Refine
Document Type
Language
- English (2)
Is part of the Bibliography
- no (2)
Institute
Begutachtungsstatus
- peer-reviewed (1)
Objectives
Germany has one of the highest numbers of coronary angiography worldwide. Nevertheless, we find a great variation between German regions for both, diagnostic coronary angiographies and percutaneous coronary interventions (PCI). We assume that this variation is not only due to variation in morbidity but also reflects factors such as guideline adherence, physician-patient communication and access to care.
In this mixed method project, we aim to first describe the variation of coronary angiographies and PCIs in different German regions. Secondly, we explore current practices and motives for (non-)adherences to guidelines in the diagnostic process of patients with suspected CHD. Based on these results, we consequently plan to develop a complex intervention (treatment pathway) to improve guideline adherence and thus appropriateness of coronary angiography.
Method
The projects will be organized according to the recommendation of the Medical Research Council for the development and evaluation of complex interventions. The first study will use descriptive methods based on routine data of three German Health Care Insurances and registry data to describe status quo and associated factors of coronary angiography. The second study will use qualitative methods to understand barriers and facilitators of guideline adherence and medical decision making. Furthermore, we will discuss variations in care and thereby identify implementation targets for the planned treatment pathway. Based on this information, we will develop local treatment pathways in four selected regions. Relevant peers will develop the local pathway in group discussions. Using this bottom-up approach, we directly address implementation challenges.
The use of chatbots based on large language models (LLMs) and their impact on society are influencing our learning experience platform Hochschul-Assistenz-System (HAnS). HAnS uses machine learning (ML) methods to support students and lecturers in the online learning and teaching processes [1]. This paper introduces LLM-based features available in HAnS which are using the transcript of our improved Automatic Speech Recognition (ASR) pipeline with an average transcription duration of 45 seconds and an average word error rate (WER) of 6.66% on over 8 hours of audio data of 7 lecture videos. A LLM-based chatbot could be used to answer questions on the lecture content as the ASR transcript is provided as context. The summarization and topic segmentation uses the LLM to improve our learning experience platform. We generate multiple choice questions using the LLM and the ASR transcript as context during playback in a period of 3 minutes and display them in the HAnS frontend.