Jadwal Sholat

Memuat jadwal sholatโ€ฆ

Kedokteran & Kesehatan editorial

Open AccessOA2023

Artificial Intelligence and Public Health: Evaluating ChatGPT Responses to Vaccination Myths and Misconceptions

This study evaluates ChatGPT's responses to the WHO's 11 vaccination myths, finding that GPT-4.0 outperforms GPT-3.5 in correctness, clarity, and exhaustiveness, but both versions risk misleading non-expert users. The improved accuracy of the paid version raises ethical concerns about information access equity.
G. Deiana; M. Dettori; A. Arghittu; A. Azara; G. Gabutti; P. Castigliaยท Vaccinesยท 2023ยท DOI 10.3390/vaccines11071217

The core problem

The integration of artificial intelligence (AI) tools like ChatGPT into healthcare contexts is a subject of intense debate. This study focuses on vaccination, a critical public health topic where misinformation can have severe consequences. The authors aim to assess the quality of ChatGPT's responses to common vaccination myths and misconceptions, as defined by the World Health Organization (WHO). The research questions center on the correctness, clarity, and exhaustiveness of the AI-generated answers, comparing the free version (GPT-3.5) and the paid version (GPT-4.0). The study also explores the potential risks and ethical implications of using AI for health information, especially for non-expert users.

Innovation

The WHO's 11 "myths and misconceptions" about vaccinations were administered as prompts to both GPT-3.5 and GPT-4.0. Two independent expert raters evaluated the responses qualitatively and quantitatively. The evaluation criteria were Correctness, Clarity, and Exhaustiveness, with reference to the WHO's myth and misconception statements. Inter-rater agreement was assessed using Cohen's kappa (K), with a significance level of p < 0.05. The responses were scored, and differences between the two GPT versions were analyzed. The study did not involve human participants, so ethical approval was not required.
Introduction
The integration of artificial intelligence (AI) tools like ChatGPT into healthcare contexts is a subject of intense debate. This study focuses on vaccination, a critical public health topic where misinformation can have severe consequences. The authors aim to assess the quality of ChatGPT's responses to common vaccination myths and misconceptions, as defined by the World Health Organization (WHO). The research questions center on the correctness, clarity, and exhaustiveness of the AI-generated answers, comparing the free version (GPT-3.5) and the paid version (GPT-4.0). The study also explores the potential risks and ethical implications of using AI for health information, especially for non-expert users.

Why it matters

The study demonstrates that, if appropriately questioned, AI tools like ChatGPT can be useful aids in healthcare. However, when consulted by non-expert users without expert medical advice, these tools are not free from the risk of eliciting misleading responses. The improved accuracy of the paid version (GPT-4.0) raises ethical concerns regarding the social divide in information access, as those who can afford paid versions receive more reliable information. The authors emphasize the need for caution and further research to ensure equitable and safe use of AI in public health. The findings suggest that while AI can support health communication, it should not replace professional medical advice.

Who should read this

Pembaca Research Digest dan praktisi rumpun ini.

Opening member contentโ€ฆ