Title Page
ABSTRACT
국문 초록
Contents
NOMENCLATURE 18
CHAPTER 1. INTRODUCTION 20
1.1. Background 20
1.1.1. Coronavirus disease 201 20
1.1.2. Brief introduction to Nepal 20
1.1.3. Epidemiology and burden of COVID-19 in Nepal 22
1.1.4. Previous studies on COVID-19 25
1.1.5. Spatial epidemiology and its importance 26
1.2. Objectives of the study 27
CHAPTER 2. MATERIALS AND METHODS 28
2.1. Data collection 28
2.2. Study variables 29
2.2.1. Dependent variable 29
2.2.2. Independent variables 30
2.3. Statistical analysis 36
2.3.1. Non-spatial analysis 36
2.3.2. Spatial analysis 37
2.4. Ethical consideration 41
CHAPTER 3. RESULTS 42
3.1. Descriptive statistics 42
3.2. Non-spatial analysis 45
3.2.1. Univariate negative binomial regression analysis 45
3.2.2. Multivariable negative binomial regression analysis 46
3.2.3. Model validation 49
3.3. Spatial analysis 50
3.3.1. Spatial distribution of COVID-19 incidence 50
3.3.2. Neighborhood structure of Nepal 51
3.3.3. Spatial autocorrelation 51
3.3.4. Getis Ord Gi* Clustering of COVID-19 incidence 54
3.3.5. Bayesian hierarchical spatial regression analysis using R-INLA 55
3.3.6. Sub-analysis of COVID-19 incidence using the date of initiation of nationwide COVID-19 vaccination in Nepal as the cut point 65
CHAPTER 4. DISCUSSION 70
4.1. Spatial analysis of COVID-19 incidence 70
4.2. Access to healthcare facilities and COVID-19 incidence 71
4.3. Household crowding and COVID-19 incidence 72
4.4. Gini coefficient and COVID-19 incidence 73
4.5. Unemployment rate in males and COVID-19 incidence 73
4.6. Prevalence of obesity in females, diabetes mellitus, and COVID-19 incidence 74
4.7. Comparison of the determinants associated with COVID-19 before and after the initiation of the nationwide vaccination campaign 75
4.8. Limitations of the study 76
4.9. Strengths of the study 77
CHAPTER 5. CONCLUSION 78
REFERENCES 80
SUPPLEMENTARY MATERIALS 89
Table 1. List of study variables 33
Table 2. District-level characteristics of the outcome and explanatory variables in the study population. 43
Table 3. Results of the univariable negative binomial regression analysis 45
Table 4. Results of the multivariable negative binomial regression analysis 47
Table 5. Final model selected via backward elimination based on AIC. 48
Table 6. Global Moran's I statistics to check for spatial autocorrelation 51
Table 7. Results of the fixed effects in the Besag ICAR model using INLA 56
Table 8. Results of the random effects in the Besag ICAR model using INLA 57
Table 9. Results of the fixed effects in the BYM model with pc prior using INLA 58
Table 10. Results of the random effects in the BYM model with pc prior using INLA 59
Table 11. Results of the BYM model with flat prior using INLA 60
Table 12. Results of the random effects in the BYM model with flat prior using INLA 61
Table 13. Comparison of the DIC and WAIC values of the models 61
Table 14. Results of the Bayesian hierarchical spatial analysis of COVID-19 incidence expressed as relative risk and 95% credible intervals. 64
Table 15. Results of the fixed effects in the BYM model with pc prior using INLA (for COVID-19 incidence before vaccination started) 65
Table 16. Results of the random effects in the BYM model with pc prior using INLA (for COVID-19 incidence before vaccination started) 66
Table 17. Results of the Bayesian hierarchical spatial analysis of COVID-19 incidence expressed as relative risk and 95% credible intervals (for COVID-19 incidence before... 67
Table 18. Results of the fixed effects in the BYM model with pc prior using INLA (for COVID-19 incidence after vaccination started) 68
Table 19. Results of the random effects in the BYM model with pc prior using INLA (for COVID-19 incidence after vaccination started) 68
Table 20. Results of the Bayesian hierarchical spatial analysis of COVID-19 incidence expressed as relative risk and 95% credible intervals (for COVID-19 incidence after... 69
Figure 1. Map of Nepal with (a) provincial and (b) district-level administrative divisions. 21
Figure 2. Epidemic curve of COVID-19 in Nepal. 23
Figure 3. Chronology of key events of COVID-19 in Nepal. 24
Figure 4. Histogram of the cumulative COVID-19 incidence. 36
Figure 5. Model validation by DHARMa package in R. 49
Figure 6. Spatial distribution of COVID-19 incidence per 100,000 population in Nepal. 50
Figure 7. Moran's scatter plot for COVID-19 incidence in Nepal. 53
Figure 8. Getis-Ord Gi* clustering of COVID-19 incidence in Nepal. 54
Figure 9. Interactive map for neighborhood structure of Nepal at district level. 55
Figure 10. Model validation for the BYM model using residuals versus fitted plot. 62
Figure 11. Observed COVID-19 cases versus that predicted by the INLA-BYM model. 63
Figure 12. COVID-19 cases predicted by the INLA-BYM model versus error in the prediction. 63