This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

P E R S P E C T I V E S
E L E C T R O N I C  P R E S C R I B I N G
W E B  E X C L U S I V E   

25 May 2004  
  

The Role Of Standards In Electronic Prescribing

The variety of existing e-health functions and data sources require standards
that do not benefit one stakeholder at the expense of others.

By W. Ed Hammond

ABSTRACT:

The focus on preventing medical errors has advanced the arguments for widespread implementation of electronic prescribing. The choice of systems as well as the variation in functionality is large. Value and return on investment depend on the functionality provided. The paper by Douglas Bell and colleagues defines the functionalities that are required and desirable to ensure patient safety and quality of care. Health data standards are a prerequisite for the interoperability to support electronic prescribing. This paper discusses some of the barriers and problems in producing and adopting those standards.


The report To Err Is Human, released by the Institute of Medicine (IOM) in 2000, drew the public’s attention to medical error.¹ Even though medication errors constitute only 7 percent of medical errors, they have received the most attention. The reason for this focus is that the problem is definable, visible, and, ultimately, fixable. Legislation has been introduced that will require the use of electronic prescribing; however, depending on the functions supported by such systems, they may address only some of the problems. Also, to remove the potential for errors, any solution adopted for electronic prescribing must address all facets of prescribing, dispensing, and administering medications in all sites, not just errors committed in the inpatient setting, which was the IOM report’s focus.

The paper by Douglas Bell and colleagues addresses these requirements by defining the functionalities required from an electronic prescribing system to support patient safety and health outcomes, helping patients manage costs, maintaining patients’ privacy, and promoting clinicians’ acceptance.² An underlying condition of success in electronic prescribing is the role and necessary use of standards. The IOM, in recognition of this need, published Patient Safety: A New Standard for Care, which defines the data standards necessary to support patient safety and the electronic systems that support and enable patient safety.³ Here I briefly discuss what standards are necessary to support electronic prescribing and the difficulties in producing those standards.

Need for standards. Although almost any use of automation diminishes the risk of error, the elimination of medical errors requires a level of functionality that includes a complete list of all medications a person is taking, including prescription, over-the-counter, and herbal medications, and access to the person’s diagnoses, height and weight, certain laboratory data, social environment, age, mental acuity, and potentially other data elements. These elements are aggregated from a number of sites, so data sharing must be accommodated.

Generally, standards are required for defining and identifying all of the needed data elements; standards for the unique identification of all participants; standards for defining and sharing knowledge related to electronic prescribing; and standards for the interchange of messages. Additional standards that affect electronic prescribing include standards for defining clinical guidelines, clinical protocols, and care maps and functional standards for the electronic health record. To ensure interoperability with other e-health applications, a common set of data elements including name, definition, data type, terminology, units, and value sets must be adopted.

Obstacles to standard setting. Health data standards have primarily been created by a consensus process dominated by vendors. Vendors do not view standards as strategic to their future; they view standards as a controlling factor that potentially limits their market share. Standards permit users to change from one vendor to another with relative ease and lower costs. For vendors, working with a standard-developer organization is an exercise in damage control. Each vendor has a proprietary interest in the standard and is motivated to produce a standard that comes as close as possible to matching what that vendor currently does. The resulting standard is frequently a compromise, with the ambiguity one would expect, and each vendor then implements the standard as it favors its own system.

Further ambiguity is introduced as a standard is promulgated among users. The balloting process includes the total community of vendors and users, and the number of non-vendors must equal the number of vendors to prevent vendor bias in the production of the standard. The solutions to these obstacles include a model-based approach to data standards that removes compromise and a certification process that ensures compliance with the standard. The creation of standards should be as balanced as possible and favor no particular group of stakeholders or vendors.

From the user perspective, the obstacle exists in the form of the cost of change and the belief that standards will not support necessary customization. The cost of change has both financial and behavioral modification aspects. Also, most users will not pay for a change if what exists appears to meet current demands. The solution lies in educating users and in creating the business case for savings resulting from change. Standards should reduce the cost of change and prevent being “locked in” to any vendor’s system.

Health data standards also must be a partnership between the public and private sectors, which must be equal and active participants. Government-mandated standards in a world as complex as health care are likely to be unacceptable and not likely to be transparently useable in both public and private settings. The cost, for example, of separate systems (and standards) for reimbursement and patient care can no longer be supported. Standards, including terminology, must work for both groups, and this requires both sectors to work together and adopt common standards. It is important to note that the actual creation of the standards is a lesser problem (as long as all stakeholders are represented) than the acceptance and implementation process.

An interesting problem exists between the real world (what providers actually order) and what the standards require to be explicitly specified. A clinician may actually just write “penicillin” on a prescription pad and depend on the pharmacist’s understanding of what else is required. Electronic prescribing will require behavior change on the part of such clinicians, who are usually consistent in their prescribing habits. Preempting with default ordering data will save time in entering prescriptions. Exactly what is the balance between displaying all possible side effects and possible interactions and just the important ones is a difficult call. Too many interruptions in the prescribing process will make a system unusable.

Additional challenges. Additional issues must be addressed in which common agreements must be defined and accepted. For example, if a patient has multiple medication lists, which list is “the” list? How will changes to one list modify the others? When does a specialist become not involved in a patient’s care and cease receiving updates? What will the mechanism for entering over-the-counter medications into a patient’s drug list be, and who will be responsible, the patient or some other party? How will medicines prescribed by several providers and purchased from more than one pharmacy (including outside a person’s area of residence) be reconciled on a drug list? What is the infrastructure necessary to capture all of this information? The solution lies in a single, aggregated medication list—a patient-centric approach—with a primary care gatekeeper.

The good news is that standards are keeping pace with the requirements of electronic prescribing. However, this is a process that must continue indefinitely, as data standards are developed along with better and better electronic solutions for managing the complex data involved in patient care.

NOTES

1. L.T. Kohn, J.M. Corrigan, and M.S. Donaldson, eds., To Err Is Human: Building a Safer Health System (Washington: National Academies Press, 2000).
2. D.S. Bell et al., “Recommendations for Comparing Electronic Prescribing Systems: Results of an Expert Consensus Process,” Health Affairs, 25 May 2004,
content.healthaffairs.org/cgi/content/abstract/hlthaff.w4.305 (25 May 2004).
3. Institute of Medicine, Patient Safety: A New Standard for Care (Washington: National Academies Press, 2003).

W.Ed Hammond (hamm0001{at}mc.duke.edu) is a professor emeritus in the Departments of Community and Family Medicine and of Biomedical Engineering and an adjunct professor in the Fuqua School of Business, Duke University, in Durham, North Carolina.

Read related papers by: Douglas Bell et al., David Brailer, and a Jonathan Javitt.

DOI: 10.1377/hlthaff.W4.325
©2004 Project HOPE–The People-to-People Health Foundation, Inc.