Health information quality assessment using artificial intelligence: Quality dimensions from healthcare professionals’ perspective

Recent research has shown a growing interest in the automatic assessment of health information quality on the internet. However, there is a lack of universally applicable guidelines for machine learning and deep learning practitioners to use when evaluating health information. This study seeks to address the gap by empirically identifying a set of tangible guidelines for assessing health information. Drawing from existing literature, we identified 18 criteria and collaborated with specialist doctors to convert these criteria into questionnaires. These questionnaires were then distributed through various social media platforms, including Facebook, WhatsApp, and Email, resulting in 253 responses from six Arab countries with high search volumes for health information. Our analysis revealed that the 18 criteria could be categorized into three subcategories: source quality criteria, treatment quality criteria, and content trustworthiness criteria. Each subcategory plays a crucial role in establishing the trustworthiness of the source of health information, ensuring the quality of treatment, and maintaining the general trustworthiness of the content. Furthermore, we ranked these criteria based on their perceived importance to health information quality as determined by doctors and caregivers. Our findings suggest that these dimensions are highly correlated with health information quality and can serve as valuable tools for both healthcare professionals and machine learning practitioners.