Mbarara University of Science and Technology Institutional Repository (MUST-IR)
MUST-IR preserves research output from the MUST Community

Student Forms and Guides
Download the documents needed for self-archiving your research in MUST IR.
Self-Archiving Consent Form
Download, complete and sign the consent form.
Download Consent Form (PDF)Student Self-Archiving Guide
Follow the steps for submitting research to MUST IR.
Download Student Guide (PDF)Communities in MUST-IR
Select a community to browse its collections.
- This community will contain collections from the general Administrative staff
Recent Submissions
Item type: Item , One size does not fit all: Citizen science models reveal context-specific drivers of schistosome host snails in sub-Saharan Africa(Ecological Informatics, 2026-10-31) Valderrama-Bhraunxs, Noelia; Larissa, Bonifacio; Huyse, Tine; Greter, Helena; Mbaihondoum, Jacob; Batil, Annour Adoum; Loon, Emiel Van; Rompaey, Anton Van; Madinga, Joule; Twongyirwe, Ronald; Liesbet, JacobsHuman schistosomiasis transmission depends on compatible freshwater snail hosts, yet fine-scale information on their occurrence remains limited in many endemic settings where routine malacological surveillance is logistically challenging. Citizen science can expand monitoring coverage and frequency, but ecological inference re quires attention to observation processes, class imbalance and context dependence. We analysed visit-level Biomphalaria spp. presence/absence records from citizen-science networks in Uganda, the Democratic Republic of the Congo (DRC) and Chad. Using a harmonised workflow, we validated and filtered records, delineated monitoring sites, extracted environmental covariates, fitted country-specific binomial GLMMs and evaluated predictions using complementary cross-validation schemes. Across countries, conditional R2 substantially exceeded marginal R2, indicating that occurrence was strongly structured by persistent site-level heterogeneity not captured by the measured predictors. Retained predictors and effect sizes differed among countries, consistent with context-dependent host-environment relationships. A negative association between sampling experience and occurrence persisted after accounting for calendar time, indicating that participation history captured broader changes in the observation process than observer learning alone. Predictive performance depended strongly on the prediction target. Models showed substantial skill for repeated observations at previously monitored sites (Brier Skill Score: 0.47–0.54), but little or no skill at previously unsampled sites ( 0.09 to 0.02). In the low-prevalence DRC and Chad datasets, probability-scale evaluation and calibration were highly important because high discrimination could coexist with predictions concentrated near zero. Overall, citizen science provided useful context-specific predictions of Biomphalaria spp. occurrence at established monitoring sites, supporting repeated fine-scale surveillance while cautioning against extrapolation beyond monitored networks.Item type: Item , Psychoactive Drug Use and Academic Motivation among Secondary School Students in Buhweju District(Mbarara University of Science and Technology, 2024-09) Asasira, YobuThe study aimed to establish the association between socio-demographic characteristics and psychoactive drug use, as well as the differences in academic motivation among secondary school student psychoactive drug users and non-users in Buhweju District. The study objectives were to establish the association between socio-demographic characteristics and psychoactive drug use, establish the differences in intrinsic and extrinsic academic motivation across psychoactive drug use, and assess amotivation levels between psychoactive drug users and non-users. Data was collected using self-administered, semi structured questionnaires, and processed and analyzed using SPSS Version 27. Analysis was done using Chi-square, logistic regression, and independent t-tests. The study was conducted in eight randomly and purposively selected schools in Buhweju District, with a sample of 350 secondary school students who responded to the questionnaires. Findings revealed that Anglican students had higher odds of using psychoactive drugs compared to Catholic students (OR = 1.6, 95% CI [1.00, 2.4], p = 0.049). Additionally, students with no peer drug use observation had increased odds of engaging in psychoactive drug use (OR = 1.9, 95% CI [1.19, 2.9], p = 0.006). Intrinsic and extrinsic academic motivation showed no significant differences between drug users and non-users {t (343) = 0.8, p = 0.420}, {t (343) = -0.293, p = 0.770} respectively. However, psychoactive drug users exhibited higher levels of amotivation compared to non-psychoactive drug users {t (343) = 3.28, p = 0.001}. Conclusions included: Anglican student compared to Catholics, and those who did not observe peers using psychoactive drugs were more likely to use psychoactive drugs. Additionally, there was no significant difference in both intrinsic and extrinsic academic motivation between drug users and non-users, however higher amotivation were noticed among psychoactive drug users than non-users. Recommendations emphasized collaboration of schools with religious leaders, implementing peer education, and offering counseling and mentorship programs and personalized support for students showing high levels of amotivation.Item type: Item , The Interplay among Instructional Strategies, Learners' Geometry Attitudes, and Van Hiele Levels of Geometric Reasoning: Evidence from Secondary Schools in Rwenzori Sub-Region, Uganda.(Mbarara University of Science and Technology, 2025-08) Ndungo, IssaThis study addresses persistently low achievement in geometry among Ugandan secondary school leaners, alongside limited technology integration and underuse of learner-centered instructional strategies. Anchored in the pragmatic paradigm, Van Niele theory of geometric reasoning, and constructivist learning theory, it examined the interplay between instructional strategies, learner attitudes, and progression through the Van Miele levels in transformation geometry in the Rwenzori sub-region. A quasi-experimental pre-test—post-test design was implemented over six weeks with 483 Senior Three learners from six secondary schools (three rural, three urban), comparing Conventional Van Hiele Phased Instruction (CVHPI, n = 245) and Technology-Enhanced Van Hiele Phased Instruction using GeoGebra (TVHPI, n = 238). This is among the earliest mixed-methods studies in Uganda to compare these approaches, providing evidence-based insights for curriculum reforms. The first objective investigated changes in learners' attitudes toward geometry using a validated 22-item scale (a = 0.69—0.78). Wilcoxon Signed-Rank and Mann-Whitney U tests, alongside a Difference-in-Differences OLS model, revealed significant attitude gains in both groups (p < 0.001). However, TVEIPI yielded consistently greater improvements, especially in affective domains (interest and confidence) and behavioral aspects (engagement and metacognition) (13, 0.24, p < 0.001). The second objective assessed geometric reasoning through the Van Hiele Geometry Achievement Test (a = 0.781). Both groups progressed significantly through Van Hiele levels (p <0.001), but Generalized Ordered Logit results indicated that TVHPI learners were nearly twice as likely as CVHPI learners to achieve higher reasoning levels (OR = 1.94, p < 0.001). A significant interaction effect (15 = 0.503, p — 0.034) reinforced the superior cognitive gains associated with technology integration. The third objective explored whether attitudes predicted or mediated Van Hiele progression using a Difference-in-Differences—Generalized Ordered Logit model. Neither pre-instruction (ß = 0.0155, p = 0.911) nor post-instruction attitudes (B = 0.0889, p = 0.631) significantly predicted progression, and attitude change did not mediate the instructional effect (ß = —0.0997, p 0.565). Instructional strategy remained the dominant predictor, though a marginal interaction (ß = —0.6210, p = 0.075) suggested potential for stronger attitudinal influence in CVHPI contexts. The fourth objective qualitatively examined learner experiences through semi-structured interviews with 48 purposively selected learners. Thematic analysis revealed two overarching themes: Challenges and Support Needs, and Instructional Effectiveness and Learner Satisfaction. TV HPI learners commended GeoGebra for its dynamic visualization for boosting engagement, conceptual understanding, and problem-solving confidence. CVHPI learners valued accessible local visual aids but noted slower mastery and heavier reliance on repeated practice. Common challenges included time constraints, multi- step problem difficulty, and limited resources. Collectively, the findings affirm the Van Hiele theory's value in structuring geometry learning and underscore the constructivist advantage of technology-enhanced visualization for attitudinal and cognitive gains. The study culminated in the Ndungo Framework, the Geometry Pedagogical Improvement Cycle (GeoPlC), a blended, teacher-guided model combining the Strengths Of CVHPI and TVHPL It concludes that integrating tools like GeoGebra into Van Hiele Phased Instruction significantly enhances reasoning progression and attitudes. Recommendations include testing GeoPlC effectiveness, embedding technology-enhanced geometry pathways in the national curriculum, revising teacher training to emphasize competency-based, technology-supported pedagogy, and conducting further research on long-term effects and scalability across mathematical domainsItem type: Item , Interactive marketing practices and market performance of soft drink manufacturing enterprises in Kigali city-Rwanda.(Mbarara University of Science and Technology, 2023-03) Nimusima, PereezThe purpose of the study was to establish the relationship between interactive marketing practices (IMP) and the market performance of soft drink manufacturing enterprises (SDMEs) in Kigali- city. It examined the mediating role of competitive advantage (CA) and the moderating effect of competitive environment (CE) on the association between IMP and market performance (MP). A multi-theoretical approach was applied in the direction of social exchange theory, dynamic capability theory, network theory and five forces model. The study was guided by critical realism paradigm and was cross-sectional in nature. The quantitative data were obtained from a sample of 322 SDMEs operating in Kicukiro, Gasabo and Nyarugenge districts of Kigali City. The researcher first stratified SDMEs by legion (i.e. districts as shown above) and by soft drink clusters and then selected each firm using simple random sampling from each Stratum. CEOs were purposively selected for the qualitative study, whereas marketing; communication/public relations; finance; accounting and sales managers were selected for the quantitative study using simple random sampling. Questionnaire and interview guide were used to collect data and for data extraction and reduction, principle Component Analysis and varimax rotation were used. Zero-order Pearson correlation and hierarchical regression techniques were employed for data analysis. The study results revealed a positive significant relationship between IMP and MP. Detailed analysis on demographic characteristics of unit of analysis indicate that traditional communication channels take precedence over interactive communication channels among SDMEs in Kigali City-Rwanda. CA was confirmed to be a true mediator, while CE moderated the relationship between IMP and MP. This study recommends that SDMEs disseminate their marketing information using interactive communication channels; put more attention On selecting strategic alternatives that better match their offers with market needs and to fully attend to their key industry forces in their competition management strategies and crafting environment-informed strategies to better enhance their performance. Having taken a cross-sectional design and looked at only one 'P' i.e. promotion to predict MP and ignored Other marketing mix components which are also important in overall value creation processes, future studies could investigate same relationships across time by adopting a longitudinal design, and measure variables differently Or attend to other "PS" and measure how they influence MP in another study context.Item type: Item , Validation of an AI-Driven Sputum Analysis Platform for Accurate and Reliable Tuberculosis Detection in South-Western Uganda(Medical Devices: Evidence and Research, 2026-10-15) Tumwesigye, Habert; Atwine, Daniel; Atwine, Raymond; Birungi, Abraham; Birungi, Caroline; Munguciada, Esther; Nabimanya, Phionah; Akandwanaho, Joseph; Ochama, Benard; Mitala, YekosaniBackground: Ziehl–Neelsen (ZN) sputum smear microscopy remains an accessible method for detecting acid-fast bacilli (AFB), but its interpretation is dependent on specimen quality and trained microscopy personnel. Artificial intelligence (AI) may provide a microscopy reading aid that supports consistent identification of AFB-positive and AFB-negative smear images. Objective: To evaluate transfer-learning deep learning models for classification of AFB-positive and AFB-negative digitized ZN- stained sputum smear images generated in a Ugandan regional referral laboratory. Methods: A retrospective laboratory-based cross-sectional study used 512 archived digitized ZN-stained sputum smear images from Mbarara Regional Referral Hospital. The dataset contained 256 AFB-positive and 256 AFB-negative images and was partitioned by stratified sampling into training (n=358), validation (n=77), and an internal hold-out test set (n=77). ResNet-50 V2, DenseNet-121, and Swin Transformer-Tiny models were trained using ImageNet-pretrained weights in PyTorch. Performance was assessed using sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV), F1-score, and ROC-AUC. Results: On the internal test set, Swin Transformer-Tiny achieved 97.40% accuracy, 97.37% sensitivity, 97.44% specificity, 97.37% PPV, 97.44% NPV, 97.37% F1-score, and ROC-AUC 0.9939. DenseNet-121 achieved 88.31% accuracy, 84.21% sensitivity, 92.31% specificity, and ROC-AUC 0.9730. ResNet-50 V2 achieved 85.71% accuracy, 81.58% sensitivity, 89.74% specificity, and ROC-AUC 0.9555. Wilson 95% confidence intervals were calculated for the threshold-based metrics. Conclusion: The models demonstrated promising internal performance for classification of microscopy-defined AFB status. Swin Transformer-Tiny had the highest observed performance, but the small, balanced, single-centre internal test set and use of routine microscopy labels rather than an independent microbiological reference standard limit clinical interpretation. The platform should therefore be considered an AI-assisted microscopy reading aid rather than an autonomous tuberculosis diagnostic system. Prospective, multicentre evaluation using patient/specimen-level separation, independent microbiological reference standards, and multiple blinded readers is required before clinical deployment.Item type: Item , Numerical investigation of the impact of neural stem cells on brain tumors(Medicine, 2026) Bubu, Pius Erheyovwe; Eze, Val Hyginus Udoka; Ukagwu, Kelechi John; Sankarapandian, Vidya; Odoma, Saidi; Adie, Awafung Emmanuel; Mirembe, Nansamba FridahMalignant brain tumors, including glioblastoma and medulloblastoma, remain highly treatment-resistant. Neural stem cells (NSCs) are being investigated as tumor-tropic therapeutic vectors, but the temporal interaction between tumor and NSC populations requires quantitative characterization before model-based treatment hypotheses can be interpreted reliably. This study aimed to develop and numerically evaluate a reproducible 2-population model of tumor–NSC dynamics and to quantify the influence of NSC persistence and NSC-mediated tumor killing on simulated tumor burden. A normalized system of 2 coupled nonlinear ordinary differential equations was used: dT/dt = rT(1 − T) − βTN and dN/dt = αN(1 − N) − γTN. The baseline simulation used r = 0.20000 day−1, α = 0.05000 day−1, β = 0.35000 day−1, γ = 0.30000 day−1, T0 = 0.50000, and N0 = 0.40000. Explicit Euler integration was performed with Δt = 0.01 day over 20 days. A descriptive nonlinear regression fit, Fourier spectral analysis, 1-parameter sensitivity analyses, and joint α–β sensitivity analysis were applied to the simulated trajectory. In the baseline simulation, tumor density decreased from 0.5000 to a minimum of 0.4772 at day 2.59 (4.57% below the initial value) and subsequently increased to T20 = 0.8273, while NSC density declined from 0.4000 to N20 = 0.0247. The descriptive regression reproduced the simulated tumor trajectory with R2 = 0.9973 and root mean square error = 0.00590. Increasing α from 0.05 to 0.10 day−1 reduced T20 by 12.6% relative to baseline, and increasing β from 0.35 to 0.50 day−1 reduced T20 by 19.9%. Jointly setting α = 0.10 and β = 0.50 day−1 yielded T20 = 0.4427, 46.5% below the baseline terminal tumor density. The Fourier spectrum was dominated by the zero and low-frequency components, with no distinct oscillatory peak over the 20-day window. The model indicates that NSC persistence and NSC-mediated tumor killing materially influence terminal tumor burden, but the baseline parameterization produces only transient early suppression followed by tumor regrowth as the NSC population declines. The results are population-level, model based estimates rather than tumor-type-specific clinical efficacy measures. Tumour-specific calibration, spatial migration, immune interactions, and experimental validation are required before translational conclusions can be drawn.Item type: Item , Implementation status of infection prevention and control and healthcare-associated infection surveillance in Zambian public hospitals: a WHO IPCAF tool assessment(Antimicrobial Stewardship & Healthcare Epidemiology, 2026) Mudenda, Steward; Chizimu, Joseph Yamweka; Njung’e, Joseph; Mutila, Malambo; Daka, Victor; Hangoma, Jimmy; Mweso, Oliver; Mufwambi, Webrod; Hakayuwa, Clyde Moono; Mwaba, Martha; Mutwale, Ilunga; Champo, Morreah; Lubanga, Adriano Focus; Sakalimbwe, Loveness; Mugenyi, Nathan; Musonda, Linda; Muhimba, Zoran; Lukwesa, Chileshe; Chibi, Mpela; Gardner, Priscilla Nkonde; Mwadetsa, Sandra Diana; Kasanga, Maisa; Mupila, Zachariah; Kamboyi, Harvey; Fwoloshi, Sombo; Wesangula, Evelyn; Muma, John Bwalya; Sinyange, Nyambe; Chilengi, RomaBackground: Healthcare-associated infections (HAIs) remain a major global public health challenge, with a disproportionately higher burden in low- and middle-income countries (LMICs), including Zambia, where infection prevention and control (IPC) systems face resource and capacity constraints. Effective IPC programs are essential for reducing HAIs, limiting the emergence and spread of antimicrobial resistance (AMR), and improving patient safety. This study assessed the implementation status of IPC programs in Zambian hospitals. Materials and methods: A descriptive multicenter cross-sectional survey was conducted from June 1–30, 2025, in 18 public hospitals across Zambia. The World Health Organization Infection Prevention and Control Assessment Framework (IPCAF) tool was used to evaluate IPC implementation across eight core components. Descriptive statistics and Welch’s independent-samples t-test were used to compare IPC implementation levels across hospitals. Results: The overall mean IPC score was 558 out of 800, indicating an intermediate level of implementation. Half of the hospitals (50%, n = 9) achieved advanced implementation, 39% (n =7) intermediate, and 11% (n=2) basic levels. Tertiary hospitals performed better than secondary hospitals (mean score 622 vs 507). The highest-scoring components were IPC programs (86%) and guidelines (88%), while the lowest were workload, staffing, and bed occupancy (60%) and IPC education and training (63%). Despite established IPC committees and guidelines, gaps remained in staff training, HAI and AMR surveillance, and infrastructure. Conclusion: Most Zambian hospitals have achieved intermediate to advanced IPC implementation. However, sustained investment in workforce capacity, surveillance systems, and infrastructure is required to strengthen IPC practices and support national AMR containment efforts.Item type: Item , Global Mental Wealth: A Multidimensional Analysis of Social, Financial, and Emotional Well-Being in 92 Countries(Social Indicators Research, 2026-10-30) Ruggeri, Kai; Valenzuela-Catarí, Guillermo; Polemiti, Elli; Abderahaman, Galla; Adawi, Taha R.; Ayebare, SantsonGovernments and organizations increasingly endorse the importance of mental health in all forms of policy, though emphasis in measurement still remains on indicators such as GDP, depression, and other common disorders. In this study of 92 countries (n = 53,799), we assess the mental health of individuals around the world alongside measures of financial and social well-being. Findings include a trend of languishing mental health across many groups, particularly among younger adults and within higher-income countries. While positive trends are observed for education and income, a number of concerning trends also emerge, particularly the poor results for young adults and generally moderate results for employed individuals. In sum, we find an alarming but perhaps unsurprisingly low level of positive mental health in a substantial number of countries, but also identify specific economic, social, and other indica tors that policymakers could target in improving the well-being of the populations they serve.Item type: Item , Frontline recognition and specimen referral readiness for selected priority zoonoses among health workers in southwestern Uganda(Discover Public Health, 2026-11-13) Migisha, Richard; Mutesi, Charity; Asiimwe, Derrick; Byaruhanga, Aggrey; Muhwezi, Abraham; Kavuma, Prichard D.; Achiro, Sarah; Monje, Fred; Walekhwa, Abel W.; Ario, Alex R.Background Southwestern Uganda is a livestock-intensive region vulnerable to recurrent zoonotic disease out breaks. Because confirmation of several priority zoonoses depends on higher-level or reference laboratories, preparedness requires both frontline recognition and effective specimen referral. We assessed frontline recognition, specimen-referral readiness, and selected indicators of One Health coordination and cross-sectoral information exchange among health workers. Methods We conducted a cross-sectional survey of 345 health workers in Mbarara, Kazo, Kiruhura, and Ibanda districts. Participants were purposively recruited from 54 facilities based on roles in recognition, specimen handling, referral, or reporting of suspected zoonoses. Recognition was summarized using an 11-item recognition index covering causative agents and transmission/exposure pathways. The primary outcome was self-reported absence of specimen-referral linkage, defined as a respondent answering ‘No’ to the question, ‘Do you refer samples elsewhere for testing?’ We estimated prevalence ratios (PRs) using modified Poisson regression with facility-clustered, small-sample-corrected robust standard errors (CR2), with generalized estimating equations used to assess robustness. Results Almost all respondents had heard of the selected zoonoses (99.7%), and the median recognition index was 9/11 (interquartile range [IQR] 8–10). Overall, 48/345 (13.9%) respondents reported no specimen-referral linkage. This was more common among those working at peripheral facilities than at HCIVs/hospitals (20.1% vs. 6.8%; aPR = 2.74, 95% CI 1.24–6.04; p = 0.016). Among respondents reporting referral, 294/296 (99.3%) with feedback data reported receiving results. Among 280 with referral-to-result interval data, the median interval was 10 days (IQR 7–14). Common barriers were inadequate diagnostic resources (312/345; 90.4%), unreliable transport (259/345; 75.1%), and inadequate intersectoral collaboration (256/345; 74.2%). Conclusions Frontline recognition was generally strong, but self-reported specimen-referral linkage was less consistently reported at peripheral facilities. Strengthening specimen transport, referral tracking, information exchange, and subnational human-veterinary-public health coordination may improve connectivity with existing diagnostic networks.Item type: Item , Explainable hybrid deep learning for automated cervical cytology classification(2026-07-27) Wasswa, William; Ware, AndrewAccurate and explainable automated analysis of cervical cytology remains a major challenge for artificial intelligence (AI)-assisted cervical cancer screening, particularly in low-resource settings where access to expert cytotechnologists is limited. Although deep convolutional neural networks (CNNs) have substantially improved classification accuracy, their limited interpretability hinders clinical trust and routine adoption. Conversely, neuro-fuzzy systems provide transparent decision-making but lack the ability to learn complex cytomorphological representations directly from image data. Here, we present PapsAI XNet, a novel explainable hybrid deep learning framework that integrates attention-guided CNN feature extraction with an Adaptive Neuro-Fuzzy Inference System (ANFIS) to simultaneously achieve high diagnostic accuracy and interpretable decision-making. PapsAI XNet incorporates a Channel Attention Module and feature normalization strategy to enhance morphological feature discrimination and stabilize downstream fuzzy inference. A compact ANFIS rule base initialized through data-driven clustering enables transparent reasoning while avoiding the rule explosion commonly associated with high-dimensional neuro-fuzzy systems. The framework was evaluated on the publicly available Herlev dataset comprising seven cervical cytology classes and benchmarked against ResNet-18 and MobileNetV2 using accuracy, sensitivity, specificity, precision, F1 score, receiver operating characteristic (ROC) analysis, computational complexity, and explainability analyses. PapsAI XNet achieved an overall accuracy of 97.8%, sensitivity of 96.4%, specificity of 98.6%, precision of 97.1%, and F1-score of 96.7%, significantly outperforming both benchmark CNN architectures. The greatest improvements were observed for mild and moderate dysplasia, where gradual morphological transitions frequently challenge conventional deep learning models. Class wise area under the ROC curve exceeded 0.97 for all abnormal cytological categories. Furthermore, learned Gaussian membership functions, fuzzy-rule surfaces, and rule activation patterns provided clinically meaningful explanations linking deep morphological features to diagnostic outcomes while maintaining a lightweight computational footprint suitable for real-time deployment. These findings demonstrate that integrating attention-guided deep feature learning with interpretable neuro-fuzzy reasoning enables accurate, explainable, and computationally efficient cervical cytology classification. PapsAI XNet provides a practical framework for trustworthy AI-assisted cervical cancer screening and represents an important step toward clinically deployable explainable AI for digital cytopathology, particularly in resource-constrained healthcare systems.Item type: Item , School inspection and provision of quality Education in Secondary Schools in Uganda: a case of Kasese district(Mbarara University of Science and Technology, 2025-08) Mathembele, WilsonThis study investigated the influence of school inspection on provision of quality education in secondary schools in Uganda, A case of Kasese District. The study was guided by three objectives; to establish the effectiveness of school inspection, to examine the influence of school inspection on learners' achievement, and to investigate the influence of school inspection on school improvement, and two theories, Scientific Management and Human Relations. Convergent parallel mixed methods design was used in the study. Stratified proportionate random sampling for schools, stratified sampling for teachers, and purposive sampling procedures for Headteachers, DJS, and DEO, were used to select 326 participants who responded to a self-administered questionnaire and participated in a semi-structured interview schedule with a response rate of 86.7%. A pre-test and EFA were conducted to validate the instrument. Descriptive statistics such as means and standard deviation were used to summarize quantitative data, and inferential statistics such as Spearman correlation, regression and ANOVA were used to test the relationships and hypotheses. On a scale of 1-5, the mean score for Effectiveness of School Inspection was M=2.88, Mean for Learners Achievement based on UCE results 2018-2022 was Mean=2.06, and mean for school improvement based on the infrastructure, teaching and learning resources, and learning environment was Mean=2.58. Qualitative data was analysed using thematic content analysis. Findings revealed that; school inspection in secondary schools was moderately effective (Mean 2.76) with flaws in limited coverage of schools, time for inspection, and involvement Of the PTA and BOG members. The correlation revealed that school inspection moderately influences learners' achievement, with a moderate positive correlation (FO.325*•, p=.000) between school inspection and learners' achievement based On the UCE results (2018-2022), and therefore the null hypothesis which stated that school inspection has no statistically significant influence on learners' achievement was rejected. School inspections positively influence school improvement. with a strong moderate positive correlation between school inspection and the school improvement, thus, the null hypothesis which stated that school inspection has no statistically significant influence on school improvement was rejected, hence, school inspections moderately influence the quality of education. The study recommends adoption of comprehensive and holistic approach to school inspection to enhance its effectiveness. School inspection should be integrated with other key educational factors such as enhanced teachers' professional development programs, ensuring adequate resource allocation, encouraging parental and community involvement to maximize their impact on learners' achievement. School inspection processes be properly structured and conducted to maximize their role as drivers of school improvement through strengthening the methodologies to incorporate a combination of qualitative and quantitative assessment tools followed by adequate resource allocation. A similar study involving more participants, districts, and other educational factors could be conducted.Item type: Item , Does exposure to malaria prevention messages influence insecticide-treated net use through knowledge of malaria? Evidence from a causal mediation analysis among women in Uganda(Frontiers in Public Health, 2026) Mwebesa, Edson; Agbi, Delight Mawufemor; Natuhamya, Charles; Musinguzi, Benson; Izudi, Jonathan; Kerich, Gregory; Mwangi, AnnExposure to malaria prevention messages is associated with increased use of insecticide-treated nets, but the mechanisms remain unclear. This study assessed whether comprehensive malaria knowledge mediates the relationship between exposure to malaria prevention messages and insecticide-treated net (ITN) use among women of reproductive age in Uganda. We analyzed data from women aged 15–49 years from the 2018/2019 Uganda Malaria Indicator Survey (UMIS). Using a counterfactual causal mediation framework, the study decomposed the total e ect (TE) of exposure to malaria prevention messages into the natural direct e ect (NDE) and the natural indirect e ect (NIE), adjusting for socio-demographic confounders to identify whether malaria knowledge mediated the e ect of exposure to malaria prevention messages on ITN use. Among 8,868 participants (8,231 non-refugee women, 637 refugee women), 6,143 (69.3%) reported using an ITN on the night before the survey, 3,334 (37.6%) had been exposed to malaria prevention messages, and 5,427 (61.2%) had comprehensive knowledge of malaria. Exposure to malaria prevention messages significantly increased ITN use by 6.5 percentage points TE = 0.065, 95% confidence interval [CI]: 0.034, 0.094, p < 0.001. The NDE of malaria prevention message exposure on ITN use remained significant [0.064, 95% CI: 0.034, 0.092, p < 0.001], while the NIE of malaria prevention messages through comprehensive malaria knowledge was not statistically significant [0.001, 95% CI: −0.002, 0.005, p = 0.462], accounting for approximately 1.5% of the total e ect. Exposure to malaria prevention messages is associated with increased ITN use primarily through direct behavioral pathways rather than through knowledge acquisition. Communication strategies should therefore integrate motivational and behavioral reinforcement components alongside informational content to maximize public health impact.Item type: Item , Co-utilisation of recycled ceramic waste and sawdust in refractory bricks: thermal–mechanical trade-offs and resource recovery potential(International Journal of Sustainable Engineering, 2026) Ndyabijuka, Pathias; Atwijukire, Ernest; Mutanda, Hosea Eridadi; Najjobyo, Josephine; Gumoshabe, Joyce Kayanja; Mbabazi, JamesThe growing demand for low-cost and eco-friendly refractory materials has generated interest in locally available waste materials as alternatives to conventional refractories. This study investigates recycled ceramic waste and sawdust in refractory bricks, assessing their trade-off between mechanical strength and thermal insulation. Five formulations were prepared by maintaining clay at 35 wt% and varying the ceramic waste-to-sawdust ratio before firing at 1100 °C. The bricks were evaluated for water absorption, compressive strength, thermal conductivity and dimensional stability at 1100 °C. Water absorption increased from 7.8% to 22.4% as sawdust increased from 0 to 20 wt%, while compressive strength and thermal conductivity decreased from 18.6 to 6.7 MPa and 0.92 to 0.38 W/m·K, respectively. These trends are consistent with the proposed pore-forming effect of sawdust during firing. The formulation containing 55% ceramic waste, 35% clay and 10% sawdust provided the most balanced performance, with 12.8 MPa compressive strength, 0.63 W/m·K thermal conductivity, 13.9% water absorption and no visible deformation or cracking at 1100 °C. The findings demonstrate a thermal–mechanical performance envelope from redistributing a fixed 65% waste-derived fraction between ceramic waste and sawdust, indicating resource-recovery potential while recognising the need for further environmental and service-level assessment.Item type: Item , Clinical characteristics and outcomes among patients admitted with viral hemorrhagic fever at Mbarara Regional Referral Hospital, Uganda(IJID Regions, 2026) Muhindo, Rose; Ssewanyana, Ernest; Kayongo, Badru; Mutekanga, Andrew; Mugabi, Francis; Twesigye, Precious; Lwasa, Jeremiah; Wandera, Nelson; Olds, Peter K.; Ivers, Louise C.Objectives: Viral hemorrhagic fevers (VHFs), such as Rift Valley fever and Crimean-Congo hemorrhagic fever, pose serious public health concerns in endemic regions, such as Uganda, where outbreaks occur sporadically. Despite this, there are limited data on the clinical characteristics and outcomes of affected patients in rural settings. Methods: This is a retrospective descriptive study of 36 patients with laboratory-confirmed VHF who were treated at the isolation unit of Mbarara Regional Referral Hospital in southwestern Uganda during the outbreaks that occurred between 2018 and 2024. Data on demographics, clinical presentation including laboratory parameters, complications, and outcomes were extracted from the patients’ medical records and analyzed using descriptive statistics. The Welch’s t -test was used to describe the difference in length of hospital stay between the survivors and patients who died of VHF. Results: Of the 36 patients included, 88.9% (n = 32) were male, with a median age of 35 (interquartile range [IQR] 28-42.5) years. A total of 69.4% (n = 25) of the patients interacted directly with animals or animal products in their occupations. The most common presenting symptoms included bleeding tendencies (28 [77.8%]), fever (23 [63.9%]), and vomiting (19 [52.8%]). The median duration of symptoms was 5 days (IQR 4-7). Laboratory results revealed severe kidney impairment, with a median creatinine of 10.3 mg/dL (IQR 7.46-13), elevated aspartate aminotransferase of 1014 (IQR 173-2570) IU/l and alanine aminotransferase of 795.9 (IQR 233-1634) IU/l, prolonged international normalized ratio (median 1.65 [IQR 1.5-1.7]), hyperbilirubinemia (mean serum bilirubin 5.7 [SD 1.2]), and thrombocytopenia (median platelet count 43,000 [IQR 16,000-65,000]). The overall case fatality rate was 41.7% (n = 15), with a mean time from admission to death of 2.40 [SD 2.75] days. There was a statistically significant difference in the mean length of hospital stay between the survivors and those who died in the hospital (mean difference 15.07 days; 95% confidence interval 11.57-18.57; P < 0.001). Conclusion: In southwestern Uganda, patients with VHFs present with severe illness characterized by hemor- rhagic, hepatic, and renal complications. The high case fatality rate indicates the need for early detection and comprehensive supportive care interventions, such as hemodialysis.Item type: Item , Ceftriaxone-by-default in surgical prophylaxis across the East African community: a measurement problem(Frontiers in Antibiotics, 2026) Samar, Rashid Abdullahi; Mugenyi, Nathan; Ochola, Lucy; Mohamed, Jamal Abade; Sigat, Hussein Iman; Abdirizak, Ali Yusuf; Mikomangwa, Wigilya P.; Ndihokubwayo, Armstrong; Kunjok, David Majuch; Uwamungu, Schifra; Banzi, Nathan; Tuhamize, BarbraSurgical antibiotic prophylaxis across the East African Community is dominated by ceftriaxone, a WHO Access, Watch, and Reserve (AWaRe) Watch-group third generation cephalosporin not recommended for routine prophylaxis. The practice recurs across regional point prevalence surveys, yet consumption surveillance cannot distinguish prophylactic from therapeutic use. Drawing on antimicrobial-use surveys, surgical-prophylaxis studies, resistance surveillance, and medicine-availability reports, we argue that four interacting failure modes sustain it: substitution of first-line agents, prolongation of postoperative prophy laxis, unnecessary combination, and empirical spillover from ward prescribing. Resistance among surgical-site isolates is already high enough to threaten the value of prophylaxis itself. Prophylaxis is measured for duration, documented indication, and guideline concordance, and agent choice is captured in point prevalence surveys, but none of this reaches the national consumption totals that govern procurement, and no indicator records the supply constraint determining the agent or the resistance it selects. We propose four measures, one requiring routing rather than building and three genuinely absent, that would convert an ungoverned practice into a governable one.Item type: Item , Beyond evacuation: building epidemic- resilient global surgery partnerships in an era of recurrent Ebola outbreaks(BMJ global health, 2026) Oyania, Felix; Baluku, Moris; Tilahun, Eden BelayGlobal surgery has emerged as a major domain within global health, with partner ships between high- income countries (HICs) and low- income and middle- income countries (LMICs) supporting workforce development, research collaborations, educational exchanges and infrastructure investments. These efforts are central to addressing the estimated five billion people worldwide who lack access to safe, affordable and timely surgical care.1–3 However, the growth of global surgery has largely been built on architectures that assume mobility and stability. Recurrent Ebola outbreaks in Uganda and the DRC reveal how fragile these architectures can be when confronted with sustained crises.Item type: Item , Barriers and facilitators to utilisation of KMC among preterm babies on CPAP(BMC Health Services Research, 2026) Ediamu, Tom Didimus; Ayesiga, Innocent; Nakibuuka, Victoria; Nazziwa, Ritah A.; Idro, Richard; Nakacubo, Sheba Gitta; Nalwanga, Lillian Edith; Nampijja, Cissy; Nakyanzi, JanipherIntroduction Preterm births are a leading cause of death and illness, with rates rising in Uganda. CPAP helps with respiratory problems, and kangaroo mother care (KMC) stabilises temperature, encourages bonding, and supports health. Yet, KMC for preterm infants on CPAP is rarely used in countries like Uganda. This study explores challenges and enablers for KMC in these cases. Methods The study used a cross-sectional design with qualitative methods at St. Francis Hospital Nsambya. NICU healthcare workers were conveniently sampled and interviewed until saturation. Data was transcribed, coded, and analysed thematically in Atlas.ti Version 9. Results The study included 8 respondents with an average age of 34 years (range: 26–48 years). Barriers and facilitators were categorised into factors related to parents, babies, health workers, and health systems. A total of 22 themes and 20 subthemes emerged, covering both barriers and facilitators. Key barriers included issues such as mothers being sick, misinformation about KMC and CPAP, lack of maternal knowledge, negative attitudes towards KMC, and maternal fear. Additional barriers involved critically ill babies, poor attitudes and limited knowledge among health workers, absence of protocols, staffing shortages, equipment limitations, and timing of KMC. Facilitator themes highlighted paternal involvement and awareness efforts, such as education and counselling to improve maternal knowledge; fostering positive health worker attitudes; training; incentives; functional systems; access to information; early CPAP application; availability of KMC spaces in neonatal units; and integrating KMC into routine quality improvement projects. Conclusion and recommendation Multiple barriers categorised as health worker, baby, maternal, and health system factors still hinder the effective implementation of KMC among preterm babies on CPAP. Understanding these factors and developing category-specific interventions can enhance the utilisation of KMC, thereby improving outcomes for preterm babies.Item type: Item , Arboreal signals in the skeletal proportions of the Gorilla foot(Journal of Anatomy, 2026) Greiner, Elliot G.; Ruff, Christopher B.; Lukova, Andrea; Nkwasibwe, Aventino; Muhire, Thadee; Tibenda, Emmanuel; Massey, Jason S.; Junno, Juho- Antti; Syeda, Samar M.; Noheri, Jean Bosco; Nizeyimana, Fred; Tierra, Smiley Evans; Muvunyi, Richard; Babaasa, Dennis; Robbins, Martha M.; McFarlin, Shannon C.; Kivell, Tracy L.Postcranial skeletal proportions in living apes have been linked to patterns of arboreal behaviors within and between populations, offering a comparative baseline for interpreting arboreal behavior in fossil hominoids. We quantified metatarsal and proximal phalangeal lengths across all five pedal rays in Gorilla, comparing western lowland gorillas (Gorilla gorilla gorilla), low- and high- elevation Grauer's gorillas (Gorilla beringei graueri), and two ecologically distinct mountain gorilla (Gorilla beringei beringei) populations: those of Bwindi Impenetrable National Park, Uganda, and the Virunga Massif (Volcanoes National Park, Rwanda). The Bwindi sample, assembled from individuals recovered by the Mountain Gorilla Skeletal Project between 2017 and 2025, represents the first substantial pedal dataset for this population. After separately scaling element lengths by estimated body mass and hindlimb length, Bwindi gorillas' third and fourth metatarsal lengths fell consistently between those of the more arboreal western lowland and highly terrestrial Virunga gorillas, while proximal phalangeal and hallux lengths varied little across populations. Additionally, Virunga gorillas showed longer second metatarsals compared to other populations, a feature we link to their strongly terrestrial locomotor habits. In comparing relative manual and pedal element lengths from the same individuals, metacarpals were relatively longer than metatarsals across all populations in rays 2–5, with this pattern reversed in the hallux; among these, only the second ray showed a significant interpopulation difference, with western lowland gorillas exhibiting relatively longer metacarpals than all other groups. Though interpopulation differences in the lengths of pedal elements were generally weaker than those found in the manual elements of the same individuals, these findings suggest that habitat- associated variation in foot morphology can be detected between populations of gorilla subspecies, underscoring the value of fine- grained, population- level skeletal comparisons for refining locomotor inference in the hominoid and hominin fossil record.Item type: Item , A Rare Accessory Clavicular Muscle With Clavicular Origin and Insertion in the Right Posterior Cervical Triangle: A Cadaveric Case Report(Cureus, 2026) Tuhumwire, Caleb; Nakidde, Gladys; Kamoga, RonaldVariations in cervical and supraclavicular musculature have been documented, including the cleido occipital, levator claviculae, levator scapulae, sternocleidomastoid (SCM), and trapezius muscles. In this report, we describe a rare accessory clavicular muscle with both right clavicular origin and insertion, identified during dissection of an approximately 35-year-old male cadaver. The muscle was located in the right posterior cervical triangle, originating from the superior surface of the clavicle immediately medial to the trapezius insertion and inserting on its superoposterior surface, posterior to the lateral margin of the clavicular head of the sternocleidomastoid. The muscle had close anatomical relationships with the anterior and posterior branches of the supraclavicular nerve and the external jugular vein (EJV). This rare anatomical variant has potential clinical relevance and may influence local neurovascular relationships. Recognition of such variants may improve anatomical interpretation and reduce diagnostic errors.Item type: Item , A Critical Review of Deep Learning Methods for Trustworthy Crop Yield Prediction(Discover Artificial Intelligence, 2026) Taremwa, Danison; Ahishakiye, Emmanuel; Obbo, Aggrey; Kisozi, Paul Kategaya; Kaggwa, FredDeep learning is now central to crop-yield prediction (CYP), but reported gains are too often attributed to architectural novelty rather than to the data regime and evaluation design that actually determine performance. This structured critical review treats CYP as an evidence chain rather than a leaderboard problem. Unlike prior surveys that mainly catalogue architectures or application areas, this review contributes a practical evaluative framework that links architecture choice to data regime, target definition, and validation design, helping practitioners judge model suitability beyond reported accuracy alone. It asks whether each model family matches the structure of the available data, whether explanations are validated rather than merely visualized, and whether published benchmarks support claims of transferability and deployment readiness. Across structured genotype-environment-management tables, satellite time series, UAV phenotyping, and multimodal pipelines, the literature does not support a universally superior architecture. Feedforward networks remain competitive when inputs are already distilled into informative agronomic descriptors; convolutional and recurrent models are strongest when spatial or temporal structure is explicit; and hybrid, attention-based, and multimodal systems are most persuasive when they solve a genuine alignment problem across modalities. The field is more mature in model innovation than in benchmark realism. Reported gains are frequently confounded by heterogeneous targets, leakage-prone splits, weak baselines, limited calibration analysis, sparse external validation, and inconsistent reporting across reviews. Explainability studies face a similar problem: SHAP, LIME, saliency maps, Grad-CAM, and attention mechanisms are useful only when fidelity, stability, and agronomic plausibility are evaluated together. For an AI audience, the central lesson is methodological rather than architectural: trustworthy deployment will depend less on deeper networks alone than on stronger benchmark ecosystems, uncertainty-aware reporting, realistic transfer tests, and clearer multimodal design.