Disparities in engagement with an artificial intelligence-based chat intervention post-appendectomy.

Publication/Presentation Date

9-1-2026

Abstract

PURPOSE: Artificial intelligence (AI)-based text messaging, or "chat," in post-appendectomy care has been shown to decrease preventable emergency department visits. As digital health grows, it is important to assess the impact of technology-based interventions across patients with diverse backgrounds and socioeconomic risk factors. The purpose of this study was to evaluate engagement with an AI-based texting platform for patients with risk factors related to social determinants of health (SDoH).

METHODS: An AI chat was designed with input from pediatric surgeons and specialty nurses to address postoperative concerns after appendectomy. It was implemented from March 2022 to December 2023. Caregivers or patients could respond to chats. Mann-Whitney U tests compared continuous variables, and Fisher's exact tests compared categorical variables. An ordinal logistic regression model was fit to determine which patient demographics and SDoH-related risk factors (area deprivation index [ADI] and four social vulnerability index [SVI] themes) were associated with chat engagement.

RESULTS: Of the1023 chat eligible post-appendectomy patients, 497 caregivers/patients did not engage, 229 caregiver/patient engaged with the chat once, and 297 engaged with more than one chat. Independent factors associated with decreasing engagement were increasing patient age (odds ratio, 95% confidence interval; 0.96 [0.93-0.99]), higher SVI socioeconomic score (0.28 [0.16-0.49]) and having Medicaid insurance (0.68 [0.52-0.88]). However, higher SVI housing & transportation score (2.08 [1.26-3.43]) was associated with increased engagement.

CONCLUSION: Lower chat engagement was observed among non-White patients and those with lower socioeconomic status. These are the same groups who may be at higher risk for complications. Future work should focus on optimizing feasibility and accessibility of technological interventions for patients with SDoH-related risk factors.

Volume

61

Issue

9

First Page

163238

Last Page

163238

ISSN

1531-5037

Disciplines

Business Administration, Management, and Operations | Health and Medical Administration | Management Sciences and Quantitative Methods

PubMedID

42276356

Department(s)

Administration and Leadership

Document Type

Article

Link to Full Text

Share

COinS