The Nelson Mandela African Institution of Science and Technology
Institutional Repository
Research Collection
Select to browse Research collections.
Recent Submissions
Machine Learning Techniques to Predict Fetal Nutritional Status
(The Korean Society of Medical Informatics, 2026-07-31) Mduma, Neema; Laizer, Hudson
Objectives
Malnutrition remains the leading cause of child mortality in Tanzania, with over 34% of children under 5 years of age affected by stunting and approximately 5% experiencing acute malnutrition. This study aimed to develop a machine learning model to predict fetal nutritional status using maternal and clinical data, thereby enabling early risk identification for health workers and parents and facilitating timely intervention. To enhance practical applicability, the model was deployed within a mobile application to provide accessible, real-time predictions that support prompt clinical and behavioral responses.
Methods
Using a dataset collected in Tanzania, the performance of multiple binary classification algorithms—logistic regression, multi-layer perceptron, random forest, extreme gradient boosting, and light gradient boosting machine (LightGBM)—was compared using the geometric mean and F-measure. These models were trained on clinical data from 11,703 pregnant women to predict fetal nutritional status based on maternal and clinical variables.
Results
The results indicated that the LightGBM algorithm achieved the best overall performance in predicting fetal nutritional status. The most influential predictors included maternal age, weight, fetal age, hemoglobin level, number of meals per day, medical history, and education level. Additionally, 93% of respondents reported satisfaction with the application’s predictive functionality, supporting its potential utility for early intervention in low-resource settings.
Conclusions
These findings highlight the potential of data-driven approaches to address public health challenges in maternal and child health. The proposed model may enable healthcare providers to make timely, informed decisions that improve maternal and fetal outcomes, ultimately contributing to the reduction of child malnutrition in Tanzania.
Epidemiology of bovine tuberculosis in Tanzania: Systematic review and meta-analysis
(PLOS Neglected Tropical Diseases, 2026-07-06) Anatory, Yohana; Mmbaga, Blandina; Lyimo, Beatus; Buza, Joram
Introduction
Bovine tuberculosis (bTB) is an important animal disease in the world, with both economic and zoonotic implications. The disease is still endemic in developing world, such as Tanzania due to lack of rigorous bTB control measures. Despite its significant economic and public health importance in Tanzania, its epidemiological information is still inconsistent.
Methods
To address this, we carried out a systematic review and meta-analysis of published articles on bTB epidemiology in Tanzania from 1993 to 2024. Eleven epidemiological themes were examined: prevalence, distribution, incidence, transmission, risk factors, reservoirs, diagnostics, outbreaks, surveillance, genetic diversity, and control measures. Data were synthesized using narrative analysis and random-effects meta-analysis where sufficient homogeneous data existed.
Results
Thirty-eight studies met inclusion criteria, with studies addressing bTB prevalence exhibiting high heterogeneity (I² > 90%) and the pooled prevalence of 1.0% (95% CI: 0.01-0.02) in both livestock and wildlife. Mbeya and Iringa regions experienced the highest prevalence, whereas Dodoma and Lake Zone had the lowest prevalence. Large herd size, communal grazing, older cattle, and proximity to wildlife were consistent risk factors. African buffaloes were the most important wildlife reservoirs, sharing strains with cattle at the livestock-wildlife interface. The Mycobacterium bovis spoligotype SB0133 was the most common lineage. The tuberculin test and culture were the most frequently used diagnostic approaches. Outbreak, incidence and surveillance studies were scarce and fragmented.
Conclusion
Although bTB is endemic in Tanzania, it has received little attention with geographical and thematic gaps still existing especially in incidence and surveillance. There is an urgent need for national strategy that integrates public health, livestock and wildlife systems. Also, molecular diagnostics, organized surveillance, outbreak investigations, and risk-based control measures should be among the priority areas.
Pesticide-specific case fatality and clinical outcomes among poisoning cases in Tanzania: a retrospective cross-sectional study
(Frontiers, 2026-07-23) Mwezi, Raphael; Bukalasa, Joseph; Otieno, Fredrick; Kombo, Rosevera; Ochola, Grivin; Muro, Florida; Honest, Anicet; Urasa, Sarah; Davis, Mark; Eddleston, Michael; Dekker, Marieke; Howlett, William; Vianney, John-Mary
Background: Pesticide poisoning remains a major public health challenge in low- and middle-income countries, particularly where highly hazardous pesticides (HHPs) remain accessible and toxicovigilance systems are limited. Although Tanzania has recently initiated regulatory actions targeting HHPs, evidence regarding pesticide-specific mortality patterns and clinical outcomes remains scarce. This study assessed clinical outcomes, pesticide-specific case fatality, and mortality risk among pesticide poisoning cases reported in selected Tanzanian hospitals.Methods: A retrospective cross-sectional study was conducted using hospital records from Regional Referral Hospitals and District Hospitals in Arusha, Kilimanjaro, Morogoro, Iringa, and Mwanza regions between January 2017 and November 2024. Cases were identified through hospital poisoning records, emergency department registers, admission records, and medical files. Information on demographic characteristics, exposure type, pesticide category, and clinical outcomes was extracted using structured electronic forms. Descriptive statistics summarize poisoning characteristics and outcomes. Associations between demographic factors and clinical outcomes were assessed using chi-square tests. Binary logistic regression was used to identify predictors of mortality, and adjusted odds ratios (aORs) with 95% confidence intervals (CIs) were reported.Results: A total of 635 pesticide poisoning cases were included in the analysis. Most patients were male (64.1%) and aged 16–35 years (61.3%). Intentional exposure accounted for 70.2% of cases, while unintentional exposure represented 20.3%. Most patients fully recovered (67.9%; 95% CI: 64.1–71.4), whereas 18.4% (95% CI: 15.5–21.6) died following pesticide poisoning. Paraquat poisoning demonstrated the highest case fatality (93.3%), followed by glyphosate (45.5%) and tefluthrin (33.3%). Compared with lower-risk pesticide exposures, paraquat poisoning was associated with markedly increased odds of death (aOR = 21.5; 95% CI: 9.4–48.2), while glyphosate poisoning was associated with elevated mortality risk (aOR = 5.2; 95% CI: 1.8–14.6). Pesticides with case fatality exceeding 10% contributed 87.3% of all recorded deaths.Conclusion: Mortality following pesticide poisoning in Tanzania was strongly concentrated among a small number of highly hazardous pesticides, particularly paraquat. These findings support continued restriction and removal of highly hazardous pesticides, strengthening of pesticide-specific surveillance systems, improved clinical toxicology capacity, and enhanced poisoning prevention strategies. Continued monitoring will be important to evaluate the public health impact of recent pesticide regulatory reforms in Tanzania.
Emerging rainfall risks to common bean in Sub-Saharan Africa: impacts and gender-inclusive adaptation
(Frontiers, 2026-07-16) Kalemera, Sylvia Monica; Venkataramana, Pavithravani; Mbega, Ernest; Assefa, Teshale; Ijumulana, Julian; Ochieng, Justus; Rubyogo, Jean Claude
Common bean (Phaseolus vulgaris L.) has increasingly become an important cash crop in SSA. Africa is the second-largest producer of common bean, contributing about 25% of global output and US$3.84 billion annually in Sub-Saharan Africa. However 50% of current bean-growing areas could become unsuitable by 2050 and yield is expected to reduce by 30–40% as temperatures rise, with disproportionate impacts on women, who account for more than 60% of producers largely due to persistent social and structural barriers. This systematic review of 79 studies (2007–2026) synthesizes evidence on climate risks and adaptation pathways in common bean systems. While breeding programs have released varieties tolerant to heat, drought, and poor soils, varieties bred for excessive-rainfall tolerance and explicit spatial adaptation remain absent from research agendas. Neighboring countries have already released and traded broadly adapted varieties based on regional market potential, but the lack climate-resilient cultivars constrains scalable climate resilience. Women, who dominate production and small-scale trade, are particularly vulnerable. We therefore call for urgent realignment of agricultural research with evolving climate and market realities and propose two actionable priorities: 1. Integrate excessive rainfall adaptation into breeding and climate models. (2) Shift breeding strategies toward a more regionally coordinated approach to enhance adaptation of varieties with regional market potential; 3. Strengthen seed systems and grain markets for market-relevant varieties through a gender-responsive regional approach.
Aesthetics over health: stakeholder perceptions of residential window design, indoor ventilation, disease risk, and energy use in urban Dar es Salaam, Tanzania: a mixed-methods study
(Frontiers, 2026-07-29) Mgaya, Zakia; Mawala, Jesca; Tarimo, Felista; Msuya, Ibrahim; Mlacha, Yeromin; Mpolya, Emmanuel; Lwetoijera, Dickson
Background: Residential window design is a primary determinant of natural ventilation in urban housing; however, the factors driving window design choices in rapidly urbanizing sub-Saharan African cities remain poorly understood. In Dar es Salaam Tanzania, sliding aluminum windows have become increasingly dominant in the residential housing market, yet evidence on how stakeholder perceptions of these designs relate to indoor ventilation adequacy, disease risk, and household energy use remains scarce.Method: This mixed-methods study was conducted in Dar es Salaam Tanzania, between July and October 2025. A structured questionnaire was administered to 611 participants (496 household respondents and 115 hardware store vendors), followed by two focus group discussions with housing stakeholders. Descriptive statistics (frequencies and proportions) summarized participant and household characteristics. Associations between window designs and self-reported diseases and use of artificial cooling systems were examined using chi-square tests and logistic regression, with statistical significance set at a p-value < 0.05. Focus group data were analyzed using thematic analysis, and a desktop review was done to assess the presence of multisectoral ventilation guidance.Results: Sliding aluminum windows dominated the market (57% of sales) and household installations (47%), driven primarily by aesthetic considerations (46%). Awareness of ventilation guidelines was low across all stakeholder groups (18%). Limited airflow was significantly associated with greater reliance on mechanical fans (χ2(1) = 5.72, p = 0.017). Large household size was the strongest predictor of both malaria (OR = 2.65, 95% CI: 1.61–4.39, p < 0.001) and respiratory infections (OR = 3.74, 95% CI: 1.17–12.0, p = 0.026). Use of gas as cooking fuel was protective against respiratory infections (OR = 0.35, 95% CI: 0.14–0.85, p = 0.021). The desktop review revealed no enforceable benchmarks linking window design to ventilation performance, disease prevention, or energy efficiency in Tanzanian building regulations.Conclusion: Market dynamics and structural constraints shape window design choices in urban Dar es Salaam more than health-oriented performance standards. Bringing public health ideas into housing rules, creating better ventilation guidelines for specific areas, and raising community awareness could improve indoor air quality and reduce reliance on energy-intensive solutions in rapidly growing cities.