Spatio-temporal patterns of neurodegenerative disease hospitalizations in mainland Portugal

BACKGROUND: Neurodegenerative diseases are an increasing concern for the aging population worldwide. In Portugal, as in many other developed countries, the population is aging rapidly. Understanding temporal and spatial patterns is of utmost relevance to help manage the burden these diseases place on the healthcare system.

METHODS: In this retrospective study, we analyzed over 500,000 hospitalizations discharged between 2000 and 2016. We used the empirical Bayes method to compute the smoothed age-standardized hospitalization rates for each neurodegenerative disease, for all hospitalizations per year, and across administrative divisions (districts and municipalities). We then searched for data clusters using both global and local spatial autocorrelation methods based on the Moran index.

RESULTS: A steady increase in age-standardized hospitalization rates was observed throughout the study period. Statistically significant global spatial autocorrelation was found when considering all diseases per municipality (Moran's I = 0.010, p-value < 0.001). In addition, when considering the districts, only Alzheimer's disease, dementia, and basal ganglia disorders did not show significant spatial autocorrelation. When considering municipalities, all diseases showed significant positive global spatial autocorrelation.

CONCLUSION: Temporal analysis showed increasing age-standardized hospitalization rates over time, likely reflecting the aging population in Portugal. The spatial analysis showed significant clustering, which may reflect geographic differences in hospitalization practices, access to care, population structure, or other contextual factors. Through this study, we hope to enlighten future research by providing insights into the anticipated spatio-temporal patterns.