Human-Computer Interaction

Digitalisierungszentrum Präzisions- und Telemedizin (DZ.PTM)


Overview

This project serves to gradually establish the structure of the Digitalisierungszentrums Präzisions- und Telemedizin (DZ.PTM) Würzburg – Bad Kissingen as a virtual centre spread over three locations: the University Hospital, the University of Würzburg, and the ZTM in Bad Kissingen. These locations are designated to develop, test and implement digitization projects across Bavaria to support patient care and research.

The infrastructure to be set up will be used to implement a pilot project for knowledge-based service management in radiology. The project targets to reduce workload for physicians and to improve the quality of treatment with a focus on patients with rare diseases. In addition, the pilot project promotes the development of a Testbed for the early development and testing of digitization projects in the health care sector.

The DZ.PTM cooperated closely with the funding measure “BASE-Net” (electronic patient file of the Bavarian Centers for Rare Diseases) under the leadership of the University Clinics of Regensburg and Würzburg.

Goals

Overall project goals

Goals of the HCI chair

  1. Identify issues in existing workflows
  2. Prototype and evaluate mobile ways to request clinical treatment
  3. Generate ideas on how AR and VR can be used in clinical treatment

Progress

1. Identify issues in existing workflows

The first sub-goal of the project was to identify problems in the process of requesting clinical treatments and to generate ideas on how to improve on them. To do this, we followed the rapid contextual design process described by Holtzblatt, Wendel, and Wood (2004). We gathered the requirements and current problems of physicians, radiologists, surgeons, and medical assistants by conducting a contextual inquiry and an additional focus group. In total, we found nine core problems: (1) Difficulties to access patient data & requests, (2) the large number of phone calls, (3) restricted & abused access rights, (4) request status difficult to track, (5) paper notes used for patient data, (6) lack of assistance for data entry, (7) frustration through documentation, (8) IT-systems not self-explanatory, and (9) conflicts between physicians and radiologists. Most of these problems are due to inadequate digitization in the hospital.

We then generated ideas on how to improve these together with the stakeholders. One solution to tackle multiple problems at once is the mobile availability of patient data during ward rounds. This benefits physicians in multiple ways: Physicians can request treatment directly when they are with the patient. Currently, they have to finish all patients first, leading to inefficient note-taking and errors. Also, physicians are constantly interrupted by phone calls during the day, which leads to further errors. Thus, a mobile solution where treatment can be requested easily and quickly would benefit physicians. We consolidated our results in a vision for a new workflow. We published and presented these results at the Mensch und Computer 2021.

2. Prototype and evaluate mobile ways to request clinical treatment

We wanted to develop and evaluate a prototype to support the new workflow. Thus, we refined our vision into small storyboards detailing the new workflow and gathered feedback from physicians. It became apparent that two prototypes would be needed: A mobile app to support physicians’ workflow to request treatment directly when they are with the patient and a desktop app to support radiologists and students in their practical year. To achieve this, we developed and tested prototypes for both a mobile and a desktop app iteratively. Each prototype was tested with three fidelities: as a paper, wireframe, and digital prototype.

Overall, 46 participants tested the mobile prototype, and 44 participants tested the desktop prototype. Each participant was involved in the process of requesting and managing clinical treatments. Our results indicate that both prototypes achieve excellent usability while causing a low task load. Qualitative feedback was overwhelmingly positive. Afterward, we created a specification document, which we send to the Service Centrum Medical Informatics (SMI) for implementation. We presented these results at the university clinic’s GBL meeting and the CHI Conference on Human Factors in 2022.

We then focused on one task of the process: requesting treatment for patients. For this, we evaluated three ways to support a mobile workflow:

Our results show that letting physicians freely dictate is faster, has higher usability, and generates less task load. It, however, resulted in more errors and worse quality of the request. However, we found that participants preferred the structured user interface more since it allowed them to correct their mistakes.

We thus decided to develop a prototype that supported a combined way to request treatment: First, physicians can freely dictate their request. This request is then analyzed by an AI created by our partner group of Prof. Puppe. We then show the result of this analysis in a structured way and let physicians correct their request before sending it. In theory, this should increase the quality of the request and reduce errors. This user interface is already available for surgeons to test and was since iteratively improved upon. A formal evaluation of this user interface with both surgeons and medical students is planned and will be performed in 2023.

3. Evaluate how AR and VR can be used in clinical treatment

Finally, we started to investigate how to use augmented and virtual reality to support physicians during their daily work. To this end, we conducted a focus group with physicians from different departments. We found that physicians prefer augmented reality over virtual reality since it allows them to connect with their patients. The optimal device to use in a clinical context should be feature-rich but as little intrusive as possible. The tasks that physicians would most benefit from include requesting treatment, documenting wounds, providing support during and before surgery, and use in teaching.

We started the development of a digital assistant in augmented reality, which will support physicians during ward rounds. This digital assistant will be embodied and be able to request treatment, document wounds, and answer questions. We will then evaluate it together with surgeons from the university clinic.

Funding and Collaboration

Team

Publications

Legal Information