+ Scalable Conversation Measurement
Understanding healthcare communication requires scalable, meaningful and unbiased measures that tend to what actually happens during clinical interactions and how persons experience those conversations. Our curious, persistent, and creative team brings a vibrant interdisciplinary perspective to advance this science.
What actually happens? Advances in natural language processing (NLP) and machine learning (ML) methods offer extraordinary opportunities to analyze complex clinical conversations in real time while preserving confidentiality. Examples of our work include discovery of conversational story arcs, Markov Modeling of conversational information flow, self-organizing maps for intuitive clustering of conversation feature patterns, and supervised machine learning pipelines for identifying and studying conversational pauses.
How is communication experienced? Our team developed an epidemiological field measure for point-of-care use in the clinical environment that has become a national standard of patient experience in serious illness care. The “Heard & Understood” measure has been adapted for the outpatient setting, undergone national standardization among more than 3,500 seriously ill persons, translated into multiple languages, and received official endorsement by the National Quality Forum in 2022 for use in healthcare reform payment incentives.
Program for the Study of Online, Collective Knowledge and Stories Research Award
Gramling, Reblin, Rizzo, Tarbi; National Science Foundation (EPSCoR); 2023-2028
In a way that telescopes and microscopes have done for the physical and life sciences, the purpose of this project is to powerfully advance the computational social sciences for large-scale measurement of narrative in social media and inter-personal healthcare communication.
Deep Neural Networks to Identify Moments of Human Connection in Cancer Communication
Gramling (PI), Rizzo (PI); National Institutes of Health (R21); 2024-2026
Using multi-site cohort data of naturally occurring inpatient palliative care consultations, we apply state-of-the-art artificial intelligence methods (Deep Learning Neural Networks) among more than 12,000 minutes of recorded conversation to automatically identify conversational pauses that are indicative of human connection.
Detection of Cognitive Impairment in Serious Illness Conversations with Dialysis Patients
Cheung (PI), Gramling, Rizzo; NNE CTR Research Scholar Award (U54 GM115516); 2023-2024
The purpose of this award is to examine the feasibility of computational linguistic methods (lexicon and prosody) to identify persons with kidney disease who are experiencing cognitive impairment from content of naturally occurring clinical conversations.
Expression of Awe & Wonder
Manetta (PI), Tarbi (PI), Agrawal, Bhatia, Gramling, Reblin, Rizzo, Thambi; Kate Laud Research Fund
This purpose of this work is to identify the audible and visual markers of awe and wonder expression in psilocybin assisted therapy. Using linguistic, ethnographic and machine learning approaches, we will define a provisional taxonomy of feature patterns to guide further development of scalable coding systems.
+ Caregiving
Informal family caregivers provide key support to those with serious illness and are critical to the healthcare system. Our work focuses on the social context of family caregiving. Specifically, we focus on understanding interpersonal communication between caregivers, patients, and healthcare providers and developing interventions to enhance caregiver support from both informal social networks and formal services.
CARING study in Neuro-Oncology
PI: Reblin; National Cancer Institute (R01)
This project tests a social support intervention for caregivers of patients with brain tumors. The intervention includes both a web-based tool to organize and visualize social support resources, as well as manualized caregiver navigator sessions, focused on leveraging existing support services. Through this intervention, we hope to help caregivers more effectively access the support that they need, to improve both their own physical and mental well-being, and that of those they care for.
Related projects, funded by the UVMMC Auxiliary (PI: Reblin) and the Victoria Buffum Foundation (PI: Reblin) will expand this work to identify informal and formal resources for rural cancer caregivers in Vermont.
The Home Interaction Study
PI: Reblin; American Cancer Society (Mentored Research Scholar Award)
This study focuses on how couples dealing with a cancer diagnosis interacted at home as well as in directed conversations. Enrollment in this work is complete and analyses are ongoing. Our findings thus far identify that cancer-related communication between couples happened rarely at home and, when it did, varied substantially in quantity and content. Additionally, we found that communication patterns differentially impact patients and their spouses.
SNaSI: Social Networks and Social Interactions
PI: Reblin
Characteristics of social support networks can impact the availability and quality of social support, a key factor in protecting and promoting well-being. This ongoing research focuses on identifying the structural characteristics of social networks among home hospice family caregivers, young adult cancer patients and their informal caregivers, and LGBTQ+ cancer patients and their informal caregivers.
+ Story Listening
Conversational storytelling is a near culturally ubiquitous way in which people find and share meaning about life experiences together. Single event conversational storytelling has demonstrated substantial benefits for people experiencing the loneliness that some life experiences can foster. The VCL has developed a series of studies that offer these opportunities for participants to share their stories with a StoryListening Doula. A StoryListening Doula focuses specifically on inviting and listening to experiences, as perceived and narrated by participants. Doulas offer an engaged, neutral, and accepting presence, and validate through attentive listening.
StoryListening Study Home Page
The Grief StoryListening Project
PI: Gramling: The Kate Laud Story Fund and the Holly & Bob Miller Chair in Palliative Medicine
Managing social distancing and visitor restrictions during the COVID pandemic has placed substantial strain on clinicians and families caring for people who are ill and dying. These experiences can be isolating and distressing for clinicians and families who are grieving. This study focuses on understanding the aspects of the storytelling experience that are most beneficial to quality-of-life.
The Healthcare Professional StoryListening Project
PI: Gramling: The UVMMC Auxiliary
Healthcare professionals and trainees are experiencing alarming levels of burnout and demoralization. A major element of each is existential loneliness – the loss of connection to other people, to meaning, to one’s purpose and identity. This study focuses on understanding the aspects of the storytelling experience that are most beneficial for reducing existential loneliness among clinicians, healthcare staff and healthcare learners.
+ Psychedelic Assisted Therapy
The science of Psychedelic Assisted Therapy is advancing rapidly. Accumulating evidence is very promising for ameliorating human suffering of many types, including end-of-life distress. Psychedelic Assisted Therapy (PAT) involves three components that usually happen over a period of weeks: Preparation, Dosing, Integration. In each of these stages, interpersonal communication between therapist (or guide) and participant (or patient) is critically important to create a safe and effective “therapeutic container” for the participant’s experience. Our work focuses on understanding this guide-participant communication during all stages of PAT with a particular interest in the diversity of ways in which human presence and connection support participant well-being.
Human Connection in Psilocybin Assisted Therapy
PI: Gramling/Rizzo; Kate Laud Fund & Holly & Bob Miller Endowment
Our team is working to develop valid, meaningful, reliable and, ultimately, scalable approaches to measuring human connection in Psilocybin Assisted Therapy. We focus on two core concepts of particular importance to therapist training in PAT: interpersonal grounding and empathic abiding presence.
+ TeleHealth Conversations
On the way!