A Context-Specific Adapted Technology Acceptance Frame-work for Digital Health System Adoption in Malawi

Abstract Background Empirical studies examining healthcare professionals’ attitudes and behavioral intentions toward digital health systems remain scarce in Malawi. This lack of context-specific evidence limits the ability of policymakers and system developers to design and implement digital health solutions that align with users’ needs and professional realities. Addressing this gap requires an adapted context specific technology acceptance model (TAM) that integrates established theoretical constructs with factors relevant to healthcare practice in Malawi. This study aimed to develop and empirically validate context-specific adapted technology acceptance framework for digital health system (DHS) adoption among healthcare professionals in Malawi. Methods A cross-sectional quantitative study was conducted through a sample of 615 healthcare workers from public and private health facilities across Malawi. Data were collected using a structured questionnaire informed by TAM2 and TAM3 constructs which included, perceived usefulness, perceived ease of use, job relevance, subjective norm, descriptive norm, computer anxiety and intention to use DHS. The study also included context specific social factors including computer use and training. Structural Equation Modelling (SEM) was used to examine relationships among constructs and intention to use DHS. Results Results indicated that perceived usefulness (β = 0.154, p = 0.043) and perceived ease of use (β = 0.281, p < .001) were significant positive predictors of intention to use digital health systems. Computer anxiety (β = -0.129, p < .001) had a significant negative direct effect on adoption intention and an indirect positive effect through perceived ease of use (β = 0.015, p = 0.039). Social and task-related factors influenced adoption indirectly via perceived usefulness. The adapted model demonstrated good model fit (CFI = 0.979, TLI = 0.973, RMSEA = 0.046). The model also demonstrated adequate explanatory power, with R² values for the endogenous constructs ranging from 0.4 to 0.778. Conclusion The adapted TAM framework validates the original TAM and extends its explanatory power by providing a contextual relevant understanding that improve digital health system adoption in resource-limited settings. While earlier models acknowledge the role of computer anxiety, this study provides empirical evidence of both direct and indirect pathways, thereby extending existing adoption models. This study suggests that policies and digital health programs in low-resource settings should prioritize usability-focused system design and targeted capacity-building interventions to address computer anxiety, rather than relying solely on technology availability.