1
Introduction
The growth of the Internet, its widespread use, and its availability as a platform for database searching has made it possible and practical for the National Library of Medicine (NLM) to provide Internet access for many of its 40-plus online databases. Use of the World Wide Web to search the MEDLINE bibliographic database, through NLM's PubMed and Internet Grateful Med interfaces, has seen outstanding success as an access point for health professionals and for the general public.
One component of health information that continues to be at the forefront of health concerns for a wide range of scientists and health professionals, as well as the general public, is toxicology and environmental health information. To improve access to this kind of information, in early 1998, the NLM began developing an Internet Web site to provide access to the extensive portfolio of toxicology and environmental health databases located on NLM's Toxicology Data Network (TOXNET; Table 1.1). This Web site (www.toxnet.nlm.nih.gov) is the topic of this report.
STUDY BACKGROUND
NLM's work in toxicology and environmental health information dates from the mid-1960s with the establishment of the Toxicology Information Program, now the Toxicology and Environmental Health Information Program or TEHIP. In 1969, the National Academy of Sciences (NAS) began providing advice and oversight to the expanding toxicological information resources at NLM through the Toxicology Information Program Committee (TIPCOM). For over 25 years, the toxicologists, pharmacologists, chemists, and computer scientists on
TABLE 1.1 Overview of the TOXNET Databases
TIPCOM provided timely and relevant advice as NLM's toxicology and environmental health program grew. In 1995, NLM's focus turned toward examining the medical relevance of these databases and the needs of clinicians and other health professionals for this kind of information. NLM requested a study by the Institute of Medicine (IOM) on those issues, and the resulting report (IOM, 1997) addressed future directions for the NLM's Toxicology and Environmental Health Information Program to increase its utility and accessibility. Major issues addressed by that report were increasing access to and effective use of the TOXNET databases by health professionals.
In late 1997, NLM decided to make the TOXNET databases available via an Internet World Wide Web site. During the development phase, NLM requested that IOM conduct a study that would assess the usefulness and effectiveness of the new Web site. The IOM Committee on Internet Access to the NLM's Toxicology and Environmental Health Databases was formed in early 1998. The committee was charged with seeking input from health professionals on their assessment of the Web site and with providing recommendations on the refinements necessary to facilitate searches of the TOXNET Web site by health professionals, scientists, educators, and the general public. The committee consisted of eleven members with expertise in medical informatics, environmental health, family medicine, poison control, medical librarianship, health communication and education, toxicology, and pharmacology. The committee met twice in the course of the study, solicited input from a number of health professionals and other interested individuals, and invited individuals with different skill levels in database searching to conduct searches on TOXNET using the Web interface.
INPUT TO THE COMMITTEE
The committee sought the input of a number of individuals interested in the content of the TOXNET databases. Recognizing that input should be solicited from individuals covering a broad range of expertise, the committee decided to utilize the following list of potential users of toxicology and environmental health information (IOM, 1997):
-
primary care professionals (e.g., physicians, nurses, nurse practitioners, and physician assistants) and pharmacists;
-
specialists in occupational and environmental health (physicians, nurses, nurse practitioners, physician assistants, industrial hygienists, and safety officers);
-
emergency medicine and poison control center personnel (e.g., emergency room health professionals, emergency medical technicians, clinical and medical toxicologists, and specialists in poison information);
-
health science librarians and faculty at health professional schools (including medical, nursing, public health, pharmacy, and dental schools);
-
environmental health researchers and scientists (including health physicists, epidemiologists, toxicologists, and forensic practitioners);
-
patients, the general public, and community organizations (including local emergency planning committees, public librarians, educators, and advocacy and activist organizations); and
-
health professionals in local public health departments or in state and federal agencies (e.g., policy advisors, health educators, and public clinic personnel).
The committee received input through videotaped search sessions, e-mail responses, and by working through searches with colleagues and students. This input was structured to provide the committee with feedback and was not designed or conducted as a scientific study. The committee appreciates the input of the individuals (Appendix A) who searched the TOXNET databases through the new Web interface and provided comments on the problems they encountered in the search process as well as their thoughts on improving the search interface.
Videotaping was used to record a small number of individual search sessions to more fully capture the search process and identify the barriers that individuals experienced. In this way, the search process could be more carefully examined. Participants were invited by committee members and staff to participate in a 30- to 45-minute search session. In each session, the participants were first asked to fill out a short questionnaire to provide the committee with information on their occupation, use of toxicology and environmental health information, and extent of their prior Internet and TOXNET experience (Appendix B). The participants were then asked to search the TOXNET Web site to answer two predetermined questions:
-
What is the potential for vinyl chloride to cause liver cancer in humans?
-
What are the potential health effects of working as a dry cleaner?
In addition to searching for information to answer the two questions posed by the committee, most participants searched the TOXNET site for information on topics of their own interest. Their queries were frequently related to subjects that they were currently studying or to subjects they knew well and wanted to examine the nature and extent of information in TOXNET. Topics included pfiesteria, tobacco smoke, aspartame, estradiol, malathion, and calcium channel blockers.
The computer screen was videotaped during the searches through a VGA video converter (no audio was recorded), and participants were asked to talk through the search process with the IOM staff member present. After the searches were complete, participants were asked to answer questions about each phase of the search process, to describe barriers to searching, and to suggest opportunities to refine the Web site. Similar videotaping processes have been used in other studies as a means to analyze and record human/computer interactions (Kushniruk et al., 1996; Wang and Tenopir, 1998).
Individuals who agreed to provide input via e-mail were given the same series of questions and the Web site address and were asked to search the TOXNET databases and provide information on their search experiences.
In all 31 individuals provided e-mail or videotaped input to the committee, 15 individuals participated in the videotaped search sessions, and 16 people provided e-mail responses. The participant group consisted of 5 occupational and environmental health professionals, 1 family medicine physician, 3 emergency medicine and poison control personnel, 7 librarians and information specialists, 5 environmental health scientists, 3 federal or state public health professionals, 4 health policy professionals, and 3 students. Participants included individuals who provided input through the workshop or responded to the questionnaire in the previous IOM study (1997) and other individuals who were familiar with the current study. When asked about their Internet expertise, 19 noted they were advanced Internet users, 10 rated themselves as having intermediate expertise, and two considered themselves as novices (having basic or minimal Internet expertise and experience). Sixteen of the individuals had previously searched at least one of the TOXNET databases. The search experiences and comments of the participants and the committee are described in Chapter 2.
Committee members also received input on the interface by asking students and colleagues to search the databases via the Web interface. Working through a search with a colleague or student and discussing the search experience was helpful to committee members as they performed searches themselves and assessed the search interface. The committee then discussed the input received along with their own search experiences.
OVERVIEW OF THE REPORT
This report presents the committee's assessment of the TOXNET Web site and provides the committee's recommendations. Chapter 2 summarizes the input received by the committee and the committee's assessment of the current Web site. In Chapter 3, the committee presents its recommendations for short-and long-term improvements to the TOXNET Web site, and Chapter 4 discusses recommendations for raising awareness of this valuable information resource. The committee hopes that its recommendations for refining the