Doctoral Degrees ((SACEMA) South African Centre for Epidemiological Modelling and Analysis)

Permanent URI for this collection

https://scholar.sun.ac.za/handle/10019.1/109561

Browse

Recent Submissions

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Modelling outbreak response intervention strategies for decision-making
    (Stellenbosch : Stellenbosch University, 2022-04) Azam, James Mba; Pulliam, Juliet R. C.; Pulliam, Juliet Rachel Crowder, 1979-; Ferrari, Matthew J.; Faulty of Medicine and Health Sciences. SACEMA: South African Centre for Epidemiological Modelling and Analysis.
    ENGLISH SUMMARY: Outbreaks of infectious diseases like measles and COVID-19 continue to threaten public health. Consequently, outbreak response decision-making is in constant need of advancements. Mathematical models of infectious diseases, which provide evidence based insights about pathogen spread and the impact of interventions, form an essential component of outbreak response decision-making. In this dissertation, I make three contributions in that regard. First, I conduct a systematic review of modelling studies, published during 1970-2019, that assessed the impact of reactive interventions on real and hypothetical outbreaks of human vaccine-preventable diseases and foot-and-mouth disease (FMD). I extract data including the author affiliation type (academic institution, governmental, and non-governmental organizations), whether there was an author based in the country studied, impact of vaccination, model characteristics, and modelling practices. I use the author affiliation types to group papers into two collaboration types namely, purely academic (papers with only academic author affiliations) and mixed (all other combinations). I analyse time and geographic patterns and differences in model characteristics and practices overall and between the collaboration types. I find that, in the human disease literature, mixed collaborations increased in the past decade, more often included authors based in the country studied and used more complex modelling practices. These patterns could indicate an increased recognition of modelling by decision makers or increased interaction between modellers and decision-makers. Additionally, I observe some contrasting patterns between the human and FMD literature. Second, I investigate the relative logistical and epidemiological benefits of outside cold chain (OCC) delivery of measles vaccines during an outbreak in a hypothetical setting assuming vaccine cold chain challenges. I extend a transmission dynamic model to incorporate key logistical requirements of several cold chain and OCC strategies. I find that OCC delivery of measles vaccines during outbreaks could lead to shorter campaigns, high vaccination coverage, and higher cases averted. Finally, I show how the emergence of a variant strain affects the minimally sufficient response strategy needed to mitigate the outbreak of a hypothetical pathogen. I develop compartmental models that allow the introduction of more transmissible variants with the ability to escape vaccine-induced immunity. I investigate the impact on outbreak size and peak prevalence targets of implementing either vaccination alone or a combination of vaccination and non-pharmaceutical interventions (NPIs). I show that the use of vaccination alone would require a high vaccination coverage and rapid rollout speed but adding on NPIs could reduce the vaccination coverage and rollout speed required to achieve the targets while also accounting for the risk of variant emergence. In conclusion, this dissertation makes advances that could potentially: (1) initiate discussions on the impact of modelling in outbreak response and the need for increased collaboration between model developers and users, (2) lead to advocacy towards innovations in measles outbreak response vaccination, and (3) contribute to the theory of outbreak response planning, especially in the way uncertainty regarding the potential emergence of a variant is considered in outbreak response decision-making.
  • Thumbnail Image
    Item
    Extinction probabilities for tsetse (Glossina spp.) in a world of changing climate
    (Stellenbosch : Stellenbosch University, 2020-12) Are, Elisha Bayode; Hargrove, John; Faulty of Medicine and Health Sciences. SACEMA: South African Centre for Epidemiological Modelling and Analysis
    ENGLISH ABSTRACT: Tsetse (Glossina spp) transmit trypanosomiases, as sleeping sickness to humans and nagana to livestock. These continue to have negative impacts on health and wealth in the African continent. In recent years, treatment, and vector control, have helped to reduced disease burden and the World Health Organization set 2020 as a target year for eliminating the disease in humans. Tsetse populations have also declined in parts of Africa due to climate change and human encroachment. In the fight against trypanosomiases, there is a continuing need to improve understanding of tsetse population dynamics – and particularly the conditions under which tsetse populations persist, and the implications for tsetse control/eradication in a changing world. We explore here five primary objectives. Firstly, we revisit a branching process model developed for tsetse population growth and estimates for extinction probabilities. We improve the model by modifying it to work for more realistic situations where, for example, male to female sex ratios in the population are not necessarily one-to-one. We estimate extinction probabilities as a function of the probability that a deposited larva is female, and show that tsetse populations will thrive better when there are slightly more females than males in the population. We confirm that daily mortality rates 3.5% ensure eradication of closed populations of tsetse. Secondly, we simplify the mathematical derivation of earlier estimate for extinction probabilities and carry out global uncertainty and sensitivity analyses on extinction probabilities, using Latin Hypercube Sampling and Partial Rank Correlation Coefficient methods. We show that adult female mortality has the highest correlation with extinction probability. We caution that a new tsetse control method, which proposes a strategy combining Sterile Insect Techniques (SIT) with increased pupal mortality, may not offer any added benefit for tsetse eradication. Thirdly, we estimate extinction probabilities, times to extinction and growth rates as a function of temperature for tsetse populations. We provide temperature bounds for tsetse persistence, and suggest that future control efforts should consider the impact of changing climate on the distribution and abundance of tsetse populations. Fourthly, we develop a general model for tsetse population persistence, and show that previous models are special cases of our current model. While extinction probabilities are sensitive to changes in the point of the life cycle at which we count the population, the reproduction number is independent of the counting point chosen. Finally, we derive the intrinsic rate of increase for tsetse populations using the Euler-Lotka equation. We use temperature data, and tsetse population estimates from a mark–recapture exercise, to test our model’s validity, and show that our results are comparable to estimates derived from the data. We estimate the intrinsic rate of increase for tsetse populations in the neighbourhood of Rekomitjie Research Station in Zimbabwe, using as input average daily temperatures from 1960–2018. We created multiple climate change scenarios, using 2018 daily temperatures as a baseline. We predict that a warming rate of 0.08°C per-year could drive tsetse populations to extinction in the neighbourhood of Rekomitjie within the next 50 years.