Modeling antecedents of electronic medical record system implementation success in low-resource setting hospitals

Background: With the increasing implementation of Electronic Medical Record Systems (EMR) in developing countries, there is a growing need to identify antecedents of EMR success to measure and predict the level of adoption before costly implementation. However, less evidence is available about EMR success in the context of low-resource setting implementations. Therefore, this study aims to fill this gap by examining the constructs and relationships of the widely used DeLone and MacLean (D&M) information system success model to determine whether it can be applied to measure EMR success in those settings. Methods: A quantitative cross sectional study design using self-administered questionnaires was used to collect data from 384 health professionals working in five governmental hospitals in Ethiopia. The hospitals use a comprehensive EMR system since three years. Descriptive and structural equation modeling methods were applied to describe and validate the extent of relationship of constructs and mediating effects. Results: The findings of the structural equation modeling shows that system quality has significant influence on EMR use (beta = 0.32, P < 0.05) and user satisfaction (beta = 0.53, P < 0.01); information quality has significant influence on EMR use (beta = 0.44, P < 0.05) and user satisfaction (beta = 0.48, P < 0.01) and service quality has strong significant influence on EMR use (beta = 0.36, P < 0.05) and user satisfaction (beta = 0.56, P < 0.01). User satisfaction has significant influence on EMR use (beta = 0.41, P < 0.05) but the effect of EMR use on user satisfaction was not significant. Both EMR use and user satisfaction have significant influence on perceived net-benefit (beta = 0.31, P < 0.01; beta = 0.60, P < 0.01), respectively. Additionally, computer literacy was found to be a mediating factor in the relationship between service quality and EMR use (P < 0.05) as well as user satisfaction (P < 0.01). Among all the constructs, user satisfaction showed the strongest effect on perceived net-benefit of health professionals. Conclusion: EMR implementers and managers in developing countries are in urgent need of implementation models to design proper implementation strategies. In this study, the constructs and relationships depicted in the updated D&M model were found to be applicable to assess the success of EMR in low resource settings. Additionally, computer literacy was found to be a mediating factor in EMR use and user satisfaction of health professionals. Hence, EMR implementers and managers in those settings should give priority in improving service quality of the hospitals like technical support and infrastructure; providing continuous basic computer trainings to health professionals; and give attention to the system and information quality of the systems they want to implement.