Using structural equation modelling to identify factors influencing the adoption of digital tools and biometrics in healthcare delivery among community healthcare workers in Ghana

Information, Communication, and Technology (ICT) is a vital tool in combating disease, promoting individual health and well-being, and making health systems more responsive, effective, and efficient. Despite the expansion of digital health tools in Ghana, there is limited empirical evidence on the behavioral drivers that influence CHWs’ acceptance and sustained use of these technologies, particularly biometric applications for patient identification and data management. To address this gap, this study employs a structural equation modeling (SEM) approach to examine factors associated with CHWs’ behavioral intention to use biometric and digital health tools. The study utilized a cross-sectional design via a multistage stratified mixed method. Data was collected from community health workers. Consequently, their intention to accept and recommend its further use was measured using the Technology Acceptance Model with extended constructs. A total of eleven districts were sampled from two regions. The sampling frame for the CHWs survey was all facilities across all the districts in the two regions, with a sample size of 390. Descriptive analyses were done for continuous variables (median, interquartile range), whilst categorical variables were summarized as frequencies, proportions, and charts. Analyses were carried out via a four-step approach. The first was through Exploratory Factor Analysis (EFA), followed by Confirmatory Factor Analysis (CFA), then Path Analysis (PA) and Mediation Analysis (MA). The analysis revealed that perceived usefulness (β = 0.350, p < .001), normative beliefs (β = 0.149, p < .001), security (β = 0.119, p < .001), and perceived ease of use (β = 0.216, p < .001) significantly influenced behavioral intention to adopt biometric devices among community health workers (CHWs). Perceived usefulness and ease of use reduced perceived risk (β = -0.318 and β = -0.287, p < .001). Descriptive findings showed high awareness of eTracker (88.8% overall: 86.9% Eastern, 95.5% Oti) and biometrics (84.2% overall: 85.2% Eastern, 80.7% Oti), though only 19.1% of CHWs had operated biometric devices. Key benefits cited included easy patient registration (87.2%) and identification (84.2%), while challenges included lack of training (29.0%) and ease-of-use concerns (31.9%). The 7-factor EFA explained 59.7% of variance, and CFA confirmed model fit (RMSEA = 0.067, CFI = 0.919). Regional differences emerged: Eastern Region’s 7-factor solution explained 65.2% variance, while Oti’s 5-factor solution accounted for 72.2%. The study shows that CHWs’ intention to adopt biometric tools is mainly influenced by perceived usefulness, ease of use, security, and normative beliefs, while perceived risk is reduced by the tool’s utility and usability. Despite high awareness, operational experience remains low, revealing a gap between knowledge and practical use. Regional differences in adoption factors indicate the need for tailored training and implementation strategies, focusing on security concerns and making the tools more user-friendly to increase adoption. Effectively addressing these concerns can boost user confidence and encourage adoption.