Mathematical understanding of the multi-species biological networks underpinning rabies transmission dynamics in southern Tanzania
File(s)
Author(s)
Hayes, Sarah
Type
Thesis
Abstract
Each year, an estimated 59,000 people globally die from rabies, a fully vaccine-preventable disease. Most of these deaths occur in low- and middle-income countries and affect some of the world’s poorest communities. Domestic dogs are responsible for 99% of rabies transmissions to humans. “Zero by 30” is a global strategic plan to achieve zero dog-mediated human rabies deaths by 2030 and domestic dog vaccination is a key component of this plan. Across much of Africa, domestic dogs are the only species considered essential for the maintenance of rabies but there are reports of independent cycles of rabies transmission in wildlife in some areas. Understanding the local rabies transmission dynamics is vital when planning control strategies. In this thesis a variety of quantitative methods are used to improve our understanding of the multi-species networks underpinning the transmission of rabies within animals in south-east Tanzania, where wildlife contribute over 40% of the probable animal rabies cases. Widespread domestic dog vaccination campaigns were implemented during the first six years of the study enabling exploration of the impact of domestic dog rabies vaccination on animal rabies cases in this area of unusually high levels of wildlife rabies.
Between January 2011 and July 2019 data on probable animal rabies cases were collected in the thirteen districts of the Lindi and Mtwara regions of south-east Tanzania. Hospital records of animal-bite patients presenting to healthcare facilities were used as sentinels for animal contact tracing. Over the eight-and-a-half-year study period 549 probable animal rabies cases were reported. Of these 303 (55%) were in domestic dogs and 221 (40%) were in jackals.
Generalised additive models fitted to monthly incidence data demonstrated clear evidence of a decrease in incidence in probable rabies cases in both domestic dogs and jackals over most of the study period with an increase in incidence in domestic dogs observed in 2017/2018. This increase is considered likely to be due to a waning in population-level immunity following the cessation of widespread vaccination campaigns in late 2016. A common seasonal trend in monthly incidence in probable rabies cases was found for domestic dogs and jackals. The best-fitting generalised linear model fitted to district-level data on the annual incidence of probable animal rabies cases found that increased domestic dog vaccination was associated with decreased incidence of probable rabies cases in both dogs and jackals.
Probabilistic transmission trees were constructed to assess the levels of within- and between-species transmission occurring over the study period. We found evidence of sustained chains of transmission within wildlife and frequent cross-species transmission events. Using logistic regression, a statistically significant positive relationship was identified between the proportion of jackal cases in a district and the prevalence of jackals within the susceptible population.
Finally, an existing method for outbreak cluster detection was extended to incorporate group-specific distributional parameters and reporting probabilities as well as assortative transmission between species. This newly developed generalised method was applied to the south-east Tanzania data. Of the scenarios evaluated, those that best fitted the data were those in which the distributional parameters (parameters for the serial interval distribution and distance kernel) were the same across both groups (domestic animals and wildlife) but reporting probabilities for domestic animals were 2.0 to 2.5 times higher than those of wildlife and a moderate level of assortative transmission was incorporated.
The results presented throughout this thesis suggest that whilst jackals are an important part of the rabies reservoir in the study area, they are unlikely to be able to maintain rabies independently of domestic dogs. We also present evidence for the effectiveness of domestic dog vaccination in reducing animal rabies cases in all species, even in areas with high levels of rabies within wildlife. Overall, our findings support the role of domestic dog vaccination as a key component of the Zero by 30 strategic plan.
Between January 2011 and July 2019 data on probable animal rabies cases were collected in the thirteen districts of the Lindi and Mtwara regions of south-east Tanzania. Hospital records of animal-bite patients presenting to healthcare facilities were used as sentinels for animal contact tracing. Over the eight-and-a-half-year study period 549 probable animal rabies cases were reported. Of these 303 (55%) were in domestic dogs and 221 (40%) were in jackals.
Generalised additive models fitted to monthly incidence data demonstrated clear evidence of a decrease in incidence in probable rabies cases in both domestic dogs and jackals over most of the study period with an increase in incidence in domestic dogs observed in 2017/2018. This increase is considered likely to be due to a waning in population-level immunity following the cessation of widespread vaccination campaigns in late 2016. A common seasonal trend in monthly incidence in probable rabies cases was found for domestic dogs and jackals. The best-fitting generalised linear model fitted to district-level data on the annual incidence of probable animal rabies cases found that increased domestic dog vaccination was associated with decreased incidence of probable rabies cases in both dogs and jackals.
Probabilistic transmission trees were constructed to assess the levels of within- and between-species transmission occurring over the study period. We found evidence of sustained chains of transmission within wildlife and frequent cross-species transmission events. Using logistic regression, a statistically significant positive relationship was identified between the proportion of jackal cases in a district and the prevalence of jackals within the susceptible population.
Finally, an existing method for outbreak cluster detection was extended to incorporate group-specific distributional parameters and reporting probabilities as well as assortative transmission between species. This newly developed generalised method was applied to the south-east Tanzania data. Of the scenarios evaluated, those that best fitted the data were those in which the distributional parameters (parameters for the serial interval distribution and distance kernel) were the same across both groups (domestic animals and wildlife) but reporting probabilities for domestic animals were 2.0 to 2.5 times higher than those of wildlife and a moderate level of assortative transmission was incorporated.
The results presented throughout this thesis suggest that whilst jackals are an important part of the rabies reservoir in the study area, they are unlikely to be able to maintain rabies independently of domestic dogs. We also present evidence for the effectiveness of domestic dog vaccination in reducing animal rabies cases in all species, even in areas with high levels of rabies within wildlife. Overall, our findings support the role of domestic dog vaccination as a key component of the Zero by 30 strategic plan.
Version
Open Access
Date Issued
2022-01
Date Awarded
2022-10
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Donnelly, Christl
Nouvellet, Pierre
Sponsor
Engineering and Physical Sciences Research Council
Publisher Department
School of Public Health
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)