Studies of criteria-based assessment tools have demonstrated the feasibility of objectively evaluating eHealth interventions independent of empirical testing. However, current tools have not included some quality constructs associated with intervention outcome, such as persuasive design, behavior change, or therapeutic alliance. In addition, the generalizability of such tools has not been explicitly examined.
The aim is to introduce the development and further analysis of the Enlight suite of measures, developed to incorporate the aforementioned concepts and address generalizability aspects.
As a first step, a comprehensive systematic review was performed to identify relevant quality rating criteria in line with the PRISMA statement. These criteria were then categorized to create Enlight. The second step involved testing Enlight on 42 mobile apps and 42 Web-based programs (delivery mediums) targeting modifiable behaviors related to medical illness or mental health (clinical aims).
A total of 476 criteria from 99 identified sources were used to build Enlight. The rating measures were divided into two sections: quality assessments and checklists. Quality assessments included usability, visual design, user engagement, content, therapeutic persuasiveness, therapeutic alliance, and general subjective evaluation. The checklists included credibility, privacy explanation, basic security, and evidence-based program ranking. The quality constructs exhibited excellent interrater reliability (intraclass correlations=.77-.98, median .91) and internal consistency (Cronbach alphas=.83-.90, median .88), with similar results when separated into delivery mediums or clinical aims. Conditional probability analysis revealed that 100% of the programs that received a score of fair or above (≥3.0) in therapeutic persuasiveness or therapeutic alliance received the same range of scores in user engagement and content—a pattern that did not appear in the opposite direction. Preliminary concurrent validity analysis pointed to positive correlations of combined quality scores with selected variables. The combined score that did not include therapeutic persuasiveness and therapeutic alliance descriptively underperformed the other combined scores.
This paper provides empirical evidence supporting the importance of persuasive design and therapeutic alliance within the context of a program’s evaluation. Reliability metrics and preliminary concurrent validity analysis indicate the potential of Enlight in examining eHealth programs regardless of delivery mediums and clinical aims.