The COVID-19 pandemic infected many individuals and caused a high number of deaths around the world. In Brazil, it highlighted the regional healthcare disparities and caused high in-hospital mortality. This drew attention to the system s burden and raised the question that uneven access could affect in-hospital mortality during the pandemic. Therefore, this study used the Design Science Research (DSR) methodology paired with the Data Science life cycle to identify how the healthcare access affects in-hospital mortality. Access is a complex concept, lacking a consensual definition, which motivated us to create a conceptualization that encompasses its multiple characteristics. A qualitative analysis of the interaction among the layers of this new definition pointed to accessibility as the main issue. Thus, this work used public data from the Sistema Integrado de Vigilância Epidemiológica da Gripe (SIVEP-Gripe) to better understand how it affected the in-hospital mortality among COVID-19 inpatients in Brazil during the epidemiological year of 2020. This was conducted by evaluating the differences between those hospitalized in their city of residence and those that had to go to other cities. Results showed that, in this period, 26 per cent of inpatients were not hospitalized in their city of residence. Also, there was a noticeable flux of patients from towns on the periphery of the state capitals to the capitals. These peripheries presented higher in-hospital mortality and a lower level of healthcare resources, which may have motivated the movement. Besides, the development of a mixed effects logistic regression model evidenced that inpatients that left their cities of residence had a higher chance of death than those that stayed. This showed that a deficiency in accessibility caused a negative impact in in-hospital mortality.