Nicky McCreesh

Evaluation of Respondent-driven Sampling

Background: Respondent-driven sampling is a novel variant of link-tracing sampling for estimating the characteristics of hard-to-reach groups, such as HIV prevalence in sex workers. Despite its use by leading health organizations, the performance of this method in realistic situations is still largely unknown. We evaluated respondent-driven sampling by comparing estimates from a respondent-driven sampling survey with total population data. Methods: Total population data on age, tribe, religion, socioeconomic status, sexual activity, and HIV status were available on a population of 2402 male household heads from an open cohort in rural Uganda. A respondent-driven sampling (RDS) survey was carried out in this population, using current methods of sampling (RDS sample) and statistical inference (RDS estimates). Analyses were carried out for the full RDS sample and then repeated for the first 250 recruits (small sample). Results: We recruited 927 household heads. Full and small RDS samples were largely representative of the total population, but both samples underrepresented men who were younger, of higher socioeconomic status, and with unknown sexual activity and HIV status. Respondent-driven sampling statistical inference methods failed to reduce these biases. Only 31%–37% (depending on method and sample size) of RDS estimates were closer to the true population proportions than the RDS sample proportions. Only 50%–74% of respondent-driven sampling bootstrap 95% confidence intervals included the population proportion. Conclusions: Respondent-driven sampling produced a generally representative sample of this well-connected nonhidden population. However, current respondent-driven sampling inference methods failed to reduce bias when it occurred. Whether the data required to remove bias and measure precision can be collected in a respondent-driven sampling survey is unresolved. Respondent-driven sampling should be regarded as a (potentially superior) form of convenience sampling method, and caution is required when interpreting findings based on the sampling method.

Challenges in predicting the effects of climate change on Schistosoma mansoni and Schistosoma haematobium transmission potential.

Climate change will inevitably influence both the distribution of Schistosoma mansoni and Schistosoma haematobium and the incidence of schistosomiasis in areas where it is currently endemic, and impact on the feasibility of schistosomiasis control and elimination goals. There are several limitations of current models of climate and schistosome transmission, and substantial gaps in empirical data that impair model development. In this review we consider how temperature, precipitation, heat waves, drought, and flooding could impact on snail and schistosome population dynamics. We discuss how widely used degree day models of schistosome development may not be accurate at lower temperatures, and highlight the need for further research to improve our understanding of the relationship between air and water temperature and schistosome and snail development.

Evaluation of the role of location and distance in recruitment in respondent-driven sampling

BackgroundRespondent-driven sampling(RDS) is an increasingly widely used variant of a link tracing design for recruiting hidden populations. The role of the spatial distribution of the target population has not been robustly examined for RDS. We examine patterns of recruitment by location, and how they may have biased an RDS study findings.MethodsTotal-population data were available on a range of characteristics on a population of 2402 male household-heads from an open cohort of 25 villages in rural Uganda. The locations of households were known a-priori. An RDS survey was carried out in this population, employing current RDS methods of sampling and statistical inference.ResultsThere was little heterogeneity in the population by location. Data suggested more distant contacts were less likely to be reported, and therefore recruited, but if reported more distant contacts were as likely as closer contacts to be recruited. There was no evidence that closer proximity to a village meeting place was associated with probability of being recruited, however it was associated with a higher probability of recruiting a larger number of recruits. People living closer to an interview site were more likely to be recruited.ConclusionsHousehold location affected the overall probability of recruitment, and the probability of recruitment by a specific recruiter. Patterns of recruitment do not appear to have greatly biased estimates in this study. The observed patterns could result in bias in more geographically heterogeneous populations. Care is required in RDS studies when choosing the network size question and interview site location(s).

Bayesian History Matching of Complex Infectious Disease Models Using Emulation: A Tutorial and a Case Study on HIV in Uganda.

Advances in scientific computing have allowed the development of complex models that are being routinely applied to problems in disease epidemiology, public health and decision making. The utility of these models depends in part on how well they can reproduce empirical data. However, fitting such models to real world data is greatly hindered both by large numbers of input and output parameters, and by long run times, such that many modelling studies lack a formal calibration methodology. We present a novel method that has the potential to improve the calibration of complex infectious disease models (hereafter called simulators). We present this in the form of a tutorial and a case study where we history match a dynamic, event-driven, individual-based stochastic HIV simulator, using extensive demographic, behavioural and epidemiological data available from Uganda. The tutorial describes history matching and emulation. History matching is an iterative procedure that reduces the simulators input space by identifying and discarding areas that are unlikely to provide a good match to the empirical data. History matching relies on the computational efficiency of a Bayesian representation of the simulator, known as an emulator. Emulators mimic the simulators behaviour, but are often several orders of magnitude faster to evaluate. In the case study, we use a 22 input simulator, fitting its 18 outputs simultaneously. After 9 iterations of history matching, a non-implausible region of the simulator input space was identified that was times smaller than the original input space. Simulator evaluations made within this region were found to have a 65% probability of fitting all 18 outputs. History matching and emulation are useful additions to the toolbox of infectious disease modellers. Further research is required to explicitly address the stochastic nature of the simulator as well as to account for correlations between outputs.

Respondent Driven Sampling: Determinants of Recruitment and a Method to Improve Point Estimation

Introduction Respondent-driven sampling (RDS) is a variant of a link-tracing design intended for generating unbiased estimates of the composition of hidden populations that typically involves giving participants several coupons to recruit their peers into the study. RDS may generate biased estimates if coupons are distributed non-randomly or if potential recruits present for interview non-randomly. We explore if biases detected in an RDS study were due to either of these mechanisms, and propose and apply weights to reduce bias due to non-random presentation for interview. Methods Using data from the total population, and the population to whom recruiters offered their coupons, we explored how age and socioeconomic status were associated with being offered a coupon, and, if offered a coupon, with presenting for interview. Population proportions were estimated by weighting by the assumed inverse probabilities of being offered a coupon (as in existing RDS methods), and also of presentation for interview if offered a coupon by age and socioeconomic status group. Results Younger men were under-recruited primarily because they were less likely to be offered coupons. The under-recruitment of higher socioeconomic status men was due in part to them being less likely to present for interview. Consistent with these findings, weighting for non-random presentation for interview by age and socioeconomic status group greatly improved the estimate of the proportion of men in the lowest socioeconomic group, reducing the root-mean-squared error of RDS estimates of socioeconomic status by 38%, but had little effect on estimates for age. The weighting also improved estimates for tribe and religion (reducing root-mean-squared-errors by 19–29%), but had little effect for sexual activity or HIV status. Conclusions Data collected from recruiters on the characteristics of men to whom they offered coupons may be used to reduce bias in RDS studies. Further evaluation of this new method is required.

Exploring the potential impact of a reduction in partnership concurrency on HIV incidence in rural Uganda: a modeling study.

Background: A number of African countries have planned campaigns against concurrency. It will not be possible to separate the effects of a reduction in concurrency from other behavior changes when evaluating these campaigns. This modeling study explores the potential impact of an intervention to reduce partnership concurrency on HIV incidence in contemporary rural Uganda, keeping incidence of sex acts and partnerships in the population constant. Methods: Data on demography, sexual behavior, and HIV prevalence from Uganda were used to parameterize an individual-based HIV transmission model. Three baseline model scenarios were simulated, representing the best estimate of concurrency prevalence in this population, and low and high plausible bounds. Interventions that reduced concurrency by 20% and 50% between 2010 and 2020 were simulated, and the impact on HIV incidence in 2020 was calculated. Results: Data showed 9.6% (7.9%–11.4%) of men and 0.2% (0.0%–0.4%) of women reported concurrency in 2008. Reducing concurrency had a nonlinear impact on HIV incidence. A 20% reduction in concurrency reduced HIV incidence by 4.1% (0.3%–5.7%) in men and 9.2% (2.1%–16.8%) in women; a 50% reduction in concurrency reduced HIV incidence by 6.0% (1.4%–10.8%) in men and 16.2% (6.3%–23.4%) in women. Conclusions: Interventions against concurrency have the potential to reduce HIV incidence and may have a higher impact in women than in men. In rural Uganda, overall impact was modest, and this study does not provide strong support for the prioritization of concurrency as a target for behavior change interventions. However, it may be more useful in higher concurrency settings and for reducing HIV incidence in women.

The effect of increasing water temperatures on Schistosoma mansoni transmission and Biomphalaria pfeifferi population dynamics: an agent-based modelling study.

Introduction There is increasing interest in the control and elimination of schistosomiasis. Little is known, however, about the likely effects of increasing water-body temperatures on transmission. Methods We have developed an agent-based model of the temperature-sensitive stages of the Schistosoma and intermediate host snail life-cycles, parameterised using data from S. mansoni and Biomphalaria pfeifferi laboratory and field-based observations. Infection risk is calculated as the number of cercariae in the model, adjusted for their probability of causing infection. Results The number of snails in the model is approximately constant between 15–31°C. Outside this range, snail numbers drop sharply, and the snail population cannot survive outside the range 14–32°C. Mean snail generation time decreases with increasing temperature from 176 days at 14°C to 46 days at 26°C. Human infection risk is highest between 16–18°C and 1 pm and 6–10 pm in calm water, and 20–25°C and 12–4 pm in flowing water. Infection risk increases sharply when temperatures increase above the minimum necessary for sustained transmission. Conclusions The model suggests that, in areas where S. mansoni is already endemic, warming of the water at transmission sites will have differential effects on both snails and parasites depending on abiotic properties of the water-body. Snail generation times will decrease in most areas, meaning that snail populations will recover faster from natural population reductions and from snail-control efforts. We suggest a link between the ecological properties of transmission sites and infection risk which could significantly affect the outcomes of interventions designed to alter water contact behaviour – proposing that such interventions are more likely to reduce infection levels at river locations than lakes, where infection risk remains high for longer. In cooler areas where snails are currently found, increasing temperatures may significantly increase infection risk, potentially leading to new, high-intensity foci of infection.

The effect of simulating different intermediate host snail species on the link between water temperature and schistosomiasis risk.

Introduction A number of studies have attempted to predict the effects of climate change on schistosomiasis risk. The importance of considering different species of intermediate host snails separately has never previously been explored. Methods An agent-based model of water temperature and Biomphalaria pfeifferi population dynamics and Schistosoma mansoni transmission was parameterised to two additional species of snail: B. glabrata and B. alexandrina. Results Simulated B. alexandrina populations had lower minimum and maximum temperatures for survival than B. pfeifferi populations (12.5–29.5°C vs. 14.0–31.5°C). B. glabrata populations survived over a smaller range of temperatures than either B. pfeifferi or B. alexandrina (17.0°C–29.5°C). Infection risk peaked at 16.5°C, 25.0°C and 19.0°C respectively when B. pfeifferi, B. glabrata and B. alexandrina were simulated. For all species, infection risk increased sharply once a minimum temperature was reached. Conclusions The results from all three species suggest that infection risk may increase dramatically with small increases in temperature in areas at or near the currents limits of schistosome transmission. The effect of small increases in temperature in areas where schistosomiasis is currently found will depend both on current temperatures and on the species of snail acting as intermediate host(s) in the area. In most areas where B. pfeifferi is the host, infection risk is likely to decrease. In cooler areas where B. glabrata is the host, infection risk may increase slightly. In cooler areas where B. alexandrina is the host, infection risk may more than double with only 2°C increase in temperature. Our results show that it is crucial to consider the species of intermediate host when attempting to predict the effects of climate change on schistosomiasis.

Combining process-based and correlative models improves predictions of climate change effects on Schistosoma mansoni transmission in eastern Africa

Currently, two broad types of approach for predicting the impact of climate change on vector-borne diseases can be distinguished: i) empirical-statistical (correlative) approaches that use statistical models of relationships between vector and/or pathogen presence and environmental factors; and ii) process-based (mechanistic) approaches that seek to simulate detailed biological or epidemiological processes that explicitly describe system behavior. Both have advantages and disadvantages, but it is generally acknowledged that both approaches have value in assessing the response of species in general to climate change. Here, we combine a previously developed dynamic, agentbased model of the temperature-sensitive stages of the Schistosoma mansoni and intermediate host snail lifecycles, with a statistical model of snail habitat suitability for eastern Africa. Baseline model output compared to empirical prevalence data suggest that the combined model performs better than a temperature-driven model alone, and highlights the importance of including snail habitat suitability when modeling schistosomiasis risk. There was general agreement among models in predicting changes in risk, with 24-36% of the eastern Africa region predicted to experience an increase in risk of up-to 20% as a result of increasing temperatures over the next 50 years. Vice versa the models predicted a general decrease in risk in 30-37% of the study area. The snail habitat suitability models also suggest that anthropogenically altered habitat play a vital role for the current distribution of the intermediate snail host, and hence we stress the importance of accounting for land use changes in models of future changes in schistosomiasis risk.

Effect of water temperature and population density on the population dynamics of Schistosoma mansoni intermediate host snails.

BackgroundMathematical models can be used to identify areas at risk of increased or new schistosomiasis transmission as a result of climate change. The results of these models can be very different when parameterised to different species of host snail, which have varying temperature preferences. Currently, the experimental data needed by these models are available for only a few species of snail. The choice of density-dependent functions can also affect model results, but the effects of increasing densities on Biomphalaria populations have only previously been investigated in artificial aquariums.MethodsLaboratory experiments were conducted to estimate Biomphalaria sudanica mortality, fecundity and growth rates at ten different constant water temperatures, ranging from 13-32°C. Snail cages were used to determine the effects of snail densities on B. sudanica and B. stanleyi mortality and fecundity rates in semi-natural conditions in Lake Albert.ResultsB. sudanica survival and fecundity were highest at 20°C and 22°C respectively. Growth in shell diameter was estimated to be highest at 23°C in small and medium sized snails, but the relationship between temperature and growth was not clear. The fecundity of both B. sudanica and B. stanleyi decreased by 72-75% with a four-fold increase in population density. Increasing densities four-fold also doubled B. stanleyi mortality rates, but had no effect on the survival of B. sudanica.ConclusionsThe optimum temperature for fecundity was lower for B. sudanica than for previously studied species of Biomphalaria. In contrast to other Biomphalaria species, B. sudanica have a distinct peak temperature for survival, as opposed to a plateau of highly suitable temperatures. For both B. stanleyi and B. sudanica, fecundity decreased with increasing population densities. This means that snail populations may experience large fluctuations in numbers, even in the absence of any external factors such as seasonal temperature changes. Survival also decreased with increasing density for B. stanleyi, in contrast to B. sudanica and other studied Biomphalaria species where only fecundity has been shown to decrease.