Department of Medical Science of Meridian, College of Korean Medicine, Kyung Hee University, Seoul, Republic of Korea.
Department of Meridian & Acupoint, College of Korean Medicine, Dong-eui University, Busan, Republic of Korea.
J Med Internet Res. 2023 May 26;25:e42238. doi: 10.2196/42238.
Obesity is a public health issue worldwide. Conversational agents (CAs), also frequently called chatbots, are computer programs that simulate dialogue between people. Owing to better accessibility, cost-effectiveness, personalization, and compassionate patient-centered treatments, CAs are expected to have the potential to provide sustainable lifestyle counseling for weight management.
This systematic review aimed to critically summarize and evaluate clinical studies on the effectiveness and feasibility of CAs with unconstrained natural language input for weight management.
PubMed, Embase, the Cochrane Library (CENTRAL), PsycINFO, and ACM Digital Library were searched up to December 2022. Studies were included if CAs were used for weight management and had a capability for unconstrained natural language input. No restrictions were imposed on study design, language, or publication type. The quality of the included studies was assessed using the Cochrane risk-of-bias assessment tool or the Critical Appraisal Skills Programme checklist. The extracted data from the included studies were tabulated and narratively summarized as substantial heterogeneity was expected.
In total, 8 studies met the eligibility criteria: 3 (38%) randomized controlled trials and 5 (62%) uncontrolled before-and-after studies. The CAs in the included studies were aimed at behavior changes through education, advice on food choices, or counseling via psychological approaches. Of the included studies, only 38% (3/8) reported a substantial weight loss outcome (1.3-2.4 kg decrease at 12-15 weeks of CA use). The overall quality of the included studies was judged as low.
The findings of this systematic review suggest that CAs with unconstrained natural language input can be used as a feasible interpersonal weight management intervention by promoting engagement in psychiatric intervention-based conversations simulating treatments by health care professionals, but currently there is a paucity of evidence. Well-designed rigorous randomized controlled trials with larger sample sizes, longer treatment duration, and follow-up focusing on CAs' acceptability, efficacy, and safety are warranted.
肥胖是全球范围内的一个公共卫生问题。对话代理（CA），也常被称为聊天机器人，是一种模拟人与人之间对话的计算机程序。由于更好的可及性、成本效益、个性化和富有同情心的以患者为中心的治疗，CA 有望为体重管理提供可持续的生活方式咨询。
本系统评价旨在批判性地总结和评估 CA 用于体重管理的有效性和可行性的临床研究，这些 CA 具有不受限制的自然语言输入功能。
检索了 PubMed、Embase、Cochrane 图书馆（CENTRAL）、PsycINFO 和 ACM 数字图书馆，截至 2022 年 12 月。如果 CA 用于体重管理并且具有不受限制的自然语言输入功能，则纳入研究。对研究设计、语言或发表类型没有限制。使用 Cochrane 偏倚风险评估工具或批判性评估技能计划清单评估纳入研究的质量。从纳入研究中提取的数据以表格形式列出，并进行叙述性总结，因为预计存在很大的异质性。
共有 8 项研究符合纳入标准：3 项（38%）随机对照试验和 5 项（62%）无对照前后研究。纳入研究中的 CA 旨在通过教育、食物选择建议或通过心理方法进行咨询来实现行为改变。在纳入的研究中，只有 38%（3/8）报告了显著的体重减轻结果（使用 CA 12-15 周后体重下降 1.3-2.4 公斤）。纳入研究的总体质量被评为低。
本系统评价的结果表明，具有不受限制的自然语言输入的 CA 可以作为一种可行的人际体重管理干预措施，通过促进参与基于心理干预的对话来模拟医疗保健专业人员的治疗，但目前证据不足。需要设计良好的、严格的随机对照试验，纳入更大的样本量、更长的治疗持续时间和随访，重点关注 CA 的可接受性、疗效和安全性。
