Implementing carotid ultrasound and Holter monitoring through telemedicine-based training in a stroke unit in Lusaka, Zambia.

BACKGROUND AND AIMS: Stroke remains a leading cause of mortality and disability in many low- and middle-income countries, where access to diagnostic and treatment resources is often severely constrained. This pilot study investigated the feasibility of telemedicine-based training to integrate carotid artery ultrasound and Holter monitoring into routine diagnostic practices at the stroke unit of the University Teaching Hospital (UTH) in Lusaka, Zambia. METHODS: Five neurology residents at the University Teaching Hospital in Zambia, without prior clinical experience in carotid artery ultrasound, received remote online training sessions. Subsequently, they were divided into two groups: the first conducted practical examination sessions under on-site supervision of a stroke neurologist, while the second was remotely supervised via screen sharing by a stroke neurologist from Krankenhaus Nordwest in Frankfurt (KHNW), Germany. Handheld portable ultrasound probes (Butterfly IQ+) were used for the examinations. Following the training, each group of residents performed 50 extracranial ultrasound examinations in acute ischemic stroke patients at the UTH stroke unit. Each examined patient was re-examined in a separate session by an experienced stroke neurologist, who was blinded to the results of the residents' examination. The agreement between raters in the assessment of carotid stenosis was assessed using Cohen's kappa (κ), a statistical measure that evaluates interrater reliability for categorical items. Similarly, 26 stroke nurses at UTH were trained in Holter monitoring exclusively through video tutorials, without hands-on practice. They recorded 30 Holter examinations on subsequent acute ischemic stroke patients. The quality of the recordings was subsequently compared to 30 Holter recordings from consecutive patients at the stroke unit of KHNW. A cardiologist, blinded to the origin of the recordings, evaluated their quality on a scale of 1 to 10, and the results were analyzed using Welch's t-test. All participants completed multiple-choice assessments to evaluate their theoretical knowledge, along with a feedback survey on the training program. RESULTS: 50 patients underwent bilateral carotid artery ultrasound examination, split evenly between the direct and remotely supervised groups. Both groups achieved a high rate of concordance with an experienced stroke neurologist. The directly supervised group achieved 86% concordance for ICA stenosis and 88% for plaque detection, while the remotely supervised group achieved 80% and 84%, respectively. Holter recordings from UTH demonstrated higher quality than those from the stroke unit at KHNW (p < 0.01). Most participants reported enhanced confidence and knowledge, though over 60% preferred face-to-face training formats. CONCLUSION: Innovative telemedical training approaches offer a transformative solution for addressing diagnostic and infrastructure challenges in stroke care, particularly in resource-constrained healthcare settings. Comprehensive research is crucial to optimize these programs and enhance medical capabilities across diverse healthcare environments.