Training: Learning to recognize delirium with the android patient

Current patient simulators are mostly mannequins that can simulate vital functions via built-in electronics but are otherwise lifeless. An android patient could bring a breath of fresh air to training in the future. It can help hospital staff to better recognize delirium in patients and thus increase patient safety.

In this interview with MEDICA-tradefair.com, Dr. Ulf Günther talks about the development of the android patient and explains how it can be used in training.

Dr. Günther, you received the "Patient Safety" Research Award of the German Interdisciplinary Association for Intensive Care and Emergency Medicine (DIVI) and Philips for the development of an android patient. What is the idea behind your project?

Dr. Ulf Günther: We train medical staff in hospitals and intensive care units to identify states of acute confusion (delirium). We use various tools in this setting and show films or use an actress who performs signs of delirium. We contemplated whether a robot could perform this training and show participants the different types of delirium in a standardized way.

How can we envision this robot? Does it resemble a typical patient simulator?

Günther: It is similar. Trauma resuscitation or anesthesia training programs use puppets that can simulate ECG vital signs, but these tools do not have to move because they mimic intubated patients. However, for our specific purposes, our simulator must be able to blink its eyes, nod, or squeeze someone’s hand, which is why we incorporated motor skills. This turned the simulator into a robot.

How does a delirium screening test work?

Günther: You perform an attention test, for example. This also works for intubated patients. In this case, you read out a series of ten letters. Each time the patient hears the letter A, she or he must squeeze your hand. Patients must not squeeze your hand for other letters. The test requires patients to follow simple instructions such as holding up two or three fingers. Or you ask "yes" or "no" questions. We have added these capabilities to the simulator.

Are there already insights into whether and how the robot could improve training?

Günther: We conducted a study and just submitted it for publication. As part of the pilot program, we trained a small group of people. Half of the group was trained with the help of an actress, the other half by means of the robot. We subsequently compared how the test subjects picked up the learning content and found that training with the robot was not inferior to training with an actress.

What are your next steps?

Günther: The robot is currently still missing its right arm, which we need for the test. We also want to hone its motor skills and add more restlessness to the torso, which is something a person with delirium may exhibit. For now, the robot is controlled by an operator, but our goal is for it to become increasingly autonomous and able to operate independently.

We would like to take the system from prototype to product because we envision more applications aside from being a screening instrument for delirium. Autonomous robots have a bright future as a tool to train medical students and hospital care team members.