University of Kigali

Cannabis use among university students is a growing concern, particularly in demanding medical programs. We estimated prevalence, identified predictors, and compared coping strategies among medical students in Uganda. Cross-sectional survey of 318 undergraduates at Kampala International University (Western Campus). Cannabis use and hazardous/disordered use were screened with CUDIT-R (hazardous 8-11; probable use disorder ≥12, DSM-5-TR aligned). Coping was measured with the Brief COPE. Predictors were assessed using logistic regression; coping differences with the Wilcoxon rank-sum test. Cannabis use prevalence was 30.8% (n = 98); 7.6% met criteria for hazardous use and 9.4% for probable cannabis use disorder. Independent predictors of use were being separated (AOR = 12.00), being single (AOR = 3.45), Catholic faith (AOR = 2.76), and longer time at campus (AOR = 1.16 per year). Users reported higher emotion-focused and avoidant coping; problem-focused coping did not differ. Cannabis use among Ugandan medical students is common and associated with relationship status, religion, and time at campus. Coping profiles suggest greater reliance on maladaptive strategies among users. Findings support campus policies and mental-health programs that integrate substance-use screening and strengthen adaptive coping skills.

This study examined the effectiveness of the AI-powered plant classification tool, PlantNet, in enhancing Rwandan secondary school students’ conceptual understanding and engagement in plant classification. Anchored in constructivist learning theory, a sequential explanatory research design was employed involving 149 students (control = 74; experimental = 75) from both rural and urban schools, along with eight biology teachers. Data were collected through pre- and post-tests, a Likert-scale engagement questionnaire, and semi-structured teacher interviews. Quantitative data were analyzed using descriptive and inferential statistics in Statistical Package for the Social Sciences (SPSS), including t-test, while qualitative data were analyzed thematically. Results revealed a statistically significant improvement in students’ conceptual understanding and engagement in the experimental group compared to the control group (p = 0.006; p<0.01), with an effect size of Cohen’s d = 0.61, a practically meaningful (moderate to large) effect in educational research. Subgroup analysis also indicated consistent learning gains across gender and school location. Thematic analysis of teacher interviews supported the quantitative findings, highlighting enhanced student autonomy, curiosity, and confidence when using PlantNet. Although some technical challenges were reported, such as limited accuracy due to unclear images and poor internet access, teachers positively viewed the integration of AI tools into biology instruction. This study provides novel evidence on the pedagogical value of AI-powered plant identification in low-resource contexts, an area rarely explored in African secondary education. Future research should examine long-term learning retention, scalability, and integration of AI tools across other science disciplines to further advance technology-enhanced biology education.

Intervention effectiveness studies rarely empirically assess Theories of Change (ToC) to determine how an intervention worked. We examine the ToC underlying the Sugira Muryango (SM) parenting program in rural Rwanda to understand whether the intervention improved child development outcomes via changes in caregivers' behaviors to improve the home caregiving environment, as hypothesized. SM uses coaching of parents to create a safe, affectionate, stimulating, and violence-free home environment. A cluster randomized trial enrolled 1049 families with young children. SM had immediate effects on caregiver behaviors, improving scores on the Home Observation for Measurement of the Environment (HOME), harsh discipline, caregiver emotion regulation, and provision of dietary diversity. We use structural equation modeling to examine whether change in caregivers' behaviors explains intervention-related improvements in child development (Ages and Stages Questionnaire) one year after the intervention ended. Improvements in positive caregiving practices, including stimulation and early language learning as captured by the HOME, explained some of the intervention-related changes in child development, including gross motor, communication, problem-solving, and personal-social development. Increased dietary diversity explained intervention-related change in gross motor, problem-solving, and personal-social development. Change in harsh discipline and caregiver emotion regulation did not explain child outcomes. Intervention-related changes related to constructs captured on the HOME and dietary diversity were associated with changes in child development scores, but violent discipline and caregiver emotion regulation were not. Future research should examine whether these components of the intervention can be strengthened and may influence child development via other pathways, for example, via caregiver mental health.

Only very few educational mobile health applications (MHAs) on oral cancer exists. Based on existing literature, only one of such MHAs (application name: Prayaas) was created specifically for lay persons; unfortunately, the application has been critiqued to be non-inclusive and non-comprehensive. Hence, this study aimed to co-create a more comprehensive and inclusive educational MHA prototype on oral cancer through a participatory action research approach. A three-round modified Delphi technique was employed to engage stakeholders in a participatory action research process. Digital communication experts and individuals with oral cancer risk factors (n = 17) from five countries participated. Round one utilised focus group discussions to define essential application features. Rounds two and three involved iterative evaluations of successive prototypes using a validated questionnaire (Mobile Application Rating Scale). Quantitative data were analysed using descriptive statistics, with consensus defined a priori as a mean score > 2.5 on a 5-point Likert scale—indicating majority agreement on each evaluated item. Stakeholder input across three iterative rounds led to the co-creation of the “Beat Oral Cancer” prototype. Quantitative consensus was achieved across all evaluated domains, including usability, aesthetics, information quality, and perceived impact. The final prototype showed significant improvements: the second version was rated higher than the first in entertainment (+ 0.68), interactivity (+ 0.32), customizability (+ 0.64), and ease of use (+ 0.33). Key inclusive features, such as adjustable text size, were incorporated based on stakeholder feedback. This study successfully co-created a highly comprehensive, stakeholder-informed MHA prototype for oral cancer education via a modified Delphi process. The results demonstrate the value of iterative participatory design in creating digital health tools with high potential for acceptability and usability. This work establishes a methodological precedent for applying the Delphi technique in mobile health co-creation and provides a foundation for future efficacy testing and public deployment.

Mobile health (mHealth) apps are increasingly leveraged to support community health workers (CHWs) in delivering high-quality care, particularly in low- and middle-income countries. However, despite the proliferation of mHealth tools, few have been implemented at scale, partly due to limited attention to usability and acceptability among end users. In sub-Saharan Africa, mHealth tools designed for CHWs often lack systematic evaluation using validated instruments tailored to local contexts. Without such assessments, it is difficult to ensure that these tools can be integrated effectively into CHW workflows and scaled sustainably. This study aimed to adapt and validate existing mHealth usability and acceptability assessment tools to be contextually appropriate for CHWs in Rwanda. Specifically, we sought to ensure contextual appropriateness for CHWs supporting postoperative home follow-up for women after cesarean delivery. The resulting tool was designed for use in an implementation study of a novel CHW-led mHealth app. This study was conducted in the Kirehe district, Rwanda, from October 2022 to March 2023. We adapted 2 established tools-the mHealth App Usability Questionnaire and selected items from the Practitioner Opinion (Acceptability) Scale-and added new items that reflect core functions of the CHW-focused mHealth app. All items were translated into Kinyarwanda and simplified to align with CHWs' educational levels. We conducted a three-stage validation that consisted of (1) content validity testing with 8 local and international experts using a recommended content validity index threshold of >0.78; (2) face validity testing with 10 CHWs using a recommended face validity index threshold of ≥0.60; and (3) reliability testing using responses from 30 CHWs, with a Cronbach α coefficient of ≥0.70 indicating acceptable internal consistency. Of the 25 items assessed, 22 (88%) achieved a content validity index score of >0.78 for both clarity and relevance. The face validity index across all 22 items was 0.991, indicating strong comprehensibility and relevance to CHWs. Internal consistency was high: the Cronbach α was 0.86 for the mHealth App Usability Questionnaire items, 0.73 for the Practitioner Opinion (Acceptability) Scale items, and 0.87 for the newly developed questions. The final tool-named the Community Health Worker mHealth Usability and Acceptability Assessment Tool-included 22 items with strong content validity, face validity, and internal reliability. This study presents a rigorously adapted and validated tool for assessing mHealth usability and acceptability among CHWs in Rwanda. The Community Health Worker mHealth Usability and Acceptability Assessment Tool can guide future evaluations of mHealth interventions in similar contexts and serve as a model for localizing mHealth assessment tools in low- and middle-income country settings to ensure fit-for-purpose implementation.

Preterm birth, defined as delivery before 37 completed weeks of gestation, is a leading cause of neonatal illness and death worldwide, particularly in low- and middle-income countries such as Rwanda. Despite progress in maternal and newborn healthcare, preterm birth remains a major public health concern. This study aimed to determine the prevalence and determinants of preterm birth. This retrospective cross-sectional study reviewed medical records of 1,327 women who delivered at the University Teaching Hospital of Kigali between January 1 and December 31, 2024. Using total enumeration, data were collected on maternal socio-demographic characteristics, obstetric factors, and medical conditions. Descriptive statistics, bivariate analysis, and multivariable logistic regression were applied to determine the prevalence of preterm birth and its associated factors. The prevalence of preterm birth at the University Teaching Hospital of Kigali was 13.26%. Independent predictors of preterm birth included a history of preterm birth (AOR = 7.13, 95% CI: 3.08–16.49), preeclampsia (AOR = 3.02, 95% CI: 1.32–6.89), hypertension (AOR = 2.13, 95% CI: 1.04–4.35), premature rupture of membranes (AOR = 7.57, 95% CI: 4.06–14.09), and attending four or more antenatal care visits (AOR = 4.19, 95% CI: 1.84–9.56). Preterm birth remains a major public health concern in Rwanda, particularly among high-risk pregnancies referred to tertiary care centers such as the University Teaching Hospital of Kigali. These findings highlight the need for strengthening early risk detection, improving primary healthcare services, and ensuring timely referral and management of at-risk women to reduce preterm birth and its associated complications.