Electronic Health Information Exchange: Key Trends to Watch

John S. Luo, MD

Dr. Luo is Assistant Clinical Professor, Department of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles

Electronic medical records (EMRs) provide a rich medium for improving health care, with the potential benefits of disease management, electronic prescribing, physician order entry, and interfacing with personal health records. Exchange of information is one of the key elements to making EMRs beneficial to patients across the healthcare spectrum, whether at the physician’s office, hospital, or pharmacy. Presently, most patients will attest to the great difficulty in obtaining records and transferring this health information among various providers. Although the Health Insurance Portability and Accountability Act in 1996 requires the development of national standards for electronic healthcare transactions, this goal has not yet been reached. There are many reasons why physicians have been slow to invest in EMRs, but health information exchange may be one of the biggest challenges. This column will highlight the key data interchange issues to keep in mind when purchasing an EMR.

Background

In 2002, the American Medical Informatics Association (AMIA) recommended that the federal government dedicate technologic resources and medical informatics expertise to create a national health information infrastructure (NHII).¹ At the time, the immediate goal was to create the infrastructure in order to detect and track threats to public health with a long-term vision toward health information exchange and patient safety. AMIA highlighted the need for standards in healthcare data exchange as well as a common vocabulary standard, in addition to strategic planning and the creation of a secure communications network.

In May 2004, President George W. Bush appointed David Brailer, MD, as the first National Health Information Technology Coordinator.² The Office of the National Coordinator for Health Information Technology provides leadership for the development and implementation of a national interoperable health information technology infrastructure to improve the quality and efficiency of health care in both the public and private sectors. In November 2005, this office announced the award of contracts totaling $18.6 million to four groups of healthcare and health information technology organizations—IBM, Computer Sciences Corp., Accenture Ltd., and Northrop Grumman Corp.—to develop prototypes for the Nationwide Health Information Network architecture.³

Information exchange requires structured data to facilitate the transfer process; however, agreeing on a standard is difficult. Unlike the open-document standard⁴ created by the Organization for the Advancement of Structured Information Standards (a nonprofit, international consortium that drives the development, convergence, and adoption of e-business standards), healthcare technology vendors have yet to agree upon a standard for information exchange. The Healthcare Information and Management Systems Society has convened the Integration and Interoperability Steering Committee to guide the industry on allocating resources to develop and implement standards and technology needed to achieve interoperability.⁵ The Commission of Systemic Interoperability, a committee established by the Secretary of Health and Human Services, has created a report outlining the steps and recommendation toward achieving this goal. “Ending the Document Game: Connecting and Transforming Your Healthcare Through Information Technology”⁶ was released in October 2005. This report highlighted three major themes toward achieving the NHII, which included adoption, interoperability, and connectivity.

To advance adoption, the commission recommended adoption incentives, regulatory reforms, reporting on adoption gaps, assessment of workforce needs and impacts, and increasing public awareness of the need for interoperability. For interoperability, their recommendations included healthcare information product certification, data standards, standard product identifiers and vocabulary, and drug records. To achieve improved communication, the commission recommended the development of a patient authentication standard, federal privacy standard, nationwide health information network, criminal sanctions for privacy violations, and consumer protections.

Despite the demand and desire for information exchange in health care, it has been a complex process. One of the reasons for poor interoperability of an EMR has been its slow adoption by physicians. Inherent to this problem is the trade off between free-form text and a structured record system. A medical record in the form of free text is most natural to physicians since it facilitates the often non-linear thought process involved in medical decision making. An EMR often forces the physician to follow a particular pathway for data entry and retrieval, but with this structure it offers quicker access to information and ability to provide analysis.

Free text-based electronic records require additional processing by humans or computer-based expert systems so that data can be extracted and codified. For example, the Mediclass system can extract clinical encounter data from either free text or structured electronic record systems via its natural language processing and knowledge-base system.⁷ The MedLEE system was adapted to use its natural language processing to automatically generate structured encoded output from free text.⁸ These systems demonstrate the possibility to preserve the free-form text-based nature of physician documentation with reliance upon computerized expert systems to analyze and structure the data. However, these systems require specialized training data sets to analyze with 77% to 90% precision. As indicated in the Commission of System Interoperability report, one of the ways to achieve interoperability is to create a national data standard.

Data Standards

Electronic data exchange is obviously not a simple matter of sending an electronic communication from one system to another. In the practice of medicine, physicians automatically incorporate complex data, such as laboratory information, clinical history, and collateral information, toward an understanding and determination of disease states and management. A common vocabulary and terminology, such as the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition-Text Revision (DSM-IV-TR),⁹ is necessary in structuring data for interchange and storage as well as ensuring agreement about diagnosis. The Unified Medical Language System (UMLS)¹⁰ is a compendium of many different medical vocabularies with mapped structure between them. It plays a central role in helping computer systems understand the meaning of health-related information.

One of the key components to future interoperability and exchange of information is the use of standardized and structured data. The Internet is an excellent example of how structured data files and communication protocols facilitate information exchange with use of HTML. XML¹¹ expands upon the capabilities of HTML by providing more structure. It not only provides a way to define the data, but also contains the data in a database. XML is advantageous for data transfer because it is self-documenting, provides both human- and machine-readable format, and is capable of representing basic computer science data structures of trees, lists, and records. The MedLEE system implements XML as the document format to contain both data and structure.¹² XML-based medical records can be easily viewed in the most fundamental of tools, the Web browser.¹³

The Health Level 7 (HL7) Clinical Document Architecture, Release 2, based on XML, is an American Standards Institute-approved HL7 standard that can represent clinical events for the purpose of exchange.¹⁴ These data standards are certain to be incorporated into the development of the NHII.

Organizations

In addition to the data standards, the development of organizations to support data sharing will be necessary for communication. Community health information networks (CHINs) were created in the 1990s to support data sharing of regional databases as a central point of access. These networks were early attempts to create an organizational infrastructure to support planning and implementation of interoperability networks. Many of these early networks failed due to poor buy-in, concerns with data ownership, lack of trust, lack of financing, and high cost of network technology.¹⁵

In 2004, the Department of Health and Human Services released a report¹⁶ that called for the creation of the NHII as well as regional health information organizations (RHIOs). The Santa Barbara County Data Exchange (SBCDE)¹⁷ and the Massachusetts Health Data Consortium¹⁸ are examples of county and statewide efforts toward this exchange of clinical, administrative, and financial information. RHIOs are considered to be precursors to the NHII. They have the advantage of the ability to use the Internet to reach their goals, compared to CHINs which struggled with proprietary networks.¹⁹

RHIOs may have different focus, such as exchange of administrative data, clinical data, or disease surveillance.²⁰ They will have different system architectures, varying from centralized to peer-to-peer. In the centralized model, data repositories or warehouses will combine the various data sources in several locations for later distribution, whereas in the peer-to-peer model, the data remain on site but the modeal provides secure links and references to records in another location. Adherence to standards and maintenance of relationships of the various stakeholders varies in the RHIO. At the SBCDE, the initial plans for tightly regulated and mandated adherence to standards were met with resistance, and eventually a flexible and open-ended process became the fundamental methodology. The SBCDE members have no corporate structure that binds its members to one another, but all work together for the common good.

Conclusion

The goal of healthcare information exchange depends upon a variety of factors, including data and communication standards, policies, finances, and organizational efforts. Communication in the healthcare setting is vital for providing quality care as well as managing cost and improving access. There are numerous considerations in the way of achieving this interchange, such as public and private sector issues, public health, financial incentives, and legal obstacles.²¹

Education of the public, policy makers, healthcare leaders, and physicians is central toward adoption of the NHII.²² Given the complexities of healthcare information exchange, it may seem prudent to the EMR purchaser to hold off on adoption. However, delay of implementation of an EMR does not appear prudent in light of its many benefits when these interchange issues are considered.

References

1. Tang PC; AMIA Public Policy Committee. AMIA advocates national health information system in fight against national health threats. J Am Med Inform Assoc. 2002;9(2):123-124.

2. Office of the National Coordinator for Health Information Technology. Available at: www.hhs.gov/healthit/bios.html. Accessed April 9, 2006.

3. HHS Awards Contracts to Develop Nationwide Health Information Network. Available at: www.hhs.gov/news/press/2005pres/20051110.html. Accessed April 9, 2006.

4. OpenDocument Format for Office Applications (OpenDocument) v1.0. Available at: www.oasis-open.org/specs/index.php#opendocumentv1.0. Accessed April 9, 2006.

5. HIMSS. Healthcare Information and Management Systems Society. Integration and Interoperability. Available at: www.himss.org/ASP/topics_integration.asp. Accessed April 9, 2006.

6. Comission on Systematic Interoperability. Ending the Document Game: Connecting and Transforming Your Healthcare Through Information Technology. Available at: http://endingthedocumentgame.gov/index.html. Accessed April 9, 2006.

7. Hazlehurst B, Frost HR, Sittig DF, Stevens VJ. MediClass: A system for detecting and classifying encounter-based clinical events in any electronic medical record. Am Med Inform Assoc. 2005;12(5):517-529.

8. Friedman C, Shagina L, Lussier Y, Hripcsak G. Automated encoding of clinical documents based on natural language processing. J Am Med Inform Assoc. 2004;11(5):392-402.

9. Diagnostic and Statistical Manual of Mental Disorders. 4th ed, text rev. Washington, DC: American Psychiatric Association; 2000.

10. Unified Medical Language Fact Sheet. Available at: www.nlm.nih.gov/pubs/factsheets/umls.html. Accessed April 9, 2006.

11. XML. From Wikipedia, the free encyclopedia. Available at: http://en.wikipedia.org/wiki/XML. Accessed April 9, 2006.

12. Friedman C, Hripcsak G, Shagina L, Liu H. Representing information in patient reports using natural language processing and the extensible markup language. J Am Med Inform Assoc. 1999;6(1):76-87.

13. Schweiger R, Hoelzer S, Altmann U, Rieger J, Dudeck J. Plug-and-play XML: a health care perspective. J Am Med Inform Assoc. 2002;9(1):37-48.

14. Dolin RH, Alschuler L, Boyer S, et al. HL7 Clinical Document Architecture, Release 2. J Am Med Inform Assoc. 2006;13(1):30-39.

15. Overhage JM, Evans L, Marchibroda J. Communities’ readiness for health information exchange: the national landscape in 2004. J Am Med Inform Assoc. 2005;12(2):107-112.

16. Brailer DJ. The Decade of Health Information Technology: Delivering Consumer-centric and Information-rich Health Care. Framework for Strategic Action. Available at: www.hhs.gov/healthit/documents/hitframework.pdf. Accessed April 9, 2006.

17. Brailer DJ, Augustinos N, Evans LM, Karp S. Moving Toward Electronic Health Information Exchange: Interim Report on the Santa Barbara County Data Exchange. Available at: www.chcf.org/topics/view.cfm?itemID=21086. Accessed April 9, 2006.

18. Halamka J, Aranow M, Ascenzo C, et al. Health care IT collaboration in Massachusetts: the experience of creating regional connectivity. J Am Med Inform Assoc. 2005;12(6):596-601.

19. Goedert J. Are RHIOs for Real? Health Data Manag. 2006;14(2):44-62.

20. Baldwin FD. RHIOs. Healthc Inform. 2006;23(2):58-60.

21. Yasnoff WA, Humphreys BL, Overhage JM, et al. A consensus action agenda for achieving the national health information infrastructure. J Am Med Inform Assoc. 2004;11(4):332-338.

22. Stead WW, Kelly BJ, Kolodner RM. Achievable steps toward building a National Health Information infrastructure in the United States. J Am Med Inform Assoc. 2005;12(2):113-120.