KEMRI's Management Information System

by James N. Muttunga

James Muttunga is a Senior Research Officer at the Kenyan Medical Research Institute (KEMRI) and head of their information systems. He carries out consultancies for research projects in information management that have been funded by the Commonwealth Regional Health Community Secretariat. He is a biostatistician by profession.

BACKGROUND AND CONTEXT OF THIS PROJECT

The Kenya Medical Research Institute (KEMRI) was established in 1979 by the Science and Technology Act of Parliament. Its Board of Management is accountable to the Minister for Research, Science and Technology and it is primarily funded by the Kenya Government. KEMRI is mandated to conduct research in the biomedical sciences and its main objectives are to:
· cooperate with the other research organizations and institutions of higher learning in training programs and in matters of relevant research;· work with other research bodies within and outside Kenya carrying out similar research; and· cooperate with the Ministry of Health, the National Council of Science and Technology, and the Medical Science Advisory Research Committee in matters pertaining to research policies and priorities.

The Member Research Centers

KEMRI has about 1,200 staff members, of which about 450 are either scientists or technical staff. They operate through eight research centers. One is in Busia (near the Kenya/Uganda border); another is in Kisumu (in the lake Victoria region). The rest are in Nairobi: four at KEMRI headquarters and two centers near the Kenyatta National Hospital. The research centers are:
· Alupe Leprosy and Skin Diseases Research Centre (ALSDRC)· Biomedical Sciences Research Centre (BSRC)· Clinical Research Centre (CRC)
· Vector Biology and Control Research Centre (VBCRC)
· Virus Research Centre (VRC)
· Centre for Microbiology Research (CMR)
· Medical Research Centre (MRC)
· Traditional Medicines and Drugs Research Centre (TMDRC)

PROJECT DESCRIPTION

In 1987, KEMRI appointed a team to write a proposal to develop our management information system (MIS). The team suggested that KEMRI first embark on a pilot project that would:

· Study and identify KEMRI's data processing needs and requirements;
· Provide some additional hardware to facilitate immediate processing requirements; and
· Facilitate exchange of knowledge through training visits to institutions with an operational information management system.

Professor Dean Haynes, then the Deputy Director of the International Center for Insect Physiology and Ecology, an international research center located in Nairobi, helped us by identifying consultants who had the relevant experience and had carried out similar activities. The chosen consultant first reviewed KEMRI's current inventory of microcomputer hardware and software. The proposal team then reviewed the training needs for the data processing support staff, the scientists, and administrative support staff. We interviewed scientific, technical, and administrative support staff end asked about their present operational systems, the available resources for information technology, and current needs for research production.

We also arranged to train two KEMRI scientists at the Tropical Diseases Research Center at the World Health Organization (TDR/WHO) in Geneva. There, they were able to familiarize themselves with an operational management information system. The training program included working visits to each program and familiarization with the existing techniques for data capture, automation, and analysis.

We completed the study after about six months and presented the report to the KEMRI Board of Management for approval. Based on the initial study and the experience from the training visits, the management board adopted the study report and used its recommendations and identified needs as the basis for a comprehensive proposal for the development of KEMRI's MIS. The primary objectives of this project were to improve KEMRI's capability in:

· Data processing and information management, through the acquisition of relevant hardware and software;
· Literature search services, through online and CD-ROM searches;
· Desktop publishing, through provision of necessary hardware and software; and
· Human resource development, by providing training aimed specifically at scientific, technical, secretarial, library, clerical, or data entry staff.

The main project proposal underwent a number of revisions and reviews and was not approved for financial support until the second half of 1991. This project benefited from the technical input and support of several reviewers and consultants who shared their opinions, concepts, and understanding with us. The process was in itself tedious and, at times, was affected by standard bureaucratic processes resulting in substantial delays in implementation and funding. The main bottlenecks were mainly due to two factors:

· Lack of adequate information and understanding on the size, magnitude, and extent of operations at KEMRI on the part of the review teams named by the potential donors; and
· Inadequate knowledge of microcomputers systems and the costs of these processes on the part of KEMRI staff.

The hurdles were, however, overcome and the project was finally accepted and funded by the Carnegie Corporation of New York. The original budgets and planned activities were, in the end, drastically reduced to facilitate a well-phased development and implementation schedule. This allowed us to follow the strategy of implementing the system according to the primary needs of KEMRI and then building the secondary systems as needed and as resources permitted.

PROJECT EXPERIENCE AND IMPLEMENTATION

The final project provided five main initial services, which are described in some detail below.

Research Strengthening Support

The study of existing hardware revealed that computer systems of differing capacities and types had been acquired through assorted project grants. Neither the various systems nor the software were initially compatible with each other.

The Board of Management recommended that the equipment acquisition process be better coordinated in order to identify the necessary systems, select software that could be supported by the technical staff, and facilitate the exchange, sharing, and compatibility of all new systems and software. We therefore adopted a policy of standardizing all microcomputer systems to a common configuration to facilitate compatibility between machines and the exchange of data sets and software applications.

The project provided two microcomputer systems in each of the eight centers and in the secretariat for data management, analysis, and report preparation by scientific, technical and support staff. We also acquired a dot matrix printer and an uninterruptible power supply (UPS) for each center. We strengthened the central computer facilities by acquiring powerful systems that could support KEMRI's large data management and analysis projects. The project also installed appropriate software for data management, analysis, graphics, word processing, utilities, communications, and desktop publishing.

We also stated that all new projects that were expected to have massive data analysis requirements would have to include funds for the necessary equipment in their budgets. Thus a number of projects acquired the necessary support systems. Some of the research programs that benefited from this new policy are listed in Table 1.

TABLE 1 Research Programs that Included Acquisition of Computer Equipment

We initially planned to acquire the hardware and software through local firms that could provide the necessary support and maintenance. We learned, however, that the prices of these systems - even when ordered duty-free - were excessively higher than what we had budgeted. We decided on this strategy: provide, first, the new technologies that were not available directly from the local dealers. These included the CD-ROM systems for the library and the communication hardware and all the software applications that we had identified for use throughout KEMRI. We acquired the other systems after attending the first meeting of the network of Carnegie grantees in May 1992. This meeting was organized by the U.S. National Research Council and held at the African Regional Centre for Technology (ARCT) in Dakar, Senegal. During the meeting, I learned about computer manufacturing firms who were selling high quality systems at competitive prices in the United States - through a direct-mail order system - and who had already supplied some to ARCT. I used this new contact to order the new systems at prices comparable to what we had budgeted! (See Box 1.)

The equipment acquisition process has improved gradually and we have now identified other dealers who can provide equipment. The sourcing of new products is presently carried out after soliciting different quotations from a few international firms.

Documentation and Library Support

The project acquired two microcomputer systems and the hardware accessories (UPS and dot matrix printer) for use in the library and at the central computer facility for online literature searches and internal database development. Two external CD-ROM drives were installed together with Small Computer System Interface (SCSI) cards and we subscribed to MEDLINE and POPLINE databases on CD-ROM. These databases formed the core sources for literature searches performed by scientists at KEMRI. We did learn that the CD-ROM drives should be dual or multiple drives to improve on easy access and to minimize "disc swapping." Many large databases are carried on several discs and it is quite convenient to be able to search all the discs at one time. This is not possible in a single drive system.

The availability of MEDLINE on CD-ROM provided us with a unique and unexpected opportunity. For nearly three years we had been trying, without much success, to provide the Board of Management with a current list of publications and of research developments from KEMRI's scientists. The scientists had responded very poorly to our requests for copies of their published papers because they were unable to provide us with hard copies. We decided, therefore' to search MEDLINE and POPLINE - as well as our own annual medical proceedings - for information written by our own scientists. We then compiled a book of abstracts that covers the scientific output of KEMRI staff.

In early 1993, we compiled and published the first summary, covering material published between 1985-1991. In 1994, we compiled, printed, and circulated among scientists and institutions a second summary, covering 1980-1984. These abstracts now cover the period between 1980-1993, and part of 1994. They have been entered into a central database and are being edited and formatted for eventual publication on disk and in hard copy. We hope that this publication will be the first step toward the development of local databases.

Human Resource Development

The project has trained many staff groups. The type of training provided is summarized in Table 2.

Desktop Publishing

The project supports the dissemination of research results by providing desktop publishing (DTP) services. We have acquired software and a laser printer, scanner, and other accessories. DTP was initially used for the production of the proceedings of the annual scientific conference that has been convened for the last 15 years by both KEMRI and the Kenya Trypanosomiasis Research Institute (KETRI). This conference has improved in quality and has been expanded to a regional "African Health Science Congress," which is now held in different countries on a rotating basis within the African region.

TABLE 2 Training Program

The production of the conference proceedings has also improved and a new journal, The African Journal of Health Sciences, has evolved. KEMRI selects quality papers from the congress and from scientists in the region for peer review and publication in the new journal. In addition to typesetting and producing camera-ready copy for the journal and the proceedings, KEMRI DTP staff also publishes the annual report and the quarterly KEMRI News.

Email and Communications

The project has also acquired some 2400 baud modems and has installed two telephone lines for use in electronic communication among scientists in the centers, as well as among local and international organizations and associations. This has facilitated the exchange of information among local scientists and with the international scientific community.

BOX 1 Sudden Depreciation

The project suffered a nearly 40 percent loss from the sudden depreciation of the Kenya shilling against the U.S. dollar during its second year. This problem has affected all KEMRI projects and has led us to seek authority to run two external accounts, with one running as a straight U.S. dollar account and the other as local external account for all projects.

RESULTS, IMPACTS, AND BENEFITS OF THE PROJECT

Hardware and Software

KEMRI has experienced a dramatic growth in provision and use of microcomputer systems for various research activities. With the funding provided by the Carnegie Corporation, we have acquired 22 new computers. In addition, a number of systems have been acquired through the research activities supported by other donors and collaborating institutions. These systems all have similar specifications since all computer acquisitions are arranged through the technical support group.

TABLE 3 Hardware and Software Donations

Standardized software has now been acquired and installed in all the existing systems - irrespective of the funding source. A summary of these acquisitions and the donor that supported their purchase is provided in Table 3.

Usage

Figures 1 and 2 show that access by staff to computers has risen from 46 percent in 1992 to about 60 percent in 1994. Word processing is the most used application. Use of spreadsheets (Lotus 1-2-3) end presentation software (Harvard Graphics) has remained below 20 percent but the ability to use MEDLINE on CDROM rose from 30 percent in 1992 to 46 percent in 1994. Email services have been poorly used overall with user rate ranging between 4 percent and 12 percent. Eighty percent of the scientists are aware of the existence of MEDLINE on CDROM and 64 percent have conducted a search at either the Library or the central computer facility.

Figure 1: Percentage of Scientists at KEMRI with Access to Computers

Figure 2: Percentage of Scientists at KEMRI Able to Use Specific Software Applications

The KEMRI Library has, over the past few years, experienced a reduction in the number of journal subscriptions it was able to carry. With the total number of journals reduced to about 20, a shrinking budget for books and other literature services, and inter-library loan transactions taking anywhere from weeks to months, the scientific process had been rendered a myth. The introduction of the MEDLINE database on CD-ROM has transformed this process. The CD-ROM service is complemented by the traditional process of providing reprints as requests are received by the librarian. In terms of volume, the number of requests forwarded to the main library for reprints has declined from about 415 in 1990, to 208 in 1991, to 135 in 1992, to 83 in 1993, and to 45 in 1994.

Comparatively, the total requests for MEDLINE literature searches on CDROM have shown a positive increase during the past two years with about 180 requests in 1993 and rising to 240 in 1994. This in itself is an underestimate as it includes only those scientists who were assisted during the search sessions or who had their searches printed. It excludes those scientists who conducted their own searches or who downloaded the search results to floppy disks.

Other Benefits

There are many other benefits of this project. These include the following:

· The tendency to use the literature available in the MEDLINE database is contributing substantially to the production of quality papers and reports in KEMRI.
· All the reference, documentation, and library automation services, including on-line searches using CD-ROM and the development of localized databases, are wholly provided by the project.
· The training of the scientific, technical, and administrative support in use of microcomputers for various operations has been provided through the project activities.
· The project provides full support to all email, communications, and desktop publishing services that exist within KEMRI.
· Nearly 90 percent of the software available and in use in all the centers and computer laboratories has been provided by the project, inclusive of those systems that were acquired and donated by the other operational projects in the institute.

Box 2 Timing the Training Sessions

Before the project got fully under way, IDRC, which was funding one of KEMRI's large, field-based diarrhoea projects, agreed to support a training workshop for about 30 scientists in the region - of whom about 50 percent were senior scientists in KEMRI. The training course content included an introduction to microcomputers, operating systems and standard software, and online literature searching and library information systems. The data processing staff in KEMRI coordinated and facilitated these training workshops. However, since KEMRI had not yet installed computer systems, the demand for computer access that the training workshop generated could not be met. As a result, more than half of this first group needed to be retrained four years later when the equipment was more widely available.

ANALYSIS OF LESSONS LEARNED

In the course of the project implementation, we discovered several things that we should have done differently. First, the training of staff should ideally be planned to take place immediately after the acquisition of the microcomputer systems. If training is conducted prior to the availability of computers, then many of those trained will not be able to use their new skills and will need to be retrained. (See Box 2.)

Second, we learned that hardware installation should be targeted to those services that are generating data and reports. These services require "power users" who have a genuine need for the technology. Some of the applications aimed at institutional strengthening were not used as much as we expected, simply because there was not a strongly felt need for them. We also discovered that computer services located at a central facility help to promote usage and access. Gradually, the access can be decentralized while the support, training, and maintenance are centralized.

In this same vein, we learned that email services should be provided initially to those scientists and researchers who are the most motivated to communicate with colleagues in other countries or regions. Those who share a common interest in a research area or in problem-solving will readily adapt to a new technology that makes communication relatively quick and inexpensive. Their enthusiasm encourages others to try the technology.

We were pleased with our early decision to establish standards for the acquisition of both hardware and software. By setting up such procedures for coordinated procurement, we were able to avoid incompatibility problems and by selecting a standard suite of software, we were able to facilitate sharing and exchange of information and data sets.

Another part of our project that worked well was the identification of a "computer coordinator" in each department. This person acted as the link between the users, the technical support team, and the procurement officers. The coordinators were members of the oversight committee that managed the implementation and development of the MIS project.

Finally, the project should have built in mechanisms oriented toward self-sustenance. Evaluation systems need to be in place from the beginning. This involves the identification of tangible process indicators that can be used in measuring the impact the project.

CONCLUSIONS AND RECOMMENDATIONS

We are enthusiastically planning the next phase of this project. In that phase, we will do the following:

· Install a local area network to facilitate easier access to hardware and software by more staff.
· Replace the single drive CD-ROM readers with multiple-drive machines in order to improve search services. We will continue subscribing to MEDLINE and other databases.
· Promote the development of more local databases.
· Strengthen the desktop publishing operation in order to efficiently produce the new journal; put these services on a self-sustaining basis.
· Install a communication system in order to promote the use of email and to take advantage of Kenya's newly-achieved access to the Internet.
· Implement new training strategies aimed at strengthening the technical staff with sufficient skills to support and maintain the LAN and at sensitizing more of the staff to the available services.

The idea for this project dates back to 1987 and it seemed to take an incredible amount of time to actually launch it. Looking back, however, we can see that we have accomplished much since we actually began implementing the project in 1991. The project is responsible for about 80 percent of all data management, analysis, and report productions systems and services currently available at KEMRI. Furthermore, the project has helped to develop an information culture at KEMRI and this has had a positive impact on every aspect of our activities.