1991 call for nmajor changes in healthtcare, and new technologies offer the power to make ... Association (IMIA) to hold a working conference on methodological ...
Hospital Information Systems (HIS): Looking into the Future Marion J. Ball, Ed.D. Judith V. Douglas Abstract Hospital infomtation systems have to date failed to live up to the expectations of healthcare providers and informaticians. Events in the summer of 1991 call for nmajor changes in healthtcare, and new technologies offer the power to make thte computerbased patient record a reality.
Introduction This past summer saw three major events which will help to define the future of hospital information systems (HIS). The Institute of Medicine (IOM) prepared to release the recommendations of its national committee on the electronic patient record. Dr. Larry Weed updated his vision of how the computer can transform patient care. And, under the auspices of the International Medical Informatics Association (IMIA), the Working Group on Hospital Information Systems met in Gottingen, Germany; this meeting was preceded by meetings in Capetown, South Africa (1979) and Nijmegen, The Netherlands (1988). As SCAMC meets in November 1991, the content of and reactions to these events will be available for public discussion. Debate is inevitable-and welcome. The Computer-Based Patient Record The Institute of Medicine (IOM) report, which will set the course for the nation's health care system over the next 10 years, is expected to advocate replacing the traditional hospital-based medical record by the electronic patient record. The record will thus become longitudinal, carrying information over time; interdisciplinary, including information from varied health professionals; and centered on the patient rather than on an individual or an institution providing healthcare. By supporting the provision of care in multiple settings, including the doctor's office or patient's home, the electronic patient record will encourage less invasive (and less expensive) care. The 1OM recommendations are detailed in Thte ComnputerBased Patientt Record: An Essential Techtnology for Healthi Care [1]. The lengthy deliberations that resulted in those recommendations and are reflected 0195-4210/91/$5.00 © 1992 AMIA, Inc.
in the position papers by national experts now in press with the working title of The Comnputer-based Patient Record: An Informatics Imperative [2]. Underlying the electronic patient record is the problem oriented medical record, developed over 20 years ago by Dr. Larry Weed. In Knowledge Coupling [3], Weed sets forth another revolutionary concept, already being implemented by a few forward looking practitioners. This concept involves using the computer for diagnosis and management, in the process Weed calls problem knowledge coupling, which frees the health care provider from the limitations of memory and allows him to consider all possible causes of identified physical findings or problems. Clinical judgments are left to the physician, not to the probabilistic models; the patient assumes an active role throughout and retains his own record. Assessment and Outcome Studies In the fall of 1990, the International Society for Technology Assessment in Health Care (ISTAHC) joined with the International Medical Informatics Association (IMIA) to hold a working conference on methodological approaches of medical technology assessment, with special emphasis on informatics applied to medicine and health. The program stated the aim of the conference as follows: First, to develop a dialogue between the fields of Medical Informatics and Health Technology Assessment in order to share current states of progress and to build an agenda for future work and collaboration; secondly, to issue recommendations about the methodological requirements for evaluating Health Information Systems. The proceedings [4] from that conference, entitled Assessment of Medical Inzformatics Technology, will have major impacts on both fields. Evidence is clear that assessment must also extend to medical therapies. A number of studies have found that the human costs of the incorrect, suboptimal, or inappropriate use of therapeutic agents exceed the total of all other complications resulting from hospitalization in terms of morbidity and
224
mortality. Clearly, "the economic consequences are enormous [5]." Today, at the insistence of those who pay for health care (government, insurance carriers, employers), outcome studies are receiving increased attention. For all their appeal, outcome studies are difficult and possibly premature, as Weed believes [3], Whatever the case, outcomes cannot be studied unless inputs are well defined and carefully controlled. However, The coming together of expert systems software, knowledge acquisition tools and, very importantly, Hospital Information Systems (HIS) or Departmental Information Systems (DIS) that contain structured information of many types (demographic, clinical laboratory, pharmacy, imaging, etc.) now permits real-time evaluation and reevaluation of therapy in light of new data and overcomes the structural shortcomings of earlier systems. The integration of data made possible by the HIS or DIS, is ideally suited for decision-support systems which evaluate therapy [5]. The future will be much as Weed has envisioned. Computerized record systems which make the computer "a guidance system for coupling the right information to the right problems" will transform health care [6]: Thoughtful physicians will not want to be unchallenged "brokers of the health care system" when everyone-patients particularly-can thrive within a defined system of health care with corrective feedback loops based on educational principles that can be believed in, and based on powerful new tools that couple the best in current thinking with everyday actions. Expectations for HIS In 1988, the IMIA Working Group set forth basic guidelines in Towards New Hospital Infornmationz Systemns [7]. Hospital information systems (HIS), they agreed, should * Include image handling as well as information processing for both medical and administrative management * Provide interfaces to the world outside the hospital * Emphasize patient care, until now underserved.
The 1988 IMIA group affirmed Collen's basic objectives as functional requirements for HIS: * computer-stored medical records accessible to all authorized providers at all times communication across all hospital units and to all affiliated facilities * all healthcare provider functions, e.g., order entry, etc. * clinical and administrative decision making * administrative and business functions * quality assurance, accreditation, and regulatory requirements * research and education. The group did not take exception to Collen's conclusion that "The basic objectives which were sought in the 1970s have remained relatively unchanged in the 1980s." They judged that HIS, as part of the hospital infrastructure, were not easily altered or abandoned. Later, in 1990, Wiederhold and Perrault [8] noted that "The capabilities of the typical HIS lag far behind those of the most sophisticated HISs in use today." The New Technology Base The continued evolution of healthcare information systems is predicated upon technological advances, notably systems and data architecture and communication standards. Whether the highly touted Health Level 7 (HL7) or less costly communications protocols claim the market, the move away from proprietary systems and into international standards is what makes distributed computing possible. Whether or not central resources and a mainframe are involved ceases to be the issue, although many hospitals retain some dumb terminals and other vestiges of centralized computing. In the distributed model, the network becomes the computer. Linkages between different hardware and software platforms become transparent to the user, whose workstation gives access to multiple machines, databases, and applications. Thus, as the 1988 IMIA working group observed, it is no longer a strategic issue whether an HIS is "implemented on one computer configuration or on a number of interlinked configurations [7]." What is of prime importance to health care is the integration of multiple functions which these technologies make possible. Yet the promise of these new technologies can be realized only if funding, management, and institutional politics allow. New structures and new strategies must be developed to accommodate and to optimize what technology provides. The fact is that the technologies capable of supporting health care already exist. What remains is to apply them [9]. Figure 1.
225
Figure 1. Healthcare in the Year 2000 and Beyond: Summary of Healthcare Information Systems Trends Type ofFunctional Needs
Available Tools to Accomplish
Today's Level of Accomplishment
Time Needed to "4-Star" Status
Quicker and uniform access to clinical information
Database organization SMART cards
0
3-5 years
Better clinical data collection for patient
Bedside/Point of care
*0
1-3 years
Digitize clinical data
Voice data entry Knowledge bases
*0
Over 5 years
Mass data storage
Optical disk
*00
1-3 years
Mass data retrieval
Optical disk/OCR
Patient tracking
Outpatient/Physician links HIS
0 SO
1-3 years
Networking/interfacing
HL7/Open architecture
0
3-5 years
User friendly database manipulation
Executive Information System Decision Support/SQL
*0
3-5 years
Improved instrumentation data collection/treatment/ diagnostic
Offshoots of cardiology and patient monitoring systems
0
Match resources to patients needs
Nursing management
Improved use of current functionality
Retraining on current system
Critical care system
3-5 years
|
l
Over 5 years
|00
1-3 years
0
1-3 years
*Poor (0) to Excellent
(0000)
Source: T.K. Zinn, 'Confronting 2000 and Beyond: Tomorrow's Healthcare Information System," Computers in Health Care, October 1990, p. 15. Figure copyrighted 1990. Zinn Enterprises. Ltd., Chicago, Illinois.
Point of care systems, whether bedside or handheld, offer valuable new capabilities for caregivers. First, they promise to improve nursing productivity by keeping the caregiver at the patient's side. One study [10J places the amount of time saved as high as an average of 1.5 to 2 hours per nurse per shift per day. Even if this claim is inflated, these systems can eliminate potential error by allowing the caregiver to record and enter data once, at the patient's side. Moreover, by giving information access to lab results, literature searches, drug databases, and longitudinal patient data, these systems can also support decision making at the bedside. Although only about fifty of the hospitals surveyed in 1988 had some form of point of care system installed at that time, many more were considering installation. Only 18% of the respondents were not intending to install such
systems; the remainder were planning to do so within the next three years. Healthcare
professional
workstations
complement bedside systems and further refine the concept of the hospital information system. In March 1991, IMIA approved a proposal, prepared by Dr. John Silva, to formally designate a working group on the professional workstation. This group will host an invitational international working conference in the near future. Among those who have advanced the workstation concept is Shortliffe, whose vision follows: Outpatient records will be integrated with inpatient data by using the capabilities of communications networks that link hospitals with the clinics and private offices of their
226
medical staff members. At the same time, the physician-not the hospital-will retain control of the outpatient records for the private practice [11]. With access through hospital networks to national networks, including those at the National Library of Medicine, practitioners will have access to information when and where they need it, from patient history and lab results, to literature assessment programs. The networked environment depends in large part upon automated digital medical records which are supported by the new infrastructure, including networking and improved storage and retrieval capabilities (e.g., picture archiving and communications systems). Other tools [12] promise to change health care. Barcode technology can ensure positive identification of patient, specimen, medication, and even provider; automated identifiers and time stamps can document process and outcome, ensuring accountability. Voice input capabilities can free practitioners from keyboarding and illegible writing. Imaging (refer to contribution on radiology), combined with Picture Archiving and Communications Systems (PACS), can make valuable information available without the problems associated with film storage and transport. Visualization can aid in surgical procedures, and graphics can present data in readily understandable formats. Specialized databases, accessed through networks, can provide caregivers with information on clinical trials, protocols, and treatment options. The technology is there. Holography and other new applications hold promise for the next century [13]. The task is to use information technology effectively-and to use it to do the right things.
Looking to the Future The question is, what are the right things? Predictions [9] hold that the 1990s will be the era of managed care, with more than half of the United States population enrolled in managed care programs, such as preferred provider and health maintenance organizations. This shift reflects the growing emphasis on measures of performance and return on investment. For vendors, this means quantifying market factors such as customer satisfaction. For health care facilities, it necessitates moving beyond a focus on JITS features and functions to documenting success through objective indicators.
As technology becomes more sophisticated and pervasive, management strategies must also develop. With the approval of IMIA, a working group conference in the near future, led by Dr. Nancy Lorenzi, will address the organizational impact of informatics. The outcome will be to define leadership characteristics and strategies for organizational change, direction, and structures for complex healthcare organizations. References
[1] Detmer D, ed. The Computer-Based Patient Record: An Essential Technology for Health Care. In press. Washington, D.C.: National Academy Press. [2] Ball MJ and MF Collen. In press. 77Te ContputerBased Patient Record: An Informatics Imnperative. New York: Springer-Verlag. [3] Weed LL. Knowledge Coupling. 1991. New York: Springer-Verlag, 1991. [4] Flagle C, F Gremy, and S Perry, eds. 1991. Assessment of Medical Informatics Technology. Rennes, France: Editions ENSP. [5] Blaschke TF. Hospital information systems and the quality of therapeutics. 1991. Methods of Infomtation in Medicine 29:163-166. [6] Weed LL. 1981. Physicians of the future. New England Journal of Medicine 304:903-907. [7] Bakker AR, MJ Bail, JR Scherrer, and JL Willems, eds. 1988. Towards New Hospital Iniformation Systemns. Amsterdam, The Netherlands: Elsevier Science Publishers B.V. [8] Wiederhold G and LE Perreault. 1990. Hospital information systems. In Medical Informatics, ed. EH Shortliffe, LE Perreault, G Wiederhold, and LM Fagan. Reading, Massachusetts: Addison Wesley. [9] Zinn TK. 1990. Confronting 2000 and beyond. Computers in Healthcare, October, 14-18. [10] Zinn TK. Healthcare information systems as they evolve to the year 2000. 1990. Computers in Healthtcare, Market Directory, 8-13. [11] Shortliffe EH. 1991. The networked physician. In Healthcare Information Management Systems, ed. MJ Ball, JV Douglas, RI O'Desky, and JW Albright. New York: Springer-Verlag. [12] Ball MJ, JV Douglas, RI O'Desky, and JW Albright. 1991. Healthicare Infomtationi Management Systems. New York: Springer-Verlag. [13] O'Desky RI, MJ Ball, and EE Ball. 1990. Computers in health care for the 21st century. Methods of Information in Medicine 29:158-161.
227
