16 Patient-Care Systems

Patient-Care Systems

J

G. O

S

B

After reading this chapter, you should know the answers to these questions:

●

What are the four major information-management issues in patient care?

●

How have patient-care systems evolved during the past four decades?

●

How have patient-care systems influenced the process and outcomes of patient care?

●

Why are patient-care systems essential to the computer-based patient record?

●

How can they be differentiated from the computer-based patient record itself ?

16.1 Information Management in Patient Care

Patient care is the focus of many clinical disciplines—medicine, nursing, pharmacy, nutri- tion, therapies such as respiratory, physical, and occupational, and others. Although the work of the various disciplines sometimes overlaps, each has its own primary focus, emphasis, and methods of care delivery. Each discipline’s work is complex in itself, and collaboration among disciplines adds another level of complexity. In all disciplines, the quality of clinical decisions depends in part on the quality of information available to the decision-maker. The systems that manage information for patient care are therefore a critical tool. Their fitness for the job varies, and the systems enhance or detract from patient care accordingly. This chapter describes information-management issues in patient care, the evolution of patient care systems in relation to these issues, and current research. It will also show how patient care systems provide the infrastructure that determines the quality and functions of the computer-based patient record.

16.1.1 Concepts of Patient Care

Patient care is an interdisciplinary process centered on the care recipient in the con- text of the family, significant others, and community. Typically, patient care includes the services of physicians, nurses, and members of other health disciplines according to patient needs: physical, occupational, and respiratory therapists; nutritionists; psy- chologists; social workers; and many others. Each of these disciplines brings special- ized perspectives and expertise. Specific cognitive processes and therapeutic techniques vary by discipline, but all disciplines share certain commonalities in the provision of care.

In its simplest terms, the process of care begins with collecting data and assessing the patient’s current status in comparison to criteria or expectations of normality. Through

564

cognitive processes specific to the discipline, diagnostic labels are applied, therapeutic goals are identified with timelines for evaluation, and therapeutic interventions are selected and implemented. At specified intervals, the patient is reassessed, the effective- ness of care is evaluated, and therapeutic goals and interventions are continued or adjusted as needed. If the reassessment shows that the patient no longer needs care, services are terminated. This process was illustrated for nursing in 1975 (Goodwin &

Edwards, 1975) and was updated and made more general in 1984 (Ozbolt et al., 1985).

The flowchart reproduced in Figure 16.1 could apply equally well to other patient-care disciplines.

Although this linear flowchart helps to explain some aspects of the process of care, it is, like the solar-system model of the atom, a gross simplification. Frequently, for exam- ple, in the process of collecting data for an initial patient assessment, the nurse may rec- ognize (diagnose) that the patient is anxious about her health condition. Simultaneously with continuing the data collection, the nurse sets a therapeutic goal that the patient’s anx- iety will be reduced to a level that increases the patient’s comfort and ability to participate in care. The nurse selects and implements therapeutic actions of modulating the tone of voice, limiting environmental stimuli, maintaining eye contact, using gentle touch, asking about the patient’s concerns, and providing information. All the while, the nurse observes the effects on the patient’s anxiety and adjusts his behavior accordingly. Thus, the com- plete care process can occur in a microcosm while one step of the care process—data col- lection—is underway. This simultaneous, nonlinear quality of patient care poses challenges to informatics in the support of patient care and the capture of clinical data.

Each caregiver’s simultaneous attention to multiple aspects of the patient is not the only complicating factor. Just as atoms become molecules by sharing electrons, the care provided by each discipline becomes part of a complex molecule of interdisciplinary

nize that true interdisciplinary care is as different from the separate contributions of the various disciplines as an organic molecule is from the elements that go into it. The con- tributions of the various disciplines are not merely additive; as a force acting on the patient, the work of each discipline is transformed by its interaction with the other dis- ciplines in the larger unity of patient care.

For example, a 75-year-old woman with rheumatoid arthritis, high blood pressure, and urinary incontinence might receive care from a physician, a home-care nurse, a nutritionist, a physical therapist, and an occupational therapist. From a simplistic, addi- tive perspective, each discipline could be said to perform the following functions:

1. Physician: diagnose diseases, prescribe appropriate medications, authorize other care services

2. Nurse: assess patient’s understanding of her condition and treatment and her self- care abilities and practices; teach and counsel as needed; help patient to perform exer- cises at home; report findings to physician and other caregivers

3. Nutritionist: assess patient’s nutritional status and eating patterns; prescribe and teach appropriate diet to control blood pressure and build physical strength

4. Physical therapist: prescribe and teach appropriate exercises to improve strength and

flexibility and to enhance cardiovascular health, within limitations of arthritis

tions. (Source: Adapted with permission from Ozbolt J.G., et al. [1985]. A proposed expert system for nursing practice. Journal of Medical Systems, 9:57–68.

5. Occupational therapist: assess abilities and limitations for performing activities of daily living; prescribe exercises to improve strength and flexibility of hands and arms;

teach adaptive techniques and provide assistive devices as needed

In a collaborative, interdisciplinary practice, the nurse might discover that the patient was not taking walks each day as prescribed because her urinary incontinence was exac- erbated by the diuretic prescribed to treat hypertension, and the patient was embar- rassed to go out. The nurse would report this to the physician and the other caregivers so that they could understand why the patient was not carrying out the prescribed regime. The physician might then change the strategy for treating hypertension while ini- tiating treatment for urinary incontinence. The nurse would help the patient to under- stand the interaction of the various treatment regimes, would provide practical advice and assistance in dealing with incontinence, and would help the patient to find person- ally acceptable ways to follow the prescribed treatments. The nutritionist might work with the patient on the timing of meals and fluid intake so that the patient could exer- cise and sleep with less risk of urinary incontinence. The physical and the occupational therapists would adjust their recommendations to accommodate the patient’s personal needs and preferences while moving toward the therapeutic goals. Finally, the patient, rather than being assailed with sometimes conflicting demands of multiple caregivers, would be supported by an ensemble of services. Such collaboration, however, requires exquisite communication and feedback. The potential for information systems to support or sabotage this kind of care is obvious.

Although the care of individual patients is thus complex, it is far from being the total- ity of patient care. Because patients receive services from multiple caregivers, someone must coordinate those services. Coordination includes seeing that patients receive all the services they need in logical sequence without scheduling conflicts and ensuring that each caregiver communicates as needed with the others. Sometimes, a case manager is designated to do this coordination. In other situations, a physician or a nurse assumes the role by default. Sometimes, coordination is left to chance, and both the processes and the outcomes of care are put at risk. In recognition of this, the IOM recently designated coordination of care as one of 14 priorities for national action to transform health care quality (Institute of Medicine, 2002).

Delivering and managing the interdisciplinary care of each patient would seem to be

sufficiently challenging, but patient care has yet another level of complexity. Each care-

giver is usually responsible for the care of multiple patients. In planning and executing

the work of caregiving, each professional must consider the competing demands of all

the patients for whom she is responsible, as well as the exigencies of all the other pro-

fessionals involved in each patient’s care. Thus, the nurse on a post-operative unit must

plan for scheduled treatments for each of her patients to occur near the optimal time for

that patient. She must take into account that several patients may require treatments at

nearly the same time and that some of them may be receiving other services, such as

X-ray or physician’s visits, at the time when it might be most convenient for the nurse to

administer the treatment. When unexpected needs arise, as they often do—an

emergency, an unscheduled patient, observations that could signal an incipient

complication—the nurse must set priorities, organize, and delegate to be sure that at

least the critical needs are met. Similarly, the physician must balance the needs of vari- ous patients who may be widely dispersed throughout an institution. Decision-support systems have the potential to provide important assistance for both clinical and organizational decisions.

Finally, caregivers not only deliver services to patients, with all the planning, docu- menting, collaborating, referring, and consulting attendant on direct care; they are also responsible for indirect-care activities, such as teaching and supervising students, attend- ing staff meetings, participating in continuing education, and serving on committees.

Each caregiver’s plan of work must allow for both the direct-care and the indirect-care activities. Because the caregivers work in concert, these plans must be coordinated.

In summary, patient care is an extremely complex undertaking with multiple levels.

Each caregiver’s contributions to the care of every patient must take into account the ensemble of contributions of all caregivers and the interactions among them, all coor- dinated to optimize effectiveness and efficiency. Moreover, these considerations are mul- tiplied by the number of patients for whom each caregiver is responsible. Patient care is further complicated by the indirect-care activities that caregivers must intersperse among the direct-care responsibilities and coordinate with other caregivers. It is little wonder that managing, processing, and communicating data, information, and knowledge are integral and critical to every aspect of patient care.

16.1.2 Information to Support Patient Care

As complex as patient care is, the essential information for direct patient care is defined in the answers to the following questions:

●

Who is involved in the care of the patient?

●

What information does each professional require to make decisions?

●

From where, when, and in what form does the information come?

●

What information does each professional generate? Where, when, and in what form is it needed?

The framework described by Zielstorff and others (1993) provides a useful heuristic for understanding the varied types of information required to answer each of these questions. As listed in Table 16.1, this framework delineates three information cate- gories: (1) patient-specific data, which are those data about a particular patient acquired from a variety of data sources; (2) agency-specific data, which are those data relevant to the specific organization under whose auspices the health-care is provided; and (3) domain information and knowledge, which is specific to the health-care disciplines.

The framework further identifies four types of information processes that informa-

tion systems may apply to each of the three information categories. Data acquisition

entails the methods by which data become available to the information system. It may

include data entry by the care provider or acquisition from a medical device or from

another computer-based system. Data storage includes the methods, programs, and

structures used to organize data for subsequent use. Examples of standardized cod-

ing and classification systems useful in representing patient care concepts are listed in

Table 16.2. This topic is discussed in greater detail in Chapters 2, 7, and 12. Data

Table 16.1.Framework for design characteristics ofa patient-care information system with examples ofpatient-specific data,agency-specific data,and domain information and knowledge for patient care. TypesSystem processes ofdataAcquiringStoringTransformingPresenting Domain- Downloading relevant scientificMaintaining information inLinking related literature orDisplaying relevant literature specificor clinical literature or electronic journals or files,published findings;updatingor guidelines in response to practice guidelinessearchable by key wordsguidelines based on researchqueries Agency-Scanning,downloading,or Maintaining information in Editing and updating information;Displaying on request continu- specifickeying in agency policies electronic directories,fileslinking related information ously current policies and and procedures;keying in and databasesin response to queries;procedures shoring rele- personnel,financial,and analyzing informationvant policies and procedures administrative recordsin response to queries;gen- erating management reports Patient- Point-of-care entry ofdata Moving patient data into a Combining relevant data on aDisplaying reminders,alerts specific about patient assessment,current electronicsingle patient into a cue forprobable diagnoses,or diagnoses,treatments record or an aggregate action in a decision-support suggested treatments; planned and delivered,data repositorysystem;performing statistical displaying vital signs therapeutic goals patient analyses on data from manygraphically;displaying outcomespatientsstatistical results Source:Framework adapted with permission from Next Generation Nursing Information Systems,1993,American Nurses Association,Washington,DC.

transformation or processing comprises the methods by which stored data or informa-

tion are acted on according to the needs of the end user—for example, calculation of a pressure ulcer risk-assessment score at admission or calculation of critically ill patients’

acute physiology and chronic health evaluation (APACHE) scores. Figure 16.2 illus- trates the transformation (abstraction, summarization, aggregation) of patient-specific data for multiple uses. Presentation encompasses the forms in which information is delivered to the end user after processing.

Transformed patient-specific data can be presented in a variety of ways. Numeric data may be best presented in chart or graph form to allow the user to examine trends,

System Problems Interventions Goals/Outcomes

International Classification of Diseases x

NANDA Taxonomy 1 x

Current Procedural Terminology x

Nursing Interventions Classification x

Nursing Outcomes Classification x

Omaha System x x x

Home Health Care Classification x x x

SNOMED Clinical Terms x x x

Patient Care Data Set x x x

Figure 16.2. Examples of uses for atomic-level patient data collected once but used many times.

(Source: Reprinted with permission from R. D. Zielstorff, C. I. Hudgings, S. J. Grobe, and The National Commission on Nursing Implementation Project Task Force on Nursing Information Systems. Next-Generation Nursing Information Systems, © 1993 American Nurses Publishing, American Nurses Foundation/American Nurses Association, Washington, DC.)

whereas the compilation of potential diagnoses generated from patient-assessment data lends itself to an alphanumeric-list format. Different types of agency-specific data lend themselves to a variety of presentation formats. Common among all, however, is the need for presentation at the point of patient care; for example, the integration of up- to-the-minute patient-specific data with agency-specific guidelines or parameters can produce alerts, reminders, or other types of notifications for immediate action. See Chapter 17, on patient-monitoring systems, for an overview of this topic. Presentation of domain information and knowledge related to patient care is most frequently accom- plished through interaction with databases and knowledge bases, such as Medline or Micromedex (see Chapter19). Another approach is the Infobutton at New York Presbyterian Hospital. Through the Infobutton Manager, data about the patient, the provider, and the area in the clinical information system is taken into account so that context-specific knowledge is presented at the point of care (Cimino et al., 2002).

To support patient care, information systems must be geared to the needs of all the professionals involved in care. The systems should acquire, store, process, and present each type of information (patient-, agency-, and domain-specific) where, when, and how each function is needed by each professional. These systems not only support each professional’s care of individual patients but also, through appropriate use of patient- specific information (care requirements), agency-specific information (caregivers and their responsibilities and agency policies and procedures), and domain information (guidelines), such systems can greatly aid the coordination of interdisciplinary services for individual patients and the planning and scheduling of each caregiver’s work activ- ities. Patient acuity is taken into account in scheduling nursing personnel, but most often is entered into a separate system rather than derived directly from care require- ments. Integrated systems—still an ideal today—would enhance our understanding of each patient’s situation and needs, improve decision-making, facilitate communica- tions, aid coordination, and use clinical data to provide feedback for improving clinical processes.

Clearly, when patient-care information systems fulfill their potential, they will not merely replace oral and paper-based methods of recording and communicating. They will not only support but also transform patient care. How far have we come toward the ideal? What must we do to continue our progress?

16.2 Historical Evolution of Patient-Care Systems

The genesis of patient care systems occurred in the mid-1960s. One of the first and

most successful systems was the Technicon Medical Information System (TMIS),

begun in 1965 as a collaborative project between Lockheed and El Camino Hospital in

Mountain View, California (see Chapter 13). Designed to simplify documentation

through the use of standard order sets and care plans, TMIS defined the state of the

art when it was developed. Four decades later, versions of TMIS are still widely used,

but the technology has moved on. The hierarchical, menu-driven arrangement of infor-

mation in TMIS required users to page through many screens to enter or retrieve data

and precluded aggregation of data across patients for statistical analysis. Today’s users

have a different view of what can be done with data, and they demand systems that support those uses.

Part of what changed users’ expectations for patient care systems has been the devel- opment and evolution of systems that support clinical decision making. The HELP sys- tem (see Chapters 12, 13, and17) at LDS Hospital in Salt Lake City, Utah, initially provided decision support to physicians during the process of care (in addition to man- aging and storing data). HELP subsequently became able to support nursing care deci- sions and to aggregate data for research, leading to improved patient care. Other systems that provide decision support for physician order-entry, such as those developed at Partners Health Care in Boston and at Vanderbilt University Medical Center in Nashville, Tennessee, have been translated into commercial systems. By demonstrating that these systems can improve patient outcomes while containing costs, developers and vendors raise expectations in the marketplace. Still other systems, such as the order- entry system developed at Vanderbilt, are beginning to explore the effects of feedback about care effectiveness on the processes and outcomes of care.

Today, many commercially available information systems for patient care incorporate decision support, integration of information from multiple sources, care planning and documentation, organization of the clinician’s workflow, and support for care manage- ment. Both vendors of information systems and researchers in health-care enterprises are working to advance these features in systems that use the latest technologies for nav- igating and linking information. Although knowledge-based systems have been in exis- tence since the early years of biomedical informatics, we are seeing the infancy of systems that aggregate and analyze clinical data to produce new knowledge and apply it in practice. As clinical knowledge and workflow, rather than financial models, become the basis for system design, emerging generations of patient care systems appear poised to fulfill the promises of clinical informatics.

16.2.1 Societal Influences

The historical evolution of information systems that support patient care is not solely a reflection of the available technologies. Societal forces—including delivery-system structure, practice model, payer model, and quality focus—have influenced the design and implementation of patient-care systems (Table 16.3).

Delivery-System Structure

Authors have noted the significant influence of the organization and its people on the

success or failure of informatics innovations (Ash, 1997; Kaplan, 1997; Lorenzi et al.,

1997; Southon et al., 1997). As delivery systems shifted from the predominant single-

institution structure of the 1970s to the integrated delivery networks of the 1990s to the

complex linkages of the 21

Century, the information needs changed, and the challenges

of meeting those information needs increased in complexity. See Chapters 13, 14, and

15 for discussions of managing clinical information in integrated delivery systems,

in consumer-provider partnerships in care, and in the public health information infra-

structure.

Professional Practice Models

Professional practice models have also evolved for nurses and physicians. In the 1970s, team nursing was the typical practice model for the hospital, and the nursing care plan—a document for communicating the plan of care among nursing team members—

was most frequently the initial computer-based application designed for use by nurses.

The 1990s were characterized by a shift to interdisciplinary-care approaches necessitat- ing computer-based applications such as critical paths to support case management of aggregates of patients, usually with a common medical diagnosis, across the continuum

Century sees advanced practice nurses increasingly taking on functions previously provided by physicians while maintaining a nursing perspective on collabo- rative, interdisciplinary care. These changes broaden and diversify the demands for decision support, feedback about clinical effectiveness, and quality improvement as a team effort.

Physician practice models have shifted from single physician or small group offices to complex constellations of provider organizations. The structure of the model (e.g., staff model health-maintenance organization, captive-group model health-maintenance organization, or independent-practice association; see Chapter 23) determines the types

1970s 1980s 1990s

Delivery-system Single institution Single organization Integrated delivery

structure systems

Professional-practice Team nursing Primary nursing Patient-focused care model Single or small Group models for Interdisciplinary care

group physician physicians Case management

practice Variety of constellations

of physician group practice models

Payer model Fee for service Fee for service Capitation

Prospective payment, Managed Care diagnosis-related

groups (DRGs)

Quality focus Professional Continuous quality Risk-adjusted outcomes

Standards improvement Benchmarking

Review Joint Commission on Practice guidelines Organizations Accreditation of Critical paths/care maps

(PSROs) Health Care Health Employer Data

Retrospective chart Organization and Information set

review (JCAHO)’s Agenda (HEDIS)

Joint Commission for Change on Accreditation

of Hospitals’ Peer Evaluation Program Quality of Patient Care System (QUALPACS)

of relationships among the physicians and the organizations. These include issues—

such as location of medical records, control of practice patterns of the physicians, and data-reporting requirements—that have significant implications for the design and implementation of patient-care systems. In addition, the interdisciplinary and distrib- uted care approaches of the 1990s and the 2000s have given impetus to system-design strategies, such as the creation of a single patient problem list, around which the patient- care record is organized, in place of a separate list for each provider group (e.g., nurses, physicians, respiratory therapists).

Payer Models

Changes in payer models have been a significant driving force for information-system implementation in many organizations. With the shift from fee for service to prospective payment in the 1980s, and then toward capitation in the 1990s, information about costs and quality of care has become an essential commodity for rational decision-making in the increasingly competitive healthcare marketplace. Because private, third-party payers often adopt federal standards for reporting and regulation, health care providers and institutions have struggled in the early 2000s to keep up with the movement toward data and information system standards accelerated by the Health Insurance Portability and Accountability Act (HIPAA) and the initiatives to develop a National Health Information Infrastructure. See Chapter 23 for a thorough discussion of the effects of healthcare financing on health-care information systems.

Quality Focus

Demands for information about quality of care have also influenced the design and implementation of patient-care systems. The quality-assurance techniques of the 1970s were primarily based on retrospective chart audit. In the 1980s, continuous-quality- improvement techniques became the modus operandi of most healthcare organizations.

The quality-management techniques of the 1990s were much more focused on concur- rently influencing the care delivered than on retrospectively evaluating its quality. In the 21

Century, patient-care systems-based approaches—such as critical paths, practice guidelines, alerts, and reminders—are an essential component of quality management.

In addition, institutions must have the capacity to capture data for benchmarking purposes and to report process and outcomes data to regulatory and accreditation bodies, as well as to any voluntary reporting programs (e.g., Maryland Hospital Indicator Program) to which they belong. Increasingly, concurrent feedback about the effectiveness of care guides clinical decisions in real time.

16.2.2 Patient-Care Systems

The design and implementation of patient-care systems, for the most part, occurred sep-

arately for hospital and ambulatory-care settings. Early patient-care systems in the hos-

pital settings included the University of Missouri-Columbia System (Lindberg, 1965),

the Problem-Oriented Medical Information System (PROMIS) (Weed, 1975), the Tri-

Service Medical Information System (TRIMUS) (Bickel, 1979), the Health Evaluation Logical Processing (HELP) System (Kuperman et al., 1991), and the Decentralized Hospital Computer Program (DHCP) (Ivers & Timson, 1985). The Computer-Stored Ambulatory Record (COSTAR), the Regenstrief Medical Record System (McDonald, 1976), and The Medical Record (TMR) were among the earliest ambulatory care sys- tems. For a comprehensive review, see Collen (1995).

According to Collen (1995), the most commonly used patient-care systems in hospi- tals of the 1980s were those that supported nursing care planning and documentation.

Systems to support capture of physicians’ orders, communications with the pharmacy, and reporting of laboratory results were also widely used. Some systems merged physi- cian orders with the nursing care plan to provide a more comprehensive view of care to be given. This merging, such as allowing physicians and nurses to view information in the part of the record designated for each other’s discipline, was a step toward integra- tion of information. It was still, however, a long way from support for truly collabora- tive interdisciplinary practice.

Early ambulatory-care systems most often included paper-based, patient encounter forms that were either computer-scannable mark-sense format or were subsequently entered into the computer by clerical personnel. Current desktop, laptop, or hand- held systems use keyboard, mouse, or pen-based entry of structured information, with free text kept to a minimum. These systems also provide for retrieval of reports and past records. Some systems provide decision support or alerts to remind clinicians about needed care, such as immunizations or screening examinations, and to avoid contraindicated orders for medications or unnecessary laboratory analyses.

Depending on network capabilities, systems may facilitate communications among the professionals and settings involved in the patient’s care. Voice-recognition tech- nology is advancing and is beginning to permit direct dictation into the record.

Although this mode of data entry has the advantages of ease and familiarity to clini- cians, free text in the record inhibits search, retrieval, and analysis of data. Before dic- tated notes can become as useful as structured data, the entry systems will have to become able to recognize the meanings of words and their context and to store the data in databases. Although this level of intelligent processing of natural language remains in the future (See Chapter 8), systems to support ambulatory care have clearly made great strides. The best provide good support for traditional medical care.

Support for comprehensive, collaborative care that gives as much attention to health promotion as to treatment of disease presents a challenge not only to the developers of information systems but also to practitioners and healthcare administrators who must explicate the nature of this practice and the conditions under which agencies will provide it.

Patient-care information systems in use today represent a broad range in the evolu-

tion of the field. Versions of some of the earliest systems are still in use. These systems

were generally designed to speed documentation and to increase legibility and availabil-

ity of the records of patients currently receiving care. Most lack the capacity to aggre-

gate data across patients, to query the data about subsets of patients, or to use data

collected for clinical purposes to meet informational needs of administrators or

researchers. These shortcomings seem glaring today, but they were not apparent when

the very idea of using computers to store and communicate patient information required a leap of the imagination.

More recently developed systems attempt with varying success to respond to the edict

“collect once, use many times.” Selected items of data from patient records are abstracted manually or electronically to aggregate databases where they can be analyzed for administrative reports, for quality improvement, for clinical or health-services research, and for required patient safety and public health reporting. See Chapter 15 for a full discussion of public health informatics.

Some recently developed systems offer some degree of coordination of the informa- tion and services of the various clinical disciplines into integrated records and plans.

Data collected by one caregiver can appear, possibly in a modified representation, in the

“view” of the patient record designed for another discipline. When care-planning infor- mation has been entered by multiple caregivers, it can be viewed as the care plan to be executed by a discipline, by an individual, or by the interdisciplinary team. Some patient-care systems offer the option to organize care temporally into clinical pathways and to have variances from the anticipated activities, sequence, or timing reported auto- matically. Others offer a patient “view” so that individuals can view and contribute to their own records records (Pyper et al, 2002; Cimino, Patel, Kushniruk, 2002).

Doolan et al. (2003) studied information system functionality for clinical care in five sites that had won the Computer-based Patient Record Institute Davies’ Award:

●

LDS Hospital, Salt Lake City (LDSH) in 1995

●

Brigham and Women’s Hospital, Boston (BWH) in 1996

●

Wishard Memorial Hospital, Indianapolis (WMH) in 1997

●

Queen’s Medical Center, Honolulu (QMC) in 1999

●

Veteran’s Affairs Purget Sound Healthcare System, Seattle and Tacoma (VAPS) in 2000

All sites had broadscale computer-based results reporting and order entry for med- ications and other therapeutics. As compared to many other organizations, the patient care systems these sites also had some functionality related to documentation of clini- cal notes (See Table 16.4). It is noteworthy that even in these sites that are widely rec- ognized for their advanced clinical information systems, clinicians’ progress notes are not completely computer-based.

Computerized Notes

The publication of the Institute of Medicine’s reports To Err is Human (2000) and

providers for information systems that reduce errors in patient care. Information system vendors are responding by developing such systems themselves and by purchasing the rights to patient care systems developed in academic medical centers that have demon- strated reductions in errors and gains in quality of care and cost control. “Closed loop”

medication systems use technologies such as bar codes and decision support to guard

against errors throughout the process of prescribing, dispensing, administering, and

Table 16.4.Computerized Notes* LDSHWMHBWHQMC†VAPS InpatientPhysicians:admission,Physicians:problem Physicians:problem Physicians:problem Physicians:problem list allergies,list,allergies,list,allergies,list,allergies,(85%),allergies, medications,medications,medications,some medications,medications, procedures,admission note,progress in the form history,proceduresadmission,progress, discharge summaryprogress (60% ofa handover Nurses:initial procedures,discharge Nurses:initial house staff,100% summary,procedure assessment,vital summary assessment,progress,attending notes,discharge signs (general Nurses:initial assessment, vital signs,handover,physicians),summarywards),progress notes,vital medication procedures (50%),operative notes,signs (50%),operative administration discharge summarymedication notes,medication recordNurses:initial administration administration record Therapists:all notesassessment (70%),record(80%),discharge vital signssummary Therapists:most notesTherapists:all notes AmbulatoryPhysicians:problem Physicians:problem Physicians:problem Physicians:problem Physicians:problem list list,allergies,list,allergies,list,allergies,list,allergies,(85%),allergies, medications,medications,history medications,history medications,medications,history history and and physical and physical history and and physical findings, physical findings,findings (50%),findings,proceduresphysical findings,procedures proceduresvital signs,Nurses:vital signsproceduresNurses:initial Nurses:vital signsprocedures (70%)Nurses:initial assessment (80%), Nurses:vital signsassessment,vital vital signs (90%) signsTherapists:all notes Emergency departmentPhysicians:allergies,Physicians:problem Physicians:allergies,Physicians:summary Physicians:problem list medications,list,allergies,medications,note by attending (85%),allergies, discharge summarymedications summary note by physician on all medications (67%), Nurses:initial (discharge only),attending physician patientshistory and physical assessment,discharge summaryon all patientsNurses:initialfindings,progress, progress,vital signs,assessment,discharge summary transfer,medication dischargesummaryNurses:initial administration assessment,vital record,discharge signs (10%) summary *Indicates 100% oflisted function unless otherwise statement. †At QMC internal hospital ambulatory clinics and 11 out of237 affiliated physician clinics using computerized notes. Reprinted with permission.Doolan,DF,Bates,DW,& James,BC (2003).Journal ofthe American Medical Informatics Association,10(1),p.99.

recording. In other contexts, decision support systems offer “best practice” guidelines, protocols, and order sets as a starting point for planning individualized patient care;

provide alerts and reminders; use knowledge bases and patient data bases to assess orders for potential contraindications; and offer point-and-click access to knowledge summaries and full-text publications. See Chapter 20 for more information about these systems.

Many healthcare agencies have substantial investment in legacy systems and cannot simply switch to more modern technology. Finding ways to phase the transition from older systems to newer and more functional ones is a major challenge to health infor- matics. To make the transition from a patchwork of systems with self-contained func- tions to truly integrated systems with the capacity to meet emerging information needs is even more challenging (see Chapter 13). Approaches to making this transition are described in the Proceedings of the 1996 IAIMS Symposium (IAIMS, 1996) and in the Journal of the American Medical Informatics Association (Stead et al., 1996).

The cornerstone of good patient-care systems is the ability to capture clinical data in the process of care, to store the data, to aggregate them, to analyze them, and to pro- duce reports that not only describe care but also yield knowledge of quality, effective- ness, and costs that can be the basis of improved clinical processes. The key to performing these operations is data standards (see Chapters 2, 7, 12, 13, and 15). It is vital that the data standards support the care delivery and evaluation processes of the variety of healthcare professionals who participate in patient care. In this regard, the nursing profession has made substantial efforts in the creation of standardized nursing languages. More recently through the development of SNOMED CT (see Chapter 7 for more details), data reflective of the care provided others such as dentists, podiatrists, nutritionists, and physical therapists have augmented the standardized terms typically used by physicians and nurses.

If patient-care systems are to be effective in supporting better care, healthcare pro- fessionals must possess the informatics competencies to use the systems. Consequently, many are integrating informatics competencies into health science education (See Chapter 21). For example at the Columbia University School of Nursing, basic and advanced practice nursing students document their clinical encounters using personal digital assistants and receive educational content designed to meet professional stan- dards for informatics competencies for beginning and experienced nurses, respectively (Bakken et al, 2003).

To what degree do patient-care disciplines need to prepare their practitioners for roles as informatics specialists? To the degree that members of the discipline use information in ways unique to the discipline, the field needs members prepared to translate the needs of clinicians to those who develop, implement, and make decisions about information systems. If the information needs are different from those of other disciplines, some practitioners should be prepared as system developers.

16.3 Current Research

Friedman (1995) proposed a typology of the science in medical informatics. The four

categories build from fundamental conceptualization to evaluation as follows:

●

Formulating models for acquisition, representation, processing, display, or transmis- sion of biomedical information or knowledge

●

Developing innovative computer-based systems, using these models, that deliver infor- mation or knowledge to healthcare providers

●

Installing such systems and then making them work reliably in functioning healthcare environments

●

Studying the effects of these systems on the reasoning and behavior of health-care providers, as well as on the organization and delivery of health care

Following are examples of recent research on patient-care systems in each category.

16.3.1 Formulation of Models

In recent years standards development organizations (SDOs) and professional groups alike have focused on the formulation of models that describe the patient care process and the formal structures that support management and documentation of patient care. The efforts of SDOs are summarized in Chapter 7. As a complement to SDO efforts, the Nursing Terminology Summit is an informal, interdisciplinary pro- fessional collaboratory whose participants develop and evaluate formal models such as reference information models, reference terminology models, and clinical docu- ment and EHR architectures from the perspective of nursing practice. Since the first Summit Conference in 1999, the participants’ efforts have resulted in a number of significant achievements. These include among others: 1) agreement to collaborate to develop terminology models of diagnoses and interventions; 2) contribution to the development of a proposal to the International Standards Organization (ISO) from the Nursing Informatics Special Interest Group of the International Medical Informatics Association (IMIA) and the International Council of Nurses (ICN) to develop terminology models for nursing and to integrate those models with compre- hensive models for health terminologies; 3) testing of the LOINC semantic struc- ture for ability to represent standardized nursing assessments and subsequent integration of selected nursing content into LOINC; and 4) recommendations for extension to HL7 RIM in order to represent educational interventions (Ozbolt, 2000;

Hardiker, Hoy, & Casey, 2000; Bakken, Cimino, Haskell, Kukafka, Matsumoto, Chan, & Huff, 2000).

16.3.2 Development of Innovative Systems

New systems to support patient care often take advantage of information entered in

one context for use in other contexts. For example, the Brigham Integrated

Computing System (BICS), a PC-based client–server HIS developed at Brigham and

Women’s Hospital in Boston, used information from the order entry, scheduling, and

other systems to prepare drafts of the physician’s discharge orders and the nurse’s dis-

charge abstract, thus minimizing the information to be entered manually. The profes-

sionals reviewed the drafts and edited as needed (O’Connell et al., 1996). The BICS

system’s success in this and other functions led to its acquisition for commercial

deployment.

The principle of entering information once for multiple uses also drove development of the low-cost bedside workstations for intensive-care units at the University Hospital of Giessen, Germany (Michel et al., 1996). The client–server architecture combined local data-processing capabilities with a central relational patient database, permitting, for example, clinical nursing data to be used in calculating workload. These worksta- tions also combined data from many sources, including medical devices, to support the integrated care of physicians, nurses, and other caregivers.

Even as systems such as these begin to fulfill some of the promises of informatics to support patient care, research and development continue to address the demands that the complexities of patient care place on information systems. Hoy and Hyslop (1995) reported a series of projects directed toward the development of a person-based health record. They found problems with traditional approaches to automating paper-based care-planning systems that resulted in loss of data detail, inability to use data for multiple purposes, and limitations in the capacity to aggregate and query patient data.

Hoy and Hyslop (1995) recommended:

●

Making the structure of the clinical record (including the care plan) more flexible and extensible to allow summarized higher-level data, with lower-level details where appropriate

●

Simplifying the elements of that structure to make data entry and retrieval easier and more effective

Hoy and Hyslop (1995) built a prototype system to demonstrate their recommenda- tions. Like other investigators, they concluded that “the issues of language and struc- tures must be dealt with before the integration of person-based systems can be realized.” As noted in Chapter 7, significant headway has been made in this regard during the last 5 years.

At Vanderbilt University Medical Center, patient-care systems have evolved since the mid-1990s to support patient safety and quality of care in a variety of ways.

Clinical teams, assisted by specially trained clinical librarians, develop evidence-

based order sets as templates for interdisciplinary care. These order sets are instanti-

ated in Vanderbilt’s order-entry system, where they serve as the starting point for

planning and documenting each patient’s care. When a patient is admitted to the

hospital, a decision-support tool helps the physician to identify the appropriate evidence-

based order set and then to edit the template to produce an individualized plan of

care. In this way, the most current clinical knowledge provides the basis for each

patient’s care. Ozdas et al (2006) demonstrated that use of the evidence-based order

set increased physician compliance with quality indicators for treating acute myocar-

dial infarction. Other research opportunities are to explore the impact (positive or

negative) of deviations from the template order sets in the context of different patient

characteristics and comorbidities, thereby refining the evidence base and adding to

clinical knowledge. Patient care systems like this make it possible to learn from data

collected in the course of patient care about the effectiveness and safety of specific

care practices and to integrate that emerging knowledge in continual quality

improvement (Ozbolt, 2001, 2003).

16.3.3 Implementation of Systems

Higgins and associates (see Rotman et al., 1996) described the lessons learned from a failed implementation of a computer-based physician workstation that had been designed to facilitate and improve ordering of medications. Those lessons are not iden- tical to, but are consistent with, the recommendations of Leiner and Haux (1996) in their protocol for systematic planning and execution of projects to develop and imple- ment patient-care systems. As these experiences demonstrate, the implementation of patient-care systems is far more complex than the replacement of one technology with another. Such systems transform work and organizational relationships. If the imple- mentation is to succeed, attention must be given to these transformations and to the dis- ruptions that they entail. Southon and colleagues (1997) provided an excellent case study of the role of organizational factors in the failed implementation of a patient-care system that had been successful in another site. To realize the promise of informatics for health and clinical management, people who develop and promote the use of applica- tions must anticipate, evaluate, and accommodate the full range of consequences. In early 2003, these issues came to the attention of the public-at-large when a large aca- demic medical center decided to temporarily halt implementation of its CPOE system due to mixed acceptance by the physician staff (Chin, 2003). A case series study by Doolan, Bates, and James (2003) identified five key factors associated with successful implementation: 1) having organizational leadership, commitment, and vision;

2) improving clinical processes and patient care; 3) involving clinicians in the design and modification of the system; 4) maintaining or improving clinical productivity; and 5) building momentum and support amongst clinicians.

16.3.4 Study of the Effects of Systems

Many studies of the effects of patient care systems have looked at impact on process

of care. A frequent expectation of systems to support nursing care planning and docu-

mentation is that they will decrease the time required for documentation, improve the

quality and relevance of data in the record, and increase the proportion of nursing time

spent in direct patient care. Pabst and colleagues (1996) found that an automated sys-

tem designed to replace just 40 percent of manual documentation decreased the time

required for documentation by one third, or by 20 minutes per shift. Nurses using the

system spent more time in direct patient care and were more likely than were nurses

using only manual documentation to complete documentation during their shifts rather

than staying over into the next shift. Quality of documentation was not affected. Oniki

et al. (2003) demonstrated a significant decrease in charting deficiencies in the ICU fol-

lowing computer-based reminders. Adderley and associates (1997) described the bene-

fits of a phased implementation of a paperless record as related to accessibility of the

record. Verbal orders were eliminated, and progress notes were more likely to be

entered. Communications among caregivers were enhanced. Prospective, rather than

retrospective, reviews of clinical data provided concurrent assessment of patient

progress, care planning, medication use, and ancillary services. Annual cost avoidance

from using the electronic patient record was estimated at more than $300,000. Lusignan

et al. (2003) showed that feedback improved the quality of computer-based records in primary care in aspects such as linking prescriptions to diagnosis.

While improvements in processes of care are important, many patient care systems are designed with improving patient safety and outcomes in mind. Ruland (2002) reported that a handheld support system for preference-based care planning (CHOICE) not only improved the consistency of nursing care with patient preferences, but also increased patient achievement related to functional status. Bates et al. (2003) conducted a systematic review of information technologies for detecting adverse events and con- cluded that tools such as event monitoring and natural language processing can inex- pensively detect adverse events such as adverse drug events and nosocomial infections.

Wilson et al. (1997) demonstrated that the use of a shared computer-based medication record by pharmacy and nursing led to a statistically significant decrease in medication occurrences. .Chapters 12, 17, and 20 provide additional details about the effects of par- ticular types of systems used in patient care.

Improving the methods of evaluating information resources was the driving force behind the Institute of Medicine’s 1996 report on telemedicine (Field, 1996). Finding assessments of technical performance insufficient, the report recommended that evalu- ations focus on effects on patient welfare and on the processes and costs of care in com- parison to those of reasonable alternatives. Members of the committee who developed the report noted that telemedicine may be considered a subset of medical informatics and that the methods of research and evaluation applicable to telemedicine are like those applicable to other patient-care systems (also see Chapter 14).

16.4 Outlook for the Future

Patient-care systems are changing in two ways. First, legacy systems designed primarily for charge capture and other administrative functions are being replaced by systems designed to support and improve clinical practice, as well as to send clinical data to the various locations where these data are needed for practice, management, and research.

Second, systems designed to support each discipline separately are yielding to those based on integrated, interdisciplinary concepts of care. Research is continuing to develop structured clinical languages, standards, and data models; to develop innovative systems; to determine more effective and efficient ways to implement systems; and to investigate the effects of changing information resources on the processes of care and the functioning of the organization.

This environment is a fertile one for the development and growth of patient-care sys- tems. In the first decade of the 21

Century, systems that will perform all the desired functions at a high level remain just out of reach. Many factors—evolution of technol- ogy, development of standards, and societal demands among them—are converging to stimulate rapid progress. Real-world systems are extending their mastery over more and more of the functions needed to support patient care: intelligent support for clinical decisions; better organization and communication; feedback on clinical effectiveness;

linked databases for research; and administrative analyses based on pertinent clinical

data. Such tools will make available to clinicians, managers, and policy-makers the data,

information, and knowledge required for sound decisions and effective action. By complementing and extending cognitive processes, patient-care systems become an integral and essential technology for patient care.

Suggested Readings

Doolan, DF, Bates, DW, & James, BC (2003). The use of computers for clinical care: A case series of advanced U.S. sites. Journal of the American Medical Informatics Association, 10(1), 94-107.

This article describes the clinical information systems in 5 U.S. hospitals that have won the Davies Award for outstanding systems. The article describes similarities and differences in the systems and experiences at these sites and identifies factors important to successful implementation.

Ruland, CM (2002). Handheld technology to improve patient care. Journal of the American Medical Informatics Association, 9(2), 192-200.

This research study evaluates the effectiveness of a handheld application to elicit patients’ prefer- ences for functional performance. Use of the application improved the frequency with which patients’ preferences were respected.

Ida M. Androwich, Carol J. Bickford, Patricia S. Button, Kathleen M. Hunter, Judy Murphy, and Joyce Sensmeier (2002) Clinical Information Systems: A Framework for Reaching the Vision.

Washington, DC: ANA Publishing.

This concise synthesis of healthcare informatics articulates an organizing framework for clinical information systems from a professional nursing perspective, and addresses how to best design, develop, and implement them.

Questions for Discussion

1. What is the utility of a linear model of patient care as the basis for a decision-support system? What are two primary limitations? Discuss two challenges that a nonlinear model poses for representing and supporting the care process in an information system?

2. Compare and contrast “segregated” versus “integrated” models of interdisciplinary patient care. What are the advantages and disadvantages of each model as a mode of care delivery? As the basis for developing information systems to plan, document, and support patient care?

3. Imagine a patient-care information system that assists in planning the care of each patient independently of all the other patients in a service center or patient-care unit.

What are three advantages to the developer in choosing such an information archi- tecture? What would be the likely result in the real world of practice? Does it make a difference whether the practice setting is hospital, ambulatory care, or home care?

What would be the simplest information architecture that would be sufficiently complex to handle real-world demands? Explain.

4. Zielstorff et al. (1993) proposed that data routinely recorded during the process of

patient care could be abstracted, aggregated, and analyzed for management reports,

policy decisions, and knowledge development. What are three advantages of using patient care data in this way? What are three significant limitations?

5. A number of patient-care information systems designed in the 1970s are still in use.

How do the practice models, payer models, and quality focus of today differ from those of the past? What differences do these changes require in information systems?

What are two advantages and two disadvantages of “retrofitting” these changes on older systems versus designing new systems “from scratch”?

6. What are three advantages and three disadvantages of free text (including oral nar- rative entered by dictation) versus structured data for recording observations, assess- ments, goals, and plans? What is the impact of using free text on the ability to retrieve and aggregate data? Should developmental efforts focus on interpreting natural lan- guage or on creating data standards? Explain your position.

7. What are four major purposes of patient care information systems? What criteria

should be used to evaluate them? What methods of evaluation could be used to assess

the system with respect to these criteria?