CD-ROM for Health Information in Zimbabwe

by Helga Patrikios

Helga Patrikios went to school in New York and Belfast, and to university at Trinity College Dublin, where she read French and Italian language and literature. She worked as a TV documentaries researcher, teacher and editor until her marriage to a Rhodesian brought her to Harare (then Salisbury) in 1972. She took a Higher Diploma in Library, Science by correspondence with the University of South Africa in 1978, and has worked ever since at the Library of the University of Zimbabwe. She has been Medical Librarian since 1985, and has published and presented many papers on her work there. She was appointed Deputy University Librarian in 1994.

BACKGROUND AND CONTEXT OF THE PROJECT

In recent decades, throughout sub-Saharan Africa, academics, scientists and researchers have seen the decline of once flourishing institutes and universities as economic conditions worsened, and inflation and devaluation of local currencies exacerbated the chronic shortage of hard currencies. The deficiencies in traditional information sources, arising from economic, geographical and cultural factors, have worsened perceptibly in the past two decades. And while the information explosion in the North is fueled by dramatic developments in information technologies, the South must watch the gap between its own information sources and those available in the industrialized countries grow ever wider. Can that gap be narrowed or even bridged, or will the poorer countries be further marginalized in terms of technology and information access?

Fortunately, in many African countries, the advent of the personal computer has enabled libraries to leapfrog two decades of development in information technology, and move straight into effective and appropriate solutions to some of the problems of information storage and retrieval. Institutions and private companies which had never seen a mainframe or minicomputer began in the mid-1980s to acquire microcomputers - often initially only for word processing, but increasingly for compiling databases and other specific applications.

In Zimbabwe, at the Medical Library of the University of Zimbabwe (UZML), the process began with CD-ROM. This case study will describe how and how that process was set in motion, and its immediate and longer-term effects on the services of the library and on its staff and users.

Zimbabwe has a population of 11 million people, 70 percent of whom are based in the rural areas. Independence from the illegal white minority rule of Ian Smith's Rhodesia was achieved in 1980; by 1990 Zimbabwe's development programs were well advanced, particularly in the fields of education and the health services.

The colonial legacy included a centralized health service, with a concentration of sophisticated curative services for a minority in the urban centers. (See Box 1.) The goal of the new Ministry of Health was to create an extensive network of primary health care outlets in the rural areas, reallocating existing resources to establish new clinics and upgrade general and district hospitals. The success of their policies was seen in a steady improvement in health status indicators during the middle and late 1980s, including maternal and child morbidity and mortality; nutritional status; and expanded immunization programs.

Harare, the capital, is the home of the University of Zimbabwe, established in 1958, which includes the country's only Medical School - added in 1963. As well as the 500 medical students, the Faculty has departments of Nursing Science, Pharmacy and Rehabilitation, and the School is a regional center for postgraduate programs in the medical specialties. The present total of students approaches 1,000, and student intakes continue to grow.

The Medical Library was well planned from the outset, with a good collection of journals and textbooks, and a current awareness service for academics. As the country's only medical library it was in principle accessible to members of the health professions outside the University; but since it was situated in the main University Library on the campus in an outlying suburb, it was little used other than by staff and students of the Faculty. In 1978 the library moved to the spacious premises of the new medical school and teaching hospital close to the city center, where more non-university health workers began to make use of it.

In 1983, following recommendations from a World Health Organization-sponsored workshop in Tanzania, the UZML, as the only major medical library in Zimbabwe, was designated National Focal Point for Health Sciences Information Services, by both the Ministry and the University. An outreach librarian was appointed to deliver services to health workers in provincial and rural hospitals and health centers. In 1984, we carried out a survey of their information needs and priorities.1

Anecdotal evidence from rural health professionals of a dire lack of access to information - an absence of current textbooks and journals, or indeed any considerable hospital library - had prompted the survey, which had a 74 percent response rate; this evidence was emphatically confirmed by the 63 doctors and 10 senior nurses who responded. Their expressed priorities were to have core collections of books and journals at their workplaces, and 90 percent considered the lack of current medical information to be a serious difficulty or disadvantage.

Some of these needs were addressed by a Joint Health Information Committee, which we established with the Ministry of Health for this purpose - core collections of 40 textbooks and 13 manuals were provided by a Scandinavian donor for all hospitals and clinics. Adequate journal collections could not be provided to all hospitals, so we compiled instead a digest of MEDLINE abstracts on Zimbabwe's health issues, Current Health Information Zimbabwe (CHIZ), a free update for all Zimbabwe's health professionals. The abstracts were provided for us by the World Health Organization (WHO) in Geneva, who sent us the monthly MEDLINE printouts produced by the Swiss host Datastar. Datastar ran searches using a profile of Zimbabwe's major health problems, compiled with advice from former district medical officers.

We tried, in the mid- 1980s, to gain access to the online databases of the U.S. National Library of Medicine and to MEDLINE in particular - the index, with abstracts, to over 3,700 of the world's major biomedical journals, with over 8 million records. The obstacles - high and unpredictable costs in scarce foreign currency (for passwords, telecommunications, connect time and printouts) - were too great. A further disincentive were the complex command syntaxes of the databases, and the need for considerable familiarity with the medical terminology.

A less demanding more user-friendly medium was needed to access MEDLINE. In 1988, while visiting a London medical school library, we saw it - in the shape of CD-ROM, in full-color action - a technology based on laser discs that was just then beginning to penetrate European markets. The advantages of CD-ROM were immediately obvious: the powerful laser discs, used in conjunction with an ultra-friendly search engine, mounted in the homely personal computer, bringing international bibliographic databases into the most remote centers - and at a predictable cost. Unlimited online searching and training time, unlimited information retrieval, unlimited access to the almost mythical databases of the world's most advanced scientific institutions - these were suddenly within our reach.

BOX 1 Pre-Independence Health Services

Pre-Independence health services had neglected information support; medical and nursing education were historically centralized, and the only other sources of information and educational material lay in the small collections at the nursing schools. The library at the Ministry was a dusty storeroom piled high with government reports. The achievement of Independence made rapid growth and change possible not only for the health services but for health information too.

PROJECT DESCRIPTION

In mid-1988, a project proposal went off to the Carnegie Corporation of New York - for the acquisition of CD-ROM equipment with a subscription to the MEDLINE database. The general objective of the project was to improve the health information service at the UZML, with the specific objective of providing users in the Medical School and throughout Zimbabwe with MEDLINE literature searches.

The Medical Library had three professional (graduate) staff, three diplomats, and adequate secretarial and other support staff. The reference librarian had nursing training behind her, as well as her library qualifications. Most of her working hours hitherto were spent scanning the latest journal issues for individuals requiring current awareness updates - a Selective Dissemination of Information (SDI) service. We did not yet know that this task could be reduced to some minutes a month. All of us were excited by the prospect of introducing a computerized service into the Library at last.

Neither I nor any of my colleagues had ever laid hands on a computer. No one in the University Library had computer skills. Many of us were conscious of being left behind while libraries in the North were forging ahead with newer and better computer applications for information storage and retrieval, and library housekeeping systems. Among the major benefits of the project would be the familiarization of library staff with microcomputer technology.

We also counted on providing our own MEDLINE searches for CHIZ, and relieving WHO's Library in Geneva of that expensive burden. We could also revise our search strategy at will without incurring further expense or trouble for WHO.

The response from Carnegie was rapid and positive - we received a discretionary grant for a feasibility study of a MEDLENE on CD-ROM search service. The grant also enabled us to buy a personal computer, the two drives needed to search the current five-year MEDLINE file, and a two-year subscription to the MEDLINE database.

PROJECT EXPERIENCE: AND IMPLEMENTATION

What were our problems, errors, and successes in implementing the three phases of our project?

Hardware

Our initial choice of hardware was poor. Since we did not even know which questions to ask, we were guided by earlier advice from another donor's program officer, which stressed the importance of securing reliable local (IBM) maintenance. We ordered the smallest cheapest IBM: a PS2 Model 30. We were guided by excessive thrift, rather than by criteria of power and speed, of which we then knew nothing except what we learned from the conflicting advice given by local computer experts. They in turn knew nothing of the rapidly growing exigencies of CD-ROM - an unknown technology in Zimbabwe in 1988.

In reality the Harare private sector offered more than enough microcomputer support to go round; furthermore, a snippet of local lore that we absorbed has proved to be well founded: in Harare's temperate and relatively dust-free climatic conditions microcomputers, even the cheaper IBM clones, generally give little trouble. We rightly did not budget or contract for an expensive maintenance agreement with the local IBM agents - a hefty 10 percent of the equipment's cost per year, the local standard charge. (The high cost of equipment in Africa shocks buyers in the North - the addition of transport and duties almost double prices of any imported goods.)

Instead we allowed in our budget for (lower) maintenance costs and have always spent less than was budgeted on hardware and software problems. A piece of late news: recent extensive building operations next to our library have created a reddish-grey dust - clay and cement - which now thickly coats the edges and insides of our drives. They are giving occasional trouble now.

The pioneer personal computer' within three years of purchase, was unable to process the burgeoning MEDLINE database. We ran out of space on the 20 megabyte hard disk before the end of the first phase of the project - there was barely room for MEDLINE software (13 megabyte) and word processing software, and ultimately everything but MEDLINE had to be deleted. When MEDLINE needed 28 megabytes of disk space and at least one megabyte of RAM, and we found the machine's hard disk and RAM could not be upgraded, it was relegated to word processing use.

Clearly, as novices, we should have sought and found more expert advice outside Zimbabwe before deciding what to buy. (See Box 2.) It was a colleague in New York who discovered that the drives we had ordered were incompatible with the personal computer - just in time to prevent their shipment; it took several months to rectify that error; and when the equipment was all finally assembled, we found that the driver extension software was superseded, so that more weeks passed while we awaited the current version.

Professional advice provided by Carnegie, and first-hand experience, helped us in making better decisions on equipment in later phases of the project. Four years later, for instance, we were encouraged by a local supplier - and were very tempted - to network existing and new personal computers so that all workstations in the library could access all CD-ROMs. We learned, though, that not only would the proprietary networking costs of our database vendors have been beyond our long-term means, but also the absence of expertise in the library system needed to maintain the network would have left me with that responsibility. We knew that neither I nor my colleagues could muster the time or skills needed.

As time passed our local suppliers learned much more about CD-ROM; we too learned, as personal computers became cheaper, faster and more powerful in terms of clock speed, disk and memory size, that we needed to allow for rapid growth in the number and size of databases, for much more RAM, for Windows software for the databases which would' before long, be available only in Windows; for anti-virus packages to check the diskettes on which our users increasingly needed to download their search results.

Our initial choice of a vendor for the MEDLINE database (from a product range of eight competing companies) was based on advice from a large medical library in Milwaukee, which had closely examined all the options, and on journal articles which rigorously evaluated the search features of the different versions. Our chosen vendor, Ovid Technologies (formerly CD Plus), has prospered, multiplied and improved its products, and supported us with discounts. CD Plus responded, though not always promptly, to the software difficulties that arose from time to time as new features were developed. We learned to correct some of them by the simple expedient of reinstalling the software. Other software failures needed input from our suppliers. The only personal computer in which we failed to install the recent much larger DOS version of Ovid's MEDLINE is an IBM PS2 Model 70; it accepts only the Windows version of MEDLINE.

BOX 2 Relative Speed of Equipment

When, before ordering, I contacted the MEDLINE vendors in New York to ask whether our choice of equipment was appropriate for their database software, they said it would be fine, but warned that its performance would be a bit sluggish. l responded, quite seriously, that was just fine, since we too were sluggish .... It turned out that we could go off and brew tea during some of the lengthier search processes, and still find the search in progress on our return.

Delays

Our greatest frustrations came from the delays that held up the first phase by one year. There were delays in obtaining the correct drives and software; it took more time to cajole duty free certificates and import permits from the State's then ponderous bureaucratic machinery. The wasted year, 1988-1989, nonetheless saw a degree of progress in STI at UZML. We learned word processing on the personal computer, which was delivered within only a few months. Once the library had at least one functioning workstation (and by the second phase of support we had three) and could keep our search service going, we could tolerate subsequent delays. The cliff-hanging saga of the supplier who held us up in the third phase of the project by converting funds for equipment to his own use must be documented elsewhere; he delivered our goods in the end, but a year later than planned.

RESULTS, IMPACTS, AND BENEFITS OF THE PROJECT

What have CD-ROM technology and the microcomputer - the availability of in-house biomedical and health databases - brought to the library, to the Medical School, its staff and students, and to the country's health workers? Who uses MEDLINE most? Why do they want MEDLINE searches? The data gathered from our search request forms give some answers.

Numbers of MEDLINE Searches Made

The average number of searches rose from 80 searches per month (1989-1990) to 450 per month in 1995 (November). (See Figure 1.) In March 1995 a record 700 individual searches and monthly updates (SDIs) were recorded.

Figure 1: Monthly average of CD-ROM searches conducted, January 1990 to November 1995

Categories of Search Requesters

Undergraduate and postgraduate students represented 36 percent and 15 percent, respectively, of all searchers in a 6-month sample of 1,660 search request forms in 1994. (The 58 monthly SDIs that are automatically generated as updates for individuals were not categorized.) Other searchers were academic staff, nonacademic staff, or government workers, or from NGOs, parastatals, or the private sector. (See Figure 2.)

Figure 2: MEDLINE searches by user categories.

Purpose of Searches

In the same six-month period, research and assignments (in which there may be overlap) accounted for 41 percent and 13 percent, respectively; teaching 4 percent; updates and SDIs 20 percent; patient care 8 percent; publications 2 percent; in 12 percent no purpose was given. (n = 2008 searches including SDIs.)

Survey on the Use of MEDLINE and Its Abstracts

In March 1995 we carried out a survey on the use of MEDLINE by staff and students of the Medical Faculty. A questionnaire sent to 252 staff and undergraduates sought to gain information on:
· numbers of MEDLINE users and non-users;· frequency of use;· the value of searches and of abstracts;· the number of end users and the need for training in end user searching; and· reasons for non-use of MEDLINE.

We had a response rate of 50 percent (n = 252). Of 122 academic staff to whom we sent questionnaires, 49 responded. Of the 130 final-year students to whom we sent questionnaires, 67 responded. Of the total 126 respondents, 119 used MEDLINE and only seven had never used MEDLINE. (See Table 1.)

We can surmise that those who have used MEDLINE are much more likely to respond to a questionnaire on its use than those who have not used it; thus a much greater proportion of the 50 percent who did not respond may have been non-users. The reasons given by respondents who had never used the service were lack of knowledge about it and lack of time to use it. Additional survey results are given in Tables 2-5.

TABLE 1 MEDLINE Users/Non-users

TABLE 2 Frequency of Use of MEDLINE

TABLE 3 Value of Search Results

TABLE 4 Usefulness of Abstracts

TABLE 5 Completeness of Information in Abstracts

Training in MEDLINE Searching

The library has periodically offered training in MEDLINE searching to users: 49 percent of respondents had received some training. Most of these were students: only 17 (31 percent) academic staff had received training.

The survey showed that most staff and students did searches assisted by a librarian; only 25 percent of respondents (n = 29) were able or willing to do their own searches, although half had received some training. Training was thus deficient in some way for half the participants. Ninety percent (n = 87) were interested in receiving training from library staff and/or the MEDLINE tutorial program.

It is important that we enable more end users to make searches independently of library staff. Because few of our staff have a strong grasp of medical terminology we are concerned that the quality of more complex searches may be uneven. The more end users with search skills, the higher the potential relevance of search results, and the more staff time saved.

Availability of Full-text Articles

MEDLINE citations carry a note if the journal cited is held by our library, and most users reported making some effort, sometimes, to locate such journal articles. But less than half of our users attempt to obtain articles through the Library's Inter-Library Loans (ILL) service. There had been a boom in ILL requests, almost matching the growth in the numbers of searches, from 1990 to mid-1994. The service then declined sharply when ILL funds ran out.

Users have little confidence in the library's runs of journals, which are broken by frequent temporary or permanent stoppages of subscriptions; they are unwilling, too, to depend on the unreliable and very slow ILL service. When its funds run out we are dependent on the goodwill of two or three South African libraries, which cannot, of course, meet all our needs. In response to the library's attempts to improve ILL services, the University has recently approved, in principle, the introduction of realistic (almost full cost-recovery) charges, as opposed to highly subsidized minimal charges for ILL documents. While students will continue to be subsidized to some extent, researchers (academic staff) will be expected to pay for ILL services.

Immediate and Long-term Outcomes of the Service: Products and Spinoffs

Searches

MEDLINE now produces over 380 individual searches each month, as well as 58 monthly updates for individual researchers and clinicians; it also produces the monthly printouts of abstracts for our national digest Current Health Information Zimbabwe (CHIZ).

Journal selection

CD-ROM (used with the bibliographic database management program Papyrus) enables us to use MEDLINE in bibliometric studies to identify those journals which most frequently publish articles of relevance to Zimbabwe's health problems - the citations selected for CHIZ and those selected for printing out by our users. The most frequently cited titles are included in a revised core collection of most-needed journal titles.

The Bradford's Law of Scatter (illustrated in many studies including one at the National Library of Medicine, where it was found that only 300 titles of its 22,000 periodicals satisfied 70 percent of all requests reinforces the benefits of accessing journals through a document delivery service, rather than "owning" expensive subscriptions to little-used journals. The "ownership versus access" algorithm we have developed4 can be brought to bear on the selection process now that lists of most frequently cited titles can be built on objective criteria rather than relying solely on manual analyses of inter-library loans, or on the possibly idiosyncratic or unrepresentative demands of often transient specialists and other academics.

Abstracts

The survey demonstrates the dissatisfaction of users with our journal collection and inter-library loans service, and a considerable - and certainly lamentable - dependence on abstracts rather than on full-text articles, with 73 percent "usually" or "sometimes" making do with abstracts alone. How valuable can abstracts be, how complete the information they contain? No clinician, researcher, or librarian would advocate total dependence on abstracts. Many of them do not accurately reflect the main findings or conclusions of their full papers, and rarely (except in terms of the unassailable reputation of the source publication) enable a critical evaluation of a paper's real worth in terms of methodology and other criteria. A recent study on the reliability of reviews of abstracts for scientific meetings found that reviewers disagreed substantially when evaluating the same abstracts and that structured review criteria were needed.5 The proportion of structured abstracts in the medical literature is increasing, however, and enhances the value and usefulness of bibliographic databases. (See Box 3.)

Outcomes for health professionals, patients?

While published data demonstrate the impact of MEDLINE on patient outcomes in the USA6,7,8 we are not able in Zimbabwe to quantify what the direct and indirect benefits of CD-ROM databases and their abstracts are, in terms of improved efficacy of health professionals, or improved health status indicators, for instance. There were considerable gains in the health status indicators of Zimbabwe's people after the country won its Independence in 1980, but these are now being rapidly eroded by the combined effects of the AIDS pandemic, and of the economic structural adjustment programs being implemented in recent years in many African countries.

In a 1993 snap survey of opinions of senior academic staff on MEDLINE's effects the following were among their comments: "its comprehensiveness and currency [make it] revolutionary"; "it makes life so easy . . . very helpful for research and teaching"; "it plays a very important role in teaching Med students enabling them to keep up with the latest information . . . is invaluable for their clinical presentations"; "it helps me keep up with appropriate management of clinical problems"; "it is indispensable for research protocols and write ups of research projects"; "in the recent meningitis epidemic it provided essential epidemiological and serotyping data for appropriate measures for prevention, control and management"; "it plays a major role for me as a consultant teaching juniors."

Some examples (from the 1995 survey) of respondents' reactions to a question on the effects on their work of NOT having access to MEDLINE: "dramatic - my research and teaching capability would collapse without it"; one urologist declared he "would scream"; others spoke of "catastrophic", "devastating" and "drastic" effects, and "the end of the road."

BOX 3 The Importance of Abstracts

We have learned anecdotally too that our users find that abstracts can and do provide critical information in a crisis - our Drug and Toxicology Information Service, for example, inevitably has to rely on abstracts in emergencies. Abstracts inform policy and assist planning. They have had to satisfy our users for information on management of diseases and poisoning cases for decisions on drug prescribing and dosages, for literature reviews, for preparing lectures, for updating practical knowledge and skills - because the back-up of full texts is not available quickly enough. It may be deplorable but it is also true that in many contexts, including our own, the back-up texts will never be available, or will arrive too late. It must be recognized that in the poorer countries of the world an abstract is likely to be all the reader will ever see of journal articles.

Promotion of the CD-ROM Literature Search Service

We promoted the service initially only to Medical School staff and not to students, anticipating an unmanageable demand on our single workstation. Only two members of staff (out of over 100 in 21 departments to whose chairmen we sent promotional circulars) appeared at our MEDLINE open day. We assume that initially news of MEDLINE and POPLINE spread largely by word of mouth, since there was a steady increase in use after the first few weeks.

After a formal presentation of the first workstation, involving the United States Information Service (USIS), the small paragraph on the event which appeared in the daily paper showed a total lack of understanding of the technology on the part of the reporter. Our handout about CD-ROM was beyond his comprehension - we realized how important it is to recognize the impenetrability of new technologies to those who have no experience of them, and to learn how to communicate their capabilities in the simplest possible way.

We reinforced our initial low-key circulars and notices only when we had acquired two more workstations in the second phase of the project, by promotional demonstrations of MEDLINE (using a Datashow projector we acquired in the second phase of the project) at professional meetings of interest groups (such as pharmacists, physiotherapists, primary care physicians). To provide these demonstrations our reference librarian, with help sometimes from me or another colleague, had to load her car with a personal computer, four drives and a printer, drive to the meeting venue and reassemble the workstation there - a considerable undertaking. From follow-up visits and search requests from those sectors we knew the demonstrations made an impact - seeing is believing. (Recently staff in our Main Library took CD-ROM to individual faculties, and they too found that an immediate increase in use of the databases resulted.)

Demonstrations of CD-ROM searches get the most attention in our library orientation sessions for students; post-graduate students are given a longer introduction than undergraduates, and are strongly encouraged to read database tutorials and take the training which is always available from our staff. However, offers to academic staff and students of training for individuals or small groups have had little response thus far. The lack of interest in our offers of training is not surprising - very few students and staff have had the opportunity to acquire computer skills, and most still have no access to any computer other than ours. They will have to start by learning basic computer skills, and only then attempt to use the (relatively user-friendly) interface software.

Training is more likely to be accepted if it is informally given, by way of explanations to a search requestor as he helps a librarian make his search. Our understanding of the problem of low uptake of training is that, unless users practice and enlarge on newly acquired search skills on a regular basis, they forget them between visits and have to start again. Many of our users have very little spare time for acquiring new skills outside their daily work schedule.

The Medical School is constructing a new Health Sciences building and plans to incorporate considerable computer-aided learning facilities in the building. Easy access to computers will then no longer be limited to the library's CD-ROM facilities.

BOX 4 The African Index Medicus Project

Our database, Utano (the Shona word for Health), provides records for us to contribute to Who's recent database initiative, the African Index Medicus (AIM). The innovative AIM project assembles (at WHO headquarters in Brazzaville) abstracts of mainly unpublished "grey', literature collected from a growing number of African countries. WHO produces and circulates regular printed updates to libraries in Africa, and to subscribers in African institutes throughout the United States and Europe.

Other Uses of Microcomputers in the Medical Library

A database of local health literature

This project to create and compile a database of local unpublished health literature was funded by IDRC and implemented in 1991-1992, concurrently with our expanding CD-ROM service. (See Box 4.) The contract staff, the training in database use (UNESCO's Micro CDS/ISIS) and the hardware brought to the library by the project reinforced our understanding and appreciation of the potential and value of computerization, and has since offered continuing opportunities to all our professional staff to widen their computer skills. The database, after the expiry of the contracts of project staff, has continued its slow but steady growth, as two of our staff members gradually master the complexities of the ISIS software.

Email: local and international access to information

Email now enables health workers outside the capital to get access to MEDLINE searches and other material from the Harare library and beyond. In our third year of reliable email access, we now operate through Zimbabwe's upgraded telephone system, through SatelLife's local HealthNet node in Harare, which links us with other major international email networks as well as with the HealthNet users in provincial medical directorates and many of the district offices. The number of MEDLINE search requests emailed to our outreach librarian from the provinces doubled last year. We download and return search results to users by email, and can at last scan and email photocopies of articles in the library instead of mailing photocopies (which may take a week to deliver). The main limiting factor in the use of email is the still small number of health professionals with easy access to and familiarity with personal computers; but the time taken to scan and run an Optical Character Recognition program against the copy is also something of a deterrent.

Moves to bring us online into the Internet may bear fruit in the next year or two. Meanwhile email and the Grateful Med search interface give us easy and low-cost access to the MEDLARS databases of the National Library of Medicine, through the BITNIS program,9 which enables a turn-around time of 24-48 hours. The costs of our email services are modest: each of our two stations costs about $7.00 per month, less than the cost of one five-page fax.

The impact of microcomputer technologies on library staff

Many libraries in Africa, including our own, have become shelters for dwindling and aging printed resources. The effects on library staff of acquiring computer skills and delivering computerized and very current information are all the more positive in this context. The image of traditionally low-status professionals has been subtly but unquestionably enhanced by their new capabilities - doctors and nurses now request appointments for searches with library staff. . . who have something valuable to offer at last. This empowerment of librarians increases their morale at a time when economic forces have a contrary effect on the work force; computers make the job more fun for the majority of staff who enthusiastically embrace the new technologies.

The growing use of personal computers in our library imposes on us the need to re-examine our management structure; lines of authority have to be redefined. There is need for greater flexibility in job descriptions and responsibilities, need for recognition and monitoring by senior management of any radical increases or reductions in workloads brought about by automation, need for rewarding exceptional skills and providing opportunities for the development of those skills to those who show aptitude. If sustained progress in use of the new technologies is to be made, the old rigid and hierarchical structures of libraries must bend and expand to allow for new patterns and areas of growth. The application of performance indicators to services and processes is one obvious method of planning and managing the changes which automation inevitably brings in its wake.

ANALYSIS OF LESSONS LEARNED

Unquestionably CD-ROM databases, MEDLINE in particular, have brought about a radical change in the information seeking behavior of our users: unprecedented rapid and easy access to current and archival journal literature; and an increased awareness and documented increased use of the journal literature. We believe, however, that the most significant and valuable outcome of this service has been the growing awareness in its users of a range, literally a world of other

wise unknown information sources. It has been by far the greatest force for progress in information access that our users have experienced.

However, while the advent of CD-ROM has been a dramatic success in our library, and while email is the cheapest known method' pro rata, for the speedy transmission of data, there are funding and technical issues that cannot be ignored. We have learned - painlessly so far, thanks to the generosity and flexibility of our donors - that one-time set-up costs are not a reality.

Planners must face and meet the costs of maintaining and updating expensive hardware to keep pace with changes in database sizes and interfaces; they must have access to expert and unbiased information on the new technologies to avoid costly mistakes in purchasing inappropriate equipment. They must allow for the relentless growth of RAM requirements; the need for multiple workstations when multiple users want access to databases; the need for the more expensive laser printers when printing out images from full-text CD-ROMs or other electronic sources such as World Wide Web. The Internet has come to neighboring countries and is likely to reach us too before very long, with further implications for hardware, training and other costs. Growth and development in the technologies are inevitable and expensive. Institutional support and involvement become increasingly vital if the technologies are to survive and prosper in the long term in Africa.

The lack of concern of institutional planners for information supply may be attributable to negative experiences and low expectations in the libraries they have known, as much as to the failure of educational systems to include information retrieval skills in their curricula. We expect MEDLINE to break that cycle in Zimbabwe, by sensitizing students - the planners of the future - to the existence of vast and formerly unknown resources. The cost of introducing and maintaining a MEDLINE search service, represented as a proportion of the budget of a Medical School or a Health Ministry, would be insignificant. That cost calculated as a percentage of the salary of each health professional in government service would be negligible; its cost effectiveness in terms of the development of human resources and services would be obvious.

Our survey, made after five years of use of MEDLINE, shows that this database has become essential to the work of many of our academic staff: eighty per cent of respondents stated that they would elect that the library cancel ten journal subscriptions rather than the MEDLINE subscription. We are likely to have to make just such a substitute in the near future. As a first step it is essential to expose present and future generations of planners, teachers and practitioners to these databases. The fact that the library has been able to offer four or five workstations providing MEDLINE and other databases to all comers has meant quick, easy, and free access to vast quantities of medical literature, which would otherwise be inaccessible.

CONCLUSIONS AND RECOMMENDATIONS

We have realized to a considerable extent the potential of CD-ROM technology at the UZML. The primary effects of the medium have been a vastly increased access to and use of current health information sources, whether in the form of abstracts or full-text journal articles or segments from constantly updated textbooks; and along with this increase there have been very significant changes in the information seeking behavior of large proportions of academic staff and students in the Medical Faculty and of health professionals in government service and the private sector. There is evidence elsewhere, and we must to a great extent assume, that the increased use of the literature and the resulting increased knowledge base of health professionals must produce the secondary and most important benefits of improved health services in general and better patient care in particular.

Our experiences in providing CD-ROM literature search services suggest that increased and more effective use would be made of the databases, and of other computer technologies such as email and eventually the Internet, if more of our present and potential users had been previously exposed to computers, or had some computer training. Economic circumstances in Zimbabwe have hitherto prevented all but the most privileged sections of our present and potential user community from owning or having easy access to computers. However, the rate of growth in the computer industry, and growing awareness among members of the health care community of the immeasurable benefits of personal computers - to which our project has made a major contribution - suggest that a critical mass of computer users will have been reached in the longer term.

Once this critical mass has been achieved, the question of sustainability will cease to be such a worrisome issue. Institutions, health services, and individuals will have their own strong motivation to supply the material and human resources necessary to exploit and develop the new technologies.

NOTES

1. Patrikios, H.M. (1985) Socio-economic changes in developing countries: the concern of the medical librarian? In: Medical Libraries, One World: Resources, Cooperation,Services. Proceedings of the Fifth International Congress on Medical Librarianship. Tokyo: Keio University Medical Library.

2. Patrikios, M.M. (1995) Effects of MEDLINE abstracts on information use in an African medical Library. Paper presented at 7th International Congress on Medical Librarianship: Health Information for the Global Village, Washington, D.C.

3. Warren, K. (1987) The evolution of selective biomedical libraries and their use in the developing world. Journal of the American Medical Association 257(19):2628-9.

4. Patrikios, H.M. (1994, 1996) A minimal acquisitions policy for journals at the University of Zimbabwe Medical Library. In: Survival strategies in African university libraries. American Association for the Advancement of Science.

5. Rubin, H.R. et al. (1993) How reliable is peer review of scientific abstracts? Looking back at the 1991 Annual Meeting of the Society of General Internal Medicine. Journal of General Internal Medicine 8(5)255-8.

6. Lindberg, D.A., et al (1993) Use of MEDLINE by physicians for clinical problem solving. Journal of the American Medical Association 269(24):3129.

7. Wilson, S.R., et al. ( 1989) Use of the Critical Incident Technique to evaluate the impact of MEDLINE: Final report. Palo Alto, CA: American Institutes for Research.

8. Klein, M.S. et al. ( 1994) Effect of on-line literature searching on length of stay and patient care costs. Academic Medicine 69(6)489-95.

9. Search costs are currently being funded by SatelLife, Cambridge, Massachusetts.

10. Patrikios, H.M. ( 1993) A minimal acquisitions policy for journals at the University of Zimbabwe Medical Library. In: Survival Strategies in African University Libraries: New Technologies in the Service of Information. American Association for the Advancement of Science, Washington, D.C., p. 96.