| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Internal Medicine, Division of Pulmonary and Critical Care Medicine, University of Michigan Health System, Ann Arbor, MI 48109-0360.
Correspondence to: William F. Bria, II, MD, FCCP, Department of Internal Medicine, Division of Pulmonary and Critical Care Medicine, University of Michigan Health System, 1500 E Medical Center Dr, 3916 Taubman Center, Ann Arbor, MI 48109-0360; e-mail: wbria{at}umich.edu
Abstract
In part 2 of this series on applied medical informatics for the chest physician, we will examine in detail the key reasons for the current low adoption and effective use of the electronic medical record.
Key Words: computer • informatics • quality of care
In the first installment1 in this series on applied medical informatics (AMI) for the chest physician, we reviewed the structure and basic function of the electronic medical record (EMR). Even to the casual observer of the technology of the EMR in American medicine, the following two realities are manifest: (1) recently, there has been a great deal more discussion in the medical literature, in national medical organizations, and in the lay press on the benefits and risks of the EMR; and (2) there has been a great deal more talk about the EMR than has been actual use and availability of these tools in daily practice. In this, part 2 of the series on AMI, we will examine the reasons for this important and frustrating conundrum, and how to understand and realize the benefits of the EMR and avoid its drawbacks.
In his 2005 "State of the Union" message, President George Bush stated: "By computerizing health-care records, we can avoid dangerous medical mistakes, reduce costs, and improve care." Juxtaposed with this statement are the following findings from the survey of American Hospitals in 2002: of all US hospitals, only 9.6% of them (60 hospitals) have the computerized physician order entry (CPOE) component of the EMR completely available for use. In about half of these hospitals, > 90% of orders are entered with CPOE by physicians; in about another third of the hospitals, 90% of the orders are entered with CPOE by nonphysician staff.2
The causes that contribute to the low rate of EMR implementation and usage include the following:
Let us examine items 1 to 6 in more detail.
Complexity of the EMR
The elements of achieving success in an EMR selection, configuration, and implementation are complex. The three major categories are "Winning Hearts and Minds," "Technology," and "Place and Time" (Fig 1 ). Credible leadership is essential, and, as has been demonstrated repeatedly, implementing any change in medical practice usually revolves around a group leadership model. The change cannot be achieved without involvement of these pre-EMR clinical leaders. Unfortunately, in the current health-care informatics reality in this country, there is no standard EMR, no standard EMR implementation plan, and, even within an individual commercial EMR product, no standard configuration kit. Therefore, the champions of EMR need to address the following issues.
|
Technology
See installment 1 of this series.1
Place and Time
It is intuitive that the introduction of an EMR into a large academic health system will differ in many ways from its introduction into a small, community, private practice office. However, what may not be as evident is that the characteristics of these different clinical environments may have more to do with success than simply scale alone. That is, in clinical settings where there is a well-established tradition of clinical groups working cooperatively with one another (eg, a typical office practice), there is significantly greater acceptance of a change, such as the introduction of an EMR. Conversely, where there are settings in which there is a discrete group that is expected to bear the brunt of the change resulting from the EMR introduction, tragic failures have been well-documented. These failures may stem from even the perception of the possibility of losing authority or from not being involved in the configuration of the system.3
The time needed to implement an EMR is roughly proportional to the size of the organization being automated. An important consideration, however, is the profundity of the change over time. Experience over the last 25 years has shown that the introduction of an EMR system in a 
200-bed institution should be considered to take the same order of magnitude in time to achieve success as the completion of construction of a new hospital building. This comparison can be helpful in many ways. For example, the disruption that many have personally witnessed on moving to a new institutional setting provides a frame of reference for improving the likelihood that critical support systems, for instance, training and help desks, are correctly staffed and supported to avoid an actual increase in medical errors. It has recently been reported4 that these errors may be caused predominantly by the too rapid introduction of an EMR into a health-care system.
Cost of Change
The entire cost of EMR selection, configuration, and use remains high. There are many reasons to explain this aspect of why most physicians in America are without the use of a complete EMR. As of this writing, almost every aspect of the EMR in this country is nonstandardized. One major consequence of the lack of standardization is increased cost. The costs of EMR implementations (including not only software purchase, but cost of the necessary system configuration, interfacing, and total project costs) vary from hospital to hospital, vendor to vendor, and year to year.
How much do EMRs cost? The firm First Consulting Group, based in California, said it would cost $210 million in one-time start-up expenditures for the 46 Massachusetts hospitals to install them or to update existing technology to a minimum agreed-on standard. However, the consultant predicted the hospitals would save $275 million annually because the systems reduce errors and, as a result, avoid medical care for unintentional injuries to patients 5
Other than the actual cost of CPOE systems, an important consideration is where the financial incentives for adoption occur. In a one report,6 the Center for Information Technology Leadership estimated that 89% of the economic benefits of CPOE accrues to the holder of the financial risk for health care (ie, insurance companies). At a time when the financial value of EMRs and CPOE specifically are coming under increasing economic scrutiny, perhaps the consideration of where the financial value really exists for CPOE will be better realized and the incentives realigned appropriately.
Time Commitment
The extended length of time needed to plan, select, configure, implement, refine, and, finally, achieve effective use of an EMR remains one of the most important reasons for the limited availability of these systems for physicians in the United States. In a 2002 study,7 The Ohio State University reported a 7-year time period from the completion of EMR strategic planning to complete installation. Of course, this length of time may be considerably less in a smaller health-care setting; however, the importance of this report is that, irrespective of the size of the organization, the implementation of an EMR takes vision, planning, and institutional perseverance, which is a rare commodity in American health care today. By the term institutional perseverance, I mean that once organizations have decided to adopt an EMR, there must be a commitment to continued, focused, and innovative leadership in continuing the project over a period of years. The time in office of senior managers in American hospitals today, in particular the chief information officers, suggests that there may be several senior leaders within a health system before an EMR is completely implemented and effectively becomes part of the culture of an organization. However, experience has shown that expecting current leadership to be in place for the completion of an EMR project is a key element in achieving success.
Clinicians Expectation
Clinician time spent and the value perceived in the EMR remains one of the primary concerns in the United States. In more recent studies,8 physicians have definitely ranked EMRs toward the top of the solutions to decreasing the number of medical errors. However, it is important to understand that, even in the most optimistic investigations, the physician use of EMRs, in particular order entry, takes more time compared to the physical process of writing orders on an order form at this point in time.9
Even in the realm of computer hardware in support of EMR use, there are emerging data suggesting that different clinicians (ie, physicians vs nurses) obtain differing value from the location of computer terminals. In a recent study,10 the use of bedside terminals and central station desktop computers saved nurses 24.5% and 23.5%, respectively, of their overall time spent documenting during a shift. However, in the same study, the use of central station desktop computers for physician order entry was found to be inefficient, increasing the work time from 98.1 to 328.6% of a physician’s time per working shift.10
Finally, to realistically approach the issue of clinician expectations and the EMR, we need to know how our colleagues are reacting to it. For example, in a well-done recent survey study11 of the emotional aspects of CPOE, the implementation and use of systems provoked examples of positive, negative, and neutral emotional responses. However, negative emotional responses were the most prevalent, by far, in all of the observations.
Interoperability
The Oxford English Dictionary defines the word interoperability as "the ability of computer systems or software to exchange and make use of information." In the world of the EMR, this one element has possibly caused more confusion and frustration than any other. Interoperability is essentially the ability to interconnect different computer systems (eg, the laboratory, the radiology department, the pharmacy, and the medical records department) and have them work well together. Although it is possible to have this occur (Fig 2 ), it is also quite common for the connections between systems, or interfaces, to be complex and error prone, and to require as significant maintenance as any of the changes in interconnected components. The root causes for problems with interoperability include changing information technologies, competing proprietary vendor systems, and, simply, the complexity of clinical information systems.
|
In the discussion of standards and interoperability, however, has appeared the additional issue of the standardization of language. In the AMI world, this notion of standardized medical terminology usually leads to the term lexicon, or dictionary. There are many examples of standardized medical lexicons (Fig 3 ). However through the efforts of Dr. David Brailer and the Office of the National Center for Healthcare Information Technology, the lexicon SNOMED has been acquired for use throughout the United States in an effort to catalyze the effective information interchange between information systems. The manufacturers of EMRs are now rapidly incorporating this particular lexicon.13
|
It has been said that the complexity of the automation of health care is much like the automation in the aerospace industry. Although the precise complexity grading of health-care automation may never be done, the fact is that there are numerous examples of failed clinical information systems over the past 30+ years that are related in part to the poor understanding and oversimplification of the process of clinical care. One poorly understood dimension of workflow analysis in health-care information systems is group interactions. In the history of AMI, this has led to the use of systems that worked well for one group (eg, physicians, registered nurses, or pharmacists), yet were very poor at effective and safe communication among members of the health-care team. For example, CPOE systems have been created that allow the flagging of order as "STAT," yet provide no effective means of notification of nursing and other groups workers who actually have to move a STAT order forward expeditiously.
Modern systems are improving in recognition of the complexity of actual health-care workflow as a result of increased emphasis on reduction in errors and improvement in quality of care 14 A growing literature now is appearing that also recognizes the importance that the placement and availability of computer hardware varies among members of the health-care team.10
In our third and final installment of this series, we will complete our discussion of the conundrum of effective clinician use of the EMR, including issues of interoperability, clinical workflow, and competing agendas in American health care. We will also look ahead to achieving the benefits of EMR, the personal health record, and the virtual ICU.
Footnotes
Abbreviations: AMI = applied medical informatics; CPOE = computerized physician order entry; EMR = electronic medical record
Received for publication December 28, 2005. Accepted for publication January 4, 2006.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
