Publications

There is limited data on HIV evolutionary trends in African populations. We evaluated changes in HIV viral diversity and genetic divergence in southern Uganda over a twenty-four-year period spanning the introduction and scale-up of HIV prevention and treatment programs using HIV sequence and survey data from the Rakai Community Cohort Study, an open longitudinal population-based HIV surveillance cohort. Gag (p24) and env (gp41) HIV data were generated from persons living with HIV (PLHIV) in 31 inland semi-urban trading and agrarian communities (1994 to 2018) and four hyperendemic Lake Victoria fishing communities (2011 to 2018) under continuous surveillance. HIV subtype was assigned using the Recombination Identification Program with phylogenetic confirmation. Inter-subtype diversity was evaluated using the Shannon diversity index and intra-subtype diversity with the nucleotide diversity and pairwise TN93 genetic distance. Genetic divergence was measured using root-to-tip distance and pairwise TN93 genetic distance analyses. Demographic history of HIV was inferred using a coalescent-based Bayesian Skygrid model. Evolutionary dynamics were assessed among demographic and behavioral population sub-groups, including by migration status. 9,931 HIV sequences were available from 4,999 PLHIV, including 3,060 and 1,939 persons residing in inland and fishing communities, respectively. In inland communities, subtype A1 viruses proportionately increased from 14.3% in 1995 to 25.9% in 2017 (p<0.001), while those of subtype D declined from 73.2% in 1995 to 28.2% in 2017 (p<0.001). The proportion of viruses classified as recombinants significantly increased by nearly four-fold from 12.2% in 1995 to 44.8% in 2017. Inter-subtype HIV diversity has generally increased. While intra-subtype p24 genetic diversity and divergence leveled off after 2014, intra-subtype gp41 diversity, effective population size, and divergence increased through 2017. Intra- and inter-subtype viral diversity increased across all demographic and behavioral population sub-groups, including among individuals with no recent migration history or extra-community sexual partners. This study provides insights into population-level HIV evolutionary dynamics following the scale-up of HIV prevention and treatment programs. Continued molecular surveillance may provide a better understanding of the dynamics driving population HIV evolution and yield important insights for epidemic control and vaccine development.

Since the coining of the term phylodynamics, the use of phylogenies to understand infectious disease dynamics has steadily increased. As methods for phylodynamics and genomic epidemiology have proliferated and grown more computationally expensive, the epidemiological information they extract has also evolved to better complement what can be learned through traditional epidemiological data. However, for genomic epidemiology to continue to grow, and for the accumulating number of pathogen genetic sequences to fulfill their potential widespread utility, the extraction of epidemiological information from phylogenies needs to be simpler and more efficient. Summary statistics provide a straightforward way of extracting information from a phylogenetic tree, but the relationship between these statistics and epidemiological quantities needs to be better understood. In this work we address this need via simulation. Using two different benchmark scenarios, we evaluate 74 tree summary statistics and their relationship to epidemiological quantities. In addition to evaluating the epidemiological information that can be inferred from each summary statistic, we also assess the computational cost of each statistic. This helps us optimize the selection of summary statistics for specific applications. Our study offers guidelines on essential considerations for designing or choosing summary statistics. The evaluated set of summary statistics, along with additional helpful functions for phylogenetic analysis, is accessible through an open-source Python library. Our research not only illuminates the main characteristics of many tree summary statistics but also provides valuable computational tools for real-world epidemiological analyses. These contributions aim to enhance our understanding of disease spread dynamics and advance the broader utilization of genomic epidemiology in public health efforts.

Previous work has shown that primary healthcare facilities can benefit from both in-kind support (e.g., medication shipments) as well as increased cash-on-hand to spend to address service readiness gaps. However, there is limited evidence on how facility managers choose to spend available cash or how their decisions to manage their facility budgets are affected by in-kind support. Economic theory suggests that the optimal allocation of cash resources would depend on the context and constraints to how it can be spent, and expenditures would in turn affect the availability of supplies and medications. We test this theory using regression analysis on data from the Nigeria Service Delivery Indicators for Health (SDI), a health facility survey from twelve states in 2013 that included both hospitals and primary healthcare centers (PHCs). We find that facilities with financial resources available to them have higher availability of essential medicines, especially if the facility had earmarked some cash for medication expenditures. However, earmarking for other expenditure categories did not have the same effect on medication availability, which indicates that budgeting processes are an important factor in ensuring medication availability. We find that cash support had large effect (p < 0.001) on availability and that in-kind donations had a negative effect on the probability of expenditure of medications. Additionally, we find the difference between hospitals and PHCs is due to their financial situation (variables become insignificant once support variables were in regressions). Regression analyses also showed that facilities that received in-kind medications had higher availability, but this only had a significant effect in facilities that did not have cash available to spend on medications, implying that facilities are able to address their own supply needs when they have resources available to them. Thus, in-kind supplies should be targeted to facilities that cannot otherwise procure them. Overall, facilities appear to be making effective trade-offs in the context of limited resources and they should receive both cash and support for appropriate budgeting and procurement practices.

Previous studies have shown that facility autonomy, especially control over budget allocation, can have a modest positive effect on performance, but the findings depend on the context. Similarly, management practices are often cited as important contributors to facility performance, but the evidence is limited and usually qualitative. Data from the large-scale randomized evaluation of the Nigeria States Health Investment Project (NSHIP) offers an opportunity to quantitatively examine these relationships in the context of a lower middle-income country. We utilize non-parametric statistics to test for difference in means and apply regression analysis to test the hypothesis that autonomy and management affected facility performance. Our results show that facilities with greater autonomy, more budget control, and better management practices generally outperform their peers on a range of facility readiness and service delivery measures. For example, regression results found that facilities with high autonomy held on average 2.1 more outreach sessions per month than those without, and facilities with an annual business plan offered 1.8 additional outreach services. Supervision practices, such as more frequent visits and use of a quantitative checklist, are associated with 26% higher productivity and up to a 28.6% increase in equipment availability (percentage points), respectively. We conduct sensitivity analyses on our variable selection and use a random forest approach to validate that results are robust to changes in the model structure. We conclude that facility-level autonomy and especially budget control can improve primary healthcare facility readiness and service availability, even in resource-constrained contexts, Further, this can be achieved through good management practices that are reinforced through supportive supervision and routine performance monitoring to maximize the gains that result from incremental financing. This shows that these policies and practices can be critical contributors to efficiently achieving the goals of universal healthcare policies in the context of limited resources.

To improve service delivery of Nigeria’s primary health care (PHC) system, the government tested two approaches for facility-level financing: performance-based financing (PBF) and decentralized facility financing (DFF). Facilities also had increased autonomy, supervision, and community oversight. We examine how the approach, funding level, and state context affected breadth of services and structural quality. We use health facility surveys previously collected in 2014 and 2017, covering three years of implementation, in which districts were randomly assigned PBF or DFF and compared to matched districts in control states. We use log-linear regressions and non-parametric statistics to estimate the effect size of the financing approach and level of funding per capita. Service availability was highest in PBF facilities, while DFF also outperformed control on most measures. Results showed that structural readiness and service offerings both increased with more funding, especially under DFF. DFF and PBF facilities were better equipped to provide services that they claimed to offer, which was not the case for controls. Overall, PBF outperformed DFF, partially explained by funding levels. The rate of offering complimentary services followed a pattern of easiest-to-hardest to deliver. PBF and DFF both improved the breadth and structural quality of services, although DFF performance was more sensitive to funding levels. Improvements were observed at relatively low levels of funding, but larger investments were associated with better performance. Most DFF facilities exceeded the performance of higher-funded controls, implying that funding was more valuable in the context of autonomy, increased supervision, and community oversight.

Rubella infection is typically mild or asymptomatic except when infection occurs during pregnancy. Infection in early pregnancy can cause miscarriage, stillbirth, or congenital rubella syndrome. Only individuals that are still susceptible to rubella infection during child-bearing age are vulnerable to this burden. Rubella-containing vaccine (RCV) is safe and effective, providing life-long immunity. However, average age-at-infection increases with increasing vaccination coverage, which could potentially lead to increased disease burden if the absolute risk of infection during child-bearing age increases. The dynamics of rubella transmission were explored using EMOD, a software tool for building stochastic, agent-based infection models. Simulations of pre-vaccine, endemic transmission of rubella virus introduced RCV at varying levels of coverage to determine the expected future trajectories of disease burden. Introducing RCV reduces both rubella virus transmission and disease burden for a period of around 15 years. Increased disease burden is only possible more than a decade post-introduction, and only for contexts with persistently high transmission intensity. Low or declining rubella virus transmission intensity is associated with both greater burden without vaccination and greater burden reduction with vaccination. The risk of resurgent burden due to incomplete vaccination only exists for locations with persistently high infectivity, high connectivity, and high fertility. A trade-off between the risk of a small, future burden increase versus a large, immediate burden decrease strongly favors RCV introduction.

In 2020, the WHO launched its first global strategy to accelerate the elimination of cervical cancer, outlining an ambitious set of targets for countries to achieve over the next decade. At the same time, new tools, technologies, and strategies are in the pipeline that may improve screening performance, expand the reach of prophylactic vaccines, and prevent the acquisition, persistence and progression of oncogenic HPV. Detailed mechanistic modelling can help identify the combinations of current and future strategies to combat cervical cancer. Open-source modelling tools are needed to shift the capacity for such evaluations in-country. Here, we introduce the Human papillomavirus simulator (HPVsim), a new open-source software package for creating flexible agent-based models parameterised with country-specific vital dynamics, structured sexual networks, and co-transmitting HPV genotypes. HPVsim includes a novel methodology for modelling cervical disease progression, designed to be readily adaptable to new forms of screening. The software itself is implemented in Python, has built-in tools for simulating commonly-used interventions, includes a comprehensive set of tests and documentation, and runs quickly (seconds to minutes) on a laptop. Performance is greatly enhanced by HPVsim’s multiscale modelling functionality. HPVsim is open source under the MIT License and available via both the Python Package Index (via pip install) and GitHub (hpvsim.org).

Somalia is a complex and fragile setting with a demonstrated potential for disruptive, high-burden measles outbreaks. In response, since 2018, Somalian authorities have partnered with UNICEF and the WHO to implement measles vaccination campaigns across the country. In this paper, we create a Somalia-specific model of measles transmission based on a comprehensive epidemiological dataset including case-based surveillance, vaccine registries, and serological surveys. We use this model to assess the impact of these campaign interventions on Somalian’s measles susceptibility, showing, for example, that across the roughly 10 million doses delivered, 1 of every 5 immunized a susceptible child. Finally, we use the model to explore a counter-factual epidemiology without the 2019–2020 campaigns, and we estimate that those interventions prevented over 10,000 deaths.

Background Novel oral poliovirus vaccine (OPV) type 2 (nOPV2) has been made available for outbreak response under an emergency use listing authorization based on supportive clinical trial data. Since 2021 more than 350 million doses of nOPV2 were used for control of a large outbreak of circulating vaccine-derived poliovirus type 2 (cVDPV2) in Nigeria. Methods Using a bayesian time-series susceptible-infectious-recovered model, we evaluate the field effectiveness of nOPV2 immunization campaigns in Nigeria compared with campaigns using monovalent OPV type 2 (mOPV2). Results We found that both nOPV2 and mOPV2 campaigns were highly effective in reducing transmission of cVDPV2, on average reducing the susceptible population by 42% (95% confidence interval, 28–54%) and 38% (20–51%) per campaign, respectively, which were indistinguishable from each other in this analysis (relative effect, 1.1 [.7–1.9]). Impact was found to vary across areas and between immunization campaigns. Conclusions These results are consistent with the comparable individual immunogenicity of nOPV2 and mOPV2 found in clinical trials but also suggest that outbreak response campaigns may have small impacts in some areas requiring more campaigns than are suggested in current outbreak response procedures.

Access to medical treatment for fever is essential to prevent morbidity and mortality in individuals and to prevent transmission of communicable febrile illness in communities. Quantification of the rates at which treatment is accessed is critical for health system planning and a prerequisite for disease burden estimates. In this study, national data on the proportion of children under five years old with fever who were taken for medical treatment were collected from all available countries in Africa, Latin America, and Asia (n = 91). We used generalised additive mixed models to estimate 30-year trends in the treatment-seeking rates across the majority of countries in these regions (n = 151). Our results show that the proportions of febrile children brought for medical treatment increased steadily over the last 30 years, with the greatest increases occurring in areas where rates had originally been lowest, which includes Latin America and Caribbean, North Africa and the Middle East (51 and 50% increase, respectively), and Sub-Saharan Africa (23% increase). Overall, the aggregated and population-weighted estimate of children with fever taken for treatment at any type of facility rose from 61% (59–64 95% CI) in 1990 to 71% (69–72 95% CI) in 2020. The overall population-weighted average for fraction of treatment in the public sector was largely unchanged during the study period: 49% (42–58 95% CI) sought care at public facilities in 1990 and 47% (44–52 95% CI) in 2020. Overall, the findings indicate that improvements in access to care have been made where they were most needed, but that despite rapid initial gains, progress can plateau without substantial investment. In 2020 there remained significant gaps in care utilisation that must be factored in when developing control strategies and deriving disease burden estimates.