Health information seeking and digital tool engagement among diabetes patients: post-COVID-19 perspectives on trust, misinformation, and AI-Based MOH services.

Abstract

The rapid growth of digital health technologies is changing how chronic illness patients access & view their health information. During the COVID-19 pandemic there was an increase in mobile health (mHealth) service usage and an increasing number of people exposing themselves to false information via social media. To develop safe and effective health communication strategies, it is important to understand how diabetic patients navigate this new digital world. to examine the health information seeking behaviour of diabetic patients in Saudi Arabia, including the use of digital health tools; interactions with the MOH; the use of social media; and how individuals verify information they find online. A descriptive cross-sectional survey was conducted among 419 adults with diabetes. Participants were recruited through stratified random sampling to ensure variation in age, education, and digital health familiarity. Data were collected using a structured online questionnaire comprising demographic items, Likert-scale measures of digital health engagement, and perceptions of MOH-provided AI services (e.g., chatbots and self-assessment tools). Analyses were performed using SPSS and R. Structural equation modelling (SEM) examined relationships between latent constructs, while UpSet plots visualised behavioural intersections. Reliability was assessed using Cronbach's alpha. Participants demonstrated multidimensional information-seeking patterns, frequently combining physician consultation (n = 245), peer-reviewed sources (n = 213), and government-affiliated platforms (n = 187). MOH digital services, particularly the 937 consultation hotline and Sehhaty virtual clinics, were widely utilised. Trust in AI-based tools remained moderate, with most respondents viewing them as supportive rather than decision-making substitutes. SEM revealed a strong positive association between eHealth engagement and trust in AI tools (β = 0.69, p < 0.001). UpSet plots indicated common co-occurrence of cautious behaviours, including cross-source verification. Cronbach's alpha values ranged from 0.73 to 0.91, confirming acceptable to high internal consistency. Diabetes patients in Saudi Arabia appear digitally engaged yet critically evaluative, favouring authoritative sources and cautiously adopting AI-enabled services. These findings emphasise the need to strengthen digital health literacy, integrate trusted information within widely used platforms, and enhance transparency and credibility in AI-driven health systems.