Healthy Policy and Systems
History and Trends in Clinical Information Systems in the United States
Nancy Staggers, Cheryl Bagley Thompson, Rita SnyderHalpern
Purpose: To provide a synopsis of issues about clinical information systems for nurses not schooled in nursing informatics.
Organizing construct: The past, present, and future of clinical computing, including major factors resulting in the early hospital information systems (HIS) and decision support systems (DSS) in the United States, current advances and issues in managing clinical information, and future trends and issues.
Methods: Literature review and analysis.
Findings and Conclusions: The first HIS and DS.S were used in the late 1960s and were focused on applications for acute care. The change from feeforservice to managed care required a change in the design of clinical information systems toward more patientcentered systems that span the care continuum, such as the computerbased patient record (CPR). Current difficulties with CPR systems include lack of systems integration, data standardization, and implementation. Increased advances in information and technology integration and increased use of the Internet for health information will shape the future o f clinical information systems.
JOURNAL OF NURSING SCHOLARSHIP, 2001; 33:1, 7581. ©2001 SIGMA THETA TAU INTERNATIONAL.
[Key words: information systems, nursing informatics,
HIS, DSS, computerbased patient records/
During the late 20th century, information systems (IS) have become increasingly prevalent in healthcare. What were the influences that led to early advances in clinical computing? What are the current trends in clinical computing, and what can nurses expect in the future? This article is a synopsis of these issues for nurses not schooled in informatics. Although this discussion is specific to American health care facilities, the findings may apply in other countries as well.
Historical Influences of
Information Systems in Clinical Care
The development of 1S has been linked to changing needs in the healthcare industry. The U.S. healthcare industry in the 19th century was a charitable, communitybased effort to care for the sick and needy (Kissinger & Borchardt, 1996) and was comprised of standalone healthcare services for episodic, acute care. Patients' records were maintained by healthcare providers with little need to share information among providers or settings.
Throughout the 20th century, the healthcare industry expanded, fueled in large part by the federal government. In the 1950s, the emphasis was on increasing the number and quality of healthcare facilities. The HillBurton Act provided funds for this bricks and mortar development of facilities.
Then in the 19605, passage of legislation supporting Medicare and Medicaid increased access to healthcare and virtually guaranteed that healthcare settings would be amply reimbursed for their services (Kissinger & Borchardt, 1996). Throughout the era of expansion in healthcare, health records Continued to be paperbased and providercentered. Availability and assurance of a steady cash flow to inpatient facilities for providing care allowed healthcare providers, particularly physicians, to develop innovations such as new pharmacologic agents, advanced surgical procedures, and sophisticated diagnostic techniques.
The application of computer technology to clinical care was one of these innovations. Because of a stable healthcare environment through the 1960s and 1970s, and the increasing availability and lower cost for computer technology, physicians and others began investigating the use of this technology tee address healthcare needs. Early efforts were directed at developing hospital information systems and tools for decision support.
Nancy Staggers, RN, PhD, FAAN, Gamma Rho, Associate Professor; Cheryl Bagley Thompson, RN, PhD. Gamma Rho, Associate Professor; Rita SnyderHalpern, RN, PhD, C, CNAA. Gamma Rho, Associate Professor. All at University of Utah, Salt Lake City, UT. The authors thank Col. Bonnie Jennings, RN, DNSc, FAAN, for her thoughtful review of this manuscript. Correspondence to Dr. Staggers, College of Nursing. University of Utah, 10 S. 2000 E. Front, Salt Lake City, UT 84112. Email: nancy.staggers&nurs.utah.edu
! Accepted for publication August 18. 2000.
journal „t Nursing Scholarship t first Quarter Zoos 75
Clinical Information Systems in the United States
Hospital Information Systems
Hospital information systems (HIS) are large, computerized databases used to store health and administrative information. Early systems were focused on communicating orders for acute care and reporting results from ancillary departments such as pharmacy and laboratory. However, functions of HIS varied from institution to institution.
The development of one of the first systems, Technicon Medical Information Systems (now Technicon Data Systems or TDS), began in 1965 at El Camino Hospital in Mountain View, California, in conjunction with Lockheed Missiles and Space Company (Wiederhold & Perreault,1990). When TDS was first used on a patient care unit in 1971, it had capabilities for a suite of complex clinical, ancillary and administrative functions. Physicians' orders were communicated to ancillary departments, results were retrieved for laboratory tests and radiology reports, and nursing care planning and documentation were included (Barrett, Barnum, Gordon, & Pesut, 1975 ). According to a formal evaluation of the system, TDS supported substantially all information handling for nurses, physicians, and personnel in ancillary departments, including dietary, medical records, pharmacy, laboratory, radiology, respiratory therapy, and business offices (Barrett et al., 1975 ). The considerable functions provided by TDS were especially remarkable given the limited computing capabilities elsewhere in the nation.
Other HIS followed soon after (see Table 1). The HELP system at Latterday Saints (LDS) hospital in Salt Lake City, Utah, and the HIS at the National Institutes of Health Clinical Center in Bethesda, Maryland, were two of the most publicized systems, supporting a broad range of acute care functions, from order entry and results reporting to clinical care documentation. Early versions of HELP provided for bedside monitoring in intensive care units, laboratory automation, and interpretation of electrocardiograms. Beginning in the early 1970s functions were expanded to include additional portions of patients' records and decision support (Kuperman, Gardner, & Pryor, 1991). Capitalizing on the success of TDS at El Camino, the National Institutes of Health (NIH) staff also implemented a TDS HIS that had similar broad functions to support clinical care. This system was notable for its capability to track patients' problems of interest to nurses, patient care goals, and nursing orders (Cook & McDowell, 1980). The
framework developed by NIH nurses for nursing
documentation (Romano, McCormick, & McNeely, 1982) served the agency well for over 20 years.
Another well publicized information system is the Department of Defense's $1.6 billion Composite Health Care System (CHCS), developed in the 1980s as a follow up to the earlier ancillary TRIMIS systems. CHCS was expected to provide complex support for clinical care among ambulatory and acute care settings in more than 140 military clinics and hospitals. Because of funding constraints and legislation prohibiting the Department of Defense from developing computerbased patient records, CHCS was implemented with functions primarily supporting laboratory, radiology, pharmacy, and patientappointment scheduling (GAO, 1992). Capabilities for results retrieval and Email spanned the care continuum across inpatient and ambulatory settings. Unfortunately, CHCS order entry and patient appointment capabilities were usable only in ambulatory settings (General Accounting Office [GAO], 1992) and support for clinical documentation was never funded (Major Automated Information Systems Review Council [MAISRC], 1998). Similarly, the HIS at Regenstrieff (see Table 1) had support for order communication and ancillary departments, but nursing functions were not developed (McDonald, Blevins, Tierney, & Martin, 1988).
The early clinical computing efforts such as those listed in Table 1 were restricted to major tertiary care centers and large government projects. Their foci were on support to acute care, the mainstay of the 1980s when these systems were developed. Institutional HIS were developed or adapted over many years or decades. For example, the HELP system at LDS hospital in Salt Lake City, Utah, has been under gradual development since the late 1960s (Kuperman et al., 1991). CHCS was tailored for military healthcare during its 10year implementation beginning in 1988. Although functions varied, these early HIS were greatly modified to be integrated with unique care delivery methods, usually at one particular site.
Despite the success of these systems, support for clinical care and nursing practice was limited in the 1970s and 1980s (Ozbolt, Abraham, & Schultz, 1990). Hospitals' survival in this era was in feeforservice reimbursement; therefore, executives often implemented computerized billing functions first. Care functions were not a priority.
Table 1. Early Hospital Information Systems*
Date System
1965
1968
1960s
1970s
1968
1974
1970s
1970s
Technicon (TDS)
COSTAR
HELP
DHCP
TMR
Regenstrief
TriService Med. Information System (TRIMIS)
PROMIS
Hospital/Organization
El Camino Hospital
Massachusetts General
LDS Hospital
Veterans Administration
Duke Univ. Medical Center
Wishard Memorial Hospital
Department of Defense (Army, Navy, Air Force)
Med. Center Hospital of Vermont
Location
Mountain View, CA
Boston, MA
Salt Lake City, UT
United States
Durham, NC
Indianapolis, IN
Military bases worldwide
Burlington, VT
Cook 8 McDowell (1980); Gluck (1979); Kuperman, Gardner, 8 Pryor (1991); McDonald, Blevins, Tierney, 8 Martin (1988); Saba 8 McCormick (1996): and Stead 8 Hammond (1988) for
76 First Quarter 2001
/carnal of Nursing Scholarship
a
.
n the late 1980s, HIS' vendors began developing and marketing clinical applications. Major vendors such as Shared Medical Systems, HBOC, Cerner, and others offered
capabilities for care plans, nursing documentation, and other functions useful to nurses. The American Nurses' Association published guidelines for nursing information systems in the early 1990s to help guide development of nursingcentered applications (Zielstorff, Hudgings, & Grobe, 1993).
Many early systems are still in use today, and replacements are planned for the military's CHCS, LDS hospital's HELP and the HIS at the NIH Clinical Center. Major vendors often offer applications for nursing activities, but support among institutions remains uneven. For example, CHCS (MAISRC, 1998) and HIS at Registrieff (McDonald et al., 1999) still do not have computing capabilities to support clinical nursing activities. In addition to these early HIS, decision support systems were being developed in the 1980s.
Decision Support Systems
Computerized Decision Support Systems (DSS) are used in patient care decisionmaking. Haug, Gardner, and Evans (1 999) defined four categories of decision support: (a) alerting the care provider to situations of concern, (b) critiquing previous decisions, (c) suggesting interventions at the direct request of the care provider, and (d) retrospective quality assurance reviews.
Dr. Warner was one of the first people interested in using computers to support clinical decision making. As early as the 1950s, he began developing an application for use in diagnosis of congenital heart defects (Warner, Toronto, Veasey, & Stephenson, l 961). Warner's program, called Iliad, along with others such as DxExplain, Internist1, MYCIN, and QMR, were initial efforts aimed at assisting medical diagnoses (Berner et al., 1994).
Although the earliest DSS were targeted toward support for medical practice, they were also being developed to support nursing practice. The earliest nursing application was the Creighton Online Multiple Modular Expert System or COMMES. Developed in the 1970x, it assisted nurses in care planning activities. The COMMES application was evaluated among several settings and found to be useful in some locations (Thompson, Ryan, & Baggs, 1991). Later DSS prototypes were developed to assist with determining nursing diagnoses (Bradburn, Zeleznikow, & Adams, 1993; Chang & Hirsch, 1991) and training staff to help prevent incontinence in institutionalized, longterm care patients (Petrucci et al., 1992).
Like HIS the earliest DSS were focused more on medical than can nursing care. However, because of the complexities in developing DSS, applications in both disciplines were for a more circumscribed portion of the entire care process such as assistance with determining an appropriate patient care plan. Also DSS were not extensively used because they typically were kept on separate computers from the HIS. The expectation was that providers would enter all needed, data into both systems (Miller e& Geissbuhler, 1999). Few practitioners were willing tee take the time toy enter this
Clinical Information Systems in the United States
additional data. Therefore, the early DSS had their greatest use as teaching tools for nursing and medical students and were never well integrated into clinical care.
Current Trends in Clinical Computing
In the late 1980s, a shift occurred from a retrospective, feeforservice reimbursement structure to a prospective fixedcost structure. Nevertheless, healthcare inflation continued and precipitated a widespread movement toward managed care. With managed care, the traditional physicianoriented focus was shifted to a payeroriented focus emphasizing health promotion, disease prevention, and cost containment. Concurrently, questions about the appropriateness of medical decisions and the effectiveness of medical care provided momentum for the outcomes movement. To support these shifting foci, integrated healthcare delivery networks were developed by consolidating diverse types of healthcare settings. These integrated enterprises changed not only healthcare delivery structures and processes but also the types of information technology (IT) required to support them.
The Computerbased Patient Record
A computerbased patient record (CPR) is needed to effectively track patient care within managed care networks. A CPR is essential for integrated healthcare networks to be able to accomplish strategic goals of fusing business and clinical operations. The Institute of Medicine (IOM) report on improving patient records (Dick & Steen, 1991) indicated trends toward integration within the healthcare industry and the importance of having a CPR to support these new care delivery methods (HIMSS,1996). The IOM's recommendations also indicated the need for a longitudinal CPR to integrate and manage all clinical information throughout a person's lifetime in many locations. This new vision for managing clinical information was centered on patient care data across the lifespan and the increasingly important role of interdisciplinary clinical teams (Dick & Steen, 1991).
Advantages of integrated information systems were widely and enthusiastically discussed during the 1990s. Data replication and redundancy would be reduced, information consistency would be enhanced, and availability and accuracy of information would he improved. These systems would facilitate the knowledge work of clinicians, improve the quality of clinical decisionmaking, and in turn would lead to positive patient care and organizational outcomes.
In addition, the projected financial savings of computerized records are substantial. In ambulatory settings, early clinical intervention and fewer patients' visits can lead to an estimated savings of $15.3 billion a year (HIMSS, 1996). Bossard's presentation and Mulqueen's study (as cited in HIMSS,1996) indicated that implementation of a CPR could yield annual savings of $12.7 to $36 billion in hospital costs related toy decreased adverse reactions, record storage, staff time, more rapid retrieval of clinical information, and improved record review.