Implementation science offers an innovative approach to advance universal health coverage, a key element of the Sustainable Development Goals. Yet, there is limited empirical evidence on the conceptual clarity, validity, and reliability of implementation outcomes, particularly in the context of digital health booking systems in sub-Saharan Africa. This study aimed to assess the psychometric properties of measures evaluating the acceptability, appropriateness, and feasibility of students’ online health appointment system (SOHAS) in a public university in Ghana. A cross-sectional study was conducted among students at the Kwame Nkrumah University of Science and Technology, Kumasi, Ghana, using a 15-item adapted implementation outcome measure. The items were adapted for SOHAS and were also ensured to be culturally, linguistically, and contextually relevant to Ghanaian university students. Participants were recruited online via a text message containing a link to the electronic questionnaire. Both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were utilized to assess the underlying factor structure and model fit, respectively. Cronbach's alpha and item-total correlations were applied to measure internal consistency. Convergent and discriminant validity were examined using composite reliability (CR), average variance extracted (AVE), Fornell–Larcker criteria, and heterotrait–monotrait (HTMT) ratios. Cronbach’s alpha for the three constructs ranged from 0.946 to 0.977, and CR ranged from 0.949 to 0.976. EFA revealed a three-factor structure with substantial communalities and high loadings (0.716–0.969). The CFA also demonstrated a good model fit (CFI = 0.969, TLI = 0.962, RMSEA = 0.092, SRMR = 0.022). Convergent validity was strong, with AVEs ranging from 0.790 to 0.892. However, there was insufficient evidence to establish discriminant validity, as HTMT ratios between appropriateness and acceptability exceeded acceptable levels, and inter-construct correlations exceeded √AVE values. We found promising psychometric properties of the adapted Acceptability, Appropriateness, and Feasibility measures, including high internal consistency, good model fit, and strong convergent validity. However, significant overlap between acceptability and appropriateness necessitates that future studies refine these constructs conceptually and empirically, thereby improving their discriminant validity. Importantly, these validated measures provide a valuable framework for guiding evaluations of digital health implementation in similar settings.
Opoku et al. (Thu,) studied this question.