CHI+MED: Multidisciplinary Computer-Human Interaction research for the design and safe use of interactive medical devices

Patient safety is a major concern. Reliance on interactive medical devices is growing, both in clinical settings and, increasingly, for patients without direct clinical supervision. The usability and reliability of such devices is critical. For example, decimal points are a well-known source of error (e.g., .5mg misread as 5mg), yet few devices detect decimal keying errors. Considering the broader context of use, a nurse familiar with one kind of infusion pump may absent-mindedly use the same set-up procedure on a similar one, leading to incorrect dosage. These and other user programming errors cause patient deaths. Even when devices are programmed correctly, interaction difficulties raise workload and stress, increasing overall system vulnerability. For example, it is common for clinicians to turn pumps off and on to reset the state, but patient data may be lost in the process and need to be re-entered. Better interaction design, the focus of CHI+MED, will improve safety by a scientific approach to understanding and designing out latent errors. It will have very broad impact because interactive systems are encroaching everywhere in healthcare. CHI+MED complements other initiatives on the ergonomics and engineering of safer devices.CHI+MED is multidisciplinary. A combination of empirical approaches, drawn from the disciplines of HCI, computer science, psychology and the social sciences, will be applied. Empirical findings will build on, be subjected to, and be challenged by precise computational representation and reasoning. This will support integration across perspectives and generalisation of findings, as well as enforcing a distinctive rigour in all aspects of analysis. The CHI+MED approach will advance, and transform, the science of user-device interactions and the design of interactive medical devices.We will investigate which device properties are most important from safety and usability perspectives (relating to whether adverse incidents can occur, whether there are unsafe inconsistencies between device behaviours under similar conditions, etc.), and which system descriptions best support reasoning about them. We will study the causes of errors when interacting with devices and how a better science of error can improve design. This will include factors affecting slips such as stress, attention, task structure and the design of cues and feedback. We will also study the broader situational factors that provoke or mitigate interaction errors (interruptions may provoke errors; colleagues checking may mitigate them, etc.). CHI+MED will deliver a significantly better understanding of interaction design and associated cognitive and situational factors, new tools to support design and evaluation, underlying better informed standards and incident investigations, and more resilient practices.As well as addressing new scientific questions that generalise beyond the medical domain, CHI+MED will develop a dialogue with stakeholders, to raise awareness of the role of interaction design in medical errors and to develop evidence-based techniques for designing medical devices that mitigate potential errors while accommodating the cultures and processes of stakeholder groups. It will draw on research in Public Engagement and Knowledge Transfer, and explore the appropriateness and effectiveness of different approaches for different contexts. We will investigate how informed principles and tools can help developers to design more reliable and usable devices, identify factors that influence design and procurement decisions, and explore how stakeholders can be better informed about interaction design.CHI+MED involves and integrates three leading groups on a single, focused problem. It will contribute to the UK's status as an international leader in research in multidisciplinary, rigorous HCI, and make significant and informed contributions to the international agenda on design for patient safety.