Example of Involving users to improve device safety

(Lin et al 1998)

One type of medical device that has been associated with risks to patient safety is infusion pumps. This is often due to poorly designed user interfaces which place excessive strain on the user during programming. This paper describes a user-focused study that redesigned an infusion pump with the aim of reducing errors with this device, and therefore reducing the risks to patient safety. The authors used Cognitive Task Analysis (see glossary for discussion of terms in bold) to identify the psychological requirements for programming the infusion device. This involved observations of the device in user during bench tests where expert users used the device to perform certain tasks. They found a number of problems with the user interface that resulted in a high cognitive load being placed on the user. These problems included the fact that no feedback was provided to users during programming, users had to remember what data had been entered and this resulted in errors. In addition, the user interface was found to require complex programming sequences and did not provide users with a way of detecting any errors made whilst inputting data or recovering from such errors. The authors concluded that using this device in a high pressure working environment such as an intensive care unit could result in errors which could have significant implications for the safety of patients. The device interface was re-designed to address the faults and the human factors approach was continued in the evaluation of the new interface. In order to evaluate whether the task analysis and subsequent re-design had been successful Usability Tests were performed to compare the new and old interfaces. A number of users who had no direct experience of either interface were asked to perform a programming task. The new interface was found to be faster to program (improved efficiency) and it also resulted in lower mental workload ratings and fewer errors (improved safety). In addition, each of the 12 users reported that they preferred the new interface. Although this study saw a significant improvement in the design of the interface the process of evaluation and re-design was performed only once. The authors do not mention whether any further iterations were considered, however this could have improved the interface still further. This study illustrates the importance of considering the environment in which the device will be used. The authors recognised that operating the device in a stressful environment such as an intensive care unit would increase the likelihood of errors. This provided an effective rationale for identifying the points in the task where the user required extra support and providing this in the design of the new device.