Thesis defense Roel Boumans (Donders series 443)

30 September 2020

Promotors: prof. dr. M. Olde Rikkert, prof. dr. M. Neerincx (Delft University of Technology), prof. dr. K. Hindriks (Vrije Universiteit Amsterdam)
Co-promotor: dr. F. van Meulen (Eindhoven University of Technology)

Feasibility and effectiveness of social robots in acquiring patient reported outcomes from older adults

Healthcare professionals in hospitals report to have a high workload. As a consequence, certain care tasks could be left undone, such as comforting or talking with patients, educating patients, and developing or updating nursing care plans. Social robots, being robots that are able to verbally communicate with humans, may contribute to the solution for this problem by supporting healthcare professionals in specific and automatable communication tasks such as health status interviews.
This research has resulted in the design of a human-robot interaction aimed at a pleasant and effective interview on a person’s health. We first investigated the acceptability of a robot in this interviewing role among community dwelling older adults, a large group of the potential hospital population. The robot scored an average subjective usability of 8 on a scale from 0-10. Then we studied the data agreement of health data acquired by the robot interviewer and a nurse from the same group. The conclusion was that a moderate to substantial agreement could be demonstrated, in other words the acquired data were comparable. Lastly we conducted a trial among actual patients at an outpatient clinic at two hospitals, the Radboudumc and the CWZ. The intervention consisted of a social robot autonomously interviewing the patient, as part of the patient’s hospital visit under supervision of a healthcare professional. The conclusion was that both patients and their informal caregivers found the quality of care given by robot supported healthcare professionals better than the quality of care given by healthcare professionals not supported by a robot.