Language Translation of Knowledge Mobilization Resources: A Randomized Trial
试验速览
- 阶段
- 不适用
- 状态
- 已完成
- 入组人数
- 322
- 试验地点
- 1
- 主要终点
- Understanding
研究概览
简要总结
Knowledge mobilization (KM) resources are tools designed to facilitate the use of research evidence in healthcare decision-making. These resources are created in various formats - including plain language summaries, infographics, and videos - to meet the needs of diverse end-users, such as healthcare professionals, policymakers, patients, and caregivers. They are intended to be easily accessible; however, individuals whose first language is not English may have difficulty understanding them. Thus, translating KM resources into other languages is essential to support health equity and accessibility, but it is often costly and time intensive.
This study aims to explore whether artificial intelligence (AI) tools, specifically ChatGPT - an AI-based large language model developed by OpenAI - can effectively translate KM resources for members of the public whose first language is not English. The resource being evaluated offers guidance on preventing post-COVID-19 condition and has already been translated by a professional (human) translator into seven languages commonly spoken in Canada: French, Spanish, Ukrainian, Tagalog, Arabic, Chinese, and Punjabi. Using ChatGPT, AI-generated translations will be created in those same seven languages.
For this study, participants - adults living in Canada whose first language is one of the selected languages and able to read English - will be randomly assigned to review either an AI-generated or a professionally translated version of a KM resource. They will then complete a questionnaire evaluating their understanding of the resource, as well as the readability and acceptability of the translation.
This study will contribute to the Investigators' understanding of the potential use of AI for translating health information. The goal is to support equitable access to health information and promote citizen-centered care by reducing language barriers using innovative solutions.
详细描述
Background and Rationale:
KM resources play a critical role in bridging the gap between research and practice in healthcare. Effective communication of this information is essential to ensure the uptake and implementation of research evidence and health recommendations by the public.
Canada is home to a highly diverse population. Language proficiency remains a persistent challenge for many newcomers. Inadequate health communication stemming from language barriers can lead to poorer health outcomes, reduced satisfaction with care, and increased healthcare inequalities. High-quality translations of health materials are therefore essential to promoting equitable access to health information and helping reduce disparities in health outcomes. While professional translation services have long been the standard for producing linguistically accurate translated materials, they are often resource-intensive, limiting the scalability and timeliness of current translation efforts.
Recent advances in AI have generated interest in leveraging AI-powered tools to support the translation of health materials. However, concerns remain regarding the ability of AI powered tools to navigate the cultural, contextual, and emotional nuances of language that are vital in healthcare communication. No controlled trials have yet compared AI and professional translations specifically for KM resources aimed at the public. Empirical assessments of AI-generated translations are therefore essential for understanding its capabilities and limitations.
Objectives:
研究设计
- 研究类型
- Interventional
- 分配方式
- Randomized
- 干预模型
- Parallel
- 主要目的
- Other
- 盲法
- Double (Participant, Investigator)
盲法说明
Participants will be blinded to the type of translation they receive. They will be informed that the study compares different translation methods but will not be told whether the materials they view were translated by AI or by professional translators.
入排标准
- 年龄范围
- 18 Years 至 —（Adult, Older Adult）
- 性别
- All
- 接受健康志愿者
- 是
入选标准
- •• 18 years of age or older
- •Living in Canada
- •First language is one of the selected languages (i.e. French, Spanish, Ukrainian, Tagalog, Arabic, Chinese, and Punjabi)
- •Able to read and complete a questionnaire in English
- •Access to an electronic device (e.g. computer or tablet), Internet and email
排除标准
- •• Under the age of 18 years
- •Does not live in Canada
- •First language is not one of the selected languages (i.e. French, Spanish, Ukrainian, Tagalog, Arabic, Chinese, and Punjabi)
- •Unable to read and complete a questionnaire in English
- •No access to an electronic device (e.g. computer or tablet), Internet or email
研究组 & 干预措施
Group A - AI
ChatGPT
干预措施: KM Resource - AI (Other)
Group B - Professional
Human translator
干预措施: KM Resource - Professional (Other)
结局指标
主要结局
Understanding
时间窗: Baseline
Measured by participant responses to a set of 7 multiple-choice questions specifically designed to evaluate key content knowledge about post COVID-19 condition and related recommendations contained in the resource. The number of correct answers will serve as an objective indicator of understanding, assessing the ability of each translation method to preserve critical concepts.
次要结局
- Readability(Baseline)
- Acceptability(Baseline)