Unmasking pandemic patterns: decoding the COVID-19’s impact on mortality in Italy with Generalized Gamma overdispersion model

Background: Italy has been significantly impacted by the COVID-19 pandemic, particularly in its early stages, resulting in healthcare strain, societal restrictions, and disruption. Understanding the long-term effects, notably excess mortality beyond the initial peaks, remains important. Prior studies have focused on the early phase, leaving out subsequent and updated mortality trends. Method: This study analyzes Italian mortality rates from 2015 to 2023, employing Generalized Additive Models for Location, Scale, and Shape (GAMLSS), the Generalized Gamma Overdispersion model. Data analysis considered factors such as gender, age groups (under 65 and 65 or older), and geographical differences (Northern versus Central-Southern Italy) as key characteristics of the mortality trend. Results: The study identified several phases of the pandemic, characterized by a significant early 2020 mortality peak and subsequent smaller peaks. Mortality rates were higher in Northern Italy, with males and the elderly being the most affected. Overall, mortality rates increased during the pandemic, particularly among these groups, and then returned to normal levels in 2023. An increase in the overdispersion parameter, estimated via the GAMLSS model, is evident in the post-pandemic phase and persists until 2023. Conclusion: The findings highlight the complex nature of COVID-19’s impact on mortality in Italy. They reveal the temporal phases, regional disparities, and demographic vulnerabilities that contribute to the overall mortality picture. The overdispersion component indicates more significant variability and unpredictability of mortality patterns until 2023. This highlights the intricate interplay of factors, including healthcare capacity, viral mutations, and the effectiveness of public health responses. This study emphasizes the need for targeted interventions and protective measures in the most affected groups.