Learning from tragedy

by Naomi Cherne, PhD

A recent fatal tragedy at an Indian MRI facility in which an oxygen tank was allowed to be brought into an MRI room highlights the steep consequences of breakdowns in communication in the context of healthcare systems. The facts of the case are under investigation, but the initial statements reported in the press from the laypeople (i.e., non professionals) involved in the tragedy call to mind several usability principles: the importance of understanding mental models and user groups, the provision of safety information at point of use, and learning from past events.

Mental models drive expectations. The laypeople apparently expected that the magnet needed to be turned on in order to work. However, MRI magnets are always on; when an MRI magnet is stopped, helium is released in large quantities and fills the room, and the magnet can be irreparably damaged by the stopping process. Magnets are stopped only in an emergency such as a fire in the scan room or someone actively pinned by a ferromagnetic object. What does start and stop is the pulse sequence used to image the tissue in the magnet’s bore.

Understand your user groups. The laypeople describe being told that the oxygen tank could not be left outside the MRI room, and that they were not stopped when entering the room with the oxygen tank. What does a layperson know about the risks inherent to a product, safety procedures to mitigate these risks, and appropriate safety behavior?

Provide safety information at the point of use. If risks cannot be mitigated by design, then procedures, guarding, and information can be implemented as mitigations. The laypeople state that, in addition to not being stopped from entering the MRI room with the tank, they were later scolded for getting too close to the MRI machine with the metal oxygen tank. What can come into the room? How close is too close?

Learn from the past. This is not the first time that an oxygen tank in an MRI room has caused injury and death. Although rare, high profile accidents have occurred since the early 2000s. One method of addressing potential usability issues for a product is to research the accident history of that product and similar products. Some potential sources of such historical information include news items, product reviews, safety databases such as MAUDE, and internal postmarket data such as customer service or helpline records.

In this specific tragedy, the facts remain to be established. Something clearly went catastrophically wrong, and that means something critical needs to be fixed. It could be a small miscommunication, it could be a systemic safety issue. Regardless, responsible MRI facilities around the world are now turning a sharp and horrified eye to their systems, staff, and safety. Monitoring real-world relevant usability failures, applying usability principles during design, and testing the usability of designs can help improve the safe use of products.

• You can discover incorrect mental models your users bring to their use of your product, and if these have implications for safety, by conducting formative usability testing.
• You can discover what your users understand and expect about the risks of your product by conducting user interviews and simulated-use usability testing.
• You can assess the effectiveness of your procedures, guards, and safety information early in design via expert review and user interviews, and ultimately demonstrate intended behavior through simulated-use usability testing.
• You can familiarize yourself with known use events for a product category to guide the development of new products.