A Digital Serious Illness Conversation Coach - PROJECT SUMMARY
It has been shown that conversations on end-of-life care focused on a patient’s goals of care
can reduce psychological distress and anxiety and lead to improved quality of life, and
potentially prolonged life, for patients living with serious illness. The Serious Illness
Conversation Guide (SICG) is a validated and widely used framework for training the critical
communication skills needed to effectively conduct these conversations; however, there
currently exists no mechanism to easily support and continue to build these skills in a scalable
way. Evidence suggests that conversational agents, i.e., chatbots, can be used for training and
reinforcing communication skills. In this Phase I SBIR project, we propose to optimize and
assess the feasibility and acceptability of a chatbot designed to augment SICG training for
healthcare professionals (HCPs) on initiating and conducting sensitive conversations with
patients diagnosed with serious illness. We will develop and test a chatbot directly with HCPs
who are navigating serious illness conversations and/or may be expected to in the future. We
will demonstrate feasibility and acceptability of an enhanced chatbot prototype through the
following two specific aims: In Aim 1, we will create and adapt chatbot content and build an
enhanced, fully functional prototype of the chatbot with expert inputs. The chatbot will feature 3
different patient scenarios ranging in case complexity. In Aim 2, we will conduct a pilot usability
study of the prototype with 50 target end users—a diverse sample of HCPs varying in role,
specialty, and career phase who have received at least one training session in the SICG.
Following enrollment, HCPs will independently test the chatbot-led training and complete
surveys at 3 timepoints—baseline, endline, and 1-month after enrollment—to assess the
following outcome measures: pre-/post-training confidence in having serious illness
conversations, usability, and user satisfaction. Semi-structured endline interviews will be
conducted with a purposive sample of 15-20 HCPs to collect additional qualitative feedback on
user experience and recommendations for improvement. Results of this Phase I SBIR project
will inform a larger scale implementation study in potential Phase II work to evaluate the
effectiveness and cost-effectiveness of the proposed intervention to reinforce and augment
critical communication skills essential to the care of seriously ill patients.