For a high a priori speech recognition rate, creation of an adapted dictionary for our specific purpose in addition to the general dictionary of the discipline (internal medicine) appeared necessary to account for the specific vocabulary used for documentation.Ī vocabulary learning function of the speech recognition software is mandatory for a clinical speech recognition system. A characteristic of our department is the very specific purpose of the department with a focus on elite sports and a close link to sports sciences.
In our context, this translates into a high a priori speech recognition rate, so that physicians will perceive the software to be productive. When implementing a novel technology maximisation of the initial performance has generally been identified as beneficial. In our context, “short-term wins” are understood as the immediate experience of a versatile, adaptive software, which can be used productively (from the user perspective) after a relatively short period of time. In Kotter’s change management framework, it is argued that implementation of change is much facilitated by creating “short-term wins” for users. Hence, user experience was brought into central focus during the preliminary considerations when planning implementation in our department. This means that users who do not get involved with such software might not acquire the necessary skills to fully benefit from a speech recognition solution. , a change initiative concerning speech recognition deals with implementation of novel technology for which user satisfaction and objective functionality are strongly linked. In this article, we aim to provide a case study on the steps to successfully implement speech recognition software in a highly specialised university outpatient department focusing on the prearrangements that appear to be beneficial concerning productivity of the software and user motivation as the importance of this “predesign stage” has been emphasised before. As a consequence of a lack of motivation and hence, less involvement with the software, low contribution rates of speech recognition for the final text might occur, which could partially explain the heterogeneity of observed productivity gains when speech recognition is used.
It has, however, been described that introducing speech recognition software might impair user satisfaction leading to demotivation of the physicians using the software and also might impose more workload on physicians. The introduction of a speech recognition solution for medical documentation in the healthcare sector can improve productivity, as expressed by letter-turnaround-time, without remarkably impairing the quality of patient records.
0 kommentar(er)
