User satisfaction of using speech recognition in health care centers

Abstract

Solutions to various rising healthcare costs have long been sought in a variety of technologies. The technologies are designed to ease the workload of professionals to allow more time to perform other tasks. Making an entry into the patient information system is one of the most important tasks in the work of professionals and often also the most time-consuming. Constantly evolving technology offers new and powerful applications that allow professionals to develop their own work efficiency and pace. Artificial intelligence technology in particular has been paid attention to the past 10 years. The speech recognition system utilizes an artificial intelligence-based cloud service in which a professional's speech is immediately converted to text, either directly to the patient information system or to another destination from where it can be transferred to the patient information system.

The aim of the study was to find out how satisfied the users of speech recognition were and what factors influence user satisfaction. In addition, users were asked to provide suggestions for improvements to increase user satisfaction. Previous surveys on user satisfaction with speech recognition were hardly found, especially in Finland. Other parts of the world, research on speech recognition in healthcare has focused mainly on error-sensitivity and usability, which is one component of user satisfaction. The study in the thesis was conducted using a quantitative questionnaire electronically, which was sent to speech recognition users in the Saarikka and Wiitaunioni area.

Based on the results, the users were generally satisfied with the use of speech recognition. The respondents were particularly satisfied with the orientation and the logic and ease of use. Satisfaction was explicitly emphasized in the use of the software itself. Based on the results, the facilitation of speech recognition for other work and especially for patient recording was felt to a lesser extent. As development ideas, the users wanted integration back between the patient information system and speech recognition software. In addition, regular refresher training and better accuracy in identifying non-medical vocabulary were desired.

The cost-effectiveness of the use of speech recognition was identified as a topic for further research, and a more detailed study of the time spent by professionals on recording would be needed in order to better understand the benefits of using speech recognition.