Impact of prompt engineering on large language models for risk of bias assessment: a comparative study

三段论 过程(计算) 控制(管理) 风险评估 认知心理学 人工智能 代表(政治) 相似性(几何) 空白 心理学 计算机科学 自然语言处理 机器学习 荟萃分析 因果推理 计量经济学 基于案例的推理 风险管理
作者
Yu-Tao Xiong,Yuan-Chen Wang,Yuanhao Sun,Zhengzhe Zhan,Liu W,Hao-Nan Liu,Ruo-Yu Meng,Wei Tang,Chang Liu
出处
期刊:BMJ evidence-based medicine [BMJ]
卷期号:: bmjebm-2025
标识
DOI:10.1136/bmjebm-2025-114055
摘要

Objectives To evaluate the performance of large language models (LLMs) in risk of bias assessment and to examine whether prompt engineering improves their accuracy and alignment with expert reasoning. Methods We analysed 158 randomised controlled trials from 10 dental systematic reviews and their risk of bias assessments were reviewed and revised to serve as the reference standard. Two LLMs (DeepSeek-V3 and GPT-5) were evaluated under four prompting strategies, including direct command, command with reference, constrained output and formula-constrained output. The direct command served as the blank control group, simulating the approach commonly used by clinicians, whereas the other three groups employed different prompt engineering. The performance of LLMs across the seven domains of RoB-1 was evaluated using accuracy and agreement. The reasoning process of the LLMs was expressed in the form of syllogisms and its similarity to expert reasoning was assessed using MMD 2 . Results LLMs showed limited capability in risk of bias assessment under the blank control condition, with mean accuracies of 0.72 for DeepSeek-V3 and 0.65 for GPT-5. With formula-constrained prompting, the performance of both LLMs improved significantly, and the overall accuracy increased to 0.85 for both DeepSeek-V3 and GPT-5 (both vs the blank control group, p<0.001). Agreement metrics showed a similar pattern, with higher agreement under formula-constrained prompting than under the other prompting strategies (p<0.001 for both models). In addition, the syllogistic output format provided a clear representation of the reasoning process underlying risk of bias assessment. Compared with constrained output, formula-constrained prompting also produced reasoning that was more closely aligned with the reference answers, as indicated by lower MMD² values (DeepSeek-V3: 0.0765 vs 0.1239; GPT-5: 0.0548 vs 0.1068). Conclusion Prompt engineering substantially improved the performance of LLMs in risk of bias assessment. Although LLMs cannot currently replace human reviewers, they may serve as efficient and transparent tools to support this process.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
刚刚
dan1029发布了新的文献求助10
刚刚
dan1029发布了新的文献求助10
刚刚
传奇3的应助被涛1采纳,获得10
刚刚
dan1029发布了新的文献求助10
刚刚
dan1029发布了新的文献求助10
刚刚
dan1029发布了新的文献求助10
1秒前
dan1029发布了新的文献求助10
1秒前
小慕完成签到,获得积分10
1秒前
1秒前
dan1029发布了新的文献求助10
1秒前
1秒前
1秒前
尺八发布了新的文献求助10
1秒前
dan1029发布了新的文献求助10
1秒前
优秀的荔枝完成签到,获得积分10
1秒前
1秒前
dan1029发布了新的文献求助10
1秒前
科研通AI2S的应助被李宇彤采纳,获得10
2秒前
2秒前
方方发布了新的文献求助20
2秒前
MYS发布了新的文献求助10
3秒前
3秒前
3秒前
3秒前
dan1029发布了新的文献求助10
3秒前
dan1029发布了新的文献求助10
3秒前
yuhanli发布了新的文献求助10
3秒前
dan1029发布了新的文献求助10
4秒前
哈哈发布了新的文献求助10
4秒前
dan1029发布了新的文献求助10
4秒前
dan1029发布了新的文献求助10
4秒前
dan1029发布了新的文献求助10
4秒前
dan1029发布了新的文献求助10
4秒前
4秒前
4秒前
风味土豆片完成签到,获得积分10
5秒前
Jupiter 1234完成签到,获得积分10
5秒前
6秒前
6秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
John McDowells Bildungstheorie 800
Social Psychology in the Real World 600
八旗制度の研究 600
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8180713
求助须知:如何正确求助?哪些是违规求助? 9570752
关于积分的说明 21814119
捐赠科研通 7776960
什么是DOI,文献DOI怎么找? 3433298
关于科研通互助平台的介绍 2557579
邀请新用户注册赠送积分活动 2479287