4
The Road to CPR Implementation

Technological change is a complex process that is influenced by a multitude of factors, including the attributes of a technology, the users of the technology, and the environment in which the technology is used. Successful implementation of computer-based patient record systems depends on ''more than the transmission of technical details and the availability of systems" (Anderson and Jay, 1987:4). It requires an understanding of the factors that influence the development and adoption of computer technologies in health care.

This chapter first identifies the factors that could enhance or impede CPR development and use. It then presents the committee's plan for addressing these factors. The plan includes a discussion of the various organizations that have a role to play in CPR development and diffusion, the types of activities that would facilitate patient record development, how such activities might be implemented, and when such activities should take place. Chapter 5 presents the committee's formal recommendations for achieving the primary goals of the plan.

Development and Diffusion Factors

The process of technological change involves two general stages: development and diffusion. Development is the production of new capabilities or the alteration of characteristics of existing technologies; diffusion is the application of a new technology in the provision of services. These two stages do not necessarily occur in chronological sequence. User application can reveal that a technology needs further development. Alternatively,



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 138
--> 4 The Road to CPR Implementation Technological change is a complex process that is influenced by a multitude of factors, including the attributes of a technology, the users of the technology, and the environment in which the technology is used. Successful implementation of computer-based patient record systems depends on ''more than the transmission of technical details and the availability of systems" (Anderson and Jay, 1987:4). It requires an understanding of the factors that influence the development and adoption of computer technologies in health care. This chapter first identifies the factors that could enhance or impede CPR development and use. It then presents the committee's plan for addressing these factors. The plan includes a discussion of the various organizations that have a role to play in CPR development and diffusion, the types of activities that would facilitate patient record development, how such activities might be implemented, and when such activities should take place. Chapter 5 presents the committee's formal recommendations for achieving the primary goals of the plan. Development and Diffusion Factors The process of technological change involves two general stages: development and diffusion. Development is the production of new capabilities or the alteration of characteristics of existing technologies; diffusion is the application of a new technology in the provision of services. These two stages do not necessarily occur in chronological sequence. User application can reveal that a technology needs further development. Alternatively,

OCR for page 138
--> diffusion sometimes parallels the development of a new technology (NAS, 1979). Moreover, as discussed below, the factors affecting development and diffusion are interrelated: development is unlikely to occur if the conditions for diffusion are unfavorable. Barriers to Development Patient Record Definition A basic impediment to the development of CPRs and CPR systems has been the lack of a clear definition of what a CPR could and should be. An intellectual understanding of what a CPR needs to do, the range of individuals for whom it needs to function, and the expectations and performance demands of its users is an essential prerequisite to successful design of a CPR system (Teach and Shortliffe, 1981). Many computer-based information management systems are currently in place and generate and use patient data for various purposes (e.g., billing, laboratory, radiology); what is lacking is a unified concept of what constitutes a computer-based patient record system. A fully articulated definition of a CPR and CPR system should describe attributes of the record and system (i.e., content, format, and function) as a guide for system developers and users.1 It should not, however, prescribe how those attributes are to be achieved. Resolution of that question is best left to CPR system developers and vendors. Chapter 2 identifies basic CPR system requirements, but system designers require more detailed specification in certain areas. Among these are the level of performance (e.g., speed and convenience) the system must provide to achieve health care provider acceptance; the kinds of new system functions needed to justify a change in current routines of record use and the costs of implementation; the frequency with which multiple users would view the record simultaneously; the level of system security, confidentiality, and reliability required; and the level of institutional, regional, national, or international interconnectivity demanded of the system. In particular, patient record system developers require specific information about system functionality and performance to design systems that meet user needs. Understanding the diverse set of CPR user needs requires that representatives of all users be involved in a process of setting priorities for system functionality and performance. 1   As discussed in Chapter 1, the committee's definition of a CPR is an electronic patient record that resides in a system specifically designed to support users through the availability of complete, accurate patient data, alerts, reminders, clinical decision support systems, links to medical knowledge, and other aids.

OCR for page 138
--> Content and Format Standards As discussed in Chapter 3, CPR development has been, and continues to be, impeded by the lack of standards for the content and format of CPR data (McDonald and Hammond, 1989; Bradbury, 1990; Gabrieli and Murphy, 1990; Lindberg and Humphreys, 1990; Masys, 1990). To exploit the potential benefits of linking CPRs across specialties and institutions, developers must ensure the following. (1) The content of CPRs must be defined; that is, each CPR should contain a uniform core set of data elements. (2) Data elements must be named consistently; that is, some form of vocabulary control must be in place. (3) Format standards for data exchange must be developed and used. Defining a core set of data elements requires participation by representatives of all patient record user groups. (Exclusion of any group could diminish the efficiency to be gained from implementing CPR systems if, as a result, key data elements are excluded from the core data set.) However, the uniform core data set should not be so large that it requires health care professionals to collect information that does not derive directly from routine service provision. Moreover, because providers are likely to require data elements in addition to those in the uniform core data set, CPR systems should be flexible and not be limited to core data elements. Vocabulary control efforts have already led to substantial progress in developing standardized vocabularies. The use of existing controlled vocabularies in combination could cover the basic name of each patient problem and of each procedure performed by a practitioner (Lindberg and Humphreys, 1990). Inadequate support for the timely update of clinical vocabularies, however, remains an obstacle to developing better vocabulary control for the CPR. Another obstacle is user resistance: users generally have not considered the benefits of a reasonably specific, controlled vocabulary as warranting a switch to a new system—especially if that system entails higher costs for record creation and maintenance (Lindberg and Humphreys, 1990). Existing format standards focus on particular portions of the record; no single format standard exists that could be used for the entire CPR. Nor is there at present a means for establishing one standard for use by the entire health care industry. In short, format standards for data exchange need further development and a means of achieving credibility. To date, such efforts have been carried out primarily by volunteer organizations. Greater support (i.e., funding and recognition) of these efforts would help to accelerate standards development. Costs and Risks Venture capital is unlikely to be forthcoming for large-scale technological systems that require sizable investments before yielding a return in a

OCR for page 138
--> highly uncertain market (NAS, 1979). Consequently, development of a technology may require subsidization of development costs or creation of a more certain market, or both (NAS, 1979). Both of these factors apply to CPR systems: they require significant investment for development, and a high degree of uncertainty exists regarding the willingness of health care providers to purchase the systems. As discussed in Chapter 3, the committee found no comprehensive CPR systems in existence in 1990. Furthermore, no cost data are available on the monetary investment that might be required to implement such a system; evaluations of cost data related to computer applications tend to have focused on medical information systems (MISs) rather than on CPR systems. MIS development costs and time estimates, however, do convey a sense of the magnitude of CPR development costs and time. A survey of automated ambulatory care systems found that development costs ranged from $100,000 to $10 million and that development time ranged from 1 to 7 years (Henley and Wiederhold, 1975). Congress's Office of Technology Assessment (OTA) has estimated the development cost for a commercial inpatient MIS to be $25 million and the time to develop such a system to be 10 years (OTA, 1977). Costs related to the acquisition of a technology are discussed later in this chapter. A General Accounting Office (GAO) study on the use of information technology in hospitals found that commercial hospital information systems currently in use were not as comprehensive as those planned by the Department of Defense and the Department of Veterans Affairs (GAO, 1988). The GAO cited two factors that could account for the less extensive development of commercial information systems—which may itself reflect an uncertain market. The first is the potentially small market for such systems.2 The second is the low level of spending for automation in the hospital industry. The causes of this low level are uncertain, but GAO offers several possibilities: the historical lack of incentives for hospitals to minimize costs (a situation that has changed since the implementation of the prospective payment system by Medicare); the difficulties hospitals face in trying to raise initial funds for information systems; the cost savings in early systems were attributed to reductions in the work of clerical rather than clinical staff; the difficulties involved in achieving or quantifying savings (time savings in particular are likely to be fragmented); and resistance by medical personnel to the introduction of information management technology (GAO, 1987). 2   Some experts believe hospitals must have 200 or more operating beds to make optimal use of hospital information systems. According to the GAO report (1988), nearly 70 percent of community hospitals have fewer than 200 operating beds.

OCR for page 138
--> The CPR market may also be uncertain because computer applications are generally not well understood by health care practitioners (Anderson and Jay, 1987). This lack of understanding limits the demand for such products and, as a result, reduces commercial interest in developing new products. Few sources exist to help practitioners learn what a computer technology can do for them, and there is little likelihood such help will be forthcoming in the near future, given the costs associated with producing such resources. Barriers to Diffusion Technological diffusion has been analyzed in greater depth than technological development. Rogers (1987) presents five characteristics of a technology that influence its adoption: relative advantage over existing technologies (the degree to which an innovation is perceived as better than the practice it supersedes); compatibility (the degree to which an innovation is perceived to be consistent with values, experiences, and needs of potential adopters, as well as with the structure of adopting organizations); complexity (the degree to which an innovation is perceived as difficult to understand and use); feasibility3 (the degree to which one can experiment with an innovation on a limited basis); and observability (the degree to which the results of an innovation are observable to others). Other factors also affect CPR adoption and use, including the environment of the health care system; leadership; user behavior, education, and training; costs; social and legal issues; and network needs. Major concerns in these areas are briefly noted below. Environment of the Health Care System The U.S. health care system has been characterized as comprising "thousands of relatively autonomous units, centering on large hospitals, which are themselves made up of relatively autonomous divisions and departments" (Lindberg, 1979:215). Maintaining CPRs, however, "imposes requirements for greater coordination among separate ancillary services, particularly with regard to terminology" (McDonald and Tierney, 1988:3438). 3   Rogers (1987) uses the term trialability to reflect the degree to which an innovation can be experimented with on a limited basis.

OCR for page 138
--> By extension, the use of patient records that are linkable requires coordination among the institutions that are likely to contribute to or use such records. Thus, autonomy within and among provider institutions must be addressed when planning and implementing CPR systems. Disaggregation of care (i.e., the delivery of medical care by many small providers who operate independently of and in competition with one another) has significant implications for the adoption of a coordinative, systemwide technology. Such technology is often subject to nonadditive benefits—that is, the benefits of collaboration among multiple providers outweigh the benefits of individual adoption. As a result, providers have fewer incentives to acquire such technology (NAS, 1979). The reimbursement policies that are applied to providers influence their willingness and ability to acquire CPR technology. For instance, under cost-based reimbursement schemes, providers have more incentive to acquire technologies that are reimbursable than technologies that are not reimbursable. In contrast, prospective payment systems create incentives for institutions to reduce costs—and thus to acquire potentially cost-reducing technologies such as CPR systems. Under current reimbursement policies, any potential acquisition of new technology must contribute to the improvement of a provider's financial status or at least be budget neutral. It should also substantially improve patient care processes, for example, by providing clinical decision support or by giving complete record access to authorized personnel. Leadership Given the fragmented environment of the U.S. health care system, it is not surprising that at present no one organization or agency is leading the effort to establish the necessary infrastructure for national implementation of CPRs and CPR systems. National and regional organizations may be knowledgeable about the issues, but they are not consistently soliciting information from or educating their members about CPRs. Thus, despite the many aspects of CPRs that are in need of coordination, no organization has the operational responsibility and funds to establish programs and projects to set the direction for the health care industry. Overcoming this problem could be the key requirement for progress, and the committee devoted considerable attention to discussing and formulating its primary recommendation in this area (see Chapter 5). User Behavior, Education, and Training Users are more likely to accept a technology if several conditions are met: they have a stake in the system; they can use it at minimum cost; the

OCR for page 138
--> technology produces information leading to improved clinical services; that information is almost immediately available; and the technology increases their status (Young, 1987). Computing applications that do not significantly change the routines associated with the practice of medicine are also more likely to be accepted by users (Kaplan, 1987). An individual's propensity to use a technology has been attributed to specialization, fear of malpractice suits, industry promotion, a specific form of medical practice, and payment methods (Banta, 1987). An encounter with a peer who is already using a technology can also influence an individual's use (Anderson et al., 1987). Young (1987:9) suggests that problems with the acceptance of clinical computing systems relate to the way a physician organizes his or her thought processes and interacts with written aids. [U]se of the medical records is not properly appreciated. The written record is not just a repository of information; it often forms part of the doctor's thought process, so that the style of writing, the position on the paper of particular items, abbreviations, the sequence of information, use of margins, may all have an important significance for the individual practitioner—a significance which goes beyond the actual facts recorded, and which is impossible to capture in an orderly typed record or video display unit. It is the loss of these individual aspects of the medical record which causes most problems. The advantages of structured, typewritten reports commonly do not outweigh the loss of the extra information which is conveyed to the individual practitioner by the above features. Some individuals in the health care community are skeptical that computer systems can be designed to meet user performance and functional requirements. They may also doubt that a CPR system will improve an institution's ability to manage its information. Previous negative experiences with computer systems or inaccurate data generated by a system may cause some providers to actively resist the acquisition and implementation of CPRs. Whether or not an individual clinician is skeptical about a CPR system, using it will require behavioral changes. Thus, as noted earlier, some natural resistance to the CPR is to be expected. Individual institutions will need educational programs to support these changes. Educational curricula for health care professionals must also be modified to reflect the role of the CPR in the provision of health care services. Who will develop, implement, and pay for such materials (e.g., vendors or professional groups) is an area of considerable uncertainty. CPR implementation requires experts who can support CPR users, but at present only a relatively small number of individuals have the necessary expertise in medical informatics (Clepper, 1991). More people must be encouraged to enter this field and a variety of educational and training

OCR for page 138
--> programs tailored to different types of health care professionals must be developed. The committee considered the educational barrier to CPR development sufficiently important to formulate a specific recommendation to address it, as discussed in Chapter 5. In addition to education, practitioners need incentives to use CPRs to enter data and maintain patient records. Perhaps the greatest motivation for practitioners to use CPRs would be to produce evidence that CPRs can help to improve the quality of patient care and reduce the administrative burdens they currently face. As discussed in Chapter 3, at least one institution has experimented with a fee-for-data arrangement for physicians who input their own discharge summaries into an automated system. Other arrangements to encourage the use of CPRs are also feasible. For example, third-party payers could provide incentives for health care providers (including physicians) to submit claims electronically or in a computer-compatible format (e.g., diskette). Alternatively, third-party payers may reject reimbursement claims that do not contain standard data. A regulatory approach is another possibility for use in place of or in addition to an incentive structure. However, the potential side effects and costs of both incentives and constrictive regulations must be understood and carefully weighed. Costs A major factor influencing a firm's adoption of a technology is the size of the investment required relative to the size of the firm. Acquisition costs for CPR systems are likely to be substantial but are difficult to estimate.4 This difficulty arises because the purchase or lease price of a system does not reflect the total implementation cost; it excludes the cost of training and potential losses of productivity during transition to the system. Studies that have attempted to estimate total costs have tended to focus on MISs rather than CPR systems. Further, purchase or lease prices for CPR systems vary significantly, depending on the scope of functions a system offers, the size of an institution, and an institution's previous level of automation. One cost analysis of the implementation and operation of an automated ambulatory care medical record system found that the cost per patient encounter of a computer-based system was 26 percent greater than the direct costs associated with operation of a manual system. However, the manual system failed to access 18 percent of the records requested within the demand 4   A confidential survey of CPR vendors conducted by the Technology Subcommittee of this IOM study committee revealed that purchase or lease costs for a CPR system range from $2 million to $6 million for a medium-sized hospital (see the appendix to Chapter 3).

OCR for page 138
--> time, whereas the automated system provided access to 100 percent of the records requested. Quantification of the access benefit reduced the difference in costs between the two systems (Koster et al., 1987). This cost analysis illustrates another major difficulty faced by health care provider institutions deciding on acquisition of a CPR system. Decisions about whether to acquire technology are based in part on information about the benefits of the technology; once again, data on CPR system benefits are sparse. Few recent studies have analyzed actual costs and benefits. The studies that have been conducted address MISs rather than CPR systems; they also focus on projected rather than actual experience or include only those benefits that can be measured in terms of dollars and exclude such benefits as improved quality of care or reduced waiting time for patients. The costs of acquiring and operating CPRs and CPR systems will be borne primarily by practitioners and health care provider institutions. Yet the benefits of these systems will accrue to patients, practitioners, health care provider institutions, third-party payers, researchers, policymakers, and the public. These cumulative benefits of CPR systems should exceed the benefits individual practitioners and health care providers might be expected to gain. As such, CPR systems in certain respects represent a public good. Given today's strict budget constraints, health care provider institutions (including physicians' offices) must choose among alternative technologies when allocating resources. Compared with a CPR system, other technologies could offer greater monetary benefits to an individual institution, albeit lower combined or social (public good) benefits. When this situation prevails, provider institutions have fewer incentives to invest in CPRs and CPR systems. Nonclinical data users (e.g., third-party payers and researchers) could also incur costs from CPR implementation. They may need to modify existing systems or acquire new ones to be compatible with CPR standards; they may also need to revise procedures for handling computer-based data and develop training programs for personnel. The cost of CPR technology, like the cost of computer technology in general, may well decrease over time. Moreover, CPRs may reduce the costs of care enough to offset the expense of acquiring and operating CPR systems, although this remains to be proved. The committee was quite concerned about the immediate barriers to CPR implementation raised by potentially substantial costs for full acquisition, installation, and operation of CPR systems across the nation; it was also concerned about the distribution of these costs. Consequently, it formulated a specific recommendation to address these matters, which is presented in Chapter 5.

OCR for page 138
--> Legal and Social Issues The legal and social issues involved in implementing CPRs and CPR systems are formidable. State licensure requirements for hospital medical records are obsolete, ambiguous, and nonuniform (Waller, in this volume). The wide variation among states in hospital licensure requirements for medical records makes it difficult to develop CPR systems that comply with licensure laws in all 50 states; this factor in turn hinders the development of CPR formats that can be used nationally. Failure by a vendor to establish a CPR system's compliance with one or more state licensure requirements may adversely affect the system's marketability. The differences across states could be so great as to make national implementation impossible. Hospital licensure laws and regulations in some states still assume a paper patient record, which makes the legal status of CPR systems unclear. Other state laws and regulations appear to permit use of some forms of automation but not others, or the use of automation for some, but not all, patient record functions. It is not clear whether regulations requiring that records be kept in ink or be typewritten permit the creation of records electronically or the use of lasers. Further, the regulations in many states require that medical records be signed but are silent on whether a computer key or code can be substituted for a signature. Issues of record ownership, responsibility, and control must also be addressed. The physical records are the property of the provider institution at which they were created. In addition, provider institutions currently are responsible for ensuring the accuracy and completeness of record contents (Amatayakul and Wogan, 1989). As data are transmitted and shared among institutions, all with the ability to add and update information from variety of settings, the principles of ownership become blurred. It is not clear who would maintain records for routine use and maintain the structure of the record to accommodate new terminology and data elements for new diseases, treatments, tests, and approaches to health care. The locus of control of access to patient data is another unanswered question. Current laws concerning disclosure of and access to patient record information vary from state to state, further complicating the transfer of patient information across state lines. Perhaps the impediment to CPRs that is of greatest concern is the issue of privacy. The computerization of most types of record keeping, as well as the recent well-publicized cases of inappropriate access by computer hackers, has increased concerns about the misuse of personal information (Westin, 1976; Privacy Protection Study Commission, 1977; Peck, 1984; Agranoff, 1989). A system in which patient data, including sensitive information (e.g., human immunodeficiency virus test results, data on psychiatric treatment),

OCR for page 138
--> can be accessed more easily may be strongly resisted in some quarters. 5 There are several aspects to concerns about confidentiality of information and patient privacy. Ultimate responsibility for protecting the privacy of patient data that are shared among multiple users has not been defined; in addition, generally accepted standards for the protection of computer-based data do not exist (Agranoff, 1989; NRC, 1991). The consequences for breaches of confidentiality vary by state and in some cases need to be stronger than they are at present (Waller, in this volume). The National Conference of Commissioners on Uniform State laws drafted the Uniform Health-Care Information Act to address issues of patient privacy, patient access, disclosure of patient information to third parties, and transfer of such information across state lines (National Conference of Commissioners on Uniform State Laws, 1986). As of this writing, however, only one state had adopted this act (Waller, in this volume). Concerns about patient privacy affect more than the security features of CPR systems and legal remedies. A consistent personal identification number (PIN) in all patient records would facilitate record linkage across time and provider institutions (National Center for Health Statistics, 1990; Washington Business Group on Health, 1989). Nevertheless, despite its operational attractiveness to researchers and other patient record users, the PIN raises concerns about the increased potential to invade patient privacy (Washington Business Group on Health, 1989). Currently, the Food and Drug Administration (FDA) device regulations and authorities do not apply to computer products intended only for use 5   Westin's 1976 study of the impact on citizen rights of computers in the health care field found that ''most cases of actual harm involving individuals were still arising from manual records" and concluded that the main problem with the use of computer-based patient records involved "potential harm—the creation of health data systems that many health professionals, citizen groups, and individuals directly affected by such systems consider to be threats to basic rights" (Westin, 1976:xvi, emphasis in original). As noted by Lindberg (1970), however, the public and its elected legislators must have their anxieties allayed about potential misuses of data. Some observers have suggested that computer-based patient record systems will offer greater confidentiality for patient information because they can limit the information that various users can see. For example, administrative or financial personnel could be prevented from seeing sensitive diagnostic or treatment data. Further, CPR systems could provide an audit trail listing those personnel who accessed a particular record (Hard, 1990). As discussed in Chapter 3, however, existing security technology frequently has been not applied to current CPR systems. Moreover, a recent National Research Council (NRC) study on computer system security concerns concluded that several trends reflected a growing potential for system abuse. Among these trends are the proliferation of networking and embedded systems, the widespread use of databases containing personal information, and the widespread ability to use and abuse computers (NRC, 1991).

OCR for page 138
--> critical resources, it may not be the ideal agency to champion patient record development. If HCFA were to lead the CPR effort, health care providers might not be able to separate it from HCFA's other activities, particularly regulation of the Medicare program, cost containment, and related efforts. Providers might resist the CPR development initiative because of a perception that it was intended as a cost-containment mechanism rather than as a way to improve health care delivery. The Departments of Defense (DoD) and Veterans Affairs (VA) are both making significant investments in comprehensive medical information systems.17 As a result, they have a great deal of expertise in designing and implementing such systems. These departments do not seem likely candidates for leading a CPR initiative, however, because of their restricted populations and other non-health care responsibilities. In short, no existing federal organization simultaneously has the mandate, mission, credibility, and resources to take on this responsibility. Private Sector Sponsorship The committee also considered a private sector, not-for-profit membership organization, similar to the Joint Commission on the Accreditation of Healthcare Organizations, to lead the CPR development effort. A decided advantage to a purely private approach is that the conversion to computer-based records would become something championed by, rather than imposed on, these organizations and their constituents. The active involvement of such organizations might prompt them to provide core funding, and the efforts to secure funding from others might then be more effective. In addition, staff recruitment for a private organization would be facilitated by the broad-based support the organization would receive from national health care groups. This strategy has its limitations, however. Most important is that such an organization would lack the power and authority implicit in a governmental entity. Further, federal funding could be difficult to obtain unless key agencies played a major role in the organization's activities. Public-Private Commission or Consortium In theory, a public-private organization would offer the advantage of involving and being able to draw on funding and personnel resources from 17   DoD is acquiring the Composite Health Care System for installation in its 167 hospitals and nearly 600 clinics. The costs for full deployment of this system are expected to be $1.6 billion. The VA is installing the Decentralized Hospital Computer System to support its 172 hospitals and 358 outpatient facilities. The VA estimates that this system as currently defined will cost $925 million (GAO, 1988, 1990).

OCR for page 138
--> both sectors. Furthermore, because the problems with patient records affect both public and private organizations, acceptance of a solution would be more likely if both sectors were involved in the decision making process. The drawbacks of a public-private entity should not be underestimated, however. Managing the diverse interests that would be represented in such an organization presents a major challenge. Certain federal agencies already have charters that would overlap the charge to such an organization. In addition, a purely private sector organization might offer more entrepreneurial agility than a hybrid group. Perhaps the biggest drawback of a public-private organization is the inherent instability of such an approach. Lacking a federal mandate and given the less-than-immediate contribution to the profitability of private sector participants concerned with CPR development, a public-private organization may not command sufficient resources and attention to address effectively the barriers to CPRs and CPR systems. Preferred Approach The committee recognizes that the federal sector has considerable resources (including authority and knowledge) to influence CPR development. For example, HCFA can establish reimbursement mechanisms that reward providers who submit insurance claims generated by CPR systems. AHCPR is expressly mandated to improve patient data for research. The VA and DoD have gained considerable experience in CPR development and implementation. NLM has made significant contributions to the management of medical knowledge for practitioners. Nevertheless, several factors militate against a purely federal approach. First, the resources, potential change agents, and stakeholders that must be coordinated or engaged in CPR development are present in both the public and private sectors. Thus, a structure is needed that can draw from both sectors. Second, the committee believes that routine use of the CPR can be achieved most efficiently through a collaborative process that develops consensus on key issues (e.g., data and security standards, the minimum content of CPR systems) yet allows flexibility at the local level to foster innovation in the development and use of CPRs. Third, the committee believes that patient care should be the primary focus of CPR development and implementation. Practitioner use of CPR systems requires that the systems meet practitioner needs, and only if practitioners are willing to use CPR systems to capture data and to secure assistance in clinical decision making can the benefits of CPRs for moderating costs and conducting research be realized. It is essential that practitioners view the CPR as a valuable resource for improving patient care. Thus, CPR efforts must involve health care providers as well as federal agencies.

OCR for page 138
--> CPR design should not be driven solely by governmental objectives—for example, those embodied in the cost-containment and health outcomes research missions of HCFA and AHCPR, respectively. Although those agencies are likely to have significant interest in CPR development, their primary mission is not patient care. The committee concluded that a public-private approach would be optimal in the long run and, as elaborated in Chapter 5, proposes the establishment of a Computer-based Patient Record Institute (CPRI). It was also of the view that the potential base of funding in the private sector is not sufficiently solid to provide adequate support for a new organization at this time. This judgment was founded on its review of the history of CPR development and on a poll of the participants at the workshop noted earlier. As a result, the committee also concluded that a federally initiated and funded approach would be most appropriate for necessary interim activities. Immediate action is needed to advance CPR efforts and to lay the groundwork for an organization such as the CPRI that would ultimately coordinate the necessary infrastructure for a national CPR system. Many of the barriers to CPR implementation relate to lack of information; part of the interim effort thus should focus on education and evaluation. Standards development and representation of the interests of health care in the national high-speed computer network discussed earlier should also be given high-priority attention. The overall goal of such efforts would be, within five years, to turn over operational issues to a public-private organization that is supported mainly by its members. The committee noted that if the private sector failed to support CPR efforts adequately, the federal government might still be sufficiently motivated to advance the CPR unilaterally. Long-term dominance by the federal government in this area could result in an approach that was more regulatory and bureaucratic than collaborative and innovative. To preclude such an eventuality, the committee placed special importance on joint public and private sector progress. This approach is consistent with the recent General Accounting Office report on automated medical records (GAO, 1991), which made two recommendations to the secretary of Health and Human Services. First, as part of DHHS's mandate to conduct research on outcomes of health care services, the secretary should "direct the Public Health Service, through its Agency for Health Care Policy and Research, to support the exploration of ways in which automated medical records can be used to more effectively and efficiently provide data for outcomes research" (GAO, 1991:26). Second, as part of the effort to support outcomes research, the secretary should "develop a plan and budget for consideration by the Congress, to bring about the greater use of automated medical records" (GAO, 1991:26). Specific elements of such a plan could include "a national forum that

OCR for page 138
--> sets goals for automating medical information, addresses individual and organizational concerns with automated records, and identifies incentives to induce health care organizations to increase their use of automation" (GAO, 1991:26). Specific Steps for Change Agents The challenges of developing affordable CPR systems that are acceptable to users and of achieving widespread use of such systems within a decade should not be underestimated or understated. Much of the progress toward these goals is likely to occur incrementally over time and across the country. The CPRI can play an important role in tracking progress and directing future efforts, but significant contributions to CPR development and implementation can and must be made by individuals and groups other than the CPRI. A great deal of work can be accomplished at the regional, local, and institutional levels in preparation for CPR implementation. As discussed earlier in this chapter, health care professionals could support development and implementation by helping to plan and conduct research or demonstrations of CPR systems. Involvement of all kinds of CPR users is needed—especially practitioners as the primary source of data—to design systems that will meet their requirements. Professional societies could implement formal education and awareness programs as part of their membership mailings and annual meetings. They could also support conferences at which CPR users could share experiences, report on useful experiments in various settings, and meet with other professional disciplines to discuss data and function needs. Purchasers of CPR systems could actively seek systems that are able to meet basic data-exchange and security standards and offer sufficient capacity to evolve over time. Insurers could offer incentives (e.g., faster payment of claims) for data that are provided in electronic form. Health care business coalitions, chambers of commerce, and major employers could all support CPR development and implementation efforts by supporting research and pilot demonstrations as well as by developing relationships with insurers and health care provider institutions that use or support CPR systems. Federal agencies could provide substantial funding for research and development and support standards development through funding or regulatory mandate. States could serve as candidates for pilot regional studies or experimental prototypes. Health care accreditation organizations could foster CPR development,

OCR for page 138
--> acquisition, and use by setting standards for accreditation that are most effectively met through CPR systems. In their training of health care professionals (including continuing education), professional schools could both shape attitudes toward and provide the skills for using CPRs. They could also foster the development of CPR systems to the extent that the agenda of researchers is influenced by the schools' provision of space and support. Researchers could study the costs and benefits of CPR systems, at both micro and macro levels, including the impact of CPR systems on the quality and costs of care. State agencies, hospital associations, and local professional groups could establish working groups to develop a common understanding and vision of how CPRs could support their health care environment and to identify the elements of local and regional infrastructure needed to support future CPRs. Working groups could study the relationships among referring physicians and among other providers to understand their entire system of health care. They could also define needed data elements, educate local health care professionals on health care information management issues, and monitor progress in the development of standards for security and data exchange. Summary In addition to technological advances, successful implementation of CPR systems requires elimination of the barriers to development and diffusion. It also requires that the concerns of stakeholders be addressed and that potential change agents be engaged. Many impediments to the CPR arise from a lack of awareness and understanding of CPRs and their capabilities. System purchasers and users lack adequate information about the benefits and costs of CPRs. In particular, developers and vendors require more specific information about what users want from systems and what price providers would be willing to pay for systems that meet their needs. Activities aimed at improving (e.g., demonstration projects) and disseminating (e.g., education programs) available information about CPR systems constitute an important step for CPR implementation. Other impediments arise from the lack of an infrastructure to support CPR development and diffusion. Such an infrastructure comprises standards for communication of data (i.e., vocabulary control and data format standards); laws and regulations that protect patient privacy but do not inhibit transfer of information to legitimate users of data outside the clinical setting; experts trained in the development and use of CPR systems; institutional, local, regional, and national networks for transmitting CPR data; reimbursement mechanisms that pay for the costs of producing improved

OCR for page 138
--> patient care information; and a management structure (i.e., an organization) for setting priorities, garnering and allocating resources, and coordinating activities. Removal of these impediments is essential to the timely development and implementation of CPR systems. References Agranoff, M. H. 1989. Curb on technology: Liability for failure to protect computerized data against unauthorized access. Computer and High Technology Law Journal 5:263–320. Amatayakul, M., and M. J. Wogan. 1989. Fundamental considerations related to the Institute of Medicine Patient Record Project. Paper prepared for the Institute of Medicine Committee on Improving Patient Records in Response to Increasing Functional Requirements and Technological Advances. Anderson, J. G., and S. J. Jay. 1987. The diffusion of computer applications in medical settings. Pp. 3–7 in Use and Impact of Computers in Clinical Medicine, ed. J. G. Anderson and S. J. Jay. New York: Springer-Verlag. Anderson, J. G., S. J. Jay, H. M. Schweer, M. A. Anderson, and D. Kassing. 1987. Physician communication networks and the adoption and utilization of computer applications in medicine. Pp. 185–199 in Use and Impact of Computers in Clinical Medicine, ed. J. G. Anderson and S. J. Jay. New York: Springer-Verlag. Banta, H. D. 1987. Embracing or rejecting innovations: Clinical diffusion of health care technology. Pp. 132–160 in Use and Impact of Computers in Clinical Medicine, ed. J. G. Anderson and S. J. Jay. New York: Springer-Verlag. Blendon, R. J. 1988. The public's view of the future of health care. Journal of the American Medical Association 250:3587–3593. Booz-Allen and Hamilton. 1990. Response to the Strategy and Implementation Subcommittee Report to the Institute of Medicine Committee on Improving the Patient Record. Bethesda, Md. Bradbury, A. 1990. Computerized medical records: The need for a standard. Journal of the American Medical Association 61:25–35. Brannigan, V. M., and R. E. Dayhoff. 1986. Medical informatics: The revolution in law, technology, and medicine. Journal of Legal Medicine 7:1–53. Bronzino, J. D., V. H. Smith, and M. L. Wade. 1990. Medical Technology and Society: An Interdisciplinary Perspective. New York: McGraw-Hill Publishing Company. Clepper, P. 1991. Communication to committee. National Library of Medicine, Bethesda, Md. Denis, S., and Y. Poullet. 1990. Questions of liability in the provision of information services. On-line Review 14:21–32. Dowling, A. F. 1987. Do hospital staff interfere with computer system implementation? Pp. 302–317 in Use and Impact of Computers in Clinical Medicine, ed. J. G. Anderson and S. J. Jay. New York: Springer-Verlag. Food and Drug Administration. 1989. Draft Policy for the Regulation of Computer Products (photocopy). Washington, D.C. Gabrieli, E. R., and G. Murphy. 1990. Computerized medical records. Journal of the American Medical Association 61:26–31.

OCR for page 138
--> GAO (General Accounting Office). 1987. ADP Systems: Examinations of Non-federal Hospital Information Systems. IMTEX-87-21. Washington, D.C. June. GAO. 1988. Use of Information Technology in Hospitals. Statement of Melroy D. Quasney, Associate Director, Information Management and Technology Division, before the Subcommittee on Education and Health, Joint Economic Committee. T-IMTEC-88-4. Washington, D.C. May 24. GAO. 1990. Defense's Acquisition of the Composite Health Care System. Statement of Daniel C. White, Special Assistant to the Assistant Comptroller General, before the Subcommittee on Military Personnel and Compensation, Committee on Armed Services, U.S. House of Representatives. T-IMTEC-90-04. March 15. GAO. 1991. Medical ADP Systems: Automated Medical Records Hold Promise to Improve Patient Care. Washington, D.C. Gould, S. 1990a. CRS Issue Brief: Building the National Research and Education Network. Washington, D.C.: Library of Congress. Gould, S. 1990b. CRS Report for Congress: The Federal Research Internet and the National Research and Education Network. Prospects for the 1990s. Washington, D.C.: Library of Congress. Hard, R. 1990. Computers help keep patient files confidential. Hospitals 179:49–50. Henley, R. R., and G. Wiederhold. 1975. An Analysis of Automated Ambulatory Care Medical Record Systems [AARMS]: Findings. AARMS Study Group, University of California, San Francisco. June. IOM (Institute of Medicine). 1990a. Clinical Practice Guidelines: Directions for a New Program, ed. M. J. Field and K. N. Lohr. Washington, D.C.: National Academy Press. IOM. 1990b. Medicare: A Strategy for Quality Assurance, vols. 1 and 2, ed. K. N. Lohr. Washington, D.C.: National Academy Press. Kaplan, B. 1987. The influence of medical values and practices on medical computer applications. Pp. 39–50 in Use and Impact of Computers in Clinical Medicine, ed. J. G. Anderson and S. J. Jay. New York: Springer-Verlag. Koster, A., F. L. Waterstraat, and N. Sondak. 1987. Automated and ambulatory record systems: A comparative costs analysis. Journal of the American Medical Record Association 58:26–31. Krakauer, H. 1990. The uniform clinical data set. Pp. 120–133 in Effectiveness and Outcomes in Health Care, ed. K. A. Heithoff and K. N. Lohr. Washington, D.C.: National Academy Press. Lee, P. R., P. B. Ginsburg, L. R. LeRoy, and G. T. Hammons. 1989. The Physician Payment Review Commission report to Congress. Journal of the American Medical Association 261:2382–2388. Lindberg, D. A. B. 1970. A statewide medical information system. Computers and Biomedical Research 3:453–463. Lindberg, D. A. B. 1979. The development and diffusion of a medical technology: Medical information systems. Pp. 201–239 in Medical Technology and the Health Care System: A Study of the Diffusion of Equipment-Embodied Technology. Washington, D.C.: National Academy of Sciences. Lindberg, D. A. B., and B. L. Humphreys. 1990. The Unified Medical Language System and the Automated Patient Record. Paper prepared for the Institute of

OCR for page 138
--> Medicine Committee on Improving Patient Records in Response to Increasing Functional Requirements and Technological Advances. Masys, D. R. 1990. Of codes and keywords: Standards for biomedical nomenclature. Academic Medicine 65:627–629. McDonald, C. J., and W. E. Hammond. 1989. Standard format for electronic transfer of clinical data. Annals of Internal Medicine 110:333–335. McDonald, C. J., and W. M. Tierney. 1988. Computer-stored medical records: Their future role in medical practice. Journal of the American Medical Association 259:3433–3440. Metzger, M. C. 1988. Legal implications of computer-aided medical diagnosis. Journal of Legal Medicine 9:313–328. NAS (National Academy of Sciences). 1979. Medical Technology and the Health Care System: A Study of the Diffusion of Equipment-Embodied Technology. Washington, D.C.: National Academy of Sciences. NAS. 1989. Information Technology and the Conduct of Research: The Users' View. Washington, D.C.: National Academy Press. National Association of Health Data Organizations (NAHDO). 1988. NAHDO Resource Manual. Washington, D.C. March. National Center for Health Statistics. 1990. The National Committee on Vital and Health Statistics, 1989. Hyattsville, Md. National Conference of Commissioners on Uniform State Laws. 1986. Uniform Health-Care Information Act. Chicago, Ill. National Science Foundation. 1990. Report on Selected Congressional Activities . 101st Cong., 2d sess. NSB-90-117. Washington, D.C.: Office of Legislative and Public Affairs. December. Norris, J. 1991. Communication to committee. Drug Utilization Review and Point of Sale Systems. Hill and Knowlton, Waltham, Mass. March. NRC (National Research Council). 1991. Computers at Risk: Safe Computing in the Information Age. Washington, D.C.: National Academy Press. OTA (Office of Technology Assessment). 1977. Policy Implications of Medical Information Systems. OTA-H-56. Washington, D.C.: U.S. Government Printing Office. Peck, R. S. 1984. Extending the constitutional right to privacy in the new technological age. Hofstra Law Review 12:893–912. Privacy Protection Study Commission. 1977. Personal Privacy in an Information Society. Washington, D.C.: U.S. Government Printing Office. ProPAC (Prospective Payment Assessment Commission). 1990. Medicare Prospective Payment and the American Health Care System: Report to Congress. Washington, D.C. June. Rogers, E. M. 1987. Diffusion of innovations: An overview. Pp. 113–131 in Use and Impact of Computers in Clinical Medicine, ed. J. G. Anderson and S. G. Jay. New York: Springer-Verlag. Teach, R. L., and E. H. Shortliffe. 1981. An analysis of physician attitudes regarding computer-based clinical consultation systems. Computers and Biomedical Research 14:542–558. U.S. Congress. 1989. Congressional Record, vol. 135, no. 165, pt. 3. November 21. Washington Business Group on Health. 1989. Fostering Uniformity for Health Care Assessment Data Gathering. Washington, D.C.

OCR for page 138
--> Westin, A. F. 1976. Computers, Health Records and Citizen Rights. National Bureau of Standards Monograph 157. Washington, D.C.: U.S. Government Printing Office. Willick, M. S. 1986. Professional malpractice and the unauthorized practice of professions: Some legal and ethical aspects of the use of computers as decision-aids. Rutgers Computer and Technology Law Journal 12:1–32. Young, D. W. 1987. What makes doctors use computers? Discussion paper. Pp. 8–14 in Use and Impact of Computers in Clinical Medicine, ed. J. G. Anderson and S. J. Jay. New York: Springer-Verlag. Appendix: Maternal and Child Health Care and Computer-Based Patient Records As discussed in Chapter 3, much of the progress to date in implementing CPRs and CPR systems has occurred in hospitals and large multispecialty practices, particularly in health maintenance organizations. To be successful, however, CPRs and the larger computer-based systems in which they function must be useful and practical for many other types of providers, including, for instance, community-based clinics and other outpatient facilities serving either primary care or special health care needs. In this category might fall services to disadvantaged populations such as those eligible for maternal and child health care through Title V of the Social Security Act. The area of maternal and child health (MCH), both generally and as administered by the Health Resources and Services Administration (HRSA) in support of the Title V mandate, poses particularly interesting challenges and opportunities for CPRs and CPR systems. Five topics are briefly discussed here as examples. First, the challenges involve, among other matters, the capability to track populations that typically do not have regular providers of care; that move in and out of eligibility for the services; that move in and out of geographic areas (e.g., states, underserved regions) where the services are provided; and that are in many ways socially, economically, and demographically disadvantaged. In such circumstances, ease of data entry, storage, and transmission of clinical and sociodemographic information takes on special importance. CPRs can maximize these attributes, especially in comparison to traditional paper records that are often institution-specific, fragmented, illegible, and lacking in information related to social support, health education, and similar nonclinical issues. Another important factor in reaching these populations is coordination of care across many different sites. Computer-based systems will facilitate timely, accurate movement of necessary information and collaboration among those delivering both clinical and social services. Second, MCH programs are expected to reach out to obstetricians, gynecologists, and pediatricians throughout the states, to develop integrated

OCR for page 138
--> MCH service delivery programs, to improve home visit programs (including those concerned with case management), to collaborate with inpatient institutions delivering care to children with special health care needs, and to improve services to rural populations. CPR systems should make it easier to achieve these objectives by making clinical and sociodemographic data more readily available even in relatively nontraditional settings. Certainly the expectation of CPR developers and innovators is that their technologies, when broadly implemented, will offer capabilities far beyond what can be achieved with today's paper medical charts, particularly in terms of longitudinal records in which, for instance, information about immunizations and screening services must be maintained over a considerable period of time. A third major challenge facing HRSA and its MCH bureau is the required implementation of relatively recent mandates of Title V that call for development and maintenance of a data and information system. Part of this system presumably would involve the acquisition of pertinent epidemiological data that can be used by the bureau and others to improve and promote health and prevent disease among mothers, infants, children, and others eligible for MCH block grant services. To the extent that CPRs and CPR systems allow institutional providers and practitioners (e.g., physicians, visiting nurses, clinic staff) to enter data once (and only once) for both clinical and administrative purposes, they will considerably simplify the reporting requirements of these programs. Moreover, the expectation is that CPR systems ultimately will communicate with each other with the aid of a composite clinical data dictionary (CCDD); this means that states, and the providers and clinicians who operate in them, will not need to conform rigidly to a single federal data system in meeting HRSA's reporting requirements (because the federal system itself would be a party to the CCDD). A fourth way in which the movement toward CPRs and CPR systems may be important for MCH activities concerns the support of research. As a general proposition, researchers are more accustomed to the use of computers for data gathering and analysis than are clinicians, so movement toward the implementation of such systems presumably would be regarded as welcome progress by those involved in MCH research. Indeed, one specific aim of research grants in this area is to use automated systems to facilitate the management and delivery of health care for target populations. One special aspect of the research effort involves infant mortality and Medicaid (Title XIX) services. The secretary of Health and Human Services is expected to develop a national data system for linking vital statistics (birth and death) records with information on Medicaid insurance claims forms. This task clearly lies within the realm of CPR development. In a related vein, special projects may be mounted to enhance family-centered and community-based health care, both within and across states.

OCR for page 138
--> These projects could benefit greatly from a capacity to acquire large amounts of information, locate those data securely in a single site (for analysis and archival purposes), and make them available to properly authorized users for analysis (and perhaps for clinical purposes as well). The collection, storage, maintenance, and use of such information through computer-based systems, rather than through paper records (or paper records secondarily data-entered into computer files), can be expected over the long run to promote more comprehensive and more productive special projects on this and other complex topics (e.g., services to children with serious impairments and handicaps such as spina bifida, debilitating chronic illnesses, or cleft lip and palate). The fifth point relates more specifically to the potential value of communication and collaboration across federal agencies. Several federal departments have already taken steps to design and implement one or more components of a CPR system. Notable among these are the Departments of Defense (DoD) and Veterans Affairs (VA). Both departments have health care delivery responsibilities for distinct, and sometimes quite dispersed, populations that have both traditional primary care needs (e.g., for screening and prevention) and health care problems at least equivalent in complexity and severity to those facing disadvantaged MCH populations. Both departments also have considerable health research programs. Lessons from their efforts to date may prove helpful in planning computer-based systems to serve MCH needs; conversely, issues that arise in planning for and delivering services to disadvantaged MCH populations might be posed to DoD and the VA as a means of bringing difficult technical questions to their attention. In addition, the Health Care Financing Administration (HCFA) has mounted an interesting effort to develop a ''uniform clinical data set"; although it is oriented toward hospital care (and care for the elderly), its developmental history to date offers useful and perhaps cautionary lessons for others attempting to develop clinical data sets and data dictionaries. Certainly it would be important for HCFA developers to understand the special needs and perspectives of a computer-based data set oriented more toward Medicaid than toward Medicare. Finally, the Agency for Health Care Policy and Research (AHCPR) has a specific congressional mandate to work toward the development of computer databases that will serve broad clinical evaluation purposes. Part of AHCPR's interest has been specifically in computer-based record systems, including the potential for a public, or public-private, entity to undertake many technical, legal, and other tasks related to the establishment of CPRs and CPR systems nationally over the next decade. Because the MCH component of HRSA will have wide concerns about quality of care, outcomes of care, and similar issues, it would seem prudent for HRSA to work together from the outset with a sister Public Health Service agency on many of these subjects.