Virtual therapeutic assistants in outpatient aftercare - Expectations of patients and health care professionals: A survey (Preprint)

BACKGROUND To explore the expectations of patients with depression and healthcare professionals towards a socially interactive agent as a virtual therapeutic assistant in outpatient aftercare, a qualitative survey was conducted. OBJECTIVE The current analysis focus on research questions covered the appearance and role of the assistant, assistant-patient interaction (time of interaction, skills and abilities of the assistant, ways of interacting) and therapist-assistant interaction. METHODS A two-part qualitative study was conducted to explore the perspectives of the two groups (patients and care providers). As a first step, care providers (N=30) were recruited during a regional offline meeting. After a brief lecture, they were given a link and asked to take part in a semi-structured online questionnaire. Secondly patients (N=20) were recruited in a clinic and interviewed in a semi-structured face-to-face interview. RESULTS Findings included the following: The assistant should be a multimodal communicator (voice, facial expressions, gestures) and counteract negative self-evaluation. The gender of the assistant should be female or optional. However, only male was rejected due to negative associations. The assistant is seen as proactive and the patient as a passive recipient of information in terms of patient-assistant interaction. Gaps in aftercare can be filled by the assistant`s unlimited availability. However, the assistant should train the patient's autonomy to avoid dependency. Health status monitoring was viewed positively by both groups. In order to detect early warning signs of illness, a biofeedback function is desired. When appropriate to the situation, humor in the assistant is considered desirable. With regard to the desired skills of the assistant, these can be summarized in terms of providing structure and emotional support, especially warmth and competence in order to build trust. Consistency is important for the assistant to appear authentic. In terms of care providers-assisted therapist-assistant interaction, an objective measure of the patient's health status, an emergency system for suicide prevention, as well as an information tool and decision support for health care professionals (psychiatrists, psychotherapists, nurses and social workers) were identified as crucial. CONCLUSIONS Overall, the survey conducted provides innovative guidelines for developing a virtual therapeutic assistant to fill the gaps in patient aftercare.