Biomedical Informatics Training in South Africa Christopher J Seebregts1,2, Maurice Mars2, Deshendran Moodley3, Hamish SF Fraser4, Sarie P Human5, Carl Fourie3, Yashik Singh3, Milan Hajek3, Woolaganathan Pillay3, Jules-Raymond Tapamo3, Hugh Murrell3, Tulio de Oliveira6, Donnie McGrath7, and Michael L. Bennish8 1

Biomedical Informatics Research Division, Medical Research Council, South Africa; Department of Tele-Health, Nelson R Mandela Medical School, University of KwaZuluNatal, South Africa; 3School of Computer Science, University of KwaZulu-Natal, South Africa; 4Brigham and Women’s Hospital and Partners in Health, Harvard University, Boston, USA; 5Department of Health Studies, University of South Africa; 6Department of Zoology, Oxford University, UK; 7TUFTS University Medical School, Boston, USA; 8Nuffield Department of Clinical Medicine, University of Oxford, UK 2

Abstract: Informatics is essential for biomedical research. There is, however, a lack of skilled practitioners and training programmes in developing countries. We have established a cross-disciplinary Biomedical Informatics Training programme in South Africa (www.biomedinfo.org) with the support of the Informatics Training for Global Health Programme of the Fogarty International Centre, National Institutes of Health, USA. The Informatics Training Programme is one of only a few such programmes providing training in research informatics in a developing country. It offers a fulltime coursework Masters in Medical Science (Medical Informatics) for computer science graduates and a Master of Public Health with specialization in medical informatics for medical and life science graduates The Master of Public Health degree is offered either parttime through the University of KwaZulu-Natal or by distance learning through the University of South Africa. An MSc or PhD by research is also available for qualifying students. Teaching is provided jointly by the Departments of Tele-Health and Computer Sciences at UKZN and research supervision is provided by established informatics researchers. Nine trainees registered in the first year of operation and twenty-seven applicants have been received for the second year. Students have made a valuable contribution to existing projects and collaborations.

INTRODUCTION Biomedical informatics is a new discipline at the convergence of medical informatics and bioinformatics [1] [2] [3]. Combining traditional medical informatics strengths in data management, electronic medical record systems, decision support systems and knowledge representation, and incorporating new tools and techniques in bioinformatics analysis and genomic medicine, it has made for new discoveries in biomedicine, particularly in the field of infectious diseases, genomics [4] [5] [6] [7] and public health [8]. Although several informatics training programmes exist in developed countries (eg [9]), there are few programmes in developing countries, despite the substantial need. This probably results from the lack of persons with cross-disciplinary skills in biomedical science and research coupled with knowledge and proficiency in computer science and/or information systems to staff such programmes.

In 2002, we established medical informatics training in South Africa via a part-time programme leading to a Masters degree in Medical Informatics in the Department of Tele-Health at the University of KwaZulu-Natal (UKZN). This programme has been established with the financial support of the International Training in Medical Informatics (ITMI) Programme of the Fogarty International Centre, National Institutes of Health (FIC-NIH), USA (www.ukzn.ac.za/telehealth). In 2004 we expanded the training programme to include training in biomedical informatics (www.biomedinfo.org). This programme was established with the support of the continuation grant of the ITMI programme from FIC-NIH, which was broadened in scope and renamed the Informatics Training for Global Health (ITGH) Programme. NIH institutes in addition to FIC contributing to the ITGH programme include the US National Human Genome Research Institute, the National Institute of Biomedical Imaging and Bioengineering and the National Library of Medicine

(http://www.fic.nih.gov/programs/training_grants/it gh/index.htm). The same consortium that had one of the original ITMI awards from FIC also won in competitive recompetition the successor ITGH award. This consortium was led by Tufts University, with the prime South African partner being the University of KwaZulu-Natal. Partner institutions on the grant have been the South African Medical Research Council (MRC), Harvard University, the Massachusetts Institute of Technology (MIT) and TUFTS-New England Medical Centre. This consortium proposed to develop a biomedical informatics training programme at UKZN, building on the previous part-time Masters Programme in Medical Informatics at UKZN (www.ukzn.ac.za/telehealth). The first programme to be established was a fulltime taught Masters in Medical Science (MMedSc) over a period of eighteen months. Soon after initiation of this Masters degree programme, the teaching was expanded to include a Masters in Public Health (MPH) with specialization in medical informatics coordinated by the Department of TeleHealth and the School of Family and Public Health, Nelson R Mandela School of Medicine at UKZN or, by distance learning, through the Department of Health Studies at the University of South Africa (UNISA; (http://www.unisa.ac.za/Default.asp?Cmd=ViewCo ntent&ContentID=14073). UNISA is the largest distance-based university in the Southern hemisphere and provides the required infrastructure for distance-based education that can be expanded to many other countries, particularly in sub-Saharan Africa. The MPH programmes also extend the scope of the Programme to include students with a medical or life sciences background. A second ITGH grant in South Africa was awarded to a consortium of researchers and trainers from Stanford University and SANBI. The StanfordSANBI Biomedical Informatics Programme (http://stanford.sanbi.ac.za/) emphasizes bioinformatics and short courses while the emphasis in the UKZN programme is on medical informatics and a taught Masters degree. However, the joint interest in biomedical informatics presents substantial opportunity for collaboration. The two groups (UKZN and Western Cape) are already working on one large collaborative research project (see www.saturn4hiv.org).

through appropriate didactic content and experiential learning through existing research projects, either in the student’s current work place or provided by one of the faculty members. Qualifications The Programme qualifications:

presently in

Medical

offers

three

1.

MMedSc UKZN;

Informatics

at

2.

MPH (Medical Informatics) at UKZN or UNISA, and;

3.

MSc or PhD by research in the Department of Computer Science at UKZN.

MMedSc and MPH students are registered once at the beginning of each year while MSc or PhD students by research may register at any convenient time during the academic year. Registration for the MPH programme only occurs every alternate year although students may register for the computer modules in any year. Didactic Programme The didactic part of the Programme comprises a selection from the following modules. The three core modules offered by the Department of Tel-Health are: 1.

Introduction to Physiology Biochemistry (16 credits)

2.

Quantitative Skills for Informatics 1 (8 credits)

Medical

3.

Quantitative Skills Informatics 2 (8 credits)

Medical

for

and

The three core modules offered by the Department of Computer Science are: 1.

Medical Information Systems (16 credits)

2.

Electronic Medical Records (16 credits)

3.

Medical Artificial Intelligence (16 credits)

The six elective modules offered by the Department of Computer Science are: 1.

Bioinformatics (16 credits)

2.

Cryptography and Network Security (16 credits)

3.

Programming 1 (16 credits)

4.

Programming 2 (16 credits)

METHODS

5.

Databases (16 credits)

The UKZN programme is focused on the development of research informatics capacity

6.

Object Oriented Programming Software Design (16 credits)

and

During 2007, additional modules will be developed including a specialist “Introduction to Biomedical Informatics” module. MMedSc in Medical Informatics The MMedSc (Masters of Medical Science) in medical informatics is a fulltime taught Masters offered jointly by the Department of Tele-Health and the School of Computer Science at UKZN over a minimum of 18 months. The entrance requirement is an honors degree or four-year degree equivalent as well as undergraduate training in computer science up to at least final (third) year undergraduate degree level. In order to be awarded the degree of MMedSc, a student needs to accumulate a total of 196 credits from a combination of the didactic programme (132 credits) and the research project (64 credits). In order to earn the 132 didactic credits, students need to take the three core modules offered by the Department of Tele-Health (32 credits) and the three core modules offered by the School of Computer Sciences (48 credits). The remaining credits are earned by taking elective modules. As from 2008, the research project component will be increased to half the total points awarded (96 points) with the remaining half (96 points) to be made up from the didactic courses. This will ensure that the degree is in line with international best practice for taught Masters programmes and also allow students and trainers to attract additional funding from national funding bodies that require a minimum of 96 points to provide Masters bursaries (scholarships). MPH (Medical Informatics) The MPH (Medical Informatics) is offered jointly by the Departments of Tele-Health and Computer Science at UKZN and the Department of Health Studies at UNISA. The UKZN MPH is a part-time Programme at Nelson R Mandela School of Medicine in Durban while the UNISA Programme is distance-based. The UNISA MPH programme offers two streams: a computer science steam that emphasizes software development including C++ and Java programming and an information systems stream that concentrates on visual programming tools and database development. Conversion from MPH to MMedSc In exceptional cases students enrolled on the MPH programme may be permitted to convert to the MMedSc degree with approval of the members of the Programme Curriculum Committee. The general criterion for allowing students to convert between programmes is performance in the first and second year computer science modules and the

student’s general aptitude for computing. Conversion is only available for MPH students enrolled in the computer science stream. MSc (Medical Informatics), by research The Programme in conjunction with the Department of Tele-Health and the School of Computer Sciences at UKZN offers MSc and PhD (Medical Informatics) by research for applicants with approved backgrounds. Typically, successful applicants will have already undergone significant didactic training in computer science and will have been accepted by the relevant department in which the research is to be conducted. Research projects are either provided by the Department or via the student’s own working environment. Applicants The Programme encourages applications from medical informatics professionals working in existing research programmes/environments in subSaharan Africa and previously disadvantaged groups within South Africa. A limited number of stipends are available. Selected students may be sponsored to spend time at collaborating US institutions. RESULTS Programme Growth Since its inception the programme has experienced rapid growth both in the number of trainees and in the number of training programmes offered, reflecting the demand for informatics training in sub-Saharan Africa. Nine trainees enrolled during the first year of the grant. In the second year the number of trainees has increased to 27 with a majority of applicants from countries in sub-Saharan Africa other than South Africa. In addition to the existing complement of lecturers, the Programme is developing capacity to sustain the Programme in future years. The teaching programme will be supplemented in 2007 with faculty with strong expertise in bioinformatics and biomedical informatics. Experiential Training Experiential training is a critical component of the Programme both because it is a requirement for the research project and also because it prepares students for the working environment. The programme has drawn on a network of researchers to offer students leading edge projects such as (i) development of open source electronic medical record software, and; (ii) development of

advanced algorithms for analyzing HIV genetic sequences. Collaborative Open Source Software Development Software development is a central activity of most informatics programmes and the South African Biomedical Informatics Training Programme presently participates in the collaborative OpenMRS (Medical Record System) software development project (www.openmrs.org; [10]). The OpenMRS consortium has developed and implemented an open source electronic medical record application predominantly for HIV patient and treatment management in developing country settings. There are several advantages to participation in collaborative, open source software development projects, including the following: Firstly, students gain valuable experience in the principles of open source software development. With the limitations on resources in developing countries, open source is expected to be a major paradigm for the development of appropriate developing country software and applications. Secondly, valuable teaching and experiential learning is also obtained from other developers in the open source network. This is particularly valuable considering the scarcity of skilled lecturing staff in developing countries, particularly in computer science and software deveopment. Thirdly, students are encouraged to contribute code developed as part of coursework and research projects to the open source code base. This teaches students to develop robust working code and also contributes to the code base of OpenMRS. Our students and collaborators are presently driving the development of handheld application extensions to OpenMRS that will enhance our previous work using handheld computers [11] [12]. This is both an innovative informatics research and development project as well as a useful application for health research data collation and management. In terms of this latter use, the project will contribute to research in other areas through efficient data management for patient and treatment monitoring, general health research data collection and surveys. Computational Analysis of Resistant HIV Genomes Another biomedical informatics research project students have participated in is the development of advanced algorithms and machine learning techniques for predicting drug resistant phenotypes from genotypic mutation patterns in HIV. HIV drug resistance testing is rarely conducted in the management of patients on antiretroviral therapy in South Africa even though it would be useful to

monitor the emergence of drug resistant phenotypes in Southern Africa. By bringing together clinical and epimemiological observations with genotypes collected, from genetic studies and then analyzing associations with machine learning techniques, we hope to better understand genetic drivers of the HIV-1 subtype C epidemic that is prevalent in southern Africa. This is a joint project by the Southern African HIV Treatment and Resistance Network (SATuRN; www.saturn4hiv.org), a project originally initiated by the Stanford-SANBI ITGH Program. Collaborations In addition to the collaboration with the SATuRN network and the Stanford-SANBI ITGH Programme, described above, the South African Biomedical Informatics Training Programme has also established a valuable collaboration, with the Masters Programme in Telemedicine at UKZN. The collaboration with Telemedicine will result in the development of student projects in image processing and the utilization of telemedicine infrastructure for data management and teleeducation. Tele-Education The Biomedical Informatics Training Programme has also implemented a tele-education service to lecture students. This service will enable lecturers to participate in the program from any of three sites in South Africa and increase the potential to attract and keep high quality lecturers. If required, we may expand this service using bandwidth implemented for telemedicine applications.

University of South Africa MRC Pretoria

University of Pretoria

VideoConferencing Stellenbosch University MRC Cape Town University of Cape Town

ISDN

MRC Durban

University of KwaZulu-Natal

University of the Western Cape

Figure 1. Tele-Education Facility CONCLUSION Significant demand exists for biomedical informatics training in sub-Saharan Africa. The collaborative Biomedical Informatics Training Programme supported by the ITGH programme of the FIC of NIH is expected to play a significant

role in meeting this demand. During 2007, the biomedical informatics component of the project will be enhanced through the involvement of key personnel with domain expertise in bioinformatics and computer science and the development of specific biomedical informatics teaching material. The programme has also established its primary base at the MRC, which has become the primary grant holder. This is of note because it is the only one of the ITGH grants in which the primary grant holder is in a developing country rather than in the United States, but also by holding the grant at the MRC it allows for greater future integration of other tertiary educational institutions in South Africa. ACKNOWLEDGMENTS The South African Biomedical Informatics Training Program is funded by an Informatics Training for Global Health grant from the Fogarty International Centre, National Institute of Health, USA: “Developing Informatics Capacity in South Africa” (Grant Number 1 D43 TW007004-01), supplemented by in-country funding from the MRC and UKZN. Additional funding for student projects was provided by two grants from the ACACIA and Connectivity Africa Programmes of the Canadian International Development Research Centre (IDRC): “Developer Network and Open Source PDA Software for Health Data Collection” (Grant Number 101974-001) and “Development and Interpretation of a Medical and Bioinformatics Database to Support and Evaluate the AntiRetroviral Therapy Rollout Programme in the Free State Province, South Africa” Grant Number 102411-001). References 1.

2.

3.

4.

Shortliffe EH, Perreault LE, Wiederhold G, Fagan LM. Medical Informatics. Computer Applications in Health Care and Medicine. Second edition. New York: Springer-Verlag; 2001. Shortliffe EH and Cimino JJ. Biomedical Informatics. Computer applications in Health Care and Biomedicine. Third edition. New York: Springer-Verlag; 2006. Seebregts, C. Biomedical Informatics. African Institute of Biotechnology Science and Technology 2006; 16: 4-5 Maojo V, Martin-Sanchez F, Billhardt H, Iakovidis I, Kulikowski C. Establishing an Agenda for Biomedical Informatics. Methods Inf Med 2003, 2: 121-125

5.

Altman RB. Bioinformatics in Support of Molecular Medicine. AMIA Annual Symposium. Orlando: 1998. p. 53-61. 6. Altman RB. The Interactions Between Clinical Informatics and Bioinformatics. JAMIA 2000;7(5):439-43. 7. Martin-Sanchez F, Maojo V, Lopez-Campos G. Integrating Genomics into Health Information Systems. Methods Inf Med 2002;41:25-30. 8. Maojo V, Martin-Sanchez F. Public Health Implications of Bioinformatics. Yearbook of Medical Informatics 2004:137-43. 9. Aronsky D, Aliferis C, Johnson K, Lorenzi N, Miller R. Biomedical Informatics Training Program at Vanderbilt University. Yearbook of Medical Informatics 2004:165-74. 10. 17. Mamlin, B., Biondich, PG , P. Biondich, B. Wolfe, H. Fraser, D. Jazayeri, C. Allen, J. Miranda, and W. Tierney, Cooking Up An Open Source EMR For Developing Countries: OpenMRS - A Recipe For Successful Collaboration. Proc AMIA Symp 2006: 529533. 11. Zwarenstein M, Seebregts CJ, Mathews C, Fairall L, Flisher AJ , Seebregts C et al. Handheld Computers For Survey and Trial Data Collection in Resource-Poor Settings: Development and Evaluation of PDACT, a Palm™ Pilot Interviewing System. Manuscript submitted for publication, Int J Med Info 2006. 12. Jaspan HB, Flisher AJ, Myer L, Mathews C, Seebregts CJ, Berwick JR, Wood R, Bekker LG. Methods for collecting sexual behaviour information from South African adolescents - a comparison of paper versus personal digital assistant questionnaires. J Adolescent Health. 2006. Accepted. Address for correspondence Dr Chris Seebregts, Biomedical Informatics Research Division, Medical Research Council, PO Box 19070, Tygerberg 7505, South Africa, E-mail: [email protected]. Details of the informatics training programmes, are available from the Project Administrator, Ms Shivani Naidoo, Tel: +27 31 260 4543, or by email at [email protected].