UWCScholar

This repository serves as a digital archive for the preservation of research outputs from the University of the Western Cape

Communities in UWCScholar

Select a community to browse its collections.

Now showing 1 - 5 of 11

Recent Submissions

Detection of substances in human dentine using liquid chromatography tandem mass spectrometry
(University of the Western Cape, 2026) Smith-Jacobs Jami Ashleigh; Kabbashi Salma; Chandler Susan
Substance use disorders remain a major public health and forensic concern globally and within South Africa, particularly in the Western Cape. Methamphetamine, methaqualone, and heroin remain highly prevalent and have devastating effects on communities of the Cape Flats, with links to violence, crime and fatalities. Conventional forensic toxicology matrices such as blood and urine are limited by short detection windows and rapid postmortem degradation. Mineralised dental tissues, particularly dentine, may provide a stable alternative matrix capable of preserving evidence of long-term substance exposure. However, research investigating isolated human dentine as a toxicological matrix remains scarce. To date, research reporting the detection of methaqualone in human dentine is notably absent. Furthermore, existing dentine toxicology research has been restricted to postmortem conditions, with no published studies utilising dentine from living participants or comparisons between active and past substance users to evaluate long-term substance retention within the matrix. Aim: The core aim of this exploratory study was to evaluate the feasilibility of isolated human dentine as a matrix for detecting methamphetamine, methaqualone, and heroin metabolites in active and past substance users in the Western Cape, South Africa, using Ultra-Performance Liquid Chromatography Tandem Mass Spectrometry (UPLC-MS/MS). Materials and methods: A total of 23 dentine samples were analysed from two participant cohorts: 18 past substance users were recruited from the U-Turn Homeless Ministry and 5 active substance users were recruited from the South African National Council on Alcoholism and Drug Dependence (SANCA). Dentine was isolated from extracted teeth, pulverised, solvent extracted, and analysed using UPLC-MS/MS.
Adverse Drug Reactions in Paediatric Patients in Eswatini: Reporting Patterns and Characteristics.
(University of the Western Cape, 2026) Sithole Chengetai Memory; Coetzee Renier
Background: Adverse drug reactions (ADRs) in paediatric patients remain a major public health concern, particularly in low- and middle-income countries where pharmacovigilance systems are still developing. This study aimed to describe the reporting patterns, demographic characteristics, and clinical features of ADRs in children in Eswatini. Methods: A retrospective descriptive study was conducted using ADR reports submitted to the Eswatini Pharmacovigilance Centre from January 2020 to December 2023. Data on patient demographics, drug classes, routes of administration, ADR types, severity and outcomes were analysed using descriptive statistics and chi-square tests. Results: Of the 1452 ADR reports received, 137 (9.4%) involved paediatric patients with mean and median ages of 12.65 ± 4.8 years and 15 years, respectively (IQR: 11–17). Most ADR reports occurred in patients aged 11–18 years (75.9%; n = 104). Female patients accounted for 62.8% (n = 86) of the paediatric ADR reports. Regional reporting was uneven, with most reports originating from the Hhohho region (28.5%), through the Expanded Programme on Immunisation (EPI) Call Centre. COVID-19 vaccines accounted for half of all paediatric ADRs (47.5%), particularly the Pfizer vaccine (33.6%), followed by antiretroviral agents (25.6%) and anti-tuberculosis drugs (11.7%). Common reactions include headache, fever, rash and anaemia. Most ADRs were reported after intramuscular (50.4%) or oral (46.0%) administration of the drug. Most (84.7%) were non-serious; however, 12 (8.8%) were severe, with one fatality linked to procaine penicillin. Younger children (0–5 years) experienced a higher proportion of severe ADRs (p = 0.01). Half of the patients (50.1%) recovered fully, 11.7% were still recovering, 10.2% had not recovered and 26.3% had an unspecified outcome. Conclusion: Paediatric ADRs in Eswatini represented less than 10% of all reports, with a predominance among adolescents and females. COVID-19 vaccines and antiretroviral drugs are the leading causes of these reactions. Although most reactions were mild, the occurrence of severe and fatal ADRs highlighted the need to strengthen pharmacovigilance systems, improve regional reporting consistency, and enhance safety monitoring of paediatric drug use in Eswatini.
Exploring the factors influencing the decline of learners electing accounting at the high school level within the Western Cape
(University of the Western Cape, 2026) Simons Shameema; Williams Badrunessa
This study aims to explore the factors contributing to the decline in the number of learners completing grade 12 with accounting as an elective subject. To achieve this aim, this study is structured around three objectives: first, to identify the factors contributing to learners' decisions not to choose accounting as a subject during the FET phase; second, to explore the reasons for the drop-out rates of learners who initially selected accounting as a subject during the FET phase, but did not complete grade 12 with accounting; and third, to investigate the rationale for schools discontinuing accounting as a subject. Using a qualitative research design, the study examines data from various participant categories. These categories include: (i) learner participants (grade 9 learners choosing subjects, grade 10 learners who chose not to elect accounting, and learners who initially elected accounting but later dropped it), (ii) accounting teachers, and (iii) school principals. This comprehensive approach allows for rich, nuanced data to be extracted from this study. The findings show that learners’ career aspirations, perceived subject-specific competence, and confidence in mathematics influence both the decision to elect accounting and, in some cases, the decision to drop it. For learners, their experiences with subject teachers affect whether they continue with a subject. Perceptions of accounting as a high-workload, rigid, and intellectually demanding subject further deter sustained learner participation. At the school level, the study reveals that although accounting is viewed as a valuable subject, its delivery is hindered by resource constraints and inadequate teacher preparedness and professional development. Based on the findings, the study suggests improving teacher training, addressing staffing and resource challenges and increasing learner engagement with accounting-related career pathways. It is also suggested that student feedback be solicited about their experiences with teachers, ensuring this is considered as a developmental tool.
Customer Personality Analysis using Machine Learning Techniques
(University of the Western Cape, 2026) Ntolo Siyambonga; Davies Clint
To evaluate customer data and identify common patterns and qualities among a group of consumers, unsupervised learning is a machine learning technique that can be used. Companies may build customer groups with different traits by utilizing clustering algorithms to group like consumers. This enables them to target each segment with their marketing and sales initiatives. The study investigates customer segmentation and profiling using machine learning techniques, which has been referred to as customer personality analysis in this mini-thesis, adopting a quantitative research methodology using unsupervised machine learning techniques on a retail marketing dataset. Unsupervised learning algorithms are effective tools for businesses wanting to better understand and service their consumers since they can be trained on vast volumes of datasets. Customers have traditionally been analyzed manually by either marketing or sales teams; they would utilize their knowledge and experience to spot recurring trends and patterns among customers. The development of data mining and machine learning tools, however, has made it feasible to automate this procedure by utilizing algorithms that can scan massive volumes of data and find common patterns and qualities among clients. The researcher applied K-Means and Agglomerative Clustering, after pre-processing the data, to identify distinct customer segments. The researcher’s SHAP analysis showed that the customer's monetary expenditure on wine products is a significant contributor or determining factor to the campaigns they are inclined to accept. The results demonstrate the viability of these methods for targeted marketing. The results reveal segments characterized by age, education, and purchasing channels. The researcher recommends that businesses adopt these analytical techniques because these will lead to improved customer understanding
Sensory Augmentation for the Visually Impaired - A Novel Navigation System.
(University of the Western Cape, 2026) Naidoo Nathan Lyle
Recent reports by the World Health Organization (WHO) estimate that over 2.2 billion people worldwide are affected by visual impairment (VI), with prevalence up to four times higher in low- and middle-income regions. Independent mobility is essential for quality of life, yet traditional mobility aids such as white canes, guide dogs, and human assistance remain constrained by limited environmental awareness and functionality. This research investigates how Electronic Travel Aids (ETAs) can provide accurate, timely, and comprehensive environmental information to support safe navigation. This study proposes a novel multi-node ETA framework comprising three wirelessly connected wearable nodes integrated into a glasses, belt, and cane platform. Guided by Orientation and Mobility principles and Hall’s proxemic theory, the framework distributes sensing across the body to maximise environmental coverage while reducing blind spots. The system integrates ultrasonic sensors, LiDAR, inertial measurement units, computer vision, and artificial intelligence to provide real-time obstacle detection, object classification, spatial interpretation, and multimodal audio and haptic feedback. The methodology comprises hardware evaluation, sensor placement optimisation, system implementation, and experimental validation. Experimental evaluation used a custom-designed autonomous test rig replicating human posture and cane motion under controlled laboratory conditions, enabling objective and repeatable assessment. Static and dynamic experiments evaluated power consumption, obstacle detection accuracy, object classification, latency, inertial stability, and system responsiveness. Results demonstrate reliable obstacle detection across head-, torso-, and ground-level regions, end-to-end response times below the 500 ms design objective, and detection accuracy within predefined performance targets. The autonomous evaluation platform further contributes a repeatable methodology for benchmarking ETA systems before human participant studies. The research contributes a novel multi-platform wearable ETA architecture, a usercentred sensor placement framework grounded in spatial perception theory, and an autonomous evaluation methodology that advances the design and objective validation of next-generation assistive navigation systems for individuals with visual impairment.