CHI+MED logo banner

Infusion Pumps: Human-Computer Interaction for Medical Devices

Key Points

Working with the US regulator, the Food and Drug Administration (FDA), we have:

  • Identified previously unnoticed design issues in the data entry system of commercial infusion pumps
  • Investigated how infusion pumps are actually used in the real work environment
  • Developed specialised tools that can be used within a regulatory framework to analyse the user interface of existing devices, and design a new generation of safer and more usable medical devices


Human-Computer Interaction (HCI) is the discipline of design and evaluation of interactive systems. It combines theoretical and applied perspectives with the aim to reduce the likelihood and consequences of use errors, which is one of the top causes of patient harm in healthcare.

Within CHI+MED, we pursued multiple complementary research themes to learn more about medical devices and how people design, buy and use them in the real world. Our research benefits from our continuous close collaboration with patients and their carers, nurses and other medical practitioners, manufacturers who create medical devices, NHS staff who purchase them, and regulatory bodies who oversee patient safety. The main CHI+MED research theme include:

  • Improving Data Entry. We have investigated a range of different use errors that can occur during data entry tasks, such as filling out patient records, and programming infusion pumps.
  • Understanding Normal Practice. We have investigated how real device users perform tasks in real-world situations, and how use error can be avoided on a daily basis.
  • Tools for safe usable devices. We have developed specialised tools and techniques to analyse usability and safety aspects of existing devices, as well as to support design and procurement of new medical devices.
  • Moving from blame culture to learning culture. We have investigated the public perceptions of human error with the aim to better understand how to develop constructive learning attitudes in healthcare systems.


We developed:

  • Tools for early evaluation of the safety and usability of data entry software, and validated them using detailed realistic models of infusion pump software.
  • Tools for rapid prototyping of user interface software based on formal (mathematical) models, and techniques for mechanised verification of medical user interface software against safety and usability requirements.
  • Training material that can be used by hospitals to improve their procurement decisions.
  • Case studies demonstrating how medical devices are used in the real world by users.
  • Guidelines that can help developers to identify important contributing factors that facilitate procedural tasks and decrease the likelihood of mistakes.
  • Case studies exploring how hospitals report adverse events, and ways to improve error reporting systems.