JMIR Publications

ABSTRACT

Background:

Collecting information on the medical history of a patient is an important step during the diagnosing process. Besides the interrogation by the physician, computerized questionnaires are used to collect the data. To facilitate interaction, implementation of digital medical interview assistants (DMIA) using conversational user interfaces (CUI) gain in interest.

Objective:

The aim of this research is to assess patient’s and physician’s perceptions towards a DMIA with CUI. Beyond, we want to understand how such DMIA can be used in real-world context, what issues and barriers exist in their usage.

Methods:

We developed a web-based DMIA with CUI (referred to as AnCha for anamnesis chatbot) as a research prototype in a participative and iterative development process. We conducted a pilot trial in a practice for general medicine. Patient perceptions were collected and physicians were interrogated regarding usefulness of collected information.

Results:

31 patients were approached, and 9 participants were included in the pilot trial; 3 conversation protocols were used by the physicians to prepare for the encounter. Participants spanned all age groups from digital natives (n=5), and digital workers (n=3) to digital seniors (n=1). Patients can easily interact with AnCha and are willing to provide information to the digital tool. They recognize benefits while using the dialog system compared to the existing process. Important insights into practical implementation and integration into practice workflows could be gained.

Conclusions:

Providing information on complaints and medical history before the actual encounter is considered useful. In order to be supportive for physicians, information has to be made available in a sufficient time frame before the encounter. Future work has to assess in particular whether AnCha is also well accessible for digital seniors.