AI-enabled Communication to Advance Research Engagement (AI-CARE)

NCT07788898 | Not Yet Recruiting

• 2 other identifiers: 2026-17824UM1TR004400
• Study Type: interventional
• Target: 250
• Locations: 1 country
• Sites: 1

Brief Summary

The objective of AI-CARE is to train researchers and community members to use artificial intelligence (AI) to create plain-language summaries of research studies and to determine whether a co-creation process for developing these summaries improves understanding and relevance of the research, confidence in AI and in institutions, and AI literacy. The hypothesis is that AI-generated research summaries, co-developed with researchers and community partners, will significantly improve confidence in research and institutions, understanding and accessibility of the research, and appropriateness of the summary compared with standard researcher-generated materials, and that the educational and co-creation process will build community members' AI literacy and confidence in using AI, as well as researchers' readiness to use and comfort with using AI for communicating research.

Conditions

AI Literacy; Research Communication

Keywords

AI Literacy; Research Communication

Outcome Measures

Primary Outcomes (2)

• Confidence in medical research and researchers - Aim 3

  This is the primary outcome measure for both Aim 2 and Aim 3. This outcome measure will be assessed using subscale 1 of the PoRT scale. PoRT is a validated, self-report measure that assesses confidence in medical research and researchers. Subscale 1 contains 9 items, with response options for each item ranging from 1 (strongly disagree) to 5 (strongly agree); thus the subscale ranges from 9 to 45. Higher scores indicate higher confidence.

  Immediately following review of research summary, up to 1 hour

• Confidence in medical research and researchers - Aim 2

  This is the primary outcome measure for both Aim 2 and Aim 3. This outcome measure will be assessed using subscale 1 of the PoRT scale. PoRT is a validated, self-report measure that assesses confidence in medical research and researchers. Subscale 1 contains 9 items, with response options for each item ranging from 1 (strongly disagree) to 5 (strongly agree); thus the subscale ranges from 9 to 45. Higher scores indicate higher confidence.

  Immediately before the co-creation process and immediately following the co-creation process, up to 1 hour

Secondary Outcomes (6)

• Perceived appropriateness of research summary - Aim 2

  Immediately before the co-creation process and immediately following the co-creation process, up to 1 hour

• Institutional confidence - Aim 2

  Immediately before the co-creation process and immediately following the co-creation process, up to 1 hour

• Research comprehension - Aim 2

  Immediately before the co-creation process and immediately following the co-creation process, up to 1 hour

• Perceived appropriateness of research summary - Aim 3

  Immediately following review of research summary, up to 1 hour

• Institutional confidence - Aim 3

  Aim 3: Immediately following review of research summary, up to 1 hour

Study Arms (4)

Standard AI prompt - Aim 3

EXPERIMENTAL

Participants receive plain-language research summaries generated using a standardized AI prompt without researcher prompt engineering or community co-design.

Behavioral: Experimental: Standard AI prompt

Researcher prompt engineering - Aim 3

EXPERIMENTAL

Participants receive summaries generated after researchers complete AI prompt-engineering training and develop optimized prompts.

Behavioral: Experimental: Researcher prompt engineering

Community co-created AI summaries - Aim 3

EXPERIMENTAL

Participants receive summaries developed through an iterative community-researcher co-creation process following AI training and structured co-design sessions.

Behavioral: Community co-created AI summaries

Co-creation cohort - Aim 2

EXPERIMENTAL

Participants will be divided into three groups, each assigned 1-2 research summaries, and will engage in structured co-creation sessions. The goals of will be to: 1) define readability and appropriateness, 2) review AI-generated drafts produced by the summary engine, and 3) provide feedback to refine prompts and outputs. Several iteration cycles will be conducted as needed, with documentation of any changes and the rationale for making decisions. In this way the team will develop a modular AI "summary engine" that transforms research products (e.g., manuscripts) into draft community-facing materials using co-created prompts and templates. Co-creation will be conducted during one or two facilitated sessions. Each session will include structured review of drafts, guided discussion, and documented decisions regarding readability, transparency, visuals or graphics, etc. Iteration cycles will continue until pre-specified stopping criteria are met.

Behavioral: AI literacy training and co-design

Interventions

Community co-created AI summaries BEHAVIORAL

Participants review AI-generated summaries refined through iterative collaboration between researchers and community members using community-engaged co-design principles.

Community co-created AI summaries - Aim 3

Experimental: Researcher prompt engineering BEHAVIORAL

Participants review AI-generated summaries generated after researchers complete AI prompt-engineering training and develop optimized prompts.

Researcher prompt engineering - Aim 3

Experimental: Standard AI prompt BEHAVIORAL

Participants review plain-language research summaries generated using a standardized AI prompt without researcher prompt engineering or community co-design.

Standard AI prompt - Aim 3

AI literacy training and co-design BEHAVIORAL

The co-designed summaries which result from this process will become one of the three intervention conditions evaluated in Aim 3.

Co-creation cohort - Aim 2

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: Yes
• Age Groups: Adult (18-64), Older Adult (65+)

You may qualify if:

• English-speaking individuals at least 18 years of age

You may not qualify if:

• Individuals unable to provide informed consent
• Participants who have had formal training in AI prompt engineering
• The eligibility criteria are common to both Aims 2 and 3.

Sponsors & Collaborators

• Albert Einstein College of Medicine: lead
• National Center for Advancing Translational Sciences (NCATS): collaborator

Study Sites (1)

Albert Einstein College of Medicine

The Bronx, New York, 10461, United States

Location

Related Publications (6)

• Shoemaker SJ, Wolf MS, Brach C. Development of the Patient Education Materials Assessment Tool (PEMAT): a new measure of understandability and actionability for print and audiovisual patient information. Patient Educ Couns. 2014 Sep;96(3):395-403. doi: 10.1016/j.pec.2014.05.027. Epub 2014 Jun 12.

  PMID: 24973195 BACKGROUND

• Snyder HR, Miyake A, Hankin BL. Advancing understanding of executive function impairments and psychopathology: bridging the gap between clinical and cognitive approaches. Front Psychol. 2015 Mar 26;6:328. doi: 10.3389/fpsyg.2015.00328. eCollection 2015.

  PMID: 25859234 BACKGROUND

• Marafino BJ, Dudley RA, Shah NH, Chen JH. Accurate and interpretable intensive care risk adjustment for fused clinical data with generalized additive models. AMIA Jt Summits Transl Sci Proc. 2018 May 18;2017:166-175. eCollection 2018.

  PMID: 29888065 BACKGROUND

• O'Connor MA, Erasmus JH, Randall S, Archer J, Lewis TB, Brown B, Fredericks M, Groenier S, Iwayama N, Ahrens C, Garrison W, Wangari S, Guerriero KA, Fuller DH. A Single Dose SARS-CoV-2 Replicon RNA Vaccine Induces Cellular and Humoral Immune Responses in Simian Immunodeficiency Virus Infected and Uninfected Pigtail Macaques. Front Immunol. 2021 Dec 21;12:800723. doi: 10.3389/fimmu.2021.800723. eCollection 2021.

  PMID: 34992610 BACKGROUND

• Kreuter MW, Wray RJ. Tailored and targeted health communication: strategies for enhancing information relevance. Am J Health Behav. 2003 Nov-Dec;27 Suppl 3:S227-32. doi: 10.5993/ajhb.27.1.s3.6.

  PMID: 14672383 BACKGROUND

• Holzer JK, Ellis L, Merritt MW. Why we need community engagement in medical research. J Investig Med. 2014 Aug;62(6):851-5. doi: 10.1097/JIM.0000000000000097.

  PMID: 24979468 BACKGROUND

Study Officials

• Jessica Kahn, MD, MPH

  Albert Einstein College of Medicine

  PRINCIPAL INVESTIGATOR

Central Study Contacts

Jessica Kahn, MD, MPH

CONTACT

718-430-4317; jessica.kahn@einsteinmd.edu

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: RANDOMIZED
• Masking: DOUBLE
• Who Masked: PARTICIPANT, OUTCOMES ASSESSOR
• Masking Details: Aim 2: N/A Aim 3: Participants are blinded to how the research summaries were developed. The summaries are presented in identical formats without identifying the development method. Study personnel performing the primary analyses remain blinded to treatment assignment until completion of the primary analyses.
• Purpose: OTHER
• Intervention Model: PARALLEL
• Sponsor Type: OTHER
• Responsible Party: SPONSOR

Model Details: Aim 2: Single group: participants are prospectively assigned to an intervention (AI literacy training and co-design) to evaluate the effect of the intervention on health-related behavioral outcomes. Because only one response is allowed for allocation and number of participants, please note that allocation is N/A and number of participants is 40. Aim 3: Parallel: participants are randomized 1:1:1 to one of three summary types to compare the effect of the intervention on health-related behavioral outcomes. Because only one response is allowed for allocation and number of participants, please note that allocation is randomized, and number of participants is 210.

Study Record Dates

• First Submitted: August 24, 2026
• First Posted: August 27, 2026
• Study Start (Estimated): September 1, 2026
• Primary Completion (Estimated): September 1, 2028
• Study Completion (Estimated): September 1, 2028
• Last Updated: August 27, 2026

Record last verified: 2026-08

Locations

US (1)