Greg Walton

Any strategy for success in health care must be developed with the goal of improved quality (of both clinical care and business performance) foremost in mind. A competitive market and shrinking reimbursement have placed efficient business practices at the forefront of many health care professionals’ attention, while the rise of consumerism and recent reports from the Institute of Medicine on medical errors and other health care quality shortcomings have intensified interest in improving the quality of patient care.

Clinical information systems (CIS) offer, perhaps, the best set of tools for achieving these quality goals and are at the heart of health care information technology. Initially developed as outgrowths of existing financial information systems, CIS can potentially deliver opportunities to improve outcomes and reduce errors, as well as to control costs by realizing a host of efficiencies in clinical data entry and patient care.

CIS in Context

H. Stephen Lieber

A comprehensive CIS should not be viewed as one stand-alone system or technology, but as a conglomerate of systems, medical equipment, and technologies working together. The linchpin of a CIS is the elusive computer-based patient record.

One barrier to the effective design and implementation of CIS and the computer-based patient record is the lack of a clear definition of these terms and the functions that they encompass. Generally speaking, a CIS may be defined as an automated system serving “as a tool to inform clinicians about tests, procedures, and treatment in an effort to improve quality of care through real-time assistance in decision making and to increase efficiency and decrease unnecessary utilization.”1

On the other hand, the computer-based patient record is generally defined more broadly (see table 1), including not only clinical data and decision support, but maintaining “information about an individual’s lifetime health status and health care (across multiple episodes of care).”1

Perhaps the most convenient thumbnail distinction between CIS and the computer-based patient record is that a CIS focuses on care, while the computer-based patient record focuses on health (not only across settings and episodes of care, but outside institutional walls, as well). For example, a computer-based patient record, with links to appropriate databases, can be an effective component of a system to combat the public-health effects of bioterrorism.

Another useful framework for thinking about CIS and the computer-based patient record was developed by a consulting group.2 Using this framework, systems that interact with computer-based patient record systems to facilitate patient care can be classified as:

• context systems (such as registration systems and managed care systems) that contribute information to computer-based patient record (CPR) systems;

• cooperating systems (such as ancillary clinical systems, pharmacy systems, and scheduling systems) that both provide information to and receive information from computer-based patient record  systems; or

• subscriber systems (such as financial and reporting systems) that require information from computer-based patient record systems.

Computer-based patient record systems, themselves, ideally are capable of handling clinical documentation and data capture, data display, data storage, clinical work flow, clinical decision support, knowledge management, security, communication, and processing.3

Much as the health care system itself is segmented, but becoming conglomerated; so, too, must the various information systems currently in place in diverse settings be conglomerated to represent one consistent, cohesive data set. There must be easy access to reliable information, and systems must address government-mandated security and confidentiality requirements. Strategies to achieve these goals must be developed based on a firm understanding of current CIS capabilities, as well as on an understanding of coming developments that will allow an organization to capitalize on these developments most efficiently.

Current CIS Implementations

Reports are mixed about current implementations of CIS, in general, and the computer-based patient record, specifically. According to the 2001 HIMSS Leadership Survey,4 64% of respondents indicated that a CIS is the application that will be most important to their organizations over the next 2 years, and 66% of respondents said that their organizations are moving forward or have moved forward to implement a computer-based patient record solution set.

Greg Walton

Only 13% of respondents, however, said that their organizations have a fully functioning computer-based patient record, and the computer-based patient record ranked fourth among systems that respondents said would be most important to their organizations over the next 2 years. The consulting group reports that only 5% of clinicians “use any type of computing device at the point of care.”3

As for the systems themselves, current computer-based patient record systems are useful, but evolving, and they are still short of the goal of sophisticated systems that can truly assist clinicians in making patient-care decisions. The consulting group3 has defined five generations of computer-based patient records (table 2), and its research indicates that current systems are, at best, of the second generation (facilitating some access to, and documentation of, clinical data, but providing little decision support).

In light of the health care industry’s intense efforts to reduce the number of medication errors, one key component of a computer-based patient record is worth highlighting: computerized physician-order entry (CPOE). CPOE systems are “electronic prescribing systems that intercept errors when they most commonly occur-at the time medications are ordered.” Such a system can, among other benefits, provide prompts that warn against the possibility of drug interaction, allergy, or overdose; provide accurate, up-to-date information about new drugs; and provide drug-specific information.5 The key to this concept is having the physician using the device for all orders, whether they are for medications or not. The benefits of CPOE use include cutting out low-valued, manual steps while bringing the power of computers to physicians, other caregivers, and the organization’s clinical work flow.

Despite the potential benefits of such systems, very few provider organizations have implemented CPOE systems (and those have often been implemented in limited settings). Recognizing the potential benefits, many organizations have made implementing CPOE and other computer-based patient record applications a high priority, despite the cost and challenges involved. The Leapfrog Group6 recently developed an evaluation tool that should help organizations select and implement CPOE systems.

CIS Challenges

While the foundation seems to be in place for the development of a comprehensive CIS, obstacles remain to be overcome. Perhaps the central challenge facing CIS implementation and improvement is the necessity of successfully communicating a cohesive, ideal vision of the system to all stakeholders (including vendors, end users, and organizational implementation teams), accompanied by the need to inculcate and nurture the required degree of stakeholder support. Not only the vision, but the strategies to be implemented to realize that vision, must be clearly articulated and communicated, and organizational leaders must be united in their vision. Most important, statements regarding their vision must be backed up with action and initiative. Stakeholders at all levels must see concrete evidence of leadership’s commitment to the new system to ensure their commitment in response.

The full commitment of one key project leader is necessary to the successful implementation of a comprehensive CIS. This individual must be given full accountability for the success of the project, as well as the support and authority necessary to achieve that success.7 Such a project leader must understand both the vision that the leadership team has for the system and the individual culture of the organization. Project-management needs include the formation of a committee to represent every group of stakeholders. This will secure their cooperation, as well as ensuring that no key operational concerns are overlooked. Time frames are, of course, always an issue. Insofar as it is possible, those responsible, on the ground level, for implementing the CIS must provide an accurate, detailed picture of the tasks involved and the time required. They must be allotted the time to complete these tasks, without facing pressure to cut corners in order to meet deadlines. Delayed implementation dates, while never desirable, are usually much less costly than reengineering systems after implementation to solve problems.

H. Stephen Lieber

Electronic data sharing of any kind raises concerns of security and confidentiality. These concerns are particularly prominent for clinical data, given its very personal nature. Existing challenges in protecting security while allowing for increased ease of data retrieval have been significantly complicated by the Health Insurance Portability and Accountability Act of 1996 (HIPAA). Specific legal and ethical issues vary from state to state, from specialty to specialty, and (in fact) from caregiver to caregiver. Incorporating these variations into a cohesive, comprehensive CIS presents a considerable challenge.

The communication of both large-scale and small-scale issues to all concerned is essential to the success of any CIS. In addition to the timely and accurate dissemination of necessary information to users, feedback from users and stakeholders must be actively sought and acknowledged (and, when appropriate, acted upon and incorporated). In addition to undermining the organizational cohesion necessary to implement large-scale system overhauls, ignoring such feedback means that the organization runs the very real risk of neglecting the issues of information technology functionality necessary for a system to be a truly effective tool in day-to-day operations. A communication plan should be designed and implemented concurrently with project design and implementation. In fact, the communication plan should be considered a key part of the overall implementation strategy. Telephone hotlines are useful tools for ensuring that information flows both ways, and they offer the added advantage of validating user concerns and feedback. Communication with staff physicians is particularly important; any communication plan must incorporate physician leadership to ensure that business goals are balanced with patient care goals, as well as to cultivate physician cooperation with, and commitment to, the new system.

Another key challenge in CIS implementation is user training. Hands-on practice that is appropriate for the previous experience level of potential users is a necessity. Skills learned too far in advance of day-to-day use are seldom retained; training must be timely in order to be effective. The availability of real-time troubleshooting as new users adapt to the system is also essential. Perhaps nothing obstructs successful implementation more than a frustrated staff losing valuable work time trying to solve information technology problems when a help desk or information technology support staff could easily solve most problems in a matter of minutes (while educating staff on how to solve such problems themselves). Physician education is, again, very important; physician-education plans must visibly respect both the full schedules of most physicians and their unique perspective as the primary drivers of health care delivery.

Rapid changes in technology present an additional challenge to CIS implementation, but technology also provides the most promising means to address many of the challenges already discussed. Most pharmacological systems presently in use, for instance, do not allow direct entry of physician orders or direct transmission of those orders to pharmacies. Moreover, supporting clinical decision support particularly for medication alerts will challenge our industry’s best systems. Simplifying user entry is the key to increasing clinician participation. Internet-based applications and communications technologies continue to develop rapidly; voice recognition is gaining acceptance as its continued development begins to address many early but infamous usage problems.

Many ambulatory care locations lack the systems needed to collect the clinical information required for a comprehensive CIS. Most health care is now delivered in locations other than hospitals, and information collected in these locations is not generally available to systems in acute care environments (and vice versa). This increases the likelihood of errors in treatment and the probability that efforts and costs are being duplicated unnecessarily. A concerted effort must be made to bring data entry and access to the point of patient care, wherever that might be.

Lack of standardization in data vocabularies and other data elements is a large roadblock to the functionality of a CIS. For the comparability of individual patient records, vocabulary must be standardized to allow true comparison. For large-scale clinical research, comparability across large numbers of patient records will only be possible under a standardized vocabulary.

Perhaps the larger barrier to extracting the rewards of CIS has been the very limited integration of clinical devices into effective work flow. Given the vast number of industry suppliers and the spectrum of clinical equipment that providers use, this challenge will endure for some time. As the integration of patients, equipment, caregivers, and processes improves, however, great strides in quality improvement and cost reduction can be expected.

Computer-based patient records continue to be a goal seemingly just out of the reach of current CIS development. While computer-based patient records can be considered to be the heart of a CIS system, offering the potential benefits of clinical data sharing and improvements in the quality of care, clinical performance, and compliance, they have yet to be implemented successfully on a large scale.8


While partial information systems (such as pharmacy systems, clinical documentation systems, and clinical data repositories) already exist, much work remains to be done to integrate these tools into a comprehensive CIS that will improve patient care while reducing overall clinical costs. In addition, clinician acceptance and use, particularly among physicians, will be a key part of the success of the overall CIS implementation. This acceptance and use is, however, largely dependent on the functionality and accessibility of the CIS in question. In the near term, acceptance will depend more and more on the ability of organizations to make effective changes in work flow and clinical processes as a way to raise quality and lower costs. The solution lies in a cohesive vision for the system, realized through a well-planned, step-by-step development and implementation process. That process must take into account the strengths and weaknesses of information systems currently in use, as well as the potential strengths and weaknesses of the tools, equipment, and systems currently in development.


Greg Walton is vice president and CIO, Carilion Health System, Roanoke, Va, and chair, Healthcare Information Management Systems Society (HIMSS) board of directors.

H. Stephen Lieber is president and CEO, HIMSS


  1. Rognehaugh A, Rognehaugh R. Healthcare IT Terms. Chicago: HIMSS; 2001.
  2. Gartner Group. A Restatement of Gartner’s CPR Definition. Stamford, Conn: Gartner Group; 2000.
  3. Gartner Group. CPR Generations: An Update. Stamford, Conn: Gartner Group; 2001.
  4. HIMSS. 12th Annual HIMSS Leadership Survey. Chicago: HIMSS; 2001.
  5. Leapfrog Group. Computer Physician Order Entry Fact Sheet. Washington, DC: Leapfrog Group; 2000.
  6. Kilbridge P, Welebob E, Classen D. Overview of the Leapfrog Group Evaluation Tool for Computerized Order Entry. Washington, DC: Leapfrog Group; 2001.
  7. Brady M, Hassett M. Clinical Informatics. Chicago: HIMSS; 2000.
  8. Newell LM. Introduction to Computer-Based Patient Records: Selected Articles from the Proceedings of the 2000 and 2001 Annual HIMSS Conferences. Chicago: HIMSS; 2001.
  9. Institute of Medicine. The Computer-based Patient Record: An Essential Technology for Health Care. Washington, DC: IOM; 1991.