Publication Details
Abstract
Efficient documentation in healthcare is vital for service quality, yet electronic health records (EHRs) often impose time, cognitive, and usability burdens on physicians. In Uzbekistan, challenges in manual EHR entry, typographical errors, and copy-pasting hinder clinical productivity and data integrity. While global studies highlight these limitations, there is limited empirical research in post-Soviet healthcare settings on the role of automatic speech recognition (ASR) in overcoming them. This study evaluates the effectiveness of ASR tools in improving documentation time, accuracy, and medical data richness within Uzbek clinical environments. Using a mixed-methods quasi-experimental design across two medical institutions, ASR reduced documentation time by 41%, decreased typographical errors by 17.6%, and increased the volume of recorded medical data by 28%. Physicians also reported a 7.5% decline in copy-paste behavior and noted improved satisfaction and workflow efficiency. This research provides the first quantified national estimate of ASR’s potential in saving over 108 million hours annually, translating into approximately $292 million in cost reductions for Uzbekistan’s healthcare system. ASR integration not only boosts operational efficiency but also enhances patient safety and clinical decision-making through richer, error-reduced documentation. The study supports piloting ASR in diverse medical settings and embedding digital literacy in healthcare education to ensure adoption across age groups, contributing to broader digital transformation in emerging healthcare systems.