Important first steps have been taken in applying NII technology to the information needs of population-based public health. Nonetheless, a much stronger connection between the NII initiative and the public health community will be needed to make the kinds of scenarios described earlier in this paper a reality.
As we move forward, it is important to remember that while the NII has an important role to play in addressing public health's information problems, it is not, by itself, a "silver bullet." A broad range of barriers currently hamper the public health community in meeting its information needs. Some are unique to public health, such as the fragmented organization of public health programs and information systems, the separation of public health from the bulk of the medical care system, the declining level of financing for essential public health services, and the paucity of informatics exposure in public health training. Others, such as the lack of nationally uniform policies related to privacy, data standards, unique identifiers, and data sharing affect many sectors, but are critical in a field like public health, which depends on bringing together many different types of data from diverse sources. The challenge, then, is to identify key issues that must be addressed in order for population-based public health to realize the benefits of the NII and to devise effective strategies for making progress on these issues.
In the discussions before and during the April meetings, five major barriers to enhancing public health applications of the NII emerged, above and beyond basic resource constraints and the limited appreciation by both the public and policymakers of the importance of population-based public health. These barriers, which are discussed further below, include:
- a lack of nationally uniform policies to protect privacy while permitting critical analytic uses of health data;
- a lack of nationally uniform, multipurpose data standards that meet the needs of the diverse groups who record and use health information;
- insufficient awareness of the applicability of NII technologies in meeting the information needs of population-based public health;
- a public health workforce that lacks essential information technology skills; and
- organizational and financing issues that make it difficult to integrate information systems or bring potential partners together.
Public concern about the privacy of health data is a critical issue -- especially in an era of computerized medical records and Internet communications -- and state public health officials strongly support a nationally uniform framework that would protect the privacy of individuals while permitting critical analytic uses of health data. Without such a framework, some state governments may protect privacy through policies that prevent data linkage, hampering the development of logically integrated public health information systems. In states without sufficient or clear laws regarding privacy, health care providers and consumers may be reluctant to provide identifiable information for state-wide databases.
Since about 50 percent of the U.S. population lives near a state border, the development of reasonably comprehensive state public health information systems often requires exchange of data across one or more states. Variations in state protections of health data may preclude data sharing across state lines or involve unacceptable risks to the privacy of data that are transmitted. At the very least, the absence of nationally uniform policies to protect the privacy of individually-identifiable health information requires time-consuming, state-by-state negotiations in order to reach agreements that permit integration of data, as well as efforts to reassure the public that their data will receive reasonable protection. The absence of generally accepted "best practices" for preserving the privacy and security of health data in automated, networked systems or for linking and anonymizing health data in secure environments complicates these processes.
8.2 Health Data Standards
The lack of nationally uniform, multipurpose standards for the structure, content, and transmission of automated health data creates a "Tower of Babel" that seriously impairs the development of integrated information systems to support population-based public health. In this environment, states wishing to move forward must promulgate their own standards, negotiating with major health care payers and providers, as well as other parties whose data are critical to meeting health information needs. These duplicative efforts are time-consuming and costly, and risk the adoption of standards that are more suitable for paying claims than for meeting the needs of clinicians, researchers, and the public health community. Moreover, they inevitably lead to differing state standards that impede the collection of regional and national health statistics, and complicate public health surveillance in the many major metropolitan areas that cross state lines.
Achieving nationally uniform standards for health data that support population health is complicated by several factors. To date, there has been little direct public health participation in the standards development process, and few in the public health community are well informed about national efforts to develop consensus standards for health data. At the same time, there is uneven knowledge about projects that could lead to standardized nomenclatures for multipurpose health information systems in the near future. For example, with the expansion of the UMLS into a health vocabulary, a nomenclature may soon be available which can record data in integrated information systems in the same detailed and meaningful form in which it is entered in computerized patient and public health records. Because the UMLS Metathesaurus makes it possible to map terms in the health vocabulary to multiple classification systems, information recorded once at the point of service can be used for many different purposes. Integrated information systems designed with the flexibility to accept information in this form will be able to capitalize on the increasing use of the UMLS by the clinical and public health communities.
8.3 Awareness of NII Applicability
The low level of public health participation in broad-based NII grant programs is symptomatic of a lack of appreciation, both on the part of the NII and public health communities, of the benefits and applicability of NII technologies to population-based public health. This lack of understanding extends to the commercial information technology sector and works against creative thinking about how commercial products might be applied to the full range of public health information problems.
Multiple factors are probably at work here, including the excessively high visibility of clinical applications of the NII, a lack of information about NII grant programs among the public health community, a poor understanding of population-based public health among the NII and health care informatics communities, insufficient local investment to allow public health agencies to connect to the NII, and a lack of public appreciation of the importance of information in carrying out essential public health services. In addition, public health agencies have not been tied into the usual channels of distribution of information about NII programs. Because they have not received relevant training, many public health professionals lack the technical expertise to prepare competitive proposals for advanced information technology research or demonstration projects.
Collaboration of state and local health departments with university-based medical informatics or general computer science groups, advanced computing and telecommunications groups, the private sector, or other elements of state government is essential to apply for grants and implement applications. Some successful examples of these collaborations include the partnership of the Georgia INPHO project with the Georgia Center for Advanced Telecommunications Technology and the work of the California Health Department with Pacific Bell. To date, however, such collaboration has been relatively rare. In part, this may be due to the traditional focus of medical informatics on process automation and the public health focus on data. As computer-based patient record systems generate more data and public health officials seek to automate an array of processes, the interests of the two groups become congruent.
8.4 Workforce Training
Professionals who are unfamiliar with (or have limited access to) information technology and existing decision support and communication tools relevant to public health responsibilities cannot argue effectively for increased allocation of resources to improve the public health information infrastructure. They are unlikely to take full advantage of technology that is available to them or contribute innovative ideas for applying the information infrastructure to population health.
Facilitating advanced public health applications of information technology will require a cadre of public health professionals with advanced informatics training. It also demands significant improvements in the basic technology literacy of the general workforce in public health, and ongoing training to continuously update information skills.
8.5 Organizational and Financing Issues
The fragmented organization of public health programs plays an important role in limiting applications of the NII to population health. Public health programs are frequently funded through categorical grants, which are often associated with categorical information systems dedicated to the reporting needs of each particular program. Because these systems speak different languages (in terms of vocabulary, software and data standards), it is extremely difficult to link information across programs, even when such linkages could substantially reduce administrative work, enhance customer service, and strengthen analytic capacity. Equally important, many grants prohibit the use of categorical funds for developing or maintaining information systems that benefit other programs in addition to the one for which the funds were appropriated. As long as these restrictions (and their concomitant political and managerial barriers) apply, it is difficult for states to move forward with logically integrated information systems that could give their public health officials and policymakers a more holistic view of public health problems and resource needs.
Intergovernmental and public/private partnerships are key to public health applications of the NII for a variety of reasons, including the range of expertise that is required, the need for information from diverse sources, community-wide benefits and use, and demands for a broad base of financing. Currently, however, most states and communities have neither the policy framework nor the structural mechanisms to bring potential partners in health-oriented information systems together. Without adequate incentives to collaborate or a forum for social interaction, feasible NII projects to support population health may not get started or may falter before implementation.