Knowledge-Based Systems for Strengthening African Health Systems

Wendgounda Francis Ouedraogo

and Andreas Nürnberger

Faculty of Computer Science, Otto-von-Guericke University Magdeburg, Universitätsplatz 2 39106 Magdeburg, Germany

Keywords: Knowledge-Based Systems, Expert System, Healthcare, Health System, Sub-Sahara Africa.

Abstract: The set of difficulties characterizing health systems in Sub-Saharan Africa (SSA) are caused by many factors,

among others, the very limited number of specialists, the predominance of paramedical personnel and micro-

health centers, the poor distribution of large medical centers, the almost absence of continuing training and

missing means of maintaining and updating knowledge. These difficulties, however, find a response in the

development of targeted intelligent solutions such as expert systems capable of providing continuous assis-

tance to professionals and structures of health systems in African countries and transforming them into effec-

tive organizations. Unfortunately, the design of such systems still face many challenges, e.g., they have to

promote circulation and simplified access to current targeted information (e.g. current knowledge) that

healthcare professionals need in their daily lives and they have to take into account local needs to be better

adapted and useful for the heterogenous group of users. This article provides a short overview of the current

urgent needs of the health systems in SSA, motivates requirements on targeted digital support technology and

discusses first prototypical solutions to motivate possible research directions.

1 INTRODUCTION

Should African countries invest in research and de-

velopment of knowledge-based systems (KBS) to

sustainably strengthen their health systems? Faced

with insufficient human resources (e.g. few special-

ists), infrastructure problems (e.g. predominance of

micro-health centers) and organization (e.g. virtual

absence of continuing training), sub-Saharan African

countries are trying more actively to introduce tech-

nological solutions into medicine (MSHP, 2022;

TdH, 2019). The delicate nature of healthcare deci-

sions and tasks makes it a complex field that most of-

ten requires expert knowledge. Tolmie and Du Plessis

highlighted more than 25 years ago the apparent ad-

vantages of KBS for medicine in developing coun-

tries and in Africa, allowing us to legitimately ask the

question of knowing even today whether it is a ques-

tion of a luxury or a necessity to use these intelligent

systems in medicine in sub-Saharan Africa (SSA)

(Chikhata & Chivivi, 2017; Kendal & Creen, 2007;

Tolmie & Du Plessis, 1997). KBS, which have proven

themselves for years in solving complex problems by

providing coherent solutions to repetitive decisions,

processes and tasks, appears to be a highly recom-

https://orcid.org/0000-0002-1860-3526

https://orcid.org/0000-0003-4311-0624

mendable solution for strengthening the capacities of

African health systems (Ajanaku & Mutula, 2018;

Dey & Rautaray, 2014).

In order to illustrate our position on the possible

contribution and research challenges of KBS in re-

solving the difficulties that the health systems in SSA

are currently facing, we first discuss the current health

needs in SSA, give a short overview of possible con-

tributions of knowledge-based systems for healthcare

and finally briefly present results of a project that was

initiated in Burkina Faso with the outcome of proto-

typing an intelligent digital solution focused on spe-

cific challenges. The assistance system “@san” facil-

itates the sharing of targeted knowledge between par-

amedical personnel and with specialists in sub-Sa-

haran African areas, through the storage and retrieval

of collected expert knowledge.

2 AFRICAN HEALTH NEEDS

The weakness of sub-Saharan African (SSA) health

systems is exposed through the management of trop-

ical diseases (e.g. malaria, meningitis) throughout the

year and the responses provided to various epidemics

Ouedraogo, W. and Nürnberger, A.

Knowledge-Based Systems for Strengthening African Health Systems.

DOI: 10.5220/0012691900003690

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 26th International Conference on Enterprise Information Systems (ICEIS 2024) - Volume 1, pages 779-783

ISBN: 978-989-758-692-7; ISSN: 2184-4992

779

and/or pandemics (e.g. Lassa Fever, Ebola, Covid-

19). The growing health care needs constitute a set of

pressure factors on SSA health systems. This is due

to the absence of an adequate system to facilitate shar-

ing of health information, the lack of health personal

(> 85% of paramedical personnel), and the limited

number of public health infrastructures, of which

about 80% are poorly equipped micro-health centers

(MSHP, 2022). Among solutions to face these chal-

lenges, the most commonly explored in the last two

decades is the introduction of information and com-

munication technologies (ICT) in the health systems

(TdH, 2019; Watkins et al, 2018).

If medicine is recognized as a knowledge-inten-

sive field in which the adoption of new processes is

undertaken with great caution, it is imperative to think

of intelligent solutions to overcome the above-men-

tioned difficulties in SSA (Moahi & Bwalya, 2018).

Telemedicine – so far the technological solution that

has received the most attention thanks to its varied

successes – does not adequately solve these problems

due to the need for continuously available expert as-

sistance. Strengthening the autonomy of actions of

health workers through the use of KBS is a solution

to explore, given the advantages and means of indi-

rect use of expert knowledge proposed (e.g., storage

in documents). Guaranteeing competent health per-

sonnel – implying the need for better sharing of med-

ical knowledge – appears obligatory for the develop-

ment of the capacities of health structures in SSA and

therefore of health systems in more general (Ajanaku

& Mutula, 2018; TdH, 2019). It amounts to trans-

forming health systems into learning organizations

thanks to innovative solutions such as intelligent tar-

geted knowledge-based systems.

3 KNOWLEDGE-BASED

SYSTEMS IN HEALTHCARE

Knowledge-Based System supports decision-making

based on available knowledge and understanding the

context of the data processed. Although there are

other types of KBS (e.g. systems that are making use

of genetic algorithms, data mining, etc.), we focus in

this article on expert systems, which are often consid-

ered synonymous with KBS in the domain of health.

Capable of imitating human reasoning processes in

decision-making, an expert system is the embodiment

of a knowledge-based component of an expert skill

(e.g., medical expertise) in a computer. It aims to of-

fer intelligent advice or an intelligent decision accord-

ing to the request made to deal with the case presented

(Achmadi et al., 2018; Kendal & Creen, 2007). Ex-

pert systems are designed to compensate for the una-

vailability of specialists by providing access to their

knowledge and experience. Like the healthcare pro-

fessional who (re)uses up-to-date knowledge on pa-

tient healthcare, the knowledge contained in these

systems can be (re)used whenever necessary and eas-

ily updated (Dey & Rautaray, 2014; Nkuma-Udah et

al., 2018). Thus, an expert system facilitates the diag-

nosis of a disease by matching the patient's symptoms

with the existing rules in the knowledge base passing

through the inference engine.

Expert systems make major contributions such as

continuous learning following the evolution of sci-

ence, continuous improvement of healthcare perfor-

mance with a focus on the patient, evaluation of alter-

natives for elimination issues related to healthcare ad-

ministration, improving working relationships be-

tween healthcare professionals and effective manage-

ment of expert knowledge (Nkuma-Udah et al., 2018;

Khan MK, Munive-Hernandez, 2018; WHO, 2006).

Since the first expert systems, the best known of

which is MYCIN, used to diagnose and treat infec-

tious diseases caused by antibiotics, this field has con-

tinued to gain importance in medicine. Many pro-

grams have emerged to assist decision-making in spe-

cific areas: Computer-Assisted DIAGnostic (CA-

DIAG) which is a consultation system capable of as-

sisting in depth with the differential diagnosis and

possibly the therapeutic process in internal medicine

(Milan et al., 1997), clinical decision support system

(CDSS) which provides specific knowledge and in-

formation to facilitate health care, Aidoc Medical for

accident management stroke, pulmonary embolism,

cervical fracture, intracranial hemorrhage, intra-ab-

dominal free gas and accidental pulmonary embo-

lism, etc. (Cheichk et al., 2022; Kendal & Creen,

2007; Musen et al., 2021). Recent progress in AI has

also influenced KBS and allowed them to establish

themselves in crucial sectors such as medicine. These

solutions are, however, designed to meet the needs of

developed countries, not taking into account the real-

ities of SSA countries. The main issues are the differ-

ent diseases treated by these systems (e.g., absence of

a focus on tropical diseases), that the technologies

used almost always requires an internet connection,

the type of knowledge base used (e.g., requiring avail-

ability of experts for document linking) and the miss-

ing insertion of local African expertise (e.g., pharma-

copoeia, alternative medicine for certain diseases).

The initiatives of recent years (e.g., Iconic visual

ontology, KBS for African Traditional Herbal Medi-

cine) (Devine et al., 2022; Kouame, 2018) showing

the need to implement specific expert systems have

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

780

motivated the @san project, a mobile assistance ap-

plication which allows nurses and midwives to access

useful and up-to-date knowledge to carry out their

daily tasks in rural and remote areas of SSA. In addi-

tion to the components of an expert system, where ex-

pert knowledge is identified, acquired and stored in

the form of documents, @san also offers the possibil-

ity of active acquisition of knowledge between health

workers through direct exchanges in thematic forums.

4 USEFULNESS OF

KNOWLEDGE-BASED

SYSTEMS FOR AFRICAN

HEALTH SYSTEMS

The usefulness of Knowledge-Based Systems for

health is in SSA is especially based on their ability to

respond in a practical way and to tackle quite different

challenges. Most important, such systems should

make it possible to develop the skills of human re-

sources, to facilitate the close collaboration between

different levels of healthcare professionals, to guaran-

tee their autonomy in the management of patient care,

and to reduce the various costs while thinking about

sustainable development.

Difficulties of organizing continuing training (e.g.

cost, impassable roads, and insecurity) pose a major

challenge to health systems in sub-Saharan Africa.

An expert system should offer health professionals in

remote rural areas the possibility of accessing rele-

vant targeted information indirectly and in near real

time (e.g. via contextualized thematic forums). It also

should allow users to be attentive to advances in med-

icine (the state of knowledge) and therefore to contin-

ually improve the efficiency of the health system in

more general. Access to current specialized

knowledge (experts) has a positive impact on the

skills of health workers, particularly paramedical staff

(nurses and midwives) responsible for managing mi-

cro-health centers in rural areas without a doctor.

Fairly accurate decision-making helps reduce uncer-

tainties about diseases and provide safer treatments in

response to the urgent health needs of populations

(Cheikh et al., 2022; Nkuma-Udah et al., 2018).

Special exemption allows paramedical staff

(nurses and midwives) responsible for managing mi-

cro-health centers in rural areas to carry out tasks nor-

mally reserved for doctors (e.g., medical prescrip-

tions, patient transfer) (MSHP, 2022). Although their

training is much lower than that of physicians and

they are often engaged in these complex tasks (im-

portant decisions) many times a day, these healthcare

professionals rarely have training opportunities to ac-

quire new or updated knowledge (Abdulwadud et al.,

2019; TdH, 2019). By providing access to new

knowledge without constraints of time or location, an

expert system should tackle the difficulties of contin-

uous learning. In this way, healthcare professionals

will be able to reduce errors in their tasks and thus

avoid tragedies for many patients (e.g., delays in pa-

tient transfers) (Achmadi et al., 2018; Nkuma-Udah

et al., 2018). The systems should also provide modes

to work offline in order to provide continuous support

during internet outages or to avoid high online costs

in rural areas.

Sustainability may not be the hottest topic in

healthcare yet, but it is of paramount importance to

developing countries. For compliance with the pro-

posals of the UNDP Sustainable Development Goals

of 2015 and those of the Astana Declaration in 2018,

SSA health systems must evolve significantly in effi-

ciency and capacity by trying to obtain the maximum

positive results with minimal resources

(Chotchoungchatchai et al., 2020; Mehra & Sharma,

2021; Scherenberg, 2012; UNDP., 2015; WHO,

2018). Sustainability in health care encompasses a

harmonious combination of economic, health, and so-

cial dimensions. One possible way to achieve quick

improvements is through the use of targeted expert

systems. They could help to rationalize expenses by

reducing costs related to the organization of training,

travel to share targeted information, investments in

printing manuals, etc. Due to the digitalization of the

information processing process, environmental as-

pects must also be observed, which will have a strong

impact on the excessive use of mass paper for circu-

lars in cases of emergency, the sharing of research re-

sults, etc. Investing in digital infrastructure to facili-

tate data storage not only reduces the costs associated

with printing on paper and updating information but

also optimizes the wide portfolio of training and in-

formation sessions for all minimal changes or new

scientific advances. These aforementioned ad-

vantages are in line with the 2005-2009 strategic ori-

entations defined by the WHO Regional Office for

Africa regarding knowledge management (WHO,

2006).

4.1 A Prototypical Assistance System

In order to motivate the challenges of targeted expert

systems for the SSA health system, we present results

of a project that was initiated in Burkina Faso. The

main outcome was an assistance system called @san.

As a central operating element, @san includes an in-

formation retrieval system which models the logic of

Knowledge-Based Systems for Strengthening African Health Systems

781

reasoning of health professionals (mimicry of logical

thinking) in the search for targeted information to

guide their decision-making. When receiving a pa-

tient, the health worker proceeds by questioning and

observing clinical signs to make a diagnosis and cir-

cumscribe the disease using a combination of symp-

toms. It is important to note that from a general per-

spective in the medical field, that all symptoms have

the same importance (MSHP, 2022). Therefore, a

Boolean retrieval model was chosen for the IR mod-

ule implementation in @san. Each term in a docu-

ment collection implicitly defines the set of docu-

ments in which the term appears (exact match re-

trieval). The three basic operators of Boolean algebra

are conjunction (AND), disjunction (OR), and nega-

tion (NOT) (Manning et al., 2009; Pohl, 2012).

A documentary database with a focus on tropical

illnesses was chosen for the storage of targeted infor-

mation concerning each disease in @san. This has the

advantage of facilitating the indexing of useful infor-

mation, thus linking – for example – the symptoms to

the diseases for which there are clinical signs. Each

disease (e.g. malaria, fever, meningitis) is dedicated

to a file, which can undergo variations according to

the criterion of gender ("pregnant" woman, man) or

age (infant, child, adult, old person). During research,

a symptom whose presence disturbs the process and

does not allow us to find a disease is removed and can

be the subject of research with a view to creating new

knowledge. The interface was designed for easy in-

teractive use as a smart phone app. An online and of-

fline mode functions of @san is planned to make it

possible to regulate the constraints linked to the train-

ing offer such as costs, organization, and availability

of experts and the workload of health workers. The

platform developed was made available to end users

for a test phase.

Figure 1: @san information retrieval page on a mobile de-

vice.

Figure 1 shows 2 images of the user interface of

@san in a disease diagnosis activity. As a health pro-

fessional proceeds in his logic of reasoning, improba-

ble diseases are eliminated and a circumscription fol-

lows as the number of symptoms identified increases.

4.2 Limits

Although the prototype gives quite encouraging re-

sults and reinforces the idea that an expert system is a

solution to explore it is merely an initial step that can

be used to raise awareness and show potentials of

such systems for the health system. More research

projects are necessary, e.g., improvements must be

made to properly satisfy the recovery of targeted in-

formation (knowledge). As challenges to be over-

come, we can note among others for the moment the

rather incomplete information in the knowledge base

leading to non-satisfying fulfilment of information

needs, the format of presentation of the information

retrieved, and access to information in offline mode

to circumvent the difficulty of unstable internet con-

nections in SSA. Furthermore, the usability needs to

be improved and integrations of existing knowledge

resources needs to be done.

5 SUMMARY

Knowledge-based Systems (KBS) have proven them-

selves for decades and have established themselves

comfortably in the field of medicine. With recent pro-

gress in artificial intelligence, these technological so-

lutions have gained more visibility and are becoming

more and more essential. Despite its great advantages,

KBS seems to have been forgotten with regard to Af-

rica where we still find a large deficit in terms of re-

search initiatives. The solutions previously proposed

in developed countries cannot be used on a large scale

in Africa without undergoing certain adaptation be-

cause their development did not take into account lo-

cal realities and needs. Access to timely, useful and

targeted information on tropical diseases (e.g. ma-

laria, meningitis and Ebola) via a knowledge-based

system is what end users in SSA need. An expert sys-

tem must, however, also take into account the diffi-

culties of internet connection in Africa, the cost of its

creation and its large-scale deployment, the type of

knowledge base for the representation of knowledge

and its updating, the availability of knowledge ex-

perts, the technological habits and skills of end users.

@san's design observed these aspects in Burkina Faso

by enlisting the help of potential end users (nurses and

midwives) in its design and implementation. In this

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

782

sense, there is an urgent need for cooperation on re-

search projects and funding guidance to make health

systems in SSA effective learning organizations.

ACKNOWLEDGEMENTS

I greatly thank Neris Technologies SARL for the col-

laborations in the design and development of @san.

Thanks to Ms Golane / Ouedraogo Adam Ariane as

well as to all the other nurses and midwives who par-

ticipated freely and voluntarily in the process.

REFERENCES

Abdulwadud, O., Nyareza, S., Mthoniswa, P.J., Nonde, L.,

Mlasha, F., Lifuka, E. (2019). Knowledge of Cochrane,

evidence-based medicine and the Cochrane Library at

Defence Force Medical Services: a baseline cross-sec-

tional survey among military health personnel in Zambia.

Medical Journal of Zambia; 46(1), 1-9.

Achmadi, S., Mahmudi, A., Gita, A.N. (2018). Expert Sys-

tem Design to Diagnose of Virus Infection Disease in

Children with Certainty Factor Method. Journal of Sci-

ence and Applied Engineering; 1(2), 88-95.

Ajanaku, O.J., Mutula, S. (2018). The relationship between

knowledge management and nursing care performance.

South African Journal of Libraries and Information Sci-

ence; Vol. 84, No. 2, 39-51.

Cheikh, B.A., Gorincour, G., Nivet, H., May, J., Seux M.,

Calame P., et al. (2022). How artificial intelligence im-

proves radiological interpretation in suspected pulmo-

nary embolism Imaging. Informatics and Artificial Intel-

ligence Volume 32, pages 5831–5842.

Chikhata, F., Chivivi, O.O. (2017). The Significance of

Knowledge Management in Improving the Quality of

Health Care in the Public Sector in Zimbabwe. Interna-

tional Knowledge Sharing Platform. 7(5): p. 16-30.

Chotchoungchatchai, S., Marshall, I.A., Witthayapipopsakul,

W., Panichkriangkrai, W., Patcharanarumol, W.,

Tangcharoensathien, V. (2020). Primary health care and

sustainable development goals. PubMed Central; 98(11),

792–800.

Devine, S.N.O., Kolog, E.A., Atinga, R. Toward a

Knowledge-Based System for African Traditional

Herbal Medicine: A Design Science Research Approach.

Front. Artif. Intell 2022; 5:856705, 1-15.

Dey, M., Rautaray, S.S. (2014). Study and analysis of data

mining algorithms for healthcare decision support sys-

tem. International Journal of Computer Science and In-

formation Technologies; 5(1), 470-477.

Kendal, S.L., Creen, M. (2007). An introduction to

knowledge engineering. Springer..

Nkuma-Udah, I.K., Chukwudebe, A.G.,Ekwonwune, N.E..

(2018). Medical Diagnosis Expert System for Malaria

and Related Diseases for Developing Countries, E-

Health Telecommunication Systems and Networks,

Vol.7 No.2.

Khan, M.K., Hernandez, M. (2018). Assessing quality man-

agement system at a tertiary hospital in Oman using a hy-

brid knowledgebased system. International Journal of

Engineering Business Management; 10, 1-13. 13.

Kouame, A. (2018). Système de gestion de la médecine tra-

ditionnelle dans une plateforme web social et sémantique

: une approche basée sur une ontologie visuelle. Univer-

sité Gaston Berger, Saint Louis, Sénégal. [Ph.D. disser-

tation]

Manning, C., Raghavan, P., Schütze, H. (2009). Introduction

to Information Retrieval. Cambridge University Press; 1-

65, 151-175.

Mehra, R., Sharma, K.M. (2021). Measures of Sustainability

in Healthcare. Sustainability Analytics and Modeling;

Vol 1, 1-15

Milan, D., Petr H., Phuong H.N.. (1997). CADIAG-2 and

MYCIN-like systems. Artificial Intelligence in Medicine

Volume 9, Issue 3, Pages 241-259.

Moahi, H.K., Bwalya, J.K. (2018). Knowledge Sharing for

Healthcare and Medicine in Developing Countries: Op-

portunities, Issues, and Experiences. IGI Global.. [disser-

tation]

MSHP - Burkina Faso. Annuaire statistique 2021. (2021).

Available from: https://www.sante.gov.bf/fileadmin/an-

nuaire_2021_mshp.pdf (accessed 26 Aug 2023)

Musen, A.M., Middleton, B., Greenes, A.R. (2021). Clinical

Decision-Support Systems. Biomedical Informatics,

Spinger. pp 795–840

Pohl, S. (2012). Boolean and Ranked Information Retrieval

for Biomedical Systematic Reviewing. The University of

Melbourne, Victoria, Australia.. [Ph.D. dissertation]

Scherenberg, V. (2012). Nachhaltigkeit im Gesundheits-

wesen : Nur ein Modewort oder Anstoß zum Umdenken?

Dr. med. Mabuse 200; 9(1), 72-74.

TdH. IeDA: using technological innovations to improve pri-

mary healthcare. (2019). Available from:

https://www.tdh.org/de/projekte/ieda (accessed 25 Jan

2024)

Tolmie, C. J.,, Du Plessis, J.P. (1997). The use of knowledge-

based systems in medicine in developing countries: A

luxury or a necessity? Methods of information in medi-

cine; 36(02), 154-159. 12.

UNDP. (2015). Sustainable Development Goals by 2030.

Available from: https://www.undp.org/sustainable-devel

opmentgoals?gclid=Cj0KCQjw0IGnBhDUARIsAMwF

DLn29zR07vU3rrWfE2R74XHfphMpPvUY7ijN bx3lo

EO971fatibAKnEaAk7yEALw_wcB (accessed 25 Jan

2024)

Watkins, J., Goudge, A., Gòmez-Olivé, F.X., Griffiths, F.

(2018). Mobile phone use among patients and health

workers to enhance primary healthcare: A qualitative

study in rural South Africa. Soc Sci Med.; 198:139-147.

World Health Organization (WHO), Regional Office for Af-

rica. (2006). “Knowledge Management in de WHO Re-

gion Africa,” Addis Ababa, Ethiopia. Available from:

https://www.afro.who.int/sites/default/files/2017-06/AF

R-RC56-16%20Knowledge%20Management%20-%20

Final.pdf (accessed 25 Jan 2024)

World Health Organization (WHO). (2018). Available from:

https://apps.who.int/iris/bitstream/handle/10665/347879

/WHO-EURO-1978-3938-43697-61471-eng.pdf?seque

nce=1&isAllowed=y (accessed 25 Jan 2024)

Knowledge-Based Systems for Strengthening African Health Systems

783