medical informatics, and the design and implementation of high-speed communication links that would integrate existing resources and connect all faculty and students in the Faculty of Medicine.

Specifically, we recommend that the Medical Informatics Steering Committee, once appointed, get to work as promptly as possible to achieve the following five recommendations:

Recommendation 2:

That the existing infrastructure and expertise in medical informatics in the Faculty of Medicine and its affiliated hospital environments continue to be identified and assessed, and then consolidated and integrated.

Recommendation 3:

That installation of hardware and software - including networking and server systems - be extended that would increase the efficiency of research and thus improve the overall ability to attract external funding from both granting agencies and the private sector.

Recommendation 4:

That properly integrated academic information systems and hospital information systems be installed to facilitate access to scientific literature and provide common data bases for clinical research and patient care.

Recommendation 5:

That education and training programs in medical informatics be established at the undergraduate, postgraduate, and continuing levels, based on faculty expertise and an adequate infrastructure (i.e. on all of the above).

Recommendation 6:

That productive extramural collaborations be developed with other universities, provincial medical societies, departments of health, regional health boards, hospitals, and industry.

The existing budget and in-kind contributions of the Faculty of Medicine can cover a significant portion of the projected expenses of the proposed medical-informatics program, provided the use of resources---including existing faculty---is rationalized and properly planned. The balance of the budget will have to be sought from external agencies by competitive research groups. However, we believe that collaboration will provide overall savings as Nova Scotians develop an overall strategy for medical informatics. Dalhousie Medical School, its associated hospitals, Provincial Governments and the Nova Scotia medical Society will all benefit from a collaborative approach.

2. Existing infrastructure for informatics

The first proposal to establish a computing environment that would be conducive to developing an integrated medical-informatics program was submitted to the Dean of Medicine in a 1987 report entitled Access to Computing in the Faculty of Medicine. In 1990, representatives from Dalhousie University and the Halifax teaching hospitals identified the need for establishing a high-speed network linking their institutions, in a document entitled Report to the Dean of Medicine: Inter-Institutional Network for the Nova Scotia Teaching Hospitals and the Dalhousie Medical School. That report was forwarded by the Dean of Medicine to the CEOs of the Halifax teaching hospitals, who, in turn, appointed technical administrators to a committee to (merely) investigate the need for such a network.

However, in spite of the slowness with which the concept of networking percolated through layers of hospital administration, a considerable amount of work has been done towards this end by some individual faculty members, who realized that implementing a network that would encompass all of the teaching hospitals would---due to fiscal restraints---have to be done in phases. Thus, a less ambitious concept emerged: to connect first the Victoria General Hospital complex and the IWK/Grace Maternity Hospital to the Dalhousie Ethernet network, and thereby serve the clinical faculty's computing needs in research, teaching, patient care, and administration. Since 1992, this effort was supported by some clinical departments (Medicine, Anaesthesia and Radiology contributed $9,500 each), the Centre for Clinical Research ($20,000), the Dalhousie Medical Research Foundation ($31,000), and the office of the Dean of Medicine ($40,000). This investment paid for the design and construction of the Ethernet backbone for the VGH complex. Unfortunately, only a handful of users---in the Department of Radiology and the Cardiology Division of the Department of Medicine---benefit from this initiative. There are several reasons for this unsatisfactory situation; the main one is that hospital administration did not, up to this point, embrace the Ethernet concept (thus encouraging further proliferation of obsolete networking technology, requiring installation of new telephone lines with low-speed modems). This urgent matter has to be addressed by the proposed Medical Informatics Steering Committee.

The integration of Halifax hospitals into the QE II - Grace/IWK complex provides an opportunity to reexamine the issue of networking. The University and QE II must collaborate now to develop a comprehensive system that would satisfy both internal networking needs and the need to communicate with the rest of the world. Communications directors, Stewart Gray of the QE II and Ron MacLeod of the Faculty of Medicine, should work together under the guidance of the Medical Informatics Steering Committee, and build on the existing collaborations, to achieve this goal.

The Appendix 3 describes in detail the existing infrastructure for medical informatics in basic-science departments of the Faculty of Medicine.

3. Existing expertise in informatics

A strong core group is essential for the development of medical informatics in the Faculty of Medicine. Currently there are individuals and groups within Faculty who have national and international reputations in one or more of the major areas that comprise medical informatics. These existing intellectual resources are the natural nuclei of the future program in medical informatics. Consequently, it is essential that the Faculty recognize the value of this group of individuals by providing appropriate financial and time resources.

Recommendation 7:

That the Faculty of Medicine and Dalhousie Medical Research Foundation support those existing groups within the Faculty that have the potential to become the cornerstones of a comprehensive medical informatics program (and we believe completion of this account should be an early task of the Medical Informatics Steering Committee).

4. Proposed infrastructure for informatics

Despite the major investment in a medical-informatics infrastructure and considerable local knowhow, as partially described in the previous section, many clinical faculty members do not have access to the full range of communications services. We see inadequate communications as the most serious deficiency of the present infrastructure for medical informatics. Therefore, we make the following general recommendation:

Recommendation 8:

That the issue of inadequate communications within the Faculty of Medicine receives immediate attention, resulting in prompt action, of the proposed Medical Informatics Steering Committee

To address the issue of computing and information technology in the comprehensive manner, two hierarchical levels of technology have to receive immediate attention simultaneously: firstly, a local area network (LAN) that would reach all faculty and students of the Faculty of Medicine of Dalhousie University; secondly, a high-speed test network in Atlantic Canada in association with the National Test Network.

4.2 Local Area Network for the Faculty of Medicine

There are many projects being undertaken by members of the various departments of the Faculty of Medicine that require advanced computational methods and powerful hardware; such computer applications include digital signal processing, advanced statistical analysis, mathematical modelling, two- and three-dimensional image processing, and pattern recognition. Moreover, these computational requirements of some projects notwithstanding, every member of the Faculty and every student should have a high-speed connection to Internet. Therefore,we consider it a matter of high priority that:

Recommendation 9:

That the Faculty of Medicine completes within one year development of a high-speed communications network providing equal access to all members regardless of their location,and facilitating the sharing of clinical, educational and administrative information.

Dalhousie University's Ethernet network should encompass all basic-science and clinical departments of the Faculty of Medicine. We recommend a strategy that would initially build a modest local capability for high performance computation linked together by a Faculty-wide high performance communications network. High-speed regional network linkages (ACORN) feeding into the national CANARIE initiative would provide all members of the Faculty of Medicine access to the full range of available computing facilities elsewhere in Canada and in the United States.

The proposed networked computing environment at the Faculty of Medicine should be based on the concept of computing through clients and servers. This means that storage, printing, graphics, compilers, scientific-subroutine libraries, scientific desktop-publishing systems, and other services should reside on shared server systems (VAX/VMS-based servers or UNIX-based servers), which will be accessible to any user who needs the services they provide, no matter what sort of equipment he/she operates: workstation, personal computer or terminal.

Each department's computing environment should be an integral part of the campus-wide Ethernet communications network, which is operated by University Computing and Information Services (UCIS) and is connected to the world through Internet. This will facilitate collaboration with investigators at remote sites.

5. Proposed health information systems

Medical research often uses information which is collected in the normal course of treatment. Unfortunately, much of that data is not collected in a usable form. Moreover, much of the information which is used to evaluate hospital performance has major problems.

There has been little evidence of collaboration between the various institutions in Halifax or elsewhere in Nova Scotia, regarding the development and implementation of Patient-Care Information Systems (PCIS). Neither the Department of Health, nor Regional Health Boards, nor hospitals, have identified the type of information required for a useful PCIS. On the other hand, all hospitals have developed Patient Information Systems; although the latter systems are useful in collecting demographic information, they do not help clinicians and researchers in developing useful research data bases, and most of them do not provide information about health outcomes for individual patients.

Appendix 4 describes the most widely used hospital information systems in use in Canada today, and it indicates some of the problems associated with the CIHI system. Appendix 5 lists the suggestions for a reengineered health record made by a QEII Health Sciences Centre working group. The medical faculty play a major role in choosing the type of information which is included in the health record, and it must be an important collaborator as the health record is redesigned. Therefore, we recommend:

Recommendation 10:

That collaborations be developed between the various constituencies responsible for Clinical Information Systems, i.e. Department of Health, Regional Health Boards, Hospitals, and the Faculty of Medicine of Dalhousie University,and that a committee composed of representatives from each group be responsible for developing a unified approach to the health record as a tool for patient care, teaching and research.

This would allow vertically integrated approaches to health records from education through patient care through funding mechanisms.

Substantial information will be available in the normal course of care. Several researchers maintain useful databases, but this existence is not widely known.

Recommendation 11:

Consequently, the Faculty of Medicine should keep a catalogue of clinical databases which could be used by people planning research.

Recommendation 12:

The population health research unit should further develop its ability to link clinical data bases while maintaining client confidentiality and should inform and consult with researchers and clinicians.

Library Systems Integration

A key prerequisite for any integrated academic information management system in the Faculty of Medicine will be the active and willing collaboration of the Kellogg Health Sciences Library, which will be required to implement a wide range of library information and related services. These should include basic scholarly information services such as Medline and other data bases (e.g. Cancerlit). It is critical that the Kellogg Health Sciences Library be responsive to the recommendations of the Dean and the unique needs of the medical community. However, Kellogg Library does not presently report to the Dean of Medicine, but to the management of Killam Library. In order to prevent unnecessary duplication and establishment of a parallel system outside of the library system, it is important that the management of the Kellogg Health Sciences Library responds to the rapidly changing needs of the medical community and not be restricted by the structure of the library system for the rest of the university.

Recommendation 13:

That in order to ensure cooperation; and prevent the duplication of services, the Kellogg Health Sciences Library has an official reporting responsibility to the Dean of Medicine.

6. Proposed medical decision support systems

The rapid growth of medical information has made medical decision making immensely complex. It is essential that medical decision support systems be implemented to help physicians acquire the information they need for patient care, research and education. The purpose of the proposed project would be to develop a comprehensive decision-support system in the Faculty of Medicine, by bringing together multiple sources of information that reside on disparate computers and different database systems scattered throughout the Faculty of Medicine and the Dalhousie campus.

Recommendation 14:

That a system of information resources for clinical decision making and patient-care management be created at the Faculty of Medicine.

Rapid and effective access to biomedical information can be approached by linking all faculty, students and staff in the Ethernet network of distributed systems. The objective is to create a "medical decision support system" for the Faculty's education, research and patient-care programs. The purpose of this network would be to install, over the next five years, a family of databases that would be accessible electronically by faculty and students from office, laboratory or clinical settings, and from home. The sources ideally would include bibliographic, informational and clinical databases, accessible from any of a variety of recommended users' workstations, by means of an extended Dalhousie Ethernet network with LAN nodes. (Computers and computer-related technology are increasingly evident in the operating room and in the intensive-care units. In the near future, the use of computers in this environment will become much more sophisticated and it will require more powerful computing resources.)

7. Proposed education-and-training programs

7.1 Graduate studies

To establish medical informatics firmly as a scientific field at Dalhousie University, the Faculty of Medicine should---in collaboration with the Department of Mathematics, Statistics and Computing Science and with the Technical University of Nova Scotia---establish a new academic program in medical informatics, advertise it nationally, and attract bright graduate students. A degree program would facilitate the development of expertise, attract funding, and improve the Faculty's ability to recruit new faculty in the field of medical informatics.

Recommendation 15:

That, jointly with the Department of Mathematics, Statistics and Computing Science, and with the Technical University of Nova Scotia, a graduate program in medical informatics be established, after a careful review and analysis of the merits and multidisciplinary resource implications.

To bring faculty up-to-date on the latest developments in the field of medical informatics, close personal contacts with world-class researchers in medical informatics should be encouraged by Faculty of Medicine -- particularly with those at universities in the northeastern United States. These links can be developed by encouraging faculty to spend their sabbatical leaves at these institutions.

Recommendation 16:

That an educational-brokering program be initiated whereby faculty could apply for funding support for sabbatical and clinical fellowship training in medical informatics at institutions where this field is at the advanced stage, and apply that training to existing research and educational programs at the Faculty of Medicine

Students in the Master of Science in Medical Informatics program and in the Doctor of Philosophy in Medical Informatics program would be selected from graduates of schools of medicine and from students in disciplines related to clinical medicine and public health. In addition, graduates from such disciplines as physics, electrical engineering, and computing science would be also eligible. A research project in one of the four disciplines that are academically strong at Dalhousie University (biosignal processing, image processing, modelling/simulations and decision-support systems) and advanced courses would constitute the basic requirements.

7.2 Postgraduate education

The ability to use informatics effectively in making important decisions about patient care is one of the most important skills that residents should acquire. Additionally, the effective use of information retrieval stimulates interest and discussion, and may lead to increased opportunities for clinical research in the academic setting of the Faculty of Medicine.

Currently, there is no formal education process for residents in the area of medical informatics, and computer access varies widely among departments. In some departments residents must share computers with other members. In other departments, hardware and software are not available. Residency staff are unaware of the variety of instructional software that is available and have no exposure to decision-making systems.

Ideally, all residency staff would have their own personal computers. In addition, common facilities must be available for housestaff to do searches and work on research projects and teaching assignments at the hospital. Each program should have a shared computer that its housestaff can use 24 hours a day for medical informatics. Funds for the computer and software should be provided by the hospital, the department, and the Faculty of Medicine.

Formal instruction in medical informatics should begin in the first year and continue for the duration. Participation by outside resource people, such as library staff, should be encouraged. Some instruction by departmental staff in decision making within the specialty should be mandatory. In addition to resident education, this would have the added benefit of forcing at least one department member to become familiar with the instructional software available in that specialty.

Based on these findings, we make the following formal recommendations:

Recommendation 17:

That shared computer hardware and software in teaching hospitals be made available specifically for housestaff, and that housestaff be encouraged to purchase personal computers for use at home, to allow literature searches and access to a variety of information data bases around the clock.

Recommendation 18:

That training in information retrieval and in the use of certain specialty-specific software be made a compulsory component of each Dalhousie-sponsored residency program and that specific guidelines should be developed for each specialty.

7.3 Continuing medical education

The Faculty of Medicine will also need to address the continuing-medical-education needs of practicing primary-care physicians and specialists. To do so will require innovative approaches, such as distance education using interactive technologies. Proposed developments in communications via fibre-optic cables will create an integrated infrastructure linking universities, hospitals, clinics, physicians' offices, libraries, and homes. With the continuing development of computer-aided learning tools and techniques, the Faculty of Medicine and other medical schools (e.g. those in the North East Medical Schools Consortium) will produce and use more computer-aided-learning software at all levels of medical education. Cost-effective continuing medical education on an entirely new level will therefore be feasible.

One major barrier to implementing advanced methods of continuing medical education is that many practicing physicians lack basic computer skills. Thus, in our opinion, Continuing Medical Education should further expand a formal program for training Nova Scotia physicians so that they can take advantage of informatics technology in their practice and in their self-development.

Informatics technology can support collaboration between clinicians and researchers in geographically disparate areas. Distance education can also be used to promote the use of informatics including decision support and knowledge of informatics methods.

Recommendation 19:

That Continuing Medical Education further expand appropriate training programs in medical informatics for practicing physicians and that this continue to be done in collaboration with the Medical Society of Nova Scotia and the Nova Scotia Department of Health

Recommendation 20:

That a proposal to establish distance education in medicine be submitted to the Council of Nova Scotia University Presidents

7.4 Undergraduate education

Medical students ideally should be required to have basic computer skills, such as ability to use word processing, spreadsheets, and data-base management systems. This basic requirement is not a distant goal: according to a survey of Dalhousie's 1994-95 incoming class, almost all students (99%) had used a computer, more than half (56) owned one, more than half (58%) had taken a computer course, the great majority (88%) had taken a course in which a computer was used for coursework, a minority of students (17%) did not feel confident that they have mastered the basics of computer usage, and all (100%) considered the possession of computer skills important in the practice of medicine. Under such circumstances, the requirement that every medical student at our institution have convenient access to a computer is merely the first necessary prerequisite for introducing medical informatics into undergraduate medical education. Therefore, as an incentive to acquire basic computer skills, the Faculty of Medicine should offer subsidized purchase of computer hardware and software to medical students.

Recommendation 21:

That all medical students have access to a personal computer and that the Faculty of Medicine make an arrangement for a group purchase of basic hardware and software, and make recommendations to incoming students regarding hardware and software configurations

In 1992, the Faculty of Medicine of Dalhousie University adopted a new case-oriented problem-stimulated (COPS) curriculum. The COPS curriculum includes a variety of "themes" integrated horizontally across disciplines throughout all four years of the undergraduate curriculum. One of these themes is medical informatics. This organizational structure helps to ensure representation of this area throughout the undergraduate curriculum, and it can extend into the postgraduate- and continuing-education areas.

Recommendation 22:

That the integration of the medical informatics theme in the COPS Undergraduate Medical Education Curriculum be completed, so that there is at least one medical informatics activity in each unit and clerkship rotation

We are convinced that implementation of our recommendations will facilitate the development of an integrated informatics program at Dalhousie University and will improve the effectiveness of teaching, research and patient care.

The proposed Medical Informatics Steering Committee will have an opportunity to develop an efficient program in medical informatics that will help reduce the overall cost of implementation by preventing redundant activities. In particular, networking will facilitate communication between faculty and improve the efficiency of clinical research.

Although many individual groups have a legitimate interest in medical informatics, the overall cost to the taxpayer will be reduced if the various constituencies, including Dalhousie Medical School, Provincial Medical Societies, Departments of Health, and Hospitals collaborate to develop a common infrastructure for medical informatics. We risk developing another tower of Babel if each constituency continues on its own course.

An integrated medical informatics program promises to open door to many exciting opportunities for Dalhousie faculty.

Recommendation 1: That the Faculty of Medicine immediately strike a Medical Informatics Steering Committee.

Recommendation 2: That the existing infrastructure and expertise in medical informatics in the Faculty of Medicine and its affiliated hospital environments continue to be identified and assessed, and then consolidated and integrated.

Recommendation 3: That installation of hardware and software - including networking and server systems - be extended that would increase the efficiency of research and thus improve the overall ability to attract external funding from granting agencies and the private sector.

Recommendation 4: That properly integrated academic information systems and hospital information systems be installed to facilitate access to scientific literature and provide common data bases for clinical research and patient care.

Recommendation 5: That education and training programs in medical informatics be established at the undergraduate, postgraduate, and continuing levels, based on faculty expertise and an adequate infrastructure (i.e., on all of the above).

Recommendation 6: That productive extramural collaborations be developed with other universities, provincial medical societies, departments of health, regional health boards, hospitals, and industry.

Recommendation 7: That the Faculty of Medicine and Dalhousie Medical Research Foundation support those existing groups within the Faculty that have the potential to become the cornerstones of a comprehensive medical informatics program.

Recommendation 8: That the issue of inadequate communications within the Faculty of Medicine receives immediate attention, resulting in prompt action of the proposed Medical Informatics Steering Committee.

Recommendation 9: That the Faculty of Medicine completes within one year development of a high-speed communications network providing equal access to all members regardless of their location, and facilitating the sharing of clinical, educational and administrative information.

Recommendation 10: That collaborations be developed between the various constituencies responsible for Clinical Information Systems, i.e. Department of Health, Regional Health Boards, Hospitals, and the Faculty of Medicine of Dalhousie University, and that a committee composed of representatives from each group be responsible for developing a unified approach to the health record as a tool for patient care, teaching and research.

Recommendation 11: That the Faculty of Medicine keep a catalogue of clinical databases for review and use by clinical researchers, and that the population health research unit develop the ability to link clinical data bases while maintaining client confidentiality.

Recommendation 12: That the population health research unit should further develop its ability to link clinical data bases while maintaining client confidentiality and should inform and consult with researchers and clinicians.

Recommendation 13: That, in order to ensure cooperation and prevent the duplication of services, the Kellogg Health Sciences Library has an official reporting responsibility to the Dean of Medicine.

Recommendation 14: That a system of information resources for clinical decision making and patient-care management be created at the Faculty of Medicine.

Recommendation 15: That, jointly with the Department of Mathematics, Statistics and Computing Science, and with the Technical University of Nova Scotia, a graduate program in medical informatics be established, after a careful review and analysis of the merits and multidisciplinary resource implications.

Recommendation 16: That an educational-brokering program be initiated whereby faculty could apply for funding support for sabbatical and clinical fellowship training in medical informatics at institutions where this field is at the advanced stage, and apply that training to existing research and educational programs at the Faculty of Medicine.

Recommendation 17: That shared computer hardware and software in teaching hospitals be made available specifically for housestaff, and that housestaff be encouraged to purchase personal computers for use at home, to allow literature searches and access to a variety of information data bases around the clock.

Recommendation 18: That training in information retrieval and in the use of certain specialty-specific software be made a compulsory component of each Dalhousie-sponsored residency program.

Recommendation 19: That Continuing Medical Education further expand appropriate training programs in medical informatics for practicing physicians and that this be done in collaboration with the Medical Society of Nova Scotia and the Nova Scotia Department of Health.

Recommendation 20: That a proposal to establish distance education in medicine be submitted to the Council of Nova Scotia University Presidents.

Recommendation 21: That all medical students have access to a personal computer and that the Faculty of Medicine make an arrangement for a group purchase of basic hardware and software, and make recommendations to incoming students regarding hardware and software configurations.

Recommendation 22: That the integration of the medical informatics theme in the COPS Undergraduate Medical Education Curriculum be completed, so that there is at least one medical informatics activity in each unit and clerkship rotation.

Appendix 1

Suggested Draft Terms of Reference Medical Informatics Steering Committee

1. Purpose
   • To promote rational development and use of information technology at Dalhousie Medical School including techniques related to communication and decision support.
   • To facilitate the development of graduate and undergraduate education in medical informatics.
   • To facilitate collaboration between pertinent groups as new information techniques are implemented and as decisions are made with regard to the collection and dissemination of information.

2. Duties:
   • To implement the recommendations outlined in the report of the Medical Informatics Review Committee.

3. Membership:
   • 3 members chosen from those active in clinical practice and research.
   • 4 from basic sciences department including one from Community Health and Epidemiology.
   • The expertise of the committee should include
     1. biomedical engineering and computer science which includes highly technical hospital based specialties such as radiology, laboratory medicine and anaesthesiology
     2. imaging and signal processing
     3. use of computers in clinical care
     4. date base management including clinical and administrative information systems

   Ex-Officio
   Administrative Manager of Medical Computing and Media Services Medical Informatics Coordinator

   The committee will invite representatives of external groups to participate with the committee on an as needed basis.

   REPORTING:
   The Medical Informatics Steering Committee will report directly to the Dean.
   
   

   
   

   Appendix 2

   Related recommendation from "Strategies for the 1990s": Faculty of Medicine of Dalhousie University"

   #2. That graduating physicians have the knowledge and skills to be responsible for, and committed to, life-long self learning.

   #3. That the faculty continue to develop computer-assisted learning systems for health education.

   #6. That the faculty plan for a continuum of medial education curricula from undergraduate through postgraduate to continuing medical education.

   #7. That the faculty strengthen its research effort in medical education with an emphasis on program evaluation.

   #8. That the faculty place continuing emphasis on the fulfilment of a commitment of clinical or basic research during residency training in all specialties and sub-specialties.

   #12. That the faculty explore the concept of developing postgraduate training programs of excellence with a national and international reputation.

   #16. That the faculty participate with the provincial licensing agencies and the medical societies to develop a validated, peer-review process complete with remedial processes for those who require them.

   #23. That efforts be increased to convince the provincial governments and hospitals of the region of the necessity of supporting and directly funding local medical research efforts and to identify new source of research funding.

   #24. That the faculty, together with others, provide adequate and appropriate support to initiate and sustain both basic science and clinical research at Dalhousie.

   #25. That the faculty ensure that physicians of the future develop specific skills in the use of new information technology.

   #34. That the faculty maintain and enhance its collaboration and working relationships with the departments of health of the three Maritime Provinces to ensure [...] the faculty can meet its commitments.

   #35. That the faculty maintain and enhance its collaboration and working relationships with affiliated teaching hospitals by establishing new mechanisms for functional collaboration on an operational basis.

   
   

   ---

   
   

   Appendix 3:

   Responses to January 13, 1995 request to identify "Priorities for Medical Informatics"
   1. Dr. Brian O'Brien, Medicine

      - medical records, with emphasis on a new medical record for the informatics age rather than just a digitization of the present form

      - continuing education at a distance

      - ransfer of images to allow consultation

      - continuing the inclusion of informatics concepts in the undergraduate and postgraduate curricula.

   2. Dr. Jean Gray, Medicine

      - Research and Evaluation Methods in Medical Informatics--and the responsibility to build in an evaluation component of the development process in order to demonstrate accountability.

      - Good set of evaluation tools is needed, as an incentive for industrial support of our programs.

   3. Dr. Jock Murray, Neurology

      - personal-information needs, utilization of data and formats within my own computer and on personally owned discs, databases, etc.

      - inks with major on-line services, Medline, Grateful Med.

      - communications through e-mail and services like Compuserve

      - Internet

      - research

   4. Dr. Grace Johnston, School of Health Services Administration
      undergraduate program (BSc) in health informatics, which would be an integration of Gloria Coughlan's program at the Camp Hill Medical Centre's School of Administration (e.g. medical records), research and computer skills at School of Health Services Administration.

   5. Dr. Mike Murphy, IWK Emergency
      opportunities for clinical faculty, such as clinical fellowships and sabbaticals in medical informatics.

   6. Dr. Martin Pusic, IWK Emergency
      educational program, such as sabbaticals in medical informatics or educational brokering, so that those faculty who wish to be trained at established medical informatics programs that are recognized internationally can do so, and then come back to Halifax and do peer training.

   7. Dr. John Finley, IWK Cardiology
      In general, the outline in VoxMeDal article on training future physicians appears appropriate. The uses of the Internet must be explored, and brainstorming sessions around the medical school and hospitals must be held to look ahead. We cannot depend on information services for this.

   8. Dr. Tim Pollak, Medicine
      Provide equitable access to the Internet for all academic faculty, regardless of where their offices are. Hospital-based faculty are governed by hospital policies, and their needs as academics are not taken into consideration, so such things as access to the Internet are filtered and restrictions imposed.

   9. Ron MacKinnon, Professor, St. Francis Xavier University
      Nursing informatics program at St.F.X. makes extensive use of the Internet. Interested in cooperation.

   10. Bryan Lynch, Professor, St. Francis Xavier University
       HUMAN--Grassroots information service sponsored by Human Resources Canada, is expanding into a support network for the St. Martha's Regional Hospital. Collaboration between regional hospital and metro hospitals can be done using the Internet for communications.

   11. Dr. Randy Giffen, Middleton, NS
       Dalhousie has an opportunity to play a leadership role in medical informatics in Canada. The Faculty of Medicine should open a Department of Medical Informatics with staff to provide undergraduate and graduate education. The postgraduate training should include master's and doctorate programs and fellowship programs for MDs. Medical informatics is changing the way health care is provided. We cannot afford not to play an active role in its development.

   12. Dr. Winston Dykeman, Atlantic Institute of Clinical Computing, NB
       In response to his cross-Canada query "Are any Canadian universities offering postgraduate programs in Clinical Informatics?", he discovered there appears to be no formal Canadian program anywhere. Strange, for a nation that struggles to develop a more efficient health system that our national vision of helping the actual caregivers develop more efficient tools is overlooked.

       With respect to developing computer modules that integrate patient care with the National Clinical Practice Guideline on Diabetes Mellitus: computerized management plans that embody a national guideline

       1. can be fun and fast to use;
       2. enable rapid assessment of patient response to treatment;
       3. enable a total practice evaluation of response to diabetes management; and,
       4. automatically produce epidemiological data sets that evaluate the effectiveness of the guideline in the first place.

   13. Beth Taylor, Med I
       Circulated survey to students on their educational needs in medical informatics and is working on an Internet "home page" so that students will learn to use the Internet in much the same way as they use the library to address an information need.

   14. Bruce Ramsey, Med III
       Access to Tupdean1, the network server for the computers in the Medical Computing Lab, from the hospitals. As clerks, and probably as residents, students have little time to run over to the Tupper Building. Until we have complete interhospital and Tupper connections, we will not really be using the full potential of telecommunications. Why are we so far behind other medical schools?

   15. Dr. Aidan Stokes, Psychiatry
       The thrust towards education should not be towards the students but should include and even perhaps be primarily directed towards faculty. While faculty may be familiar with wordprocessing and information retrieval, they are not as comfortable with graphics, database management and using the Internet.

       Computer-based learning has not attracted the interest that we thought it might, and that may reflect the unfamiliarity of the faculty rather than that of the students with such learning.

       We have to train our physicians to use computers in patient management and computer-based hospital-information systems

   16. Dr. Michael Moss, Pathology
       Opportunities for exposure to

       - electronic communication

       - access to international electronic databases

       - image processing

       - authorware training

       - rule-based decision-support software

   17. Dr. Meng Tan, Medicine

   18. Dr. Karen Joughin, Surgery, Saint John Regional Hospital, NB
       Extend the medical informatics lecture and workshop series to New Brunswick faculty.

   19. Dr. Peter O'Hanley, Emergency Services, The Moncton Hospital
       Most practicing physicians today graduated before attaining any facility in electronic information management, but realize they must attain this facility to keep current

       Non-electronic ways of introducing the spectrum of abilities for electronic presentation and manipulation of data and medical information

       Explore the possibility of establishing a simple bulletin board service available to practitioners in the Maritimes at Dalhousie, where they can ask specific questions for consultants, or as a place to touch base on guidelines and current practices (e.g. the Drugs and Therapeutics newsletter published in Halifax).

   20. Dr. Gerald Klassen, Cardiology
       We are rapidly entering the digital era. Digital information brings with it enormous advantages but also unique problems. The power of large databases in health-care research raises issues of confidentiality which needs to be addressed. As well, issues of justice occur in the maintenance of the integrity of information as it relates to both the preservation of information for future use and the detection of fraud when data is manipulated. All of the above are important in maintaining both benefit and prevention of harm to individuals, as well as caring and respect.

   
   

   ---

   
   Other correspondence
   1. Letter of December 13, 1994, to Dean Ruedy with cc: Armand Pinnard, Deputy Minister, Department of Health, requesting suggestions of ways to facilitate collaboration between health-care institutions and the university community.

   2. Letter of January 13, 1995, to Department and Division Heads, DMSS Executive, and Interns and Residents, requesting input on Priorities in Medical Informatics.

   3. Letter from D. Zitner, Chair, Camp Hill Medical Centre Medical Quality Improvement Committee, to HMRI (Health Records Management Institute) board members, re why HMRI information is crude and unreliable. Included attachment of Resource Intensity Weightings

   4. Letter of Sept. 14, 1994, from D. Zitner, Health Issues Council, to Ms. L. Dobbin, Deputy Minister, Nova Scotia Department of Health, re evaluation issues.

   5. Memorandum from D. Zitner, Camp Hill Medical Centre, to Mr. Dan Rice, Department of Health and Dr. Judy Kazmiriski, Chair, Health Issues Council, Medical Society of Nova Scotia, on the Health Information Systems Visioning Workshop held January 12-13, 1995, sent January 16, 1995.

   6. Memorandum of January 10, 1995, to IS Strategy Workshop participants from Dan Rice, DOH Project Director, with the list of NEEDS and ISSUES identified by participants at the nine focus groups held in early December. Sample Application Groups: Summary Descriptions from January 12-13, 1995, Department of Health Information Systems Visioning Workshop.

   7. Memorandum from D. Kaufman to Dean Ruedy and Dr. Karen Mann, on Department of Health Strategic Planning Session, December 11, 1994, and handout on "Information Technology and Health Care": the big picture in health care and the big picture in information technology.

   8. Comparison of CIHI, MedisGroup and CareTRACK by Property characteristic based on the installation of CareTRACK in 4D at Camp Hill Medical Centre at the beginning of March 1995.

   9. Correspondence from M. Horacek on Powerful Server.

   10. Computers in Medical Care Survey, provided by Dr. Charles Friedman, Paul Cudmore lecturer January 1994, and used at six medical schools in the U.S. (e.g. with Dr. Bill Detmer, Stanford)

   11. Outline of Medical Informatics as an Academic Discipline workshop given by Dr. Charles Friedman, January 1994.

   12. Population Health Research Unit (PHRU) Terms of Reference, January 16, 1995
   
   

   ---

   
   Articles Circulated
   1. Frisse, M. E. Medical informatics [editorial]. Academic Medicine, 1995; 70(1): 30.

   2. Frisse, M.E., Braude, R. M., Florance, V., Fuller, S. Informatics and medical libraries. Academic Medicine, 1995; 70(1): 30-34.

   3. National Library of Medicine Extramural Programs: Research Training in Medical Informatics, U.S. Department of Health and Human Services/Public Health Service/National Institutes of Health.
   
   

   ---

   
   

   Appendix 4:

   It should be apparent from the following section that there is considerable expertise within Faculty of Medicine in such important areas of informatics as biomedical-systems technology, image and signal analysis, statistical techniques and modelling, medical-decision-support systems, and education and training.

   4.1 Image and signal analysis

   IBM Scalable Parallel System SP2, which is accessible via Ethernet connections to the VGH, allows researchers at the Faculty of Medicine to deal with computationally intensive medical-imaging applications. Dr. Alan Fine in the Department of Physiology & Biophysics has developed two such applications in the field of neuroscience: optical methods with high spatial and temporal resolution for the study of the mechanisms of synaptic function and plasticity; and mechanisms and treatments of neurologic disorders such as Alzheimer's disease, Parkinson's disease, and stroke. Computational methods for multimodal merging of images of the same subject will be developed on the SP2 system. Such multimodal data sets of normal as well as diseased brains will become part of a crucial database for future investigations that will, through Ethernet links between the VG Hospital and Dalhousie University, be accessible to a large community of basic and clinical investigators.

   An example of the extensive study that involves signal analysis, is the identification of postinfarction patients who are at risk for life-threatening ventricular arrhythmias, by an assessment based on the computer analysis of 120 simultaneously recorded digital electrocardiograms. This study was performed by Dr. Milan Horacek's group on an IBM RS/6000 computer in the Department of Physiology & Biophysics, and clinical data were gathered in Cardiology Division of the Department of Medicine in the VGH (Dr. Martin Gardner) and in the Foothills Hospital of the University of Calgary (Dr. L. Brent Mitchell). Dr. Milan Horacek's laboratory has also developed a complete computer system for acquisition, processing and display---to be used in the Electrophysiology Laboratory of the VGH---which will guide a cardiologist during radio frequency ablations that are to be performed in patients with preexcitation syndromes and/or life-threatening ventricular arrhythmias. This application is of great scientific and clinical importance and can be only achieved through accessing powerful server computers linked to the Electrophysiology Laboratory by Ethernet network.

   4.2 Decision-support systems

   In the Department of Family Medicine, Drs. Donald Fay and David Zitner have developed a PC-based decision program called CareTRACK, which allows clinicians and administrators to review health status and hospital activity for each day of a patients stay. CareTRACK identifies long-term care and home services which would be required in order to discharge patients from an acute care hospital into a setting of long-term care or home care.

   In the Department of Anaesthesia, Dr. Charles E. Hope directs several projects that deal with the assessment of information from multiple sources in an integrated fashion. This is directly applicable in the operating room, where physiological data from a number of monitors, have to be considered simultaneously.

   4.3 Health-care management

   The Department of Anaesthesia is also using administrative data files to evaluate factors that affect the efficiency of health-care delivery in operating rooms. For the purposes of this study, all data processing has been carried out by means of SAS running on an IBM RS/6000 computer at the UCIS or on an IBM RS/6000 Model 590 in the Department of Physiology & Biophysics, either of which can be controlled from a remote Zenith Data Systems Z-386/20 Internet node running FTP and Telnet under OS/2.

   4.4 Education and training

   The Division of Medical Education at the Faculty of Medicine, in particular Dr. David Kaufman and Grace Paterson, contributes to the promotion of medical informatics in medical education. Their activities include: directing and planning for the integration of medical informatics into the medical curriculum; development of modules for the community-access network; and faculty development in medical informatics. In addition, the Division of Medical Education contributed to the development of the evidence-based medicine curriculum for Family Medicine, and of telemedicine projects for remote consultations between physicians.

   From the multitude of activities in various departments, we list those that are in the advanced stage.

   Information retrieval and file management

   In the Department of Family Medicine, Dr. Stewart M. Cameron purchased a CD-ROM database which will provide a medical reference library for family physicians.

   Computer-aided learning

   In the Department of Physiology & Biophysics, Dr. Dieter Pelzer used a multimedia adapter board for a PC computer in developing computer-based simulations/animations of "difficult" concepts in undergraduate medical education.

   In the Department of Pharmacology, Dr. John Downie developed a Hypercard-based Macintosh program called MacDog Lab as a substitute for live-animal laboratories.

   In the Department of Pediatrics, Dr. Martin Pusic (who is now at Johns Hopkins University) has designed tutorials for self-study.

   Dr. A.D. MacKeen in the Division of Medical Physics of the Department of Radiology has amassed a wealth of teaching films on chest radiography. He is now producing---with Drs. Cupido Daniels and A.D. McKeen---a teaching program on interpreting chest radiographs that will reside on a compact disk (CD-ROM).

   In the Department of Psychiatry, Dr. Max Michalon plans to implement an expert system in psychiatry and patient-management simulations for undergraduate and postgraduate medical students.

   The Department of Ophthalmology, Dr. Ann Hoskin-Mott developed a prototype of the computerized model of the human pupil that will help in demonstrating pupil testing in the normal and abnormal state.

   In the Division of Nephrology in the Department of Medicine, Dr. Michael West has purchased a computer-aided learning software package---which runs on both IBM and Macintosh computers---on fluid, electrolyte, and acid-base evaluation (FEABE).

   In the Department of Pediatric Cardiology, Drs. Douglas Roy and John P. Finley have developed a computer-aided package for teaching the recognition of heart sounds during auscultation.

   In the Department of Anatomy & Neurobiology, Dr. Ian Mobbs has established a workstation that is dedicated as a self-evaluation and review facility for use in the Med I COPS Human Body Unit.

   In the Division of Medical Physics in the Department of Radiology, Drs. Cupido Daniels, Sian Iles, and Gerald Schaller developed a mammography teaching file on CD-ROM.

   In the Department of Anatomy & Neurobiology Dr. Peggy Hansell has used MedPics. Pictures of typical histological tissues are displayed on the computer monitor.

   In the Department of Anatomy & Neurobiology, Drs. Howard Ellenberger and Frank Smith is also developing computer-aided drawings from digitized images for a neuroanatomy lab manual.

   In the Department of Pediatrics, Dr. Mark Ludman has purchased the Visual Genetics software package.

   Patient management

   In the Department of Pediatrics, Dr. John LeBlanc participates in the Childhood Injury Research and Prevention Program (CHIRPP) --- a surveillance system based in nineteen hospitals in Canada.

   In the Department of Surgery (Pediatric Orthopaedics), Dr. Lorne Leahey educates students in the use of computer-based information-storage-and-retrieval technology to methodically record the diagnosis, treatment, and complications for the patients under the doctor's care.

   In the Department of Family Medicine, Dr. David Maxwell is developing several extensive databases.

   In the Department of Family Medicine at the Camp Hill Medical Centre, Drs. David Zitner and Donald Fay have developed software to help detect instances of inefficient care while care is being delivered. The module is the first in a series designed to provide concurrent information about patient comfort, function, life expectancy, and the delivery of unique or nonunique hospital services.

   The big picture

   The Atlantic Canada Organization for Research Networking (ACORN) is a partnership between the government, universities and private sector, proposing a joint venture to establish a high-speed test network in Atlantic Canada in association with the Canadian Network for the Advancement of Research, Industry & Education (CANARIE Inc.).

   This will allow industry collaboration, support for the information technology sector, and support for multimedia applications which require high-speed network structures. Some of the major applications areas that can drive ACORN initially are telemedicine, distance learning, virtual libraries, a distributed "metacomputer," medical imaging, decision-support systems, personalized multimedia electronic news, and education.
   

   ---

   
   

   Appendix 5:

   Computing environment of the basic-science departments

   In contrast to the unsatisfactory situation in teaching hospitals, the computing environment of the basic-science departments of the Faculty of Medicine developed steadily (mainly due to the influx, albeit modest, of external funding) and it became an integral part of the campus-wide high-speed Ethernet network, which is maintained by University Computing and Information Services (UCIS), and is connected to the world through the Internet. The Ethernet serves most major buildings on campus, including the Sir Charles Tupper Medical Building, and the Clinical Research Centre. Users of the network gain direct access to all the campus server computers and services (including the computers of UCIS and of the Department of Physiology & Biophysics, and LAN servers in Medical Computing and Media Services), and through them to the worldwide Internet.

   Shared research-computing resources of the Faculty of Medicine, all accessible to Ethernet users, are concentrated in the Sir Charles Tupper Medical Building --partly in Medical Computing and Media Services (MCMS) and partly in the Department of Physiology & Biophysics. The former unit offers customer-oriented service, providing electronic-mail services, information storage, and software and technical support; the latter provides state-of-the-art facilities and expertise in scientific computing.

   MCMS operate Tupdean1 Novell server, MCMS1 network file server and Image server, which are used by over 600 clients, including medical students, faculty, staff, and researchers throughout Faculty of Medicine, Dentistry, Health Professions, the Grace/IWK Hospital, the Weldon Law Building (Health Law), and the Environmental Health Clinic at both the Bethune Building and the Fall River site. The average user count on the system is between 100 and 120.

   The Department of Physiology & Biophysics operates Tupphysiol1 Novell server, which can be regarded as paradigm of departmental networking for the very near future. (Available software includes RIS Reference Update, Journal Abstract service etc.) In addition, high-performance research computing is concentrated in the Department of Physiology & Biophysics; available hardware includes IBM RS/6000 Model 590 (the most powerful machine for scientific computing in the Faculty of Medicine), Stellar GS1000 graphics supercomputer, DEC MicroVAX 3400 computer, and Silicon Graphics IRIS 4D/320VGX. Other computers in the Department of Physiology & Biophysics include IBM RS/6000 Model 320H workstation, DEC Alpha Model 3000/300XL, and two MicroVAX 3100 workstations. An enormous boost to high-performance scientific computing is the recently acquired IBM Scalable Parallel System SP2, operated by UCIS and accessible to users from the Faculty of Medicine by virtue of the DMRF's contribution towards funding this system. This powerful facility is capable of delivering up to one billion calculations per second.

   Workstations specialized for specific tasks have been developed by various research groups at the Faculty of Medicine; such workstations can be replicated for other similar applications, thus reducing the cost of development. For instance, real-time acquisition and processing of physiological data is an important requirement in many research laboratories and in clinical care. Often, there are commercially available dedicated computer systems or workstations that are ideally suited for a given application (such as the Axoclamp system for membrane electrophysiologists or anaesthesia workstations by North American Drager), but in some instances front-line research requires systems that have to be custom designed. Two custom-designed state-of-the-art acquisition systems have been developed in the Department of Physiology & Biophysics for clinical applications. The first one is a digital-data-acquisition system, based on a standard PC, for reading data from standard commercial monitors. (This system takes advantage of the fact that digital data ports are now standard feature of the patient-monitoring equipment.) The second one, also based on a standard PC, has been developed for high-performance 128-channel acquisition required for electrocardiography research. (The "front end" of this system is located in the Catheterization Laboratory or in the Electrophysiology Laboratory in the VGH, and is connected by a single optical fibre to the host computer, which records continuously 256 kilobytes of data per second for as long as is desired. This acquisition system communicates over the Ethernet with the IBM RS/6000 server system in the Department of Physiology & Biophysics.)
   

   ---

   
   

   Appendix 6:

   Current Canadian clinical information systems

   Starting April 1 the collection of information for the Canadian Institute of Health Information (CIHI), formerly HMRI, will be mandatory for Nova Scotia hospitals.

   The CIHI system is unfinished and probably unreliable. Consequently, many health professionals fear that the current CIHI method will be misused and lead to unreasonable funding decisions. Communities and Medical Faculty should be concerned about an information system which seems to favour acute hospital care for healthier people but not for sick or complex patients.

   Briefly, the CIHI method measures the length of stay for hospitalized patients with a particular diagnostic label and allows each institution to compare its performance with that of other hospitals treating patients with the same label. For CIHI the most responsible diagnosis is the diagnosis which was responsible for the greatest resource use during an admission. Most responsible diagnosis is an economic label not a medical label -- it refers to the diagnosis responsible for the greatest resource use during a particular hospital stay.

   The diagnostic labels are chosen by Health Records Professionals who are required to decide which particular diagnosis was responsible for the greatest resource use. Often the label chosen by health records is the same as that chosen by medical staff; sometimes the label is different.

   The chart analysis is done after the patient has been discharged so CIHI system cannot modify care while it is provided.

   Hospital care, provided by our faculty will be evaluated using CIHI methodology. Starting April 1 all hospitals in NovaScotia were required to submit information to CIHI. Consequently, it is important that each Clinical Department become familiar with the methodology and participate with hospitals in trying to develop a superior method.

   CIHI does not address the important results of care and so cannot provide estimates of value.

   Problem premises for CIHI

   • A premise for CIHI is that the output of a hospital is the number of people discharged with a particular diagnosis. In fact, we hope, that hospital care changes health status. CIHI systems do not measure changes in health status and healthy patients with a particular set of diagnostic labels generate the same resource intensity weighting as sicker patients with the same diagnosis. Hospitals which admit healthier patients appear more efficient compared with hospitals treating sick patients with the same set of diagnostic labels.1

   • Funding by diagnostic labels means that for each hospital the least severe examples of a disease become the economic cream while the most complex cases become an economic burden.

   • The system does not raise questions of value. Costs in health care are appropriate when a patient benefits or is likely to benefit. The current CIHI system has no measure of benefit.

   • For patients the real output of hospitals are
     1. Increased life expectancy
     2. Improved function
     3. Increased comfort
   • Funding by diagnostic label does not make sense. Having a particular diagnosis is usually neither necessary nor sufficient to justify admission. For example, many patients with diabetes and/or congestive heart failure are treated as outpatients. The label congestive heart failure does not justify a hospital stay.

   • The appropriate length of stay in hospitals is closely related to the community in which that hospital is situated. CIHI does not report on community services available to participating hospitals.

     For example, the availability of home oxygen or home I.V. therapy has an important influence on length of stay. When they are unavailable we would expect lengthier stays.

   • Hospitals with astute health records reviewers appear more efficient than hospitals with less skilled health records employees. Under the new CIHI system hospitals whose health records reviewers successfully capture a large number of labels will appear more efficient than those whose health record reviewers capture fewer labels.

   • Our Medical Faculty must participate in the development and use of modern health records so that communities can properly assess the value of care. Each health record should explicitly define the benefits or losses which resulted from a hospital stay.

   • For information, Dr. Lorne Leahy of the Orthopaedic service at the IWK has developed a system which captures information about the results and complications of care.
   
   
   CIHI information may be unreliable
   • Often, health record reviewers cannot identify the diagnosis responsible for the greatest resource use. A CIHI reabstracting study done using 1988-89 data found that for medical patients health record reviewers agreed on the most responsible diagnosis 80% of the time. When more than one diagnosis was considered they agreed 40% of the time. Although reabstracting studies are planned none have been reported since. For complex medical patients physicians may not be able to identify a single condition which was most responsible for admission. Of course if forced to choose from several alternatives they will make a guess. Moreover, many administrators cannot articulate which of several medical conditions used the most resources. In many cases the task for health record reviewers is impossible.

   • CIHI corporate policy has been that each institution is responsible for data quality. Consequently, different institutions might use different coding rules.

   • Hospitals are not consistent in labelling days as "alternate level of care" (ALC) so length of stay statistics do not reflect actual days in hospital.

   • Reliable information is expensive. Most hospitals fail to include quality control procedures in the health record review process.

   • Incidentally, the newest CIHI method claims to adjust for complexity by collecting information about the number of diagnoses attributed to each patient. The new method indicates that clinical faculty at Camp Hill Medical Centre achieve substantially shorter average length of stay than the Canadian average (for patients with a similar number of diagnostic labels). Similar data is not yet available for other Dalhousie affiliated teaching hospitals.
   
   

   ---

   
   

   Appendix 7:

   Health records engineering: by Records Review Task Force of Queen Elizabeth II Health Science Centre

   This multidisciplinary, multisite group will make recommendations regarding the organization and content of health records.

   The thinking of this group is that a Health record has two major components. One component is the working papers and encounter record. The second comprises a set of conclusions regarding a patients health status, the goals of treatment and response to treatment.

   The following constitutes conclusions already reached by the Record Review Task Force. Work is ongoing and the conclusions are not final.

   1. The current health record requires modernization.

   2. A functional health record should provide information about a patients health status at admission and discharge.

   3. The health record should document important changes in health status as they occur.

   4. We agree with the new accreditation requirements which suggest that the health record must document the results of care.

   5. Determining health status requires information and conclusions about
      1. Patient function
      2. Patient comfort
      3. Disease severity (Life Expectancy)

   6. Changing health status is the main goal of health care. However, providing information about health status is also worthwhile. For example, advice that chest pain is self limited chest wall pain not cardiac is useful.

   7. Much of the written material in a chart constitutes an encounter record and working notes.

   8. The encounter record and working notes should lead to conclusions about health status and a strategy for care.

   9. The conclusions about health status and the strategy for care should be readily available and accessible to all disciplines involved in the care of a particular patient.

   10. The health record is similar to the minutes of a committee meeting. Many people participate with the common goal of improving the health status and life situation of a particular patient.

   11. The information which leads to a conclusion can come from a variety of sources and disciplines.

   12. We agree with the conclusions of the (1992) integrated charting group at the Nova Scotia Rehabilitation Centre which suggested that Duplication of Information is evident across all disciplines and that we should try to avoid duplication of effort.

   13. The conclusions reached by each discipline should be readily available to all without the need to sift through large amounts of encounter material or working notes.

   14. Conclusions can take a variety of forms Examples include
       • To cure congestive failure
       • To improve exercise tolerance so the patient can walk up one flight of stairs
       • To reduce pain
       • To involve specific community supports in order to permit a patient to live at home.
       • To teach patient skills required to get in an out of a bathtub so they can manage without additional community supports.

   15. Many disciplines can provide the information required to reach a conclusion but some conclusions might be discipline specific.

       We have not yet decided how to organize the chart so that the conclusions regarding health status and treatment are easily available but we agree that the information should be easy to obtain.

   16. Ideally, a computerized health record should be organized to provide immediate information about health status, goals of treatment and treatment modalities.

   17. Once we have described the essential chart elements other groups, with specific expertise, should better define those elements.

       For example, what demographic information is required for each patient. How will each program define health status. What is the data set required for patients in particular programs. An ophthalmology record might have the same form as a psychiatry record, but the elements of function which are recorded might be different.
       

       ---

       
       

       Appendix 8:

       WWW links to medical informatics
       1. Medical Informatics Cultural Literacy Project (MICL)

          Medical Informatics Cultural Literacy Project.
          Working copy of W. Hammond's GLOSSARY FOR HEALTHCARE STANDARDS.
          What is Medical Informatics?
          Descriptive Literature about Medical Informatics.
          

          - [596] http://dmi-www.mc.duke.edu/dukemi/essent/miclp.html

          
          
          
       2. Medical Informatics FAQ

          Archive-name: medical-informatics-faq Posting-Frequency: monthly
          Last-modified:
          1995/03/27 Frequently Asked Questions (FAQ): Medical Informatics,
          sci.med.informatics This document is intended to answer some frequently
          asked questions
          about
          

          - [596] http://www.faqs.org/faqs/medical-informatics-faq/(7K)

          
          
          
       3. Southeastern Medical Informatics Conference

          Hosted by the University of Florida Health Science Center. Saturday, June 10, 1995. A conference designed to enable networking between developers, researchers, engineers and clinicians working in Medical Informatics.

          
          
          
       4. Medical Informatics FAQ

          Newsgroups: sci.med.informatics, sci.med, sci.med.telemedicine ,sci.med.radiology, sci.med.pharmacy, sci.med.nursing, sci.engr.biomed ,sci.med.pathology, comp.lang.mumps, comp.protocols.dicom , misc.education.medical ,sci.answers , ...

          - [595] http://www.lib.ox.ac.uk/internet/news/faq/archive/medical-informatics-faq.html (8K)

          
          
          
       5. University of Missouri-Columbia Medical Informatics Group

          Go to Medical Informatics at the University of Missouri-Columbia. Other Medical Informatics Sites: Complete List of Biomedical WWW Sites. Medical Informatics and Medicine: Some Useful Links. Einet Galaxy--Medical InformaticsGroups ...

          - [595] http://www.hsc.missouri.edu/mig/(2K)

          
          
          
       6. Medical Informatics Training Environments

          This is a selected list of leading medical informatics training environments in the United States. Program director, mailing address, phone and other information is provided where known. Links to the institution may not be to the informatics ...

          - [595] http://www-smi.stanford.edu/projects/smi-web/academics/informaticsprgms.html (5K)

          
          
          
       7. University of Limburg - Dept. of Medical Informatics

          Faculties of Medicine and Health Sciences.

          
          
       8. Medical Informatics Home Page

          Medical Informatics Research Group.

          - [594] http://www.cc.gatech.edu/gvu/biovis/visual.html (3K)

          
          
          
       9. Informatics and Imaging

          The National Library of Medicine. HyperDOC. A multimedia/hypertext resource of the NLM. Educational Technology Branch. Computer-based education in the health sciences. Information on Grateful Med. Grateful Med is a program for ...

          - [594] http://www.mcl.tulane.edu/informatics/informatics.html (2K)

       Footnote
       1. A study performed for the Department of Health reviewed admissions for a hospital which seemed to have the shortest length of stay and lowest death rate for several diagnoses. The study was done in collaboration with Vangie Brown from the Department of Health and Krysti Snook - Health Records at Camp Hill Medical Centre showed that many patients did not require admission and did not receive services which were unique to a hospital.