Toxicology and Environmental Health Information Resources: The Role of the National Library of Medicine (1997)

In this concluding chapter, the committee highlights four areas that are important to the future growth and development of the TEHIP program: stable funding, the integration of the TEHIP program into NLM initiatives, the establishment of an advisory committee, and the implementation of a strong evaluation process. Additionally, the committee presents its recommendations for future directions that will provide health professionals with the tools necessary to readily access toxicology and environmental health information. The committee reiterates its belief that NLM should continue to exert its role as a library and, in doing so, expand awareness of the full range of national, state, local, international, and private-sector information resources in toxicology and environmental health (see Chapter 3). The continued challenge for the TEHIP program will be to provide easily accessible and authoritative information to an increasingly computer literate and environmentally-aware user community that includes health professionals and the general public.

The 1993 report by the NLM Long Range Planning Panel on Toxicology and Environmental Health (NLM, 1993) served as an invaluable resource to the committee. The committee found the report to be a useful basis for its work and wishes to reaffirm the planning panel's recommendations, some of which have not yet been implemented. It is the committee's hope that NLM will reexamine the planning panel's report and use its recommendations in conjunction with the recommendations of this report to strengthen the TEHIP program.

A critical factor to the success of any program is a stable level of funding adequate enough to maintain staffing levels and meet programmatic requirements. As with NLM as a whole, the TEHIP program receives funds from two sources: funds legislatively appropriated to NLM and reimbursements from other government agencies.

As discussed in Chapter 2, the TEHIP program's budgetary appropriations have remained fairly constant over the past 29 years; however, fluctuations in reimbursements from other agencies have been significant. Reimbursement funding is the result of collaborative projects with other federal agencies, and the TEHIP program is funded in large part through interagency agreements. However, changing priorities, responsibilities, and resources within the various agencies have significantly affected the TEHIP program. Between 1992 and 1993, the reimbursable budget dropped by approximately 50 percent, from $2.45 million in FY 1992 to $1.27 million in FY 1993. Since 1993, the reimbursable budget has remained at that reduced level (the FY 1995 reimbursable budget was $1.23 million).

For the TEHIP program to be responsive to the changing demands of health professionals and to improve the utility of the databases, funding must be adequate to implement changes based on the results of the user profile analysis (Chapter 4) and on committee recommendations for improving access, navigation, and program evaluation. Fluctuating funding makes it difficult to plan future activities, to develop long-range goals, and to implement necessary program changes. As noted, however, a mechanism for using limited resources is the prioritization effort based on the results of the user analysis. Nonetheless, fluctuating funding will affect staffing and the ability to keep pace with the rapidly changing trends in technology necessary to make the most useful databases even more accessible to health professionals. The committee did not discuss what appropriate levels of funding might be and suggests that NLM determine the level of commitment to the TEHIP program and develop a cost analysis based on the needed improvements to the program. Clearly, the TEHIP program requires a stable funding base that is not subject to changes in the priorities and programs of other federal agencies.

The committee has noted the need for a strong leadership role in promoting and marketing the TEHIP program. As NLM engages in new initiatives and research and development (R&D) projects that are relevant to the TEHIP

program, especially with regard to state-of-the-art computer technology, the TEHIP program leadership should urge NLM to integrate the TEHIP program into those initiatives. For example, as NLM continues to develop the Unified Medical Language System, Internet Grateful Med, and a range of training and outreach programs, TEHIP's priorities should also be included in those activities. Thus, the TEHIP program leadership should be vocal about its program priorities within NLM so that they may be incorporated into new initiatives and be addressed by NLM. Similarly, the committee encourages NLM to more fully integrate the TEHIP program's activities into broader NLM efforts in order to maximize their effects, avoid duplicative efforts, and leverage limited resources.

The mission statement of the TEHIP program has three priorities: to provide selected core information resources and services in toxicology and environmental health, to facilitate access to national and international information resources in this field, and to strengthen the information infrastructure of toxicology and environmental health. The committee believes that the TEHIP program is well-positioned to carry out its mission, which incorporates the committee's recommendation for broadening the focus of the program to encompass the full range of toxicology and environmental health information. The complexity of scientific and technical issues faced in implementing the TEHIP program requires a knowledgeable staff with a range of scientific and computer expertise. The longevity of the TEHIP program is due in large part to the high level of expertise and dedication of the Specialized Information Services Division (SIS) staff. The committee believes that the TEHIP program has demonstrated a commitment to develop, maintain, and improve environmental health databases. As TEHIP staff continue to refine and improve the databases and fulfill the mission, as interpreted broadly by the committee, they should exert their leadership role in providing environmental health and toxicology information by incorporating their program priorities into NLM activities where appropriate.

The growth and development of any program are strengthened and invigorated by advice provided by outside experts. NLM uses this mechanism through a number of boards and advisory committees. Four of the six divisions within NLM's organizational framework have standing advisory committees made up of external scientists, librarians, and health professionals (Figure 7.1). In some cases the committees have specific tasks; for example, the Literature Selection Technical Review Committee selects journals for indexing in Index Medicus and MEDLINE, and the Biomedical Library Review Committee supports the work

of the Division of Extramural Programs by reviewing grant applications under the Medical Library Assistance Act. The Lister Hill National Center for Biomedical Communications and the National Center for Biotechnology Information each have a Board of Scientific Counselors that reviews and makes recommendations on NLM's intramural research and development program (NLM, 1995).

For 25 years (1969 to 1994), SIS received policy and technical advice from an advisory committee of the National Academy of Sciences (NAS), the Toxicology Information Program Committee (TIPCOM). The committee was made up of leading toxicologists, pharmacologists, chemists, and computer scientists who reviewed the progress and growth of the toxicology and environmental health program at NLM. In 1993, NLM's focus turned toward examining the medical relevance of the TEHIP databases and their under-utilization by health professionals. NLM requested that NAS, through the Institute of Medicine (IOM), convene a planning meeting to examine health professionals' use of the TEHIP databases and suggest future evaluation activities. Subsequent to the planning meeting, NLM and IOM developed the study on which this report is based. There is currently no other advisory committee for the TEHIP program or for the other diverse activities of SIS in general.

The committee has considered the needs for continuing oversight of the TEHIP program and believes that an outside advisory committee would benefit the program. SIS has multiple responsibilities that are primarily split between two subject areas: toxicology and environmental health and HIV/AIDS. These two areas differ widely in the issues that need to be addressed and in the interested communities and advocacy groups. Therefore, the committee believes that an advisory committee should be established specifically for the TEHIP program.

A necessary component of any advisory committee is stakeholder participation. A large number of diverse groups are interested in the TEHIP program, and their participation and input would be beneficial. These groups include toxicologists (medical and clinical), health science librarians, medical informaticians, computer scientists, environmental scientists, environmental policymakers, health care professionals (including tertiary care professionals in occupational and environmental health; primary care professionals; and emergency medicine, public health, and poison control personnel), and interested members of the general public including community environmental activist organizations and public librarians.

Additionally, the committee suggests that SIS obtain input via other mechanisms on a periodic basis. This may include the use of focus groups, the formation of a liaison panel to the advisory committee composed of representatives from other federal agencies involved in environmental health issues, or the use of other evaluation mechanisms such as surveys and workshops.

To cover all relevant issues, the advisory committee should examine the issues brought to its attention by NLM, by advisory committee members, and by other interested parties. It is important that the advisory committee provide timely and relevant advice on issues, including the following:

• program content priorities, management, planning, and evaluation;¹

• resource needs;

• quality assurance and peer review;

• relationships with other government agencies and initiatives; and

• technical operations.

The outside advisory body should be constituted to facilitate a two-way flow of information and ideas. Beyond its role of providing advice to the program, there is a significant role for the advisory committee in conveying knowledge about the program and its products back to the range of communities represented on the committee. This may include presentations at professional conferences and involvement in outreach activities.

An integral component to the effectiveness of any program is a thorough and ongoing evaluation effort planned during the initial phase of each project and incorporating benchmarks to assess ongoing progress. NLM has used a range of methods and strategies to assess the effectiveness of its products, services, programs, and policies; however there is a need for expanded evaluation efforts.

Current evaluation methods include beta-testing, a process of iterative testing and refinement that uses input provided from groups of end users. Selected test sites, often NN/LM member libraries, implement the new product and collect user input. For example, the beta-test version of the LOANSOME DOC² program, software, and documentation was sent to 57 libraries and more than 600 health professionals (Wallingford et al., 1996). Additionally, user input has been sought through surveys, workshops, and focus groups. Needs assessment and baseline surveys have been used to assess the level of knowledge about NLM's products (e.g., Grateful Med) prior to an outreach effort, the impact of which is then evaluated by follow-up questionnaires. One of the new mechanisms in evaluation efforts is the use of focus groups in which small numbers of people, often 10 to 12 peers, discuss specific topics through the leadership of a

facilitator (Seiver, 1989). Focus groups of health professionals have been used to assess familiarity with NLM services and barriers to their use (Mullaly-Quijas et al., 1994).

Consideration should be given to developing and implementing an evaluation plan focused specifically on the TEHIP program. This may include the detailed user analysis (recommended in Chapter 4) and the previously mentioned advisory committee. Evaluation is often considered only when the program or product is in place or fully developed. Thus, the committee emphasizes the importance of incorporating an evaluation component into a project during its initial planning stages and early implementation. Additionally, the committee stresses the advantage of building benchmarks into the project plan to assess progress at specified stages. Evaluation studies for new TEHIP products and for training and outreach efforts should continue to use a variety of traditional approaches (including surveys and focus groups) and should also expand to use new methods (e.g., Internet access counters, online surveys). Input should be sought from health professionals in a range of work settings and with a range of expertise.

As the committee conducted this study, several programmatic issues came to the forefront. It is clear that NLM has taken a great deal of initiative in disseminating toxicology and environmental health information; however, there is a need for a stabilized funding base, an internal commitment to the TEHIP program, an involvement of the TEHIP program in broader NLM R&D efforts, an interdisciplinary advisory committee, and the development and implementation of an evaluation plan. The TEHIP program continues to make a substantial contribution to the fields of toxicology and environmental health, and the beneficial use of toxicology and environmental health information resources by health professionals and other interested user communities can be considerably increased given the necessary resources and support.

Because health professionals will continue to need accessible and authoritative sources of toxicology and environmental health information, the committee believes that it is important to provide health professionals with the tools to retrieve this information. During the course of its deliberations, the committee determined that the optimal approach would be to provide health professionals with two different options for accessing toxicology and environmental health information. Primary access would be via an online directory or ''road map" that would assist the user in identifying and connecting with the relevant TEHIP database or other (non-TEHIP) online information resource (e.g., World Wide Web sites of government agencies, directory information for environmental health organizations).³ Ultimately, the search would be conducted automatically and seamlessly without requiring the user to select specific information resources. The second access option would be a toll-free telephone number or similar single-access information center that would link health professionals (person-to-person) with a specialist who could provide consultative services on environmental health issues and concerns.

This two-part approach would require cooperative efforts between NLM, other federal agencies, and private-sector organizations. NLM's expertise and current R&D efforts in medical informatics would make it the logical agency to take the leadership role for the online directory, although the assistance of other agencies and organizations would be needed to provide directory information. The development and implementation of the second option, the single-access information center, go beyond NLM's mandate, and this recommendation should be considered by multiple federal agencies and other private-sector organizations involved in environmental health. Public-private partnerships could play a key role in providing these information services. The collaboration of NLM with universities, industry, international resources, and local, state, and federal governments could be particularly productive given the numerous information resources in this field (Chapter 3).

The committee believes that an online directory of toxicology and environmental health information resources would be most useful if it not only provided information about the available online resources in toxicology and environmental health but also assisted the user in locating and connecting to the most

relevant resource for his or her information query. Currently, it can be time consuming to search the World Wide Web for information resources on environmental health, as users may experience the frustration of massive and poorly differentiated retrievals. As recommended in Chapter 3, it is important for NLM to exert a traditional library role in comprehensively assembling, organizing, and maintaining an inventory of information resources in toxicology and environmental health. The online directory would involve an extensive compilation with links to online databases; directories of environmental health specialists; Web sites of government agencies and private-sector organizations involved in environmental health; and other similar information resources.

The TEHIP program staff has started this process by assembling an Internet Web site that links the user to other federal and international Web sites with toxicology and environmental health information. Additionally, much of the directory information is available in the current DIRLINE database. To maximize the utility of this approach, an inventory of all relevant information resources in this field should be conducted and incorporated into the data management scheme.

An important component of this road map would be the incorporation of tools such as overlying decision trees that would assist the user in locating the relevant resource. Decision points could include questions regarding the purpose for which the information is needed (e.g., patient care, research, policy development, or education), the type of information needed (e.g., bibliographic, factual, or directory), the subject matter, or the type of information needed (e.g., animal data, epidemiologic studies, or clinical consultation). Additionally, for those users requiring expert assistance, the decision tree could provide pointers to the single-access information center (described in the next section). In order to build effective decision trees, it would be important for the online directory to incorporate information beyond pointers such as subject coverage, access points, and quality indicators. Input on the key elements to be included in the decision tree structure could be provided by the TEHIP program advisory committee and by the user profile analysis.

NLM is currently funding R&D efforts on the components of expert systems that could provide this decision tree and directory technology. The Unified Medical Language System (UMLS)—one of NLM's major research initiatives—is a long-term R&D effort that aims to assist health professionals and other searchers in retrieving and integrating online biomedical information from diverse information sources. Currently, the UMLS Information Sources Map (ISM) is a promising component of a system that would analyze a searcher's information request to identify and connect the user to the appropriate information source (Clyman et al., 1993). The ISM is a database with records that describe information providers and information sources. The scope of each information source is indexed by using MeSH and other relevant terms. NLM, in a collaborative project with the Clearinghouse for Networked Information Dis

covery and Retrieval, has developed a Web-based application, Sorcerer, that accepts user queries, identifies the information sources in the ISM that are likely to have relevant information, and then connects to the source, conducts the query, and returns the results to the user (NLM, 1995). A companion software tool, Apprentice, is in development. Apprentice will enable information providers throughout the biomedical community to register their information sources with the centralized ISM database (NLM, 1995). Expert systems in development and prototype phases are incorporating the ISM. This technology, in combination with the hypertext capabilities of the World Wide Web and other advances in informatics, bodes well for the online directory to be an attainable goal in the near future.

A model for the development of this online directory is the National Environmental Data Index (NEDI), an interagency effort to provide a central Internet directory or "yellow pages" for accessing environmental data (e.g., data on air and water quality, global change, or renewable energy; see also Chapter 3). Phase I of this project is in progress and aims to provide distributed access to data systems within the federal government. Phase 2 will expand to provide links to state, local, international, and private (nonprofit and commercial) data systems (NEDI, 1996). Links between NEDI and the proposed online directory for environmental health information would be critical for providing comprehensive information to all interested user communities.

In addition to a directory of online information resources, the committee became aware that in some clinical and other situations, health professionals would prefer to consult with an expert in environmental health. Although there are several toll-free telephone numbers that provide limited information related to environmental health and some poison control centers provide environmental health services, there is no single resource that health professionals or others can call with detailed questions on these topics. Many focus group participants expressed the need for a single-access point for toxicology and environmental health information, particularly clinically-pertinent information in real time.

This idea has been discussed and recommended by previous IOM committees that examined the occupational and environmental health information needs of primary care professionals. Those committees determined that a comprehensive center providing clinical and nonclinical services through a consultation structure would meet the information needs of many health professionals (IOM, 1988, 1990). The center could provide expertise in patient care, risk assessment, and exposure reduction, including industrial hygiene, diagnostic and treatment services, and assistance in dealing with government agencies (IOM, 1988). Justifications for a single-access point include the need for rapid and authoritative

information, the scarcity of specialists in occupational and environmental health, the fragmentation of environmental health responsibilities throughout the agencies of the federal government, and the vast and disparate information resources in this area.

Poison control centers are examples of single-access points that are effectively meeting the information needs of health professionals and other user communities. Poison control centers provide 24-hour-a-day, 7-day-a-week service to health professionals and the general public primarily by responding to calls for immediate treatment management of toxic exposures. Additionally, poison control centers provide telephone follow-up and, as necessary, provide referrals for specialized health care. Forty-seven regional poison control centers are certified by the American Association of Poison Control Centers (AAPCC), and several of these have developed occupational and environmental health information services. AAPCC compiles the Toxic Exposure Surveillance System, a nationwide database on human exposure cases. Poison control centers fill a unique niche in emergency medical services, and it is estimated that at least $5 is saved in health care costs for each dollar spent on poison control center services (Durbin and Henretig, 1995; Kearney et al., 1995; Lovejoy et al., 1994). The savings are attributed primarily to decreases in the unnecessary use of ambulance and emergency department services for minimal symptoms. However, many centers are facing substantial budgetary cutbacks.

The implementation of a single-access information center would have budgetary and interagency ramifications that the committee did not have the mandate to explore. However, in its exploration of the TEHIP databases and its inquiries into the dissemination of toxicology and environmental health information—especially to health professionals—the committee realized the need for this type of information resource for toxicology and environmental health information. The committee believes that poison control centers, which are already established, would make excellent resources for providing toxicology and environmental health information to health professionals. However, the committee is particularly mindful of the budgetary considerations and of the problems that poison control centers would face if their mandate is expanded without the necessary financial resources for implementation. Although the provision of a single-access information center for toxicology and environmental health information is not within the purview of NLM, the issue should be explored because it is important to expanding the use of this information by health professionals.