Evaluating an AI-based nutrition expert platform delivered via SMS-text to support patients with cancer.

Authors

null

Marissa Lubin Buchan

Savor Health LLC, New York, NY

Marissa Lubin Buchan , Hao Tang , Susan Bratton , Sridevi Padmanabhan , David Raben

Organizations

Savor Health LLC, New York, NY, Teachers College, Columbia University, New York, NY, Department of Radiation Oncology, Aurora, CO

Research Funding

No funding received

Background: Interventions incorporating nutritional strategies can prevent and manage nutrition-related symptoms, however, due to a shortage of oncology dietitians (RD CSO) combined with healthcare access disparities, many patients do not receive the support they need, resulting in poor outcomes and quality of life (QoL). High rates of cell phone utilization among all demographics offers a unique opportunity to provide interventions using mobile technology. Methods: Launched to select groups in 2019, Ina (trademarked by Savor Health LLC)is a virtual nutrition assistant powered by an artificial intelligence (AI)-enabled expert platform, to facilitate self-management of cancer treatment side effects. The platform combines evidence, expertise, and unique patient data to deliver the personalized guidance patients would receive from an RD CSO, “on-demand” via text. This study applied the RE-AIM framework to evaluate the intervention from five dimensions. Results: Reach: The program reached 1,706 users as of 2021. Based on self-report among 1209 users, 78% are patients, 18% are caregivers, and 5% are healthcare professionals with 66% female and 34% male. Ina supports all cancer types, and top diagnoses among users are genitourinary (28%), lung (20%), gynecologic (16%), gastrointestinal (15%), and breast (11%). Disease burden is high, with 62% of users reporting they are experiencing nutrition-addressable symptoms at any given time. Effectiveness: Based on weekly patient reported outcome (PRO) surveys, 87% of respondents report Ina helps them manage symptoms, and 80% report that using Ina has improved their QoL. The cumulative likelihood to recommend is 4.1 on a 5-point scale. Adoption: Users are from over 18 cancer organizations, proving the feasibility and accessibility of this intervention. Implementation: Ina is accessible 24/7 via text, and is offered to patients in a B2B2C model. Users typically receive responses to nutrition-related questions within seconds from the AI platform (or 1-3 minutes with live RD oversight). Each active user (one that engages at least once in a given month) asks an average of 2.8 questions per month. Maintenance: The median time users remain on the platform is 156 days (range 0-884) and 57% of evaluable users stayed on the platform 6 months or longer. The platform’s database, which includes over 54,000 referenced interventions, grows daily via supervised machine learning supported by RD CSOs. When surveyed on future feature development, respondents are most interested in tele-nutrition (57%), mental health (65%), and stress management (62%). Conclusions: Initial data suggests this is a feasible and accessible tool to support cancer patients’ unique nutrition and symptom-management needs. Clinical trials are needed to validate feasibility and assess impact on clinical and QoL outcomes. Product development to integrate language and cultural preferences is ongoing.

Disclaimer

This material on this page is ©2024 American Society of Clinical Oncology, all rights reserved. Licensing available upon request. For more information, please contact licensing@asco.org

Abstract Details

Meeting

2022 ASCO Annual Meeting

Session Type

Poster Session

Session Title

Care Delivery and Regulatory Policy

Track

Care Delivery and Quality Care

Sub Track

Digital Technology

Citation

J Clin Oncol 40, 2022 (suppl 16; abstr 1569)

DOI

10.1200/JCO.2022.40.16_suppl.1569

Abstract #

1569

Poster Bd #

161

Abstract Disclosures