Mbarara University of Science and Technology Institutional Repository (MUST-IR)

MUST-IR preserves research output from the MUST Community

Communities in MUST-IR

Select a community to browse its collections.

Now showing 1 - 10 of 10

Annual Research Dissemination Conference 3
Central Administration 124
This community will contain collections from the general Administrative staff
Faculty of Applied Sciences and Technology 58
Faculty of Business and Management Sciences 113
Faculty of Computing and Informatics 142
Faculty of Interdisciplinary Studies 212
Faculty of Medicine 2310
Faculty of Science 543
Institute of Tropical Forest Conservation 126
Maternal Newborn and Child Health Institute 18

Recent Submissions

Prediction of maize yield in Uganda using CNN-LSTM architecture on a multimodal climate and remote sensing dataset
(Springer, 2026-01-29) Taremwa,Danison; Ahishakiye,Emmanuel; Obbo,Aggrey; Kisozi , Paul Kategaya; Kaggwa , Fred
Maize is a staple crop in Uganda, underpinning both food security and rural livelihoods. Accurate forecasting of maize yields is therefore crucial for guiding agricultural planning, resource allocation, and policy design. Yet traditional statistical methods are often limited by low accuracy, poor scalability, and weak integration of diverse inputs, leaving them unable to capture complex, nonlinear, and spatiotemporal dynamics of crop growth. To overcome these constraints, we developed a hybrid convolutional neural network and long short-term memory (CNN-LSTM) model. This model integrates remotely sensed climatic variables and vegetation indices with biannual maize yield records from Uganda’s Zonal Agricultural Research and Development Institute (ZARDI) zones for the period 2018–2020. Due to the scarcity of high-quality yield data, we applied the Synthetic Minority Oversampling Technique for Regression (SMOGN) alongside feature selection to balance the dataset and improve predictive robustness. The CNN-LSTM model’s ability to select features and perform extensive hyperparameter tuning enabled it to outperform baseline models. It achieved a Mean Squared Error (MSE) of 0.107 tonnes2, a Mean Absolute Error (MAE) of 0.267 tonnes, a Root Mean Squared Error (RMSE) of 0.327 tonnes, and an R2 score of 0.783. A comparative analysis revealed that the CNN + Random Forest (RF) model achieved an MSE of 0.137 tonnes2, a MAE of 0.281 tonnes, an RMSE of 0.370 tonnes, and an R2 score of 0.722. These results outperformed the standalone CNN (MSE = 0.216, R2 = 0.562) and RF (MSE = 0.211, R2 = 0.573) models, underscoring the advantage of combining spatial–temporal learning for improved predictive accuracy. Residual analysis further confirmed the model's stability, showing minimal bias and close agreement between observed and predicted yields. These findings highlight the potential for integrating spatial–temporal deep learning and ensemble methods to deliver accurate crop yield forecasts in data-limited smallholder systems. By offering a scalable framework for evidence-based farm planning and food security policy, our study demonstrated that advanced machine learning can directly support sustainable development in sub-Saharan Africa. Future research will extend the framework to incorporate Transformer architectures, high-resolution satellite imagery, and explainable AI, further enhancing accuracy, interpretability, and decision-support capacity.
IoT Adoption for Agricultural Transformation in Developing Countries: Challenges and Opportunities - A case of Uganda
(VFAST Transactions on Software Engineering, 2026-03-31) Obbo,Aggrey; Ogot,Bernard
70% of Uganda’s workforce is employed in agriculture, which accounts for 26.2% of the country’s GDP and is a fundamental pillar of the country’s economy. Nevertheless, the sector is increasingly hampered by the unpredictability of the climate, traditional farming inefficiencies, and serious information gaps that jeopardize local food security. The potential of Internet of Things (IoT) technology to address these issues through data-driven resource management and precision farming is examined in this paper. Primary data was gathered from 116 smallholder farmers in Eastern Uganda using a descriptive study approach and the Unified Theory of Acceptance and Use of Technology (UTAUT) framework. Performance Expectancy (Mean = 4.19) and Behavioral Intentions (Mean = 4.11) emerged as the strongest drivers in the quantitative results, indicating a high readiness for adoption. Additionally, 95.7% of respondents reported severe crop losses due to drought, confirming the urgent need for technical intervention. Despite this incentive, Facilitating Conditions such as technological infrastructure and government support scored far lower (Mean = 3.17), indicating a crucial structural barrier. The study offers a strategic framework for policymakers to use the suggested IoT-enabled soil productivity monitoring model to strengthen smallholder resilience and revolutionize Uganda’s agricultural productivity, concluding that although farmer intent is high, sustainable adoption necessitates targeted investment in infrastructure and technical assistance.
Dyslipidemia and Associated Factors Among a High-Risk Population for Stroke: A Cross-Sectional Study
(Stroke Research and Treatment, 2026) Mark Kaddumukasa; Scovia Nalugo Mbalinda; Josephine Nambi Najjuma; Martin Kaddumukasa; Jane Nakibuuka; Doreen Birungi; Carla Conroy; Joy Yala; Levicatus Mugenyi; Christopher J. Burant; Shirley Moore; Elly T. Katabira; Martha Sajatovic
Aims: Dyslipidemia is a major modifiable risk factor for stroke; however, it is poorly understood among patients at risk of stroke in Uganda. This study is aimed at determining the prevalence of dyslipidemia and identifying associated factors among a Ugandan sample at risk for stroke. Methods: This was a hospital based cross-sectional study conducted across three Ugandan sites. The serum lipid levels were determined following the National Cholesterol Education Program guidelines. Data were analyzed with STATA employing un invariable and multivariable logistic regression. Statistical significance was set at p < 0.05. Results: We enrolled 247 study participants with a mean age (SD) of 55.4 (12.0) years. Majority of the participants were female, n = 168 (68%). About 81% (N = 199) had elevated serum lipid levels. Sixty-one (24.7%) had elevated levels of total serum cholesterol, whereas half of female participants had abnormally low levels of HDL cholesterol. About a third (N = 82 and 84, respectively) had elevated serum LDL and triglycerides. Nearly 40% (N = 98) were obese and 23.5% had a sedentary lifestyle (N = = 58). Only 20.2% (N = 50) were receiving lipid lowering drugs. Prior family stroke history and personal history of stroke had lower odds of 58% (AOR = 0.42, 95% CI: 0.20–0.88, p = 0.022); and 64% (AOR = 0.36, 95% CI: 0.17–0.76, p = 0.008), respectively, of having dyslipidemia. Conclusions: Approximately four in five Ugandans at risk of stroke have dyslipidemia. The majority also have low HDL-c levels. Implementation of systematic screening and provision of statin therapy among those at high risk for stroke is urgently needed to reduce stroke burden in Uganda.
The application of artificial intelligence in surveillance and control for antimicrobial resistance in hospital-acquired infections
(Discover Public Health, 2026) Innocent Ayesiga; Michael Oppong Yeboah; Jonathan Mawutor Gmanyami; Ahgu Ovye; Naya Gadzama Bulus; Lillian Edith Nalwanga; Gerald Mawanda; Tom Didimus Ediamu; Lenz Nwachinemere Okoro; Juliet Sylivia Nalugya; David Tenywa; Colleen M. Farrelly; Sheba G. Nakacubo
Antimicrobial resistance (AMR) continues to undermine the effectiveness of modern medicine, with hospital-acquired infections (HAIs) as major contributors to morbidity and mortality. Traditional surveillance systems for AMR in HAIs are often fragmented, delayed, and reactive, limiting their ability to inform timely interventions. Emerging evidence demonstrates that artificial intelligence (AI) can transform AMR surveillance and control. AI can enable predictive modelling, risk stratification, and outbreak forecasting. This narrative review describes various AI applications for monitoring and evaluating antimicrobial resistance (AMR) among HAIs in hospital settings. We begin by discussing how machine learning models can predict the emergence and spread of antibiotic-resistant pathogens. This is achieved through the analysis of various datasets, including microbiological results, electronic health records, and hospital workflows. Patient-level risk scoring systems are examined, demonstrating their ability to identify individuals at increased risk of multidrug-resistant infections. We describe AI-powered early-warning systems that provide outbreak alerts, enabling timely infection prevention and control measures. Hospital applications, including predictive resistance models validated in European hospitals, such as those in Lower Saxony, Germany, and early-warning dashboards tested in Asia and Africa, illustrate the potential impact of these approaches. However, these studies vary substantially in their designs, validation methods, and implementation contexts. To ensure sustainability, we propose a roadmap for integrating AI into AMR programs. We discuss ethical, legal, and regulatory frameworks. We also discuss strategies for capacity building, cost-effectiveness, and scalability, particularly in low- and middle-income countries. Our findings suggest that AI has the potential to strengthen infection control, enhance stewardship, and provide resilient defences within hospital environments.
Third trimester maternal anemia is associated with fetal cardiovascular compromise in Southwestern Uganda
(Discover Medicine, 2026) John Semuwemba; Silvia Lopez Miranda; Yarine Leodnell Fajardo Tornes
Background: Maternal anemia is still a global public health burden during pregnancy. It impairs fetal placental perfusion leading to fetal hypoxia, that disrupts vital physiological and biochemical pathways. This affects fetal growth and developmental programming, resultant into advance. perinatal outcomes and susceptibility to various non-communicable diseases later in life. This study set out to established how clinical Doppler and other physiological cardio, cerebral and peripheral vascular parameters vary at various grades of maternal anemia. Methods: A cross-sectional study was conducted among women aged 20–35 years, with singleton third trimester pregnancies attending Mbarara Regional Referral Hospital between March 2023 and August 2023. These were grouped into non-anemic with Hemoglobin concentration≥11 g/dL, while mild anemia (Hb:10–10.9 g/dL), moderate anemia (7–9.9 g/dL), and severe anemia Hb<7 g/dL. Resistance to flow and flow distribution were assessed by Doppler-derived indices and ratios of the fetal middle cerebral artery (MCA) and umbilical artery (UA). Physiological parameters like volumetric flow were calculated from measurements. We compared means using ANOVA and evaluated the relationships between flow parameters and the severity of anemia using Spearman’s correlation coefficient. Results: We enrolled 288 participants with a mean age of 27±4.3 years. The UA resistance indices among the non-anemic were significantly lower than among the anemic and significantly escalated from among the mild to the severely anemic, for instance, UA-PI increased from (0.88±0.18) among the mild, moderate (0.92±0.24) to (1.09±0.34) severe anemia, (p<0.001), while the mean flow velocity decreased as anemia severity increased (F [3,284]=3.44, p=0.017). The mean cerebral flow velocity demonstrated an increasing trend with the severity of anemia (p=0.042). Similarly a lower mean Cerebral placental Ratio was noted in the anemic group (1.87±0.49) compared to the non-anemic group (2.15±0.53), (p<0.001), while as maternal Hb concertation reduced, the CPR significantly decreased, (r=0.31, F[3,284]=12.64, p<0.001). As the maternal hemoglobin concentration reduced, the fetal heart rate increased (r= -0.06, p=0.308). Conclusion: Anemic women exhibited higher UA flow resistance, lower Cerebral placental ratio, suggestive of altered flow patterns with increasing severity of maternal anemia. These findings suggest that maternal anemia influences umbilical artery hemodynamics, potentially responsible for compromised fetal well-being. Understanding these associations could contribute to enhanced antenatal care strategies, emphasizing the importance of monitoring maternal anemia and its potential implications on fetal cardiovascular health.
Exploring the relationship between religiosity and coping strategies among individuals with mental illness in Uganda: a cross sectional study
(BMC Psychology, 2026) Joan Abaatyo; Godfrey Zari Rukundo; Samuel Maling; Emmanuel Alol
Background: Mental illnesses are leading causes of disability worldwide, and individuals often turn to religion as a source of support and resilience. In Uganda, where religious participation is deeply embedded in society, the influence of religiosity on coping strategies among individuals with mental illnesses remains underexplored. This study examined the relationship between religiosity and coping strategies among people with mental illnesses at Mbarara Regional Referral Hospital in southwestern Uganda. Methods: A cross-sectional study was conducted among 400 adult outpatients at Mbarara Regional Referral Hospital. Religiosity was assessed using the Duke University Religion Index (DUREL), and coping strategies were measured using the Brief Coping Orientation to Problems Experienced (Brief COPE) scale. Exploratory factor analysis was used to classify coping strategies into positive and negative. Multivariable logistic regression examined associations between religiosity dimensions and positive coping while adjusting for socio-demographic and clinical factors. Results: Negative coping strategies were more prevalent (61.2%) than positive strategies (38.8%). In adjusted analyses, higher organizational religious activity was associated with increased odds of positive coping (adjusted odds ratio [aOR]=1.32, 95% CI: 1.10 to 1.57), while higher non-organizational religious activity was associated with reduced odds of positive coping (aOR=0.69, 95% CI: 0.60 to 0.81). Family history of mental illness (aOR=1.67, 95% CI: 1.06 to 2.26) and a diagnosis of bipolar disorder (aOR=1.95, 95% CI: 1.01 to 3.76) were also associated with greater use of positive coping. Conclusion: Organizational religious activity is independently associated with adaptive coping among individuals with mental illness, whereas non-organizational religious activity is linked to reduced likelihood of positive coping. The high reliance on maladaptive coping strategies underscores the need for comprehensive, culturally sensitive mental health interventions that integrate psychosocial and faith-based support. Further longitudinal research is needed to clarify causal pathways
Association between non-high-density lipoprotein cholesterol and non-alcoholic fatty liver disease among people living with HIV on dolutegravir-based antiretroviral therapy: A cross-sectional study in Southwestern Uganda
(BMC Research Notes, 2026) Charles Nkubi Bagenda; Carol Nantongo; Elastus Ssemwanga; Michael Junior Mugisa; Jesca Wafwoyo Akoth; Daniel Nzaramba; Rogers Kalyetsi; Elizabeth A. John; Jazira Tumusiime; Blendar Ainebyoona; Shadrach Omara; Vicent Mwesigye; Sylvia Achieng Lumumba; Conrad Lubwama; Benson Musinguzi; Lucas Ampaire; Rose Nassali; Herbert Itabangi; Lawrence Obado Osuwat; Brian Ssenkumba; Ronald Ouma Omolo
Objective: Non-high-density lipoprotein cholesterol, an aggregate marker of atherogenic lipoproteins, has been implicated in metabolic dysfunction and may predict non-alcoholic fatty liver disease risk. Our study investigated the association between non-high-density lipoprotein cholesterol and non-alcoholic fatty liver disease among people living with Human Immunodeficiency Virus on dolutegravir-based antiretroviral therapy in southwestern Uganda. We conducted a secondary analysis of data obtained from a cross-sectional study of 377 adults who had been on dolutegravir-based antiretroviral therapy for ≥12 months at Ruhoko Health Centre IV, southwestern Uganda. Results: Of the 377 participants, 42(11.1%; 95CI: 8.3–14.8) had non-alcoholic fatty liver disease. We observed a significant association between high non-high-density lipoprotein cholesterol and non-alcoholic fatty liver disease; second tertile (aOR=3.08, 95% CI: 1.06–8.99, p=0.039) and third tertile (aOR=4.46, 95% CI: 1.25–15.88, p=0.021). At an optimal cut-off of ≥113.4 with a sensitivity of 69% and specificity of 48%, non-high-density lipoprotein cholesterol had a significant discriminative ability; AUC=0.654(95%CI:0.561–0.747) to distinguish participants with from those without non-alcoholic fatty liver disease. Therefore, High non-high-density lipoprotein cholesterol is a potential predictor for non-alcoholic fatty liver disease.
Caregiving experiences, challenges, and coping strategies among caregivers of children with epilepsy and/or cerebral palsy in southwestern Uganda
(Discover Public Health, 2026) Angella Nanziri; Joan Abaatyo; Moses Muwanguzi; Gerald Wakweyika; Mark Mohan Kaggwa; Godfrey Zari Rukundo
Background: Caregivers of children with epilepsy, cerebral palsy (CP), or co-occurring epilepsy and CP in low-resource settings face substantial psychological, social, and economic challenges. Despite high caregiver burden in sub-Saharan Africa, limited qualitative research has explored how caregivers perceive their experiences, their children’s care needs, and the contextual factors shaping coping. This study examined caregivers’ lived experiences and coping strategies to inform interventions that strengthen caregiver well-being and capacity. Methods: A descriptive qualitative study was conducted at the paediatric neurology clinic of Mbarara Regional Referral Hospital in southwestern Uganda. Twenty in-depth interviews were held with primary caregivers of children with epilepsy, CP, or both. Purposive sampling ensured variation in caregiving duration, child diagnosis, and coping patterns. Interviews were conducted in English or Runyankore, audio-recorded, transcribed, translated, and analyzed thematically using Braun and Clarke’s six-step approach. Data collection and analysis occurred iteratively until saturation. Results: Caregivers described caregiving as intensive and continuous, with children requiring substantial assistance in daily activities, consistent medication adherence, physiotherapy, and coordinated school support. Access to hospitals, schools, and community resources was inconsistent, often leaving caregivers to bridge gaps in care. Financial strain, stigma, social isolation, and disruption of income-generating activities were pervasive challenges that compounded emotional distress. Despite these pressures, caregivers demonstrated adaptive coping strategies, drawing on hope, acceptance, faith, and social connectedness to sustain their roles. They reported emotional exhaustion and limited family or professional support, with minimal engagement in formal mental health services. At the same time, caregivers developed structured routines, environmental adjustments, and personalized problem-solving strategies to maintain stability and ensure continuity of care. Conclusion: This study demonstrates that caregiving for children with chronic neurological conditions, including epilepsy and CP, in low-resource settings is highly demanding, marked by substantial medical, emotional, social, and financial challenges. Despite these burdens, caregivers show notable resilience through faith, optimism, and adaptive coping strategies. The findings underscore the urgent need for comprehensive support, including reliable medication access, psychosocial services, assistive devices, and stigma-reduction efforts, to strengthen caregiver well-being and improve care for affected children.
The likely possibility of predicting treatment outcomes of cervical lesions using serum FOXP3 and P16INK4A as shown by a cohort study in South Western Uganda
(Discover Oncology, 2026) Frank Ssedyabane; Deusdedit Tusubira; Nixon Niyonzima; Josephine Nambi Najjuma; Thomas C. Randall; Rogers Kajabwangu; Cesar M. Castro; Hakho Lee; Joseph Ngonzi
Introduction: There is a need to assess the potential of blood-based biomarkers to detect treatment outcomes of cervical lesions. We determined the association between serum P16ink4A and FOXP3 concentrations and treatment outcomes of cervical lesions at a clinic in Southwestern Uganda. Methods: In this prospective cohort study, participants with cytologically and/or histologically confirmed cervical intraepithelial neoplasia (CIN) (n=90) and cervical cancer (CC) (n=90) were monitored for 12 months. After consent, clinical and demographic data were recorded, blood was collected, and serum P16ink4A and FOXP3 were measured (quantitative ELISA) at baseline and 12 months’ post-treatment. With multinomial logistic regression, we determined the association between treatment outcomes and serum P16ink4A and FOXP3 concentrations in STATA 17 using P-values of <0.05 as statistically significant. Results: Of the 180 participants initially enrolled, 62 returned for the 12-month follow-up assessment. At this time point, 47 participants presented with cleared lesions, 6 with persistent lesions, and 9 with progressed lesions. All participants exhibiting disease progression (n=9) were CC cases, while 82.98% (39/47) of those with cleared lesions had LSIL. For raised (>0.0545 ng/ml), relative to reduced serum FOXP3 (≤0.0545 ng/ml), the risk of progression relative to clearance of invasive cervical cancer would increase by 27.82. Also raised (>0.946 ng/ml) relative to reduced (≤0.946 ng/ml) serum P16INK4A, the risk of persistence relative to clearance of low grade cervical lesions would increase by 5.16 times, given other variables remain constant. Conclusion: Though results are not statistically significant and imprecise, serum FOXP3 and P16ink4A concentrations are likely associated with persistence and progression of cervical lesions. Their measurement may benefit the prognostic monitoring of cervical lesions.
Overcoming structural barriersto tuberculosistreatment in Africa through digital technologies
(The Lancet Microbe, 2026) Tsegahun Manyazewal; Russell R Kempker; Zelalem Temesgen; Catherine Orrell; Esther Ngadaya; Angella Musiimenta; Lilian N Njagi; Priashni Subrayen; Yimtubezinash Woldeamanuel; Hussen Mohammed; Abebaw Fekadu; Melanie J Newport; Vincent C Marconi
Millions of people still die from tuberculosis in Africa. The continent is home to 17 of the 30 countries with the highest tuberculosis burdens and 25% of all new cases worldwide. According to the WHO Global Tuberculosis Report 2025,1 the continent exceeded the first End TB Strategy milestones and recorded one of the strongest global recoveries in tuberculosis control; the efforts of African governments and their international partners have been instrumental in this progress. Despite these advances, the continent is still struggling with monitoring tuberculosis treatment adherence and completion, which directly affects cure, relapse, drug resistance, transmission dynamics, and economic wellbeing. In this Comment, we argue that evaluation of tuberculosis treatment adherence strategies in Africa, particularly digital adherence technologies (DATs), need to extend beyond biomedical treatment outcomes to include patient-centred, economic, health system, and transmission-related effects
The cost of caring: understanding occupational stress among nurses at Mulago National Referral Hospital
(Frontiers in Public Health, 2026) Edward Tumusiime; Evas Atuhaire; Eliot Namanya; Abraham Muhwezi; Steward Mudenda; Ronald Ouma Omolo; Nathan Mugenyi
Background: Occupational stress is prevalent among nurses, particularly in resource-constrained settings, where high patient volumes, limited institutional support, and societal pressures exacerbate strain. In Uganda, nurses at tertiary referral hospitals face compounded challenges, yet empirical evidence on multilevel determinants of stress is limited. This study applied the Job Demands-Resources-Support (JDRS) framework to examine health system, patient-related, and community-level factors influencing occupational stress among nurses at Mulago National Referral Hospital (MNRH). Methods: A facility-based cross-sectional study was conducted from April to June 2024, enrolling 231 nurses through availability sampling. Occupational stress was measured using a self-reported binary indicator, and explanatory variables captured job demands, control, and support per the Job Demand-Control-Support (JDCS) framework. Descriptive statistics summarized variables, while bivariate and domain-specific multivariable logistic regression analyses were used to examine adjusted associations with occupational stress with in health system, patient-related, and community-level domains. Results: Occupational stress prevalence was 64.5%. In domain-specific multivariable analyses, caregiver-related challenges were associated with higher odds of stress (aOR = 1.3, 95% CI: 1.02–1.57), while participation in continuous professional development-based programs (aOR = 0.3, 95% CI: 0.12–0.98) and lack of hospital accommodation (aOR = 0.2, 95% CI: 0.04–0.80) were associated with lower odds. Counterintuitive inverse associations were also observed for inadequate staffing and exposure to aggressive patient behavior, likely reflecting adaptation, selection effects or residual confounding. Gender-related stressors were similarly inversely associated with stress (aOR = 0.7, 95% CI: 0.52–0.88). Conclusion: Nurses at MNRH experience high occupational stress driven by complex interactions of job demands, limited resources, and societal pressures. Effective stress mitigation requires integrated interventions addressing structural capacity, professional development, and social valuation of nursing. Future research should employ longitudinal and mixed-methods designs to disentangle the dynamics of occupational stress in high-demand, resource-limited environments
Epidemiological profile of bladder outlet obstruction among adult males at a tertiary hospital in northern Tanzania
(African Urology, 2026) D Mucunguzi; JS Mbwambo; ZI Cheyo; F Bright; OJ Mbwambo; BN Ngowi
Purpose: Bladder outlet obstruction (BOO) in men frequently causes emergency visits and negatively affects their quality of life (QoL). Its prevalence increases with longer life expectancy, yet limited research exists on its frequency and aetiological patterns in our local context, highlighting a knowledge gap. Materials and methods: A three-year retrospective, hospital-based, cross-sectional study was conducted among 1 879 adult male patients who attended the urology clinic at Kilimanjaro Christian Medical Centre (KCMC) between January 2021 and December 2023. The extracted files were assessed to confirm the BOO diagnosis using clinical history, physical examination, and imaging investigations. Aetiological BOO patterns were recorded. All data were cleaned and analysed using IBM SPSS Statistics version 25. Ethical clearance was obtained from the College Research Ethics and Review Committee of the Kilimanjaro Christian Medical University College. Results: Overall, 3 464 files were extracted and reviewed between January 2021 and December 2023, of which 1 879 males (54%) had urinary BOO. Patients had a median age of 72 years (interquartile range [IQR] 64–80), and most (73.5%) presented with lower urinary tract symptoms (LUTS). The most common BOO aetiology was benign prostatic hyperplasia (BPH) (57%), followed by urethral stricture disease (USD) (20%). USD was the most common aetiology among adult males aged < 60 years. Conclusion: More than half of adult males attending our urology clinic may be experiencing BOO. The most common BOO aetiology in this population is BPH in men aged ≥ 60 years, and USD in adult men aged < 60 years. Adult men aged < 60 years should be screened for USD.
Validating the 2-minute walk test MCID for subacute stroke patients: A Pakistani multicenter cohort analysis
(PLoS One, 2026) Muslim Khan; Hafiz Shehzad Muzami; Ahmad M Osailan; Ayman Abdullah Alhammad; Samiullah Khan; Mazhar Mushtaq; Engy BadrEldin Saleh Moustafa; Theeb Naif S. Alsalem; Edward Muteesasir
Background: The 2-Minute Walk Test (2MWT) is a simple and reliable test used by clinicians to evaluate mobility gains in stroke populations. Objective: To determine the minimally clinically important difference (MCID) for the 2-Minute Walk Test (2MWT) in subacute stroke populations. Design: This was a prospective cohort study. Setting: The study was carried out across seven rehabilitation centers in Pakistan. Participants: 150 adults (<180days post-stroke), with a confirmed ischemic or hemorrhagic stroke, able to walk at least 10 meters, and Mini-Mental State Examination score ≥24. Interventions: Participants underwent 6–8 weeks of standard rehabilitation, including physical therapy (gait and balance training) and, in some centers, robotic gait training. Interventions varied by center but followed standardized gait and balance goals. Main outcome measures: The 2MWT was the primary outcome measure whereas, the Activities-specific Balance Confidence gait subscale (ABC_Gait), Global Perceived Effect (GPE) scale, and Modified Barthel Index (MBI) were the secondary outcome measures. The MCID was estimated using an anchor-based approach (ABC_Gait ≥10% improvement), validated by distribution-based methods. Results: The 2MWT MCID was 33 meters (95% CI: 30–36 meters), with 87% sensitivity and 82% specificity. A strong correlation was observed between 2MWT and ABC_Gait change scores (r = 0.68, p < 0.001). The mean 2MWT distance improved from 62.5±38.4 meters to 98.7±42.1 meters (p < 0.001). Subgroup analyses showed that MCID was consistent across age groups, stroke types, and intervention modalities. Conclusion: The 33 meters MCID for the 2MWT is a reliable and patient-centered benchmark for assessing mobility gains in subacute stroke populations. Validation in chronic stroke populations and exploration of emerging interventions like virtual reality (VR) is needed to extend the 2MWT MCID’s applicability
Mental health disorders, and associated factors among children aged 6–17 years living in Mahama refugee camp in Rwanda
(PLOS Mental Health, 2026) Allan Higiro; Dan Lutasingwa; Mickel-Ange Karamage; Everest Turatsinze; Ritah Mukashyaka; Deborah Kansiime; Odile Habimana; Angelique Uwigiciro; Elisabeth Iriza; Norbert Tuyishimire; Fred Mulisa; Egide Niyotwagira; Moses Ochora; Daniel Gatei Waweru; Aflodis Kagaba; Alain Favina
Children living in refugee camps are negatively affected by adverse mental health outcomes due to traumatic experiences and living conditions such as food insecurity. However, the mental health disorders of children living in refugee camps of Rwanda have not yet been empirically reported. This study, therefore, assessed the prevalence and associated factors of depression, post-traumatic stress disorder (PTSD), and suicidal ideation among children living in a refugee setting, with particular focus on the association with food insecurity. This was a cross-sectional study conducted among 500 children aged 6–17 years living in Mahama Refugee Camp, Rwanda. The MINI-KID tool was used to assess for mental health disorders and the Household Food Insecurity Access Scale (HFIAS) for food insecurity. Descriptive, bivariate, and multivariate logistic regression analyses were performed using STATA-17. The prevalence of depression, PTSD, and suicidal ideation were 12.6%, 3.0% and 2.2% of participants, respectively. Severe food insecurity was prevalent at 69.4% (347/500). Food insecurity was only associated with depression (aOR=1.12, 95%CI: 1.07-1.17, p<0.001). Additional factors significantly associated to depression were age (aOR=1.26, 95%CI: 1.13-1.42, p<0.001), and coming from DRC (aOR=2.83, 95%CI: 1.38-5.79, p=0.005). PTSD was associated with coming from Democratic Republic of Congo (DRC) (aOR=4.83, 95%CI: 1.55-15.04, p=0.007), while suicidal ideation was also associated with coming from DRC (aOR=6.91, 95%CI: 1.79-26.69, p=0.003) and having a disability (aOR=9.44, 95%CI: 1.83-48.77, p=0.007). The prevalence of mental health disorders among refugee children in Mahama Camp are high with depression being significantly associated with food insecurity. Integrative interventions addressing both mental health and food insecurity among children like encouraging modern agricultural practices in humanitarian settings are crucial. Supporting children living with disability according to their individual needs can further enhance their mental health.
Knowledge, attitudes and practices towards emergency contraceptive pill use among women of reproductive age in Lira District, Northern Uganda: A cross-sectional study
(Plos one, 2026) Josephine Vanessa Nakalema; Marvin Musinguzi; Deo Benyumiza; Eustes Kigongo; Edward Kumakech; Rashida Namuwaya; Joshua Oryem Opido; Nicholas Damulira; Raymond Tumwesigye; Amir Kabunga; Marc Sam Opollo
Background: Emergency contraceptives (ECs) are used to prevent unintended pregnancy after unprotected sexual intercourse. In Uganda, unsafe abortion remains a major contributor to maternal mortality, particularly in resource-limited settings such as Northern Uganda. Greater use of emergency contraceptive pills could help reduce unintended pregnancies and their related consequences. However, information on women’s knowledge, attitudes, and practices regarding emergency contraceptives remains limited. This study therefore assessed the knowledge, attitudes, and practices of women of reproductive age towards emergency contraceptives in Lira District. Method: This was a community-based cross-sectional study conducted between June and July 2023. A total sample size of 600 was estimated; however, 585 respondents were ultimately selected using multi-stage cluster sampling. Data were analyzed using SPSS version 26 at the univariate, bivariate, and multivariate levels. Results: The study had a response rate of 97.5% (585). Emergency contraceptive pill usage was at 30.6%. The mean age of participants was 28.01±7.6 years. Majority of the respondents 374 (63.93%) had good knowledge about emergency contraceptive pills Most of the respondents 386 (65.98%) had negative attitudes towards emergency contraceptive pills. The factors associated with use of emergency contraceptive pills include: having poor knowledge about emergency contraception (AOR=0.11, 95% CI 0.07–0.21, p<0.001), having positive attitudes towards emergency contraceptive pills (AOR=3.89, 95% CI 2.59–5.84, p<0.001), and being unmarried (AOR=2.85, 95% CI 1.83–4.43, p<0.001). Conclusion: Emergency contraceptive pill utilization among women of reproductive age in Lira District was modest, with about three in ten women reporting use in the previous month. Utilization was significantly influenced by knowledge, attitudes, and marital status. These findings highlight the need for interventions that strengthen knowledge, address negative attitudes, and support appropriate use of emergency contraception.
Effect of partner support on antenatal care visits among married adolescents in rural northwestern Uganda: A quasi-experimental study
(PLoS One, 2026) Saidi Appeli; Christine Chandia; Jonathan Izudi
Background: Adolescent pregnancy is associated with several challenges, and partner support is crucial. However, the causal effect of partner support on the use of maternal health services among pregnant adolescents has not been rigorously examined. We assessed the causal effect of partner support on the frequency of antenatal care (ANC) visits among married adolescents in rural Uganda. Methods: We conducted a quasi-experimental study using observational data from a cross-sectional study that involved married adolescents aged 10–19 years in rural north-western Uganda. The primary exposure was partner support, measured as a binary variable. Adolescents were considered to have received support if their spouse encouraged ANC attendance, accompanied them to visits, or provided financial or emotional support during ANC; otherwise, they were classified as having not received support. The primary outcome was the number of ANC visits, while the secondary outcome was attending four or more ANC visits. Propensity score weighting was used to ensure covariate comparability between the partner support groups (yes vs. no). Poisson regression was used to estimate the causal effect of partner support on the number of antenatal care (ANC) visits (primary outcome), while the modified Poisson regression was used to estimate the causal effect on attending four or more visits (secondary outcome). Results: Of 281 participants, 205 (72.9%) received partner support. Among participants with partner support compared to those without partner support, the frequency of ANC visits (Risk Ratio 1.15, 95% CI: 1.00–1.32) and four or more ANC visits (Risk Ratio 1.25, 95% CI: 1.01–1.52) improved. Conclusion: The study showed that partner support was associated with improvements in ANC visit frequency and attendance of four or more ANC visits, among married adolescents in northwestern Uganda. Interventions aimed at improving ANC utilization should consider engaging and educating partners, as their support positively influences ANC utilization.
Person-centered strategies for integrating TB treatment into community pharmacies for people with TB/HIV in Uganda: A human-centered design methodology study protocol
(Plos one, 2026) Jonathan Izudi; Adithya Cattamanchi; Christine Sekaggya-Wiltshire; Rachel King; Noah Kiwanuka; Amanda Sammann
Background: Community pharmacies (private retail drug shops or pharmacies) have successfully delivered antiretroviral therapy (ART) to people with human immunodeficiency virus (HIV) and could support integrated tuberculosis (TB) treatment, but the implementation strategies are unclear. To inform a planned pilot randomized trial, we aim to develop person-centered strategies for integrating TB treatment into community pharmacies targeting people with TB/HIV using a Human-Centered Design (HCD) methodology. Here, we describe the study protocol. Methods: We will employ a three-phased HCD methodology comprising inspiration, ideation, and implementation across six primary health facilities in Kampala, Uganda. Eligible participants will include people with TB/HIV, focal persons for TB and HIV, Ministry of Health officials, and community pharmacy healthcare providers. The inspiration phase will build the themes on barriers and facilitators to integrating TB treatment into community pharmacies from a qualitative study, complemented by participant observations at selected 2–3 community pharmacies to understand the care pathway of people with TB/HIV (journey mapping), including sketching the ideal pharmacy-based TB treatment pick-up. The ideation phase will use design workshops to consolidate identified themes, generate insight statements, including translating them into design opportunities, and conclude with forming low and high-fidelity prototypes. The implementation phase will comprise two rounds of prototype testing, low and high fidelity, with 12–16 participants per round, including people with TB/HIV, focal persons, Ministry of Health officials, and pharmacy health workers. Feedback on usability, desirability, feasibility, and viability will guide iterative refinement, with high-scoring prototypes in system usability surveys prioritized for trial. Discussion: Through iterative user engagement, we will adapt contextually relevant strategies that will leverage key facilitators and address barriers to TB treatment integration. Strategies demonstrating perceived usefulness, user-friendliness, high acceptability, person-centeredness, and contextual relevance will be adapted and piloted in a planned randomized trial aiming to determine feasibility, acceptability, and fidelity, including preliminary effectiveness.
Equiflow: An open-source software package for evaluating changes in cohort composition
(PLOS Digital Health, 2026) Jacob Gould Ellen; Chrystinne Fernandes; Martin Viola; Keagan Yap; Arinda Jordan; Mutesi Flavia Kirabo; João Mato; Pedro Moreira; Leo Anthony Celi
Clinical research studies routinely apply exclusion criteria and data preprocessing steps that can substantially alter dataset composition, potentially introducing hidden biases that affect validity and generalizability. This is particularly important in artificial intelligence/machine learning (AI/ML) studies where models learn patterns directly from training data. We developed Equiflow, an open-source Python package that automates creation of enhanced participant flow diagrams tracking both sample size and composition changes throughout studies. Equiflow quantifies distributional shifts at each exclusion step and generates visualizations showing how key clinical and demographic variables evolve during participant selection. In a case study of sepsis patients from the eICU database, sequential exclusions reduced the sample from 126,750–1,094 patients. Requiring non-missing troponin measurements in the final step of data processing caused substantial demographic shifts that would typically remain invisible in traditional reporting. By making compositional biases visible during cohort construction before modeling begins, Equiflow enables researchers to make informed decisions about analyses and acknowledge limitations in generalizability to their readers. This standardized, open-source approach promotes transparency in clinical research and supports development of more equitable clinical AI systems, addressing a critical need as healthcare increasingly relies on data-driven decision making.
Human Factors Shaping Cybersecurity Behavior in Work from Home Environment in Ugandan Universities
(INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS, 2026-02-19) Atuhe Aarone Mike; Akampurira Paul; Richard Ntwari
The shift to remote and hybrid work in Ugandan universities exposed new cybersecurity risks shaped by human motivation, cognitive load, and system usability challenges. As academic operations increasingly depended on digital platforms, understanding how individuals formed and enacted protective intentions within home-based work environments became critical. This study examined how human and contextual factors—including threat perception, copingappraisal, usability difficulty, cognitive load, and digital fatigue—influenced cybersecurity behaviour among staff working remotely in Ugandan universities. Guided by the Protection Motivation Theory (PMT) and supported by constructs from the Theory of Planned Behaviour (TPB), the research adopted a sequential explanatory mixed-methods design. The quantitative phase identified key motivational and contextual predictors, while the qualitative phase explored how fatigue, usability barriers, and environmental conditions shaped protective motivation. Integration was achieved through narrative comparison and joint display analysis. Quantitative findings revealed that coping confidence and usability difficulty were the most influential determinants of secure behaviour, whereas fatigue and cognitive load significantly undermined protective intentions. Qualitative narratives reinforced these patterns, highlighting themes of threat awareness, usability frustration, motivational fatigue, and uneven institutional support. The study concluded that cybersecurity behaviour in remote academic environments was driven by motivational and contextual dynamics rather than technical controls alone. Strengthening coping efficacy, reducing usability burdens, and addressing digital fatigue were identified as essential strategies for developing adaptive, human-centred cybersecurity interventions in higher education.
Integrating generative artificial intelligence in African higher education: university students’ awareness, attitudes, and use of ChatGPT in Zambia
(In Frontiers in Education, 2026) Steward Mudenda; Moses Mukosha; Ruth Lindizyani Mfune; Bernard Kathewera; Imukusi Mutanekelwa; Boris Mwanza; Webrod Mufwambi; Muchindu Hampango; Kingsley Kamvuma; Martha Mwaba; Tumelo Muyenga; Chikwanda Chileshe; Mildred Zulu; Rabecca Tembo; Florence Mwaba; Stefan Kafwimbi; Adriano Focus Lubanga; Chintu Choolwe Simweene; Shafiq Mohamed; Ngoni Veddie Muzondo; Andrew Silungwe; Mapenzi Masilani; Patience Amutuhaire; Nathan Mugenyi; Roland Nnaemeka Okoro; Jimmy Hangoma; Victor Daka; Brian Godman
Background: The rapid emergence of generative artificial intelligence (AI) tools such as ChatGPT is transforming teaching and learning practices in higher education. This study assessed the awareness, attitudes, and usage patterns of ChatGPT among university students in Zambia and examined factors associated with students’ attitudes toward the technology Methods: A multi-institutional cross-sectional study was conducted among 1,829 university students in Zambia using a structured questionnaire adapted from instruments informed by the Technology Acceptance Model (TAM). Data were analysed using SPSS version 26.0, with statistical significance set at p < 0.05. Results: Among the 1,829 participants, 81.7% were aged 18–25 years, 52.5% were female, and 92.4% were unmarried. Overall, 96.8% of the students had heard of ChatGPT, and 85.6% reported having used it before this study. Among AI-usage constructs, 74.8% of respondents perceived high risks associated with AI use, 73.3% perceived ChatGPT as easy to use, 64.0% perceived it as useful, and 56.8% reported behavioural intention to use AI tools. In multivariable analysis, students aged ≥40 years were more likely to report positive attitudes toward ChatGPT compared with those aged 18–25 years (aOR = 5.91; 95% CI: 1.23– 28.33; p = 0.026). Technology/social influence was also significantly associated with positive attitudes (aOR = 2.04; 95% CI: 1.55–2.68; p < 0.001). Conversely, perceived risks were associated with lower odds of positive attitudes (aOR = 0.57; 95% CI: 0.43–0.75; p < 0.001). Regarding ChatGPT use, perceived usefulness significantly predicted higher usage (aOR = 1.49; 95% CI: 1.17–1.89; p = 0.001), while perceived risks were associated with reduced usage (aOR = 0.61; 95% CI: 0.45–0.84; p = 0.003). Conclusion: Awareness and use of ChatGPT are widespread among university students in Zambia, with more than four out of five students reporting prior use. Perceived usefulness and social influence were positively associated with adoption, whereas perceived risks were linked to lower attitudes and reduced use. These findings highlight the need for higher education institutions to develop clear policies and integrate AI literacy into curricula to promote responsible and effective use of generative AI technologies in academic environments.