As of 22 April 2009 this website is 'frozen' in time — see the current IFLA websites

This old website and all of its content will stay on as archive – http://archive.ifla.org

IFLANET home - International Federation of Library Associations and InstitutionsAnnual ConferenceSearchContacts

60th IFLA General Conference - Conference Proceedings - August 21-27, 1994

The Internet and PERUSE: McGill's Health Sciences Information Technology for the 21th Century

Marcos Silva
Computer Services Librarian,
McGill University Libraries, McGill University,
3459 McTavish Street, Montreal, Quebec, Canada H3A 1Y1


This paper discusses the impact of Internet services and resources on health science research and teaching. It also reviews McGill University Libraries' PERUSE Project and the advent of virtual libraries. Finally, it discusses the evolving role of health science librarians and information specialists.



Rapid access to and dissemination of information are essential components of academic research and graduate studies. This is particularly true in the health sciences sector where researchers must have access to the most recently published information. Given the above, it is not surprising that an entire publication industry based on electronic information is evolving, creating new opportunities in the organization and manipulation of information and knowledge. Indeed, health sciences research and teaching community are arguably the most consistent users of new computer and telecommunication technology concerning the delivery and access of information resources. Health sciences researchers and students have had little choice in this matter; the information explosion in the health sciences has made medical decision making an extremely complex undertakin g (Broering, 1993). It is essential that they have access to an electronic infrastructure where current information and knowledge are available.

Electronic publishing and indexing have led to the development of virtual libraries, whereby temporal and spatial constraints on information are eliminated. Virtual libraries have been defined as an environment with an "universal application of advanced high speed computing and telecommunication capabilities to the access and delivery of information resources" (Rooks, 1993, p. 22). In other wor ds, the virtual library of the future will act as a portal to archives of distributed information, regardless of format or computer platform. Cornell University's Chemistry Online Retrieval Experiment (CORE), where researchers are able to access the full text and graphics of articles from their office computers Krumenaker, 1993, p. 1006), is indicative of the potential of virtual libraries.

The internet

The convergence of telecommunication technology and computers has changed the means by which medical researchers access, manipulate and interpret data. Other developments include the ability to use high speed communications, to transmit text, binary files and graphics, and to access expertise regardless of where it may be located. These developments are possible because of investment in nationa l networks such as the Canadian CA*Net and the American NSFnet which, along with other networks, are creating virtual communities of distributed information, expertise, and knowledge.

The Internet may be defined as a worldwide interconnected network of computer networks supporting a common suite of telecommunication protocols. Protocols are a set of rules that allow different computers operating under different systems to communicate with one another. Transmission Control Protocol/Internet Protocol (TCP/IP) has become the Internet de facto network standard (Lynch & Rose, 199 3, p. 12). In Québec, the link to the Internet is made via le Réseau interordinateurs scientifique québecois (RISQ). Essentially, RISQ is the regional network that connects Quebec universities and research institutions. In turn, RISQ is connected to the Canadian research network, CA*net, an initiative of the National Research Council. It interconnects all of Canada's regional network s (i.e., RISQ, BCnet, Sask#net, etc.) and acts as the national communications backbone. Connections to the American National Science Foundation Net (NSFnet) are made from Montréal to Princeton, New Jersey, Toronto to Ithaca, New York, and Vancouver to Seattle, Washington. This model, multiplied many times throughout the world, comprises the modern day Internet.

Internet protocols

Perhaps the two most widely used Internet protocols are telnet (teletype network) and FTP (File Transfer Protocol). With telnet, a researcher is able to manipulate interactively a remote computer and to access catalogues to collections housed at the U.S. National Institutes of Health, University of California or Johns Hopkins University. Furthermore, with telnet, researchers also have access to a wealth of specialized databases such as: Johns Hopkins Genetic Databases, INFAN: International Food and Nutrition and National Family Databases, and GENBANK: The National Center for Biotechnology Information. McGill is part of this growing information structure, its online library catalog, MUSE, is accessible to anyone with a connection to the Internet.

Moreover, new telecommunication protocols such as Z39.50, a U.S. ANSI standard protocol for information retrieval, makes telnet a particularly useful research tool. Essentially, under Z39.50, a user with a computer operating in a client mode can access and perform information retrieval searches on another computer operating as an information server (Lynch, 1992). That is, it allows a user to se arch a remote database using the local interface and so eliminates the need to learn the remote computer's system. Thus, the Z39.50 protocol offers a uniform common means of access to different information resources linked to a computer network or Internet (Kunze, 1992). Also, the advent of client/server systems allows for greater resource sharing among our research centres. In essence, the client is the user interface, the server indexes, locates and retrieves the data, and the protocol transmits queries and responses. Once these protocols are implemented by our research centres, any client will be able to access any server regardless of the system utilized. Even CD ROM databases could be accessed under this system. Again, the possibilities for resource sharing are exceptional; a national catalogue that indexes Quebec or Canadian medical resources is a very real possibility.

Another valuable Internet protocol is FTP (File Transfer Protocol). With FTP, a user is able to send or retrieve ASCII (text) or binary files from a remote computer to a host computer. This has become an important development because of the advent of electronic publishing. Health science students or researchers are able to retrieve the full text of journals or reports. Lastly, FTP promotes gre ater resource sharing. Research centers are increasingly establishing electronic archive sites in an effort to share and disseminate knowledge and resources. For example, Center for Biomedical Informatic in Brazil, makes available software and reports relevant to health sciences research. Medical researchers must disseminate data in the most optimum fashion possible and communicate with colleagues and other researchers. Electronic mail, perhaps the most widely used Internet service, gives researchers the ability to communicate with colleagues anywhere in the world. Given the phenomenal growth of the Internet Ä the user base is now estimated at approximately 35 million users and growing at 15% a month Ä use of electronic mail has become almost indispensable.

In order to keep abreast of developments in the field, medical researchers are able to join electronic discussion groups, or LISTSERVs, such as BIOMED L, which deals with biomedical ethics or CLINALR, which offers clinical alerts from the National Institutes of Health. Through subscription to an electronic discussion group, researchers and students are able to post or share information and knowl edge. Because of the speed of information dissemination, electronic discussion groups rapidly are becoming an official channel through which researchers announce new findings, share information, and discuss developments in the field. Also, subscription to electronic discussion groups is perhaps the most efficient method to discover new services and developments on the Internet. The growth of t he network makes it almost impossible to create up to date catalogues and indexes to its resources and services.

Finally, resource discovery tools such as Gopher, Wide Area Information Servers (WAIS), Archie, and VERONICA, make the task of locating, indexing and retrieving information far easier for the user. For instance, WAIS is a system originally developed by Thinking Machines Corporation to aid users in retrieving information over the Internet. What makes WAIS so attractive is that users are able to access any type of data: Digitized pictures, text, voice, etc., and to use the WAIS interface implemented on their host machines independently of the software mounted on the remote host (Kahle, 1991). Naturally, this is because WAIS adheres to the client/server, Z39.50 approach.

Another application that has gained popularity is the World Wide Web (WWW). The WWW can be described as a client/server information system that operates under the HyperText Transport Protocol (HTTP). Information on the WWW server is stored using HyperText Markup Language (HTML). Users, therefore, are able to navigate in a hypertext like fashion a wealth of interconnected documents that contain text, graphics, and sound. What makes WWW particularly interesting is that it is able to support other Internet protocols such FTP, telnet, and WAIS. An interesting application of WWW technology is being made by The National Center for Biotechnology Information (NCBI), which is responsible for building, maintaining, and distributing GenBank, the NIH genetic sequence database that collects all known DNA sequences from scientists worldwide. Many of its electronic resources are being made available via a WWW server. However, it has been the Internet Gopher that has revolutionized the means to access and to use Internet resources and services. The Internet Gopher, created at the University of Minnesota and named after its mascot, currently is one of the most popular and widely used public domain resource discovery tools. Gopher is a distributed information search and retrieval system that greatly simplifies navigation of the Internet. It is easy to learn and use. It is based on menus that eliminate the need to learn computer commands and allow novices to access information with little instruction.

Similar to WWW, Gopher provides intuitive access to both files (ASCII or binary) and Internet services. It functions as a search and retrieval tool, as well as an access tool for Internet TELNET sites. Help screens and information files to assist new users are easily accessible. With Gophers, however, one cannot navigate the net in a hypertext like fashion. A Gopher is a protocol composed of public domain software. Similar to WWW, it is based on client/server protocols. The client offers almost seamless navigation through "Gopherspace" using a hierarchical file system to find information. In addition, an indexing tool called VERONICA (Very Easy Rodent Oriented Net Wide Index to Computerized Archives )for Gopher menus is available. VERONICA was d eveloped to provide keyword access to information on Gopher servers. VERONICA is tightly integrated with the standard Gopher protocol, and retrieves resource information in the form of a Gopher compliant menu. From the menu, the user can select and easily retrieve the needed data.

The future

During the 1980's, the U.S. Federal government initiated a number of programs related to national and international networking. Congress, under the sponsorship of former Senator and current Vice President, Al Gore, proposed the development of a high capacity, high speed telecommunications network named the National Research and Education Network. The stated purpose of the NREN is to help the U. S. recover some of its lost international competitiveness by linking universities, research centres, education and industry and so fostering greater cooperation and communication. The High Performance Computing and National Research and Education Network Act of 1991, responsible for the creation of the NREN, was made law in December 1991.

In Canada, a similar project is called the Canadian Network for the Advancement of Research, Industry, and Education (CANARIE).The catalyst for CANARIE can be found in Industry, Science and Technology Canada's sponsoring of the Network Organization Conference in April 1991. At the Conference, an executive committee and four working groups (Business, Governance, Marketing, and Network Architectur e) were given the mandate to establish plans for the creation of a high speed network. The composition of CANARIE's working groups is reflective of its mandate to foster greater cooperation between research and industry. Among its members are found representatives from major Canadian universities, industry, and government.

Similar to NREN, CANARIE hopes to: (a) To upgrade the capabilities of our existing R&D and educational national backbone network to progressively higher bandwiths, with the objective of expanding the network to gigabit capabilities; (b) To promote the use of the network and CANARIE services in conjunction with the regional networks; (c) To establish and operate a high speed experimental test netw ork; (d) To simulate the development of commercially viable, new networking technologies, products, applications, software and services; and (e) To support the migration to operational networks of new networking technologies, products, applications, software and services as developed, tested and used on the high speed experimental test network, and on the R&D and educational network (CANARIE Asso ciates, 1992).

McGill University Libraries Response

The advent of Internet is changing radically the means to undertake research. In an excellent report written for university presidents, the President and Vice President for Academic Planning of Brown University, in conjunction with its University Librarian, argued that the information revolution will parallel the Industrial Revolution in its impact and consequences (Gregorian, Hawkins, & Taylor, 1992). Some possible benefits include: Greater interdisciplinary research, increased local or remote research collaboration, easier dissemination of and access to information, and increased participation by sectors normally excluded from the academic community. Some dangers include: Increased fragmentation of information, exclusion of disadvantaged sectors from information and expertise, and t he elimination of free and unrestricted access to information. This is the environment faced by McGill University faculty and librarians.

McGill University serves approximately twenty thousand full time students. One in four is registered in the Faculty of Graduate Studies and Research, which is comprised of twenty faculties and ten schools. Continuing Education serves many thousands more. McGill's library system contains seventeen libraries divided into four areas: Humanities and Social Sciences, Law, Physical Sciences and Engi neering, and Life Sciences. At present, there are approximately 78 professional librarians and 200 support staff.

The Life Sciences Area Libraries must offer research support to McGill University Faculty of Medicine and five affiliated teaching hospitals in the Montreal metropolitan area. It is comprised of subject libraries in Biology, agriculture, history of medicine, and health sciences. Its staff includes approximately 12 professional librarians and 31 library assistants. McGill University's response to the above trends is outlined in a report written by the Senate Committee on Computing. Of interest is the Committee's recognition that for McGill to meet its educational and research objectives, "it must be at the forefront of the evolution of knowledge and society" (McGill University, Senate Committee on Computing, 1992, p. 2). Moreover, the Committee recognized the pivotal role that libraries must play in the management of information and knowledge: "It is in the area of information management, and in particular in libraries, that the most significant changes are likely to occur in the next ten years" (1993, p. 3). The libraries, therefore, were given an explicit mandate to reexamine and restructure the information and knowledge support given to our f aculty, research, and student communities.

Even before the presentation of McGill's vision statement, the Libraries took a leadership role in the introduction of Internet services and resources. Internet seminars, for example, have been available to the entire University community since 1990. The Libraries invested scarce resources in efforts to increase user awareness of the Internet, promote user independence, and encourage greater us e of Internet resources by all sectors of the University. It was recognized that the Libraries needed to offer researchers and students local resources in addition to those found on the Internet. That is, the Libraries had to give its community access to databases and other information not freely available on the net. An additional problem was that the development or introduction of any new electronic resource has to be integrated with the existing in formation infrastructure. The growing reliance on Internet protocols by researchers means that new information products have to comply with Internet standards and protocols.

Under the direction of Fran Groen, former Head of the Life Sciences Areas Library, and current Associate Director, Public Services and Collections, a committee composed of faculty, Computing Centre staff, and librarians, was established to examine and to report on new technologies and their impact on traditional library services. Special emphasis was put on faculty and Computing Centre collabora tion and support. During its first meeting, discussion centred on ways to improve research and the use our collection. It was decided to examine the possibility of acquiring an online bibliographic system which would index journals, conference proceedings, and reports from many disciplines. A timeline of one year was set for the evaluation, acquisition, and implementation of the system. Note should be taken of the support received by the Committee from McGill's health sciences area. This is not surprising, given that the Health Sciences Area pioneered the use of many new information technologies on campus. For instance, it was the first Library Area to implement fully CD ROM sear ching in the early 1980s. As a result, the Committee understood that the chosen system had to meet the rigorous information needs of the health science research community. It had to be fully compliant with Internet protocols and standards. Furthermore, system had to offer as much searching capabilities, if not more, as the present CD ROM systems. The Committee decided to evaluate several systems, of which two were invited to give public demonstrations of their product to McGill's community. The system had to conform to certain criteria including: Client/Server philosophy, adherence to Z39.50 protocol, support for multiple platforms, and speed of retrieval. An intuitive interface and online help and tutorials were also deemed essential.

CD Plus

CD PLUS was judged as the product that best met the research needs of the McGill community. Operating under its award winning search software, OVID, CD PLUS is able to offer an intuitive and yet highly flexible interface that allows for sophisticated information searching and manipulation. OVID runs on a UNIX platform (CD ROM DOS and Windows versions are also available). A Sun Sparc 10 Model 5 1 is used as the server. The initial contract allows for 40 simultaneous users. The data is received on CD ROM disks and copied onto hard drives.

An additional factor that led to the acquisition of CD PLUS was that the databases required by the McGill community are available with the OVID system; namely MEDLINE, PsycINFO, ERIC, Compedex Plus, and the Wilson databases. The flexibility, manipulability, and intuitiveness of the OVID interface had to be exceptional; all of the above index databases were available at McGill on CD ROM formats. In no possible way could OVID offer less flexibility than the existing CD ROM databases. The McGill CD PLUS/OVID system was named PERUSE.

OVID offers two search modes: Expert Full Mode or a menu driven Easy Mode. While the Easy Mode is indeed uncomplicated and intuitive, it nevertheless offers the full searching capabilities available under Full Mode. This factor is appreciated by our reference staff because it eliminates the need for intensive training: Users are able to teach themselves the basic search commands. Consequently, learning how to search in Full Mode becomes far easier. Furthermore, users find OVID's ability to map subject searches against a controlled vocabulary particularly valuable. For instance, in MEDLINE, a subject search for cancer will be mapped against the appropriate term Neoplasms and its subcategories. Should the terminology be too recent to have subject headings, OVID offers an alternative textword search which indexes the title or abstract of the citation. In this manner, in situations where few or no citations are found, users have an alternative search method that may provide the needed information.

Finally, OVID offers users the ability to limit searches as needed. The usual boolean operators (and, not, and or) are available in addition to a host of other possible limits such as: Human, language, publication year, publication type, animal types, etc. Although these additional limits are database specific, they are able to meet the strict searching standards required by the McGill communit y. What makes OVID's database limits invaluable is their ease of use; the user is presented with an extremely powerful search engine that is exceptionally simple to learn and to use. Novices, therefore, have the ability to make full use of a complex database like MEDLINE with only a minimum of instruction and guidance. That is why reviewers say that OVID has "... the most innovative searching feature, mapping, I've seen in a long time" (Clark, 1993, p. 232). Indeed, Jacsó, in a review entirely dedicated to OVID, supports Clark by stating that he wishes "...other medi cal abstracting databases would also be available with OVID, especially a cardiology subset of OVID, and even more important, full text searchable image database versions of the medical journals with the highest impact factor ... please, just like in the magnificent OVID version of The New England Journal of Medicine ..." (1994, p. 38).

Implications for health sciences research

National policy makers cognizant of the impact that national networks comprising the Internet have on education, research, and industry, are investing resources for the upgrading of telecommunication lines and services. In the United States and Canada, plans are underway to extend network services to the general population, including the K 12 sector, public libraries, and the home (CANARIE Assoc iates, 1992; United States, Office of Science and Technology Policy, Director, 1992). Of particular interest is that national policy makers believe the health science community is one sector that will most profit from the forthcoming gigabyte networks.

Internet access significantly increases the opportunities for interdisciplinary research, collaborative work, and resource sharing. These developments have accelerated the creation of virtual libraries and research centers. Universities have been at the forefront of these trends. They were full partners in the development of Internet protocols, and they pioneered the use of the Internet in sup port of teaching and research. In fact, many of the new Internet resource discovery tools were created in universities. Archie, for instance, was created at McGill University (Deutsch, 1992) and Gopher at the University of Minnesota (Wiggins, 1993). For these reasons to exploit optimally any new information management system, it is essential that it has the potential to be integrated fully with existing network services and resources. That is, any new information system must be able to offer remote access, resource sharing, and support collaborative research and work. These are possibly the elements in an electronic virtual environment mos t needed by health sciences researchers and practitioners.

The McGill health sciences community has benefited from the use of the Internet and the implementation of CD PLUS for several reasons. First and foremost is the ability of researchers to access remotely a wealth of distributed resources, either from their offices or homes. Any campus computer workstation connected to the McGill optical fiber telecommunication backbone has access to the Internet and PERUSE. Furthermore, any networked DOS computer with TCP/IP software, such as LAN WorkPlace for DOS, can also access PERUSE or Internet sites. Finally, access via modem or SL1 card is possible by connecting to a terminal server on campus.

Second, researchers, medical practitioners associated with McGill's teaching hospitals, and students appreciate that McGill Libraries can increase significantly the number of simultaneous users on PERUSE. Depending on the availability of funding, the system can be expanded to meet any new demands placed on it. Health sciences librarians, therefore, are able to offer greater support for interdis ciplinary and collaborative work. For instance, one of the more interesting consequences of PERUSE has been greater use of McGill resources since the collection is more adequately indexed. Also, students are increasing the use of information resources that have been traditionally excluded from their domain of research; social sciences students are now searching MEDLINE to augment their research base.

Finally, because PERUSE is based on a client/server software, it is possible to link it with other resources and databases. Moreover, CD PLUS is investing in Z39.50. Arguably, users would appreciate greatly a single interface to access a multiplicity of electronic resources. This would facilitate and promote the creation of and access to remote databases and virtual libraries. The importance of the above to health sciences libraries should not be underestimated, especially since almost all major library automation vendors have announced that they will be implementing Z39.50 in either client or server mode or both. Furthermore, in September 1993, the National Library of Canada announced the availability of a suite of public domain software for the ANSI Z39.50 informati on retrieval standard (Turner, 1993). The initiative was to offer the Canadian and international information body a starting point for the implementation of Z39.50 (Turner, 1993). Therefore, McGill is able to join and contribute to the creation of virtual universities, libraries, and laboratories.


Access to and use of the Internet is indispensable to researchers in the health sciences. Indeed, it could be argued that access to the Internet is now a requisite for any university level research and work. Access to remote online catalogues and databases and the ability to retrieve or send information in the form of computer files, has revolutionized the dissemination and sharing of informati on. Local systems, compliant with Internet protocols and services such as PERUSE, offer the McGill research community support in research and education. Moreover, PERUSE presages future development to distribute worldwide information systems. Its intuitive and exceptionally flexible interface allows for complex searches to be initiated by users with no previous exposure to the system. PERUSE more e fficiently indexes the research collection and so offers the library greater research and education relevancy. Furthermore, it offers the entire McGill community unrestricted access from almost anywhere on campus, home, or abroad.

PERUSE also meets the new demands that arise from use of the Internet. If need be, PERUSE can be fully integrated with the Canadian and worldwide Internet. Moreover, CD PLUS supports new standards and protocols that have created the foundation for a global distributed collaborative virtual university and library. It has created the potential for greater cooperative projects with developing are as, particularly in the health sciences sector. Telecommunication and computer technology has changed radically the provision of information resources. Librarians increasingly are becoming knowledge managers as opposed to information providers. In the health sciences, these trends have been accentuated as a result of rigorous information and knowledge needs; the information must be accessed, disseminated, and classified in the most timely ma nner possible. The Internet and PERUSE have given McGill University Libraries the means to gain greater academic research participation and relevancy, and perhaps lead in the creation of worldwide virtual universities and libraries.


Broering, N. C. (1993). Création d'un reseau de connaissance. In Communication scientifique, nouvelles technologies et rationalisation des ressources: Un défi pour les bibliothèques universitaires, Actes du Colloque organisé à l'occasion du 25e anniversaire du Sous comité des bibliothèques, Montréal, 1 2 avril 1993 (pp. 99 106). Montr&eac ute;al, Québec: Bibliothèque nationale du Québec.

CANARIE Associates. (1992). CANARIE business plan. [Ottawa: CANARIE] [URL:ftp://unbmvs1.csd.unb.ca/pub/net/CANARIE]

Clark, K. (1993). New OVID software from CD PLUS Technologies. CD ROM Professional, 6, 230 232.

Deutsch, P. (1992). Resource Discovery in an Internet Environment. Unpublished master's thesis, McGill University, Montreal, Canada.

Gregorian, V., Hawkins, B., & Taylor, M. (1992). Integrating information technologies: A research university perspective. In What presidents need to know about the integration of information technologies on campus. [S.l.]: HEIRA. [Computer file, to retrieve this paper, send e mail to HEIRAES@CAUSE.COLORADO.EDU, with message GET HEIRA.ES1.BROWN]

Hancock, L. (1994). Internet/BITNET health sciences resources. Lawrence, KA: University of Kansas Medical Center, 1994. [URL:ftp://ftp.sura.net/pub/nic/medical.resources.3 94]

Jacsó, P. (1994). OVID/MEDLINE: A gourmet choice. DATABASE, 17(1), 31 38.

Kahle, B. (1991). An Information System for Corporate Users: Wide Area Information Servers. Cambridge, MA: Thinking Machines, 1991. [URL:/ftp://quake.think.com, /pub/wais/doc/wais corp.txt]

Krol, E. (1992). The whole Internet: User's guide & catalog. Sebastol, CA: O'Reilly & Associates.

Krumenaker, L. (1993). Virtual libraries, complete with journals, get real. Science, 260(511), 1066 1067.

Kunze, J. A. (1992). Nonbibliographic applications of Z39.50. Public Access Computer Systems Review, 3(5), 4 30.

Lynch, C. (1992). Z39.50 in plain English. (4th ed.). St. Louis, MO: Data Research Associates.

Lynch, D. C., & Rose, M. T. (Eds.) (1993). Internet system handbook. Reading, MA; Addison Wesley Pub.

McGill University, Senate Committee on Computing (1993). Information technology for McGill in the 21st century. Montreal, Quebec: McGill University Senate Committee on Computing. [Computer file, on InfoMcGill, tn3270 vm1.mcgill.ca/computing/vision]

Rooks, D. (1993). The virtual library: Pitfalls, promises, and potential. Public Access Computer Systems Review, 4(5), 22 29.

Turner, F. (1993). Z39.50 toolkit update. [URL:LISTSERV://Public Access Computer Systems Forum/PACS L@UHUPVM1.BITNET/ 10 November 1993]

United States. Office of Science and Technology Policy. Director (1992). The National Research and Education Network Program: A report to Congress. [Washington, D.C.: GPO]. [URL:ftp://expres.cise.nsf.gov/pub/fnc/nrencongr.ascii]

Wiggins, R. (1993). The University of Minnesota's Internet gopher system: A tool for accessing network based electronic information. The Public- Access Computer Systems Review, 4, 4 60.