Interactive Technology to Aid Decision Making in Cardiac Care

Abstract

This project aims to aid clinical decision making in cardiac care. Firstly, it will include a retrospective analysis of a PPCI referral dataset to identify patient referral patterns, decision making patterns and patient pathway variations as well as potential performance metrics that could be used in a dashboard. This will include a data mining approach using techniques such as anomaly detection and clustering to determine the patient archetypes and supervised machine learning to classify 30 day and one-year mortalities. This research will also assess the performance differential when using a paper-based form versus a smarter digital from, for example, how can a digital form improve the efficacy and usability in the context of PPCI. We will also assess the efficacy of different digital form designs and whether a multi-page form is more efficient when compared to a single page form. There is reason to hypothesize that a multi-page form is a better model, including the reduction in cognitive load and information as well as providing a sense of progression for the user. This research also intends to provide empirical HCI guidelines for the creation of digital forms and will build on previous work in the HCI community [1]. Guidelines are vital for providing informative best practices and for optimising the design and usability of digital forms and dashboards. We have found that there is a lack of adoption of interactive e technologies in the field of cardiac care and we hypothesise that interactive technologies can aid clinicians in improving decision making.