An investigation of ICT-based malaria intervention framework for rural communities

Abstract

Malaria remains a significant public health challenge in many sub-Saharan countries. The United Nations through member states launched Sustainable Development Goal 3.3, to end endemic malaria by 2030. Despite these concerted efforts, malaria continues to decimate people, especially in malaria-endemic countries, including Zimbabwe. Malaria predominantly affects poor rural and resource-constrained areas where it places a very high burden on communities. In addition, the outbreak of coronavirus disease 2019 (COVID-19) tenaciously challenged the progress made in the previous years to combat malaria in endemic areas by forcing the reallocation of resources devoted to fighting malaria to fight COVID-19. This caused a drastic change in prevention and control measures. Indoor residual spraying, longlasting insecticide-treated nets, and community behaviour change communication are among malaria control and prevention measures. Currently, hospitals and clinics use awareness campaigns, religious institutions, community meetings, community health workers, brochures, posters, billboards, newspapers, television, radio, and community dramas to convey malaria information. These traditional awareness strategies failed to achieve the anticipated results. More so, there is a non-existent technology-based framework for multi-sectoral linkages, collaboration, integration, and deployment of ICT-based malaria intervention in the Zimbabwean health system. This research addresses that gap by investigating a technologybased framework that supports the integration of feasible technologies to disseminate malaria information in rural communities. This study applied convergent parallel mixed methodology, quasi-experimental design, document analysis and design science research (DSR) methodology. The DSR was utilised to guide the development, refinement, and deployment of the proposed prototype. The document analysis was used to determine the most feasible technology. Also, previous malaria cases from the District of Health Information System (DHIS) were used for mapping hotspot areas and predicting malaria in hotspot wards using Quantum Geographic Information System (QGIS) and machine learning techniques, respectively. The quasi-experimental design was utilised to gather information in two phases (pre-test and post-test). The pre-test stage focused on gathering prototype user requirements before developing the artefact. The post-test phase concentrated on testing and assessing the adoption and acceptance of the proposed prototype. The acceptance and adoption of the proposed prototype was done through the modified unified theory of acceptance and use of technology (UTAUT) model. The study revealed that mobile phones, radio, television, and social media platforms were the most common ICTs used to disseminate information. Among ICTs, mobile phones are the most prominent mobile technology used for bidirectional communication and mobile money transaction in rural communities. However, the absence of policies on mobile health, technological and infrastructure barriers, poor power supply, digital illiteracy, inadequate funding, language barriers, and religious barriers were factors hindering the adoption and utilisation of ICTs in resource-constrained rural areas. The findings of this research also revealed that machine learning techniques play an imperative role in predicting malaria in hotspot wards. The study applied logistic regression (LR), decision trees (DT) and support vector machines (SVM) to predict malaria in hotspot wards. LR performed better, with an accuracy of 83%, a precision of 82%, and an F1-score of 90% using environmental data and malaria incidences. These machine learning models can assist policymakers in developing and deploying malaria early warning digital tools and optimising the distribution of resources in sporadic areas. The study modelled predictors for adopting mobile health interventions by healthcare professionals in Buhera rural community. The study utilised a modified UTAUT model and Smart-PLS to test several hypotheses. The study revealed that social influence, facilitating conditions, and effort expectancy facilitate the adoption of mobile phone-based interventions to create malaria awareness, reporting, and surveillance as well as sharing and receiving malaria data between satellite health centres. Among these predictors, facilitating conditions and effort expectancy influence health workers’ attitudes to using mobile phone-based malaria interventions. Furthermore, the study developed a mobile health framework for disseminating malaria information in resource-constrained rural communities. The proposed framework consists of surveillance activities, mobile health interventions and health facilities. This is an additional uniqueness of this study as it incorporates feasible digital technologies to disseminate health information in rural communities within Zimbabwe’s existing health system structure. This includes the Ministry of Health of Child Care (National Malaria Control Programme), Provincial Medical Office, District referral hospital, and satellite health centres. However, the study also revealed that the adoption of ICTs in rural health systems faces several impediments such as network connection barriers, inconsistent power supply, unavailability and inaccessibility of ICT infrastructure, lack of technical support and training, digital literacy, language barriers, absence of active e-health policies, insufficient funding, bureaucracy and religious barriers. There is a need to develop a mobile health framework and policy to guide the development and deployment of mobile health applications, improve ICT infrastructure and network coverage in rural communities, develop community networks to improve internet access and connectivity, promote public-private partnerships and develop robust strategies for sustainable funding of m-Health projects and applications deployed to improve access to care, especially in resource-constrained rural communities.