How Value Co-Creation Enable AI Video Interventions in Healthcare

How to cite

Li, Zhensheng. 2026. "How Value Co-Creation Enable AI Video Interventions in Healthcare." AEA RCT Registry. AEARCTR-0018265. https://doi.org/10.1257/rct.18265-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Zhensheng Li*

* Primary investigator

Registry ID
AEARCTR-0018265
DOI
https://doi.org/10.1257/rct.18265-1.1
First registered
Last updated
Status
On going
Trial dates
2026-03-01 – 2026-06-30
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/18265

Abstract

This study focuses on the impact of healthcare short-form videos on healthcare quality. It aims to design AI-generated short videos and a patient-customized AI video intervention model for healthcare delivery. We propose an intervention-based field experiment conducted in partnership with a major hospital in China and a leading AI technology firm to evaluate the impact of different interventions on health information dissemination and patient recovery outcomes.

Experimental design

This study will propose an intervention-based field experiment. Specifically, we will recruit chronic disease patients and randomly assign them to one of three experimental conditions: (1) Standard Care Control Group: Patients receive conventional follow-up care through routine clinical consultations during their recovery process, serving as the baseline comparator. (2) AI-Generated Video Group: Practitioners leverage AI video tools to generate personalized, high-fidelity digital avatars for content creation, which is then disseminated to patients. (4) AI-Co-Created Video Group: Patients actively participate in the content creation process by interacting with a video interface, enabling them to co-create AI-generated short videos with their practitioners based on individual preferences (e.g., avatar appearance, communication style). Following a one-month intervention period, we will assess specific indicators related to short-video dissemination and patient recovery.

Design details

Randomization method
The subjects were stratified and grouped by computer based on their characteristics.
Randomization unit
Individual
Clusters
The subjects will be randomly divided into 4 groups to ensure that their basic characteristics have no significant impact on the main results.
Observations
180-240 individuals.
Arms
Standard Care Control Group: 60-80 individuals. AI-Generated Video Group: 60-80 individuals. AI-Co-Created Video Group: 60-80.
Primary outcomes
Following a one-month intervention period, we will assess two categories of outcome measures. First, to capture the impact of AI-generated video interventions on operational outcomes in healthcare delivery, we will measure patient recovery status, operationalized as the 30-day hospital readmission rate. Second, to evaluate the effect of AI-enabled value co-creation on health information dissemination, we will assess patient engagement levels, measured through video view and sharing metrics, as well as the adoption rate of the intervention protocol.

Cited by

No citation data available for this registration yet.

Keywords