Quality and Accountability of Chat GPT in Health Care in Low- and Middle-Income Countries: Simulated Patient Study

被引:3
作者
Si, Yafei [1 ,2 ]
Yang, Yuyi [3 ]
Wang, Xi [4 ]
Zu, Jiaqi [5 ]
Chen, Xi [6 ,7 ]
Fan, Xiaojing [8 ]
An, Ruopeng [4 ,9 ]
Gong, Sen [10 ]
机构
[1] Univ New South Wales, UNSW Business Sch, Kensington, Australia
[2] Univ New South Wales, CEPAR, Kensington, Australia
[3] Washington Univ St Louis, Div Computat & Data Sci, St Louis, MO USA
[4] Washington Univ St Louis, Brown Sch, St Louis, MT USA
[5] Duke Kunshan Univ, Global Hlth Res Ctr, Kunshan, Peoples R China
[6] Yale Univ, Dept Hlth Policy & Management, New Haven, CT USA
[7] Yale Univ, Dept Econ, New Haven, CT USA
[8] Xi An Jiao Tong Univ, Sch Publ Policy & Adm, 28 West Xianning Rd, Xian 710049, Peoples R China
[9] NYU, Silver Sch Social Work, New York, NY USA
[10] Zhejiang Univ, Ctr Int Studies Dev & Governance, Hangzhou, Peoples R China
关键词
ChatGPT; generative AI; simulated patient; health care; quality and safety; low- and middle-income countries; quality; LMIC; patient study; effectiveness; reliability; medication prescription; prescription; noncommunicable diseases; AI integration; AI; artificial intelligence;
D O I
10.2196/56121
中图分类号
R19 [保健组织与事业(卫生事业管理)];
学科分类号
摘要
Using simulated patients to mimic 9 established noncommunicable and infectious diseases, we assessed ChatGPT's performancein treatment recommendations for common diseases in low- and middle-income countries. ChatGPT had a high level of accuracyin both correct diagnoses (20/27, 74%) and medication prescriptions (22/27, 82%) but a concerning level of unnecessary orharmful medications (23/27, 85%) even with correct diagnoses. ChatGPT performed better in managing noncommunicablediseases than infectious ones. These results highlight the need for cautious AI integration in health care systems to ensure qualityand safety
引用
收藏
页数:5
相关论文
共 10 条
[1]  
Howarth J., 2024, How many people own smartphones?
[2]   The Potential of ChatGPT as a Self-Diagnostic Tool in Common Orthopedic Diseases: Exploratory Study [J].
Kuroiwa, Tomoyuki ;
Sarcon, Aida ;
Ibara, Takuya ;
Yamada, Eriku ;
Yamamoto, Akiko ;
Tsukamoto, Kazuya ;
Fujita, Koji .
JOURNAL OF MEDICAL INTERNET RESEARCH, 2023, 25
[3]   Use of standardised patients for healthcare quality research in low- and middle-income countries [J].
Kwan, Ada ;
Daniels, Benjamin ;
Bergkvist, Sofi ;
Das, Veena ;
Pai, Madhukar ;
Das, Jishnu .
BMJ GLOBAL HEALTH, 2019, 4 (05)
[4]  
Sanders John W., 2008, Science Progress, V91, P1, DOI 10.3184/003685008X284628
[5]   Appropriateness of Cardiovascular Disease Prevention Recommendations Obtained From a Popular Online Chat-Based Artificial Intelligence Model [J].
Sarraju, Ashish ;
Bruemmer, Dennis ;
Van Iterson, Erik ;
Cho, Leslie ;
Rodriguez, Fatima ;
Laffin, Luke .
JAMA-JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION, 2023, 329 (10) :842-844
[6]   AI-based recommendation model for effective decision to maximise ROI [J].
Sellamuthu, Suseela ;
Vaddadi, Srinivas Aditya ;
Venkata, Srinivas ;
Petwal, Hemant ;
Hosur, Ravi ;
Mandala, Vishwanadham ;
Dhanapal, R. ;
Singh, Jagendra .
SOFT COMPUTING, 2023,
[7]   The quality of telemedicine consultations for sexually transmitted infections in China [J].
Si, Yafei ;
Xue, Hao ;
Liao, Huipeng ;
Xie, Yewei ;
Xu, Dong ;
Smith, M. Kumi ;
Yip, Winnie ;
Cheng, Weibin ;
Tian, Junzhang ;
Tang, Weiming ;
Sylvia, Sean .
HEALTH POLICY AND PLANNING, 2024, 39 (03) :307-317
[8]   Quantifying the financial impact of overuse in primary care in China: A standardised patient study [J].
Si, Yafei ;
Bateman, Hazel ;
Chen, Shu ;
Hanewald, Katja ;
Li, Bingqin ;
Su, Min ;
Zhou, Zhongliang .
SOCIAL SCIENCE & MEDICINE, 2023, 320
[9]   Ethical Considerations of Using ChatGPT in Health Care [J].
Wang, Changyu ;
Liu, Siru ;
Yang, Hao ;
Guo, Jiulin ;
Wu, Yuxuan ;
Liu, Jialin .
JOURNAL OF MEDICAL INTERNET RESEARCH, 2023, 25
[10]  
Xue H, 2021, PREPRINT, DOI [10.2139/ssrn.39445878, DOI 10.2139/SSRN.39445878]