Data inconsistencies, particularly in the periodicity and granularity of health event records, pose significant challenges in epidemiological modeling. Addressing these inconsistencies is crucial for accurately assessing disease risk in both spatial and temporal contexts. This paper proposes a data disaggregation methodology designed to standardize the granularity and periodicity of epidemiological data, ensuring consistency for spatio-temporal analysis. The proposed methodology is illustrated using COVID-19 data from mainland Portugal. The data released by public health authorities exhibited variations in reporting over time: initially, case counts were reported daily, later transitioning to weekly aggregates and 14-day cumulative incidence rates. Additionally, overlapping periods further complicated the time series analysis, requiring data transformation. By comparing models fitted with disaggregated data to those using observed data, we found no significant evidence that disaggregation led to poorer model fits. This suggests that the transformation process did not compromise model accuracy or reliability. The proposed method enhances our understanding of how data transformations can be applied without negatively affecting model predictions, which is crucial for accurate spatial distribution modeling and public health decision-making. These findings underscore the importance of robust disaggregation methods to ensure data consistency and provide a framework for future studies addressing similar challenges.