Introduction
The rapid rise of information and communication technologies (ICT) has reshaped health care, making digital tools central to modern practice. As generative AI expands, patients must effectively engage with electronic health (eHealth) resources to influence outcomes. The Adult Inpatient eHealth Literacy Scale (AIPeHLS) emerges to address this need—and does so with a focus on hospitalized adults within the Web 3.0 health care landscape.
eHealth literacy (eHL) was first defined by Norman and Skinner in 2006 as the ability to seek, find, understand, appraise health information from electronic sources, and apply it to health problems. Higher eHL has been linked to better self-management, adherence, and lower costs, while low eHL can delay care and worsen outcomes. The digital health momentum in China—ranging from internet hospitals to rural digitization—highlights the importance of measuring eHL among inpatients who often face acute or complex conditions and heightened vulnerability to digital barriers.
Despite a growing field of eHL instruments, there was a gap for a standardized tool tailored to inpatients in the Web 3.0 era, reflecting new challenges such as data privacy, AI-generated content, and advanced digital interactions. The Lily model anchors the AIPeHLS in six literacies—traditional, information, media, health, computer, and scientific literacy—capturing a holistic and evolving set of competencies needed to navigate modern digital health environments.
Methodology: Building the AIPeHLS
Item pool development
Researchers generated an initial pool of 53 items derived from a systematic literature review of 934 articles (Chinese and English) and from validated scales. The goal was to reflect inpatient scenarios and contemporary ICT use, including voice interactions, medication understanding, privacy considerations, data security, and bias detection in online health content. Items were aligned with the Lily model to cover foundational and advanced eHL skills in Web 1.0 to Web 3.0 contexts.
Delphi process
A two-round Delphi panel of 18 experts from 12 regions in China evaluated item importance and relevance. The process yielded high authority and consensus (Cr ~0.86; item importance >3.50, CVs <0.25). Feedback led to substantial refinement, with 44 items across six dimensions in the initial AIPeHLS.
Pilot testing and item analysis
A pilot study with 100 adult inpatients at a major hospital in Hunan used critical value analysis, item-total correlations, and exploratory factor analysis (EFA). Results showed strong discrimination and internal consistency (Cronbach’s α ranging from 0.952 to 0.975 across dimensions; total α = 0.959). A KMO value of 0.921 and a 6-factor solution explained 82.6% of variance, with all loadings above 0.40 and no cross-loadings.
Validation of the AIPeHLS
A cross-sectional validation study enrolled 532 adult inpatients from nine wards in a Grade A tertiary hospital in Hunan, using cluster sampling. Confirmatory factor analysis (CFA) supported the six-factor structure. Model fit indices were robust (GFI 0.854, CFI 0.957, RMSEA 0.048). Convergent validity was evidenced by AVE values above 0.50 and CR above 0.70; discriminant validity was confirmed by the square roots of AVE exceeding inter-dimension correlations. Content validity (I-CVI ≥ 0.78; S-CVI ≥ 0.90) and criterion validity (strong correlation with a Chinese version of eHEALS) further established the instrument’s usefulness.
Reliability was excellent overall: Cronbach’s α = 0.965; McDonald’s ω values between 0.948 and 0.971 for subscales. Split-half reliability exceeded 0.79 overall, with subscales consistently high. The final instrument comprises 44 items across six dimensions: traditional literacy, information literacy, media literacy, health literacy, computer literacy, and scientific literacy.
Discussion and Implications
The AIPeHLS fills a critical gap by providing a rigorous, inpatient-focused measure that integrates Web 3.0 competencies—such as data security, privacy awareness, and the ability to navigate AI-assisted resources. Its strong psychometric properties support its use in clinical settings to tailor interventions, improve patient engagement, and evaluate digital health interventions’ effectiveness in hospitals. Researchers can employ the scale to study relationships between eHL and health outcomes, self-management, and satisfaction with care.
Limitations and Future Directions
Limitations include the single-site sample and the use of eHEALS as a criterion standard, which may not fully capture inpatient eHL. Future multicenter studies should examine measurement invariance over time and across populations, and test-retest reliability in longer hospital stays. The AIPeHLS could be extended to diverse settings and integrated with digital health interventions to assess impact on patient-centered outcomes.
Conclusion
The AIPeHLS offers a psychometrically sound, multidimensional tool for assessing eHL among adult inpatients in the Web 3.0 era. By capturing six literacies and incorporating modern eHealth challenges, it enables clinicians, researchers, and policymakers to understand and enhance patients’ digital health competencies and engagement in hospital care.
