Data-driven models for the risk of infection and hospitalization during a pandemic: Case study on COVID-19 in Nepal

The newly emerging pandemic disease often poses unexpected troubles and hazards to the global health system, particularly in low and middle-income countries like Nepal. In this study, we developed mathematical models to estimate the risk of infection and the risk of hospitalization during a pandemic which are critical for allocating resources and planning health policies. We used our models in Nepal's unique data set to explore national and provincial-level risks of infection and risk of hospitalization during the Delta and Omicron surges. Furthermore, we used our model to identify the effectiveness of non-pharmaceutical interventions (NPIs) to mitigate COVID-19 in various groups of people in Nepal. Our analysis shows no significant difference in reproduction numbers in provinces between the Delta and Omicron surge periods, but noticeable inter-provincial disparities in the risk of infection (for example, during Delta (Omicron) surges, the risk of infection of Bagmati province is: ∼ 98.94 (89.62); Madhesh province: ∼ 12.16 (5.1); Karnali province ∼31.16 (3) per hundred thousands). Our estimates show a significantly low level of hospitalization risk during the Omicron surge compared to the Delta surge (hospitalization risk is: ∼10% in Delta and ∼2.5% in Omicron). We also found significant inter-provincial disparities in the hospitalization rate (for example, ∼ 6% in Madhesh province and ∼ 21% in Sudur Paschim) during the Delta surge. Moreover, our results show that closing only schools, colleges, and workplaces reduces the risk of infection by one-third, while a complete lockdown reduces the infections by two-thirds. Our study provides a framework for the computation of the risk of infection and the risk of hospitalization and offers helpful information for controlling the pandemic.