Muni Repository (MR)
This repository contains open access publications of Muni University Library.
Objectives:
- To digitally collect, preserve and provide electronic access to scholarly works and research output of Muni University.
- Increase the visibility and impact of our research, making it easy for researchers, students, policymakers and journalists to reference, replicate, and re-use the work.
- Issue permanent, unique and trustworthy identifiers when creating URLs to access the resource without concern that the location of the resource may change.
How to publish in Muni Repository
- Contact the library through email: libsupport@muni.ac.ug
Communities in MR
Select a community to browse its collections.
Recent Submissions
Adverse drug reaction reporting with the Med Safety app inUganda: a cluster-randomised, controlled trial
(Elsevier, 2025-10) Kiguba, Ronald; Ndagije, Helen B.; Mwebaza, Norah; Ssenyonga, Ronald; Giibwa, Lilian; Isabirye, Gerald; Owiny, Jonathan; Nambasa, Victoria; Ntale, Ismail; Atuhaire, Joanitah; Mwesigwa, Douglas; Mayengo, Julius; Walusimbi, David; Mugisa, Ian; Katureebe, Cordelia; Harrison, Kendal; Karamagi, Charles; Pirmohamed, Munir
Background: The massive roll-out of new and repurposed medicines in low-income and middle-income countries (LMICs) highlights the need for more efficient pharmacovigilance systems, including use of digital technologies. We assessed the effectiveness of the Med Safety app in improving suspected adverse drug reaction (ADR) reporting by health-care workers to Uganda's National Pharmacovigilance Centre.
Methods: This was a pragmatic, multicentre, open-label, cluster-randomised, controlled trial undertaken at health facilities (clusters), providing dolutegravir-based combination antiretroviral therapy in Uganda. Clusters were randomly assigned (1:1) to the intervention group or control group using a computer-generated simple randomisation sequence. In the intervention group, pharmacists with expertise in pharmacovigilance delivered 2 h of face-to-face training to health-care workers in clusters, regardless of their smartphone ownership, in Med Safety and traditional ADR reporting methods. The control group received the same training as the intervention group except for Med Safety training. The primary outcome was the cluster-level ADR reporting rate at the end of follow-up and was analysed in all sites that received the allocated intervention. The trial is registered with the Pan African Clinical Trials Registry (PACTR202009822379650) and is completed.
Findings: Between Aug 11, 2020 and Nov 1, 2022, 382 clusters were randomly assigned and 367 received the allocated intervention and were included in the primary outcome analysis (184 in the intervention group and 183 in the control group), with 2464 health-care workers (1211 in the intervention group and 1253 in the control group). The follow-up time for the included clusters was variable and was median 37·8 months (IQR 34·2–39·8). In the primary analysis, the intervention group had a higher mean overall ADR reporting rate of 10·6 (SD 17·4) reports per 100 000 person-months versus 5·9 (17·9) in the control group (incidence rate ratio 1·73 [95% CI 1·26–2·37]; p=0·001).
Interpretation: Med Safety increased ADR reporting rates among health-care workers in Uganda. Integrating digital technologies into pharmacovigilance systems could strengthen drug-safety monitoring in Uganda and other LMICs.
Radiation-induced degradation in the properties of pristine anddouble-walled carbon nanotube-enhancedpoly (2,5-benzimidazole) polymers for radiation shielding in the LEO
(Elsevier, 2025-09-01) Oryema, Bosco; Square, Lynndle; Ellis, Ernst
Degradation in the properties of polymer-based materials in space environments is a critical challenge for developing lightweight radiation shielding solutions. In this paper, a comparative study of the impacts of helium ion (He+) irradiation one of the ion species in the Low Earth Orbit (LEO) environment on the structural and optical properties of pristine and 1.0 wt% double-walled carbon nanotube (DWCNT)-enhanced poly(2,5-benzimidazole) (ABPBI) polymers for LEO radiation shielding applications was conducted. The two polymer categories were separately chemically prepared in the laboratory, moulded, dried, and cut into 1 cm × 1 cm pieces, and bombarded with 0.35 MeV He+ ions at varying fluences. The Ultraviolet–Visible-Near-Infrared (UV–Vis-NIR) optical analyses of the polymers following the ion bombardment revealed that He+ irradiation considerably raises the Urbach energy and decreases the optical bandgap, indicating a rise in electronic defects and structural disorder. On the other hand, the Fourier Transform Infrared (FTIR), Atomic Force Microscopy (AFM), and X-ray Diffraction (XRD) analyses revealed higher levels of structural degradation in the pristine ABPBI samples, suggesting changes brought about by irradiation-induced oxidation and chain scission processes. In contrast, the 1.0 wt% DWCNT-ABPBI composite demonstrated improved optical and structural integrity, retention, and resistance to He+ ion-induced damage. According to the results, 1.0 wt% DWCNT reinforcement reduces radiation-induced deterioration and offers more protection from energetic ion exposure in the LEO settings. Thus, this work highlights the distinct impact of He+ ion interactions with ABPBI and the effectiveness of DWCNT inclusion in improving polymer resilience, and it contributes to the fundamental understanding of the polymer composite for radiation shielding applications.
Prevalence and factors associated with neonatal hypothermia: a cross-sectional study among healthy term neonates in a peri-urban hospital in Northern Ugand
(Springer Nature, 2025-10-08) Akao, Mary Grace; Nalwadda, Gorrette; Epuitai, Joshua; Ayebare, Elizabeth; Ndeezi, Grace; Ratib, Dricile; Tumwine, James K
Background
Neonatal hypothermia is highly prevalent even in warm tropical countries. Neonatal hypothermia increases the risk of morbidity and mortality. In Uganda, the prevalence of hypothermia is not known among healthy term neonates.
Objective
To determine the prevalence of neonatal hypothermia and the associated factors in Lira Regional Referral Hospital.
Methods
Hospital-based cross-sectional study was conducted in Northern Uganda. The interviewer-administered questionnaires and direct observations used to determine the initiation of warm-chain practices after delivery for 271 newborns. The axillary temperature of neonates was measured at intervals of 10 min, 30 min, one hour, and 2 h after birth. The multivariate binary logistic regression was done. The 95% confidence interval (CI) and p-value < 0.05 used to identify factors significantly associated with neonatal hypothermia.
Results
Neonatal hypothermia was 67.6% during the first two hours postnatal. Neonatal hypothermia was 64.5% at 10 min, 81% at 30 min, 76% at one hour and 49% at two hours postnatal. Hypothermia was significantly associated with low birth weight (Adjusted odds ratio (AOR) = 2.78; 95% CI: 1.01–7.62); male sex (AOR = 1.69; 95% CI: 1.04–3.33), not drying the newborn (AOR = 3.06, 95% CI: 1.64–5.72); no skin to skin contact within five minutes postnatal (AOR = 2.17, 95% CI: 1.15–4.10); and low maternal body temperature (AOR = 2.70, 95% CI: 1.49–4.76).
Conclusions
The prevalence of neonatal hypothermia was high in the first two hours. Neonates who were more likely to have hypothermia were male, not dried properly, low birth weight, no skin-to-skin contacts, and low maternal body temperature. Proper drying of the newborn and skin-to-skin contact can reduce the burden of neonatal hypothermia. There is a need to train the midwives on proper drying of the newborn, keeping the mother warm, and the importance of skin-to-skin contact in prevention of neonatal hypothermia among male and low birth neonates.
An explainable AI framework for neonatal mortality risk prediction in Kenya: Enhancing clinical decisions with machine learning
(Science Publishing Group, 2025-09-30) Lumumba, Victor Wandera; Muriithi, Dennis Kariuki; Njoroge, Elizabeth Wambui; Langat, Amos Kipkorir; Mwebesa, Edson; Wanyama, Maureen Ambasa
Neonatal mortality remains a critical public health challenge in Kenya, with a rate of 21 per 1,000 live births—well above the SDG 3.2 target. While machine learning (ML) offers potential for risk prediction, most models lack transparency and clinical interpretability, limiting their adoption in low-resource settings. This study presents an explainable AI (XAI) framework for predicting neonatal mortality using Kenya Demographic and Health Survey (KDHS) data (N = 2,000), with a focus on model accuracy, fairness, and clinical relevance. Six ML models—Logistic Regression (LR), KNN, SVM, Naïve Bayes, Random Forest, and XG-Boost—were trained and evaluated using in-sample, out-of-sample, and balanced datasets, with performance assessed via AUC, F1-score, sensitivity, specificity, and Cohen’s Kappa. To address class imbalance and enhance generalizability, synthetic oversampling and rigorous cross-validation were applied. Post-balancing, LR achieved optimal performance (AUC = 1.0, κ = 0.98, F1 = 0.987), with SVM (AUC = 0.995) and XG-Boost (AUC = 0.982) also showing higher performance. SHAP and model breakdown analyses identified Apgar scores (at 1st and 5th minutes), birth weight, maternal health, and prenatal visit frequency as key predictors. Fairness assessments across socioeconomic subgroups indicated minimal bias (DIR > 0.8). The integration of XAI enhances transparency, supports clinician trust, and enables equitable decision-making. This framework bridges the gap between predictive accuracy and clinical usability, offering a scalable tool for early intervention. Policy recommendations include embedding this XAI-enhanced model into antenatal care systems to support evidence-based decisions and accelerate progress toward neonatal survival goals in resource-limited settings.
Value of hs‑cTnT, sST2, and Lp‑PLA2 in the classification of acute coronary syndrome
(Spandidos Publications, 2025-08-22) Peng, Hongxin; Lubanga, Nasifu; Sun, Cong; He, Bangshun; Mei, Yan‑Ping; Wang, Yishan
The present study aimed to assess the value of high‑sensitivity cardiac troponin T (hs‑cTnT), soluble suppression of tumorigenicity 2 protein (sST2) and lipoprotein‑associated phospholipase A2 (Lp‑PLA2) in the classification of acute coronary syndrome (ACS). A total of 236 patients diagnosed with ACS were enrolled in this retrospective study and were further divided into the non‑ST‑segment‑elevation (NSTE)‑ACS group (n=183) and ST‑segment elevation myocardial infarction (STEMI) group (n=53). The three biomarkers (hs‑cTnT, sST2 and Lp‑PLA2) were measured by electrochemiluminescence. The diagnostic performance of each biomarker in differentiating ACS subtypes was evaluated through receiver operating characteristic curve analysis. The DeLong test was applied to compare the discriminatory abilities of the different markers. The binary logistic regression model was employed to analyze the factors influencing ACS classification. The levels of hs‑cTnT and sST2 in males were significantly higher in the STEMI group than in the NSTE‑ACS group (P<0.05). hs‑cTnT [odds ratio (OR)=1.010, 95%CI: 1.007‑1.014] and sST2 (OR=1.022, 95%CI: 1.011‑1.033) were identified as good predictors for distinguishing STEMI from NSTE‑ACS, whereas Lp‑PLA2 (P=0.470) was not a suitable biomarker to discriminate between the two types of ACS. Additionally, the diagnostic efficacy of hs‑cTnT [area under curve (AUC=0.861)] and the combination of hs‑cTnT and sST2 (AUC=0.863) was higher than that of sST2 alone (AUC=0.833, P<0.05). In conclusion, these findings illustrated that hs‑cTnT and sST2 are promising biomarkers to classify patients with ACS. Compared with sST2 alone, hs‑cTnT and its combined detection demonstrate superior diagnostic efficiency in identifying ACS.