Applications and Recruitment Performance of Web-Based Respondent-Driven Sampling: Scoping Review.

BACKGROUND: Web-based respondent-driven sampling is a novel sampling method for the recruitment of participants for generating population estimates, studying social network characteristics, and delivering health interventions. However, the application, barriers and facilitators, and recruitment performance of web-based respondent-driven sampling have not yet been systematically investigated. OBJECTIVE: Our objectives were to provide an overview of published research using web-based respondent-driven sampling and to investigate factors related to the recruitment performance of web-based respondent-driven sampling. METHODS: We conducted a scoping review on web-based respondent-driven sampling studies published between 2000 and 2019. We used the process evaluation of complex interventions framework to gain insights into how web-based respondent-driven sampling was implemented, what mechanisms of impact drove recruitment, what the role of context was in the study, and how these components together influenced the recruitment performance of web-based respondent-driven sampling. RESULTS: We included 18 studies from 8 countries (high- and low-middle income countries), in which web-based respondent-driven sampling was used for making population estimates (n=12), studying social network characteristics (n=3), and delivering health-related interventions (n=3). Studies used web-based respondent-driven sampling to recruit between 19 and 3448 participants from a variety of target populations. Studies differed greatly in the number of seeds recruited, the proportion of successfully recruiting participants, the number of recruitment waves, the type of incentives offered to participants, and the duration of data collection. Studies that recruited relatively more seeds, through online platforms, and with less rigorous selection procedures reported relatively low percentages of successfully recruiting seeds. Studies that did not offer at least one guaranteed material incentive reported relatively fewer waves and lower percentages of successfully recruiting participants. The time of data collection was shortest in studies with university students. CONCLUSIONS: Web-based respondent-driven sampling can be successfully applied to recruit individuals for making population estimates, studying social network characteristics, and delivering health interventions. In general, seed and peer recruitment may be enhanced by rigorously selecting and motivating seeds, offering at least one guaranteed material incentive, and facilitating adequate recruitment options regarding the target population's online connectedness and communication behavior. Potential trade-offs should be taken into account when implementing web-based respondent-driven sampling, such as having less opportunities to implement rigorous seed selection procedures when recruiting many seeds, as well as issues around online rather than physical participation, such as the risk of cheaters participating repeatedly.

Using gridded population and quadtree sampling units to support survey sample design in low-income settings.

BACKGROUND: Household surveys are the main source of demographic, health and socio-economic data in low- and middle-income countries (LMICs). To conduct such a survey, census population information mapped into enumeration areas (EAs) typically serves a sampling frame from which to generate a random sample. However, the use of census information to generate this sample frame can be problematic as in many LMIC contexts, such data are often outdated or incomplete, potentially introducing coverage issues into the sample frame. Increasingly, where census data are outdated or unavailable, modelled population datasets in the gridded form are being used to create household survey sampling frames. METHODS: Previously this process was done by either sampling from a set of the uniform grid cells (UGC) which are then manually subdivided to achieve the desired population size, or by sampling very small grid cells then aggregating cells into larger units to achieve a minimum population per survey cluster. The former approach is time and resource-intensive as well as results in substantial heterogeneity in the output sampling units, while the latter can complicate the calculation of unbiased sampling weights. Using the context of Somalia, which has not had a full census since 1987, we implemented a quadtree algorithm for the first time to create a population sampling frame. The approach uses gridded population estimates and it is based on the idea of a quadtree decomposition in which an area successively subdivided into four equal size quadrants, until the content of each quadrant is homogenous. RESULTS: The quadtree approach used here produced much more homogeneous sampling units than the UGC (1 × 1 km and 3 × 3 km) approach. At the national and pre-war regional scale, the standard deviation and coefficient of variation, as indications of homogeneity, were calculated for the output sampling units using quadtree and UGC 1 × 1 km and 3 × 3 km approaches to create the sampling frame and the results showed outstanding performance for quadtree approach. CONCLUSION: Our approach reduces the manual burden of manually subdividing UGC into highly populated areas, while allowing for correct calculation of sampling weights. The algorithm produces a relatively homogenous population counts within the sampling units, reducing the variation in the weights and improving the precision of the resulting estimates. Furthermore, a protocol of creating approximately equal-sized blocks and using tablets for randomized selection of a household in each block mitigated potential selection bias by enumerators. The approach shows labour, time and cost-saving and points to the potential use in wider contexts.

Publisher Correction: Past and future spread of the arbovirus vectors Aedes aegypti and Aedes albopictus.

This Article was mistakenly not made Open Access when originally published; this has now been amended, and information about the Creative Commons Attribution 4.0 International License has been added into the 'Additional information' section.

Publisher Correction: Past and future spread of the arbovirus vectors Aedes aegypti and Aedes albopictus.

In the version of this Article originally published, the affiliation for author Catherine Linard was incorrectly stated as '6Department of Infectious Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK'. The correct affiliation is '9Spatial Epidemiology Lab (SpELL), Universite Libre de Bruxelles, Brussels, Belgium'. The affiliation for author Hongjie Yu was also incorrectly stated as '11Department of Statistics, Harvard University, Cambridge, MA, USA'. The correct affiliation is '15School of Health, Fudan University, Key Laboratory of Public Health Safety, Ministry of Education, Shanghai, China'. This has now been amended in all versions of the Article.

Past and future spread of the arbovirus vectors Aedes aegypti and Aedes albopictus.

The global population at risk from mosquito-borne diseases-including dengue, yellow fever, chikungunya and Zika-is expanding in concert with changes in the distribution of two key vectors: Aedes aegypti and Aedes albopictus. The distribution of these species is largely driven by both human movement and the presence of suitable climate. Using statistical mapping techniques, we show that human movement patterns explain the spread of both species in Europe and the United States following their introduction. We find that the spread of Ae. aegypti is characterized by long distance importations, while Ae. albopictus has expanded more along the fringes of its distribution. We describe these processes and predict the future distributions of both species in response to accelerating urbanization, connectivity and climate change. Global surveillance and control efforts that aim to mitigate the spread of chikungunya, dengue, yellow fever and Zika viruses must consider the so far unabated spread of these mosquitos. Our maps and predictions offer an opportunity to strategically target surveillance and control programmes and thereby augment efforts to reduce arbovirus burden in human populations globally.

Population mobility reductions associated with travel restrictions during the Ebola epidemic in Sierra Leone: use of mobile phone data.

Background: Travel restrictions were implemented on an unprecedented scale in 2015 in Sierra Leone to contain and eliminate Ebola virus disease. However, the impact of epidemic travel restrictions on mobility itself remains difficult to measure with traditional methods. New 'big data' approaches using mobile phone data can provide, in near real-time, the type of information needed to guide and evaluate control measures. Methods: We analysed anonymous mobile phone call detail records (CDRs) from a leading operator in Sierra Leone between 20 March and 1 July in 2015. We used an anomaly detection algorithm to assess changes in travel during a national 'stay at home' lockdown from 27 to 29 March. To measure the magnitude of these changes and to assess effect modification by region and historical Ebola burden, we performed a time series analysis and a crossover analysis. Results: Routinely collected mobile phone data revealed a dramatic reduction in human mobility during a 3-day lockdown in Sierra Leone. The number of individuals relocating between chiefdoms decreased by 31% within 15 km, by 46% for 15-30 km and by 76% for distances greater than 30 km. This effect was highly heterogeneous in space, with higher impact in regions with higher Ebola incidence. Travel quickly returned to normal patterns after the restrictions were lifted. Conclusions: The effects of travel restrictions on mobility can be large, targeted and measurable in near real-time. With appropriate anonymization protocols, mobile phone data should play a central role in guiding and monitoring interventions for epidemic containment.

Mathematical models for predicting human mobility in the context of infectious disease spread: introducing the impedance model.

BACKGROUND: Mathematical models of human mobility have demonstrated a great potential for infectious disease epidemiology in contexts of data scarcity. While the commonly used gravity model involves parameter tuning and is thus difficult to implement without reference data, the more recent radiation model based on population densities is parameter-free, but biased. In this study we introduce the new impedance model, by analogy with electricity. Previous research has compared models on the basis of a few specific available spatial patterns. In this study, we use a systematic simulation-based approach to assess the performances. METHODS: Five hundred spatial patterns were generated using various area sizes and location coordinates. Model performances were evaluated based on these patterns. For simulated data, comparison measures were average root mean square error (aRMSE) and bias criteria. Modeling of the 2010 Haiti cholera epidemic with a basic susceptible-infected-recovered (SIR) framework allowed an empirical evaluation through assessing the goodness-of-fit of the observed epidemic curve. RESULTS: The new, parameter-free impedance model outperformed previous models on simulated data according to average aRMSE and bias criteria. The impedance model achieved better performances with heterogeneous population densities and small destination populations. As a proof of concept, the basic compartmental SIR framework was used to confirm the results obtained with the impedance model in predicting the spread of cholera in Haiti in 2010. CONCLUSIONS: The proposed new impedance model provides accurate estimations of human mobility, especially when the population distribution is highly heterogeneous. This model can therefore help to achieve more accurate predictions of disease spread in the context of an epidemic.

Mapping poverty using mobile phone and satellite data.

Poverty is one of the most important determinants of adverse health outcomes globally, a major cause of societal instability and one of the largest causes of lost human potential. Traditional approaches to measuring and targeting poverty rely heavily on census data, which in most low- and middle-income countries (LMICs) are unavailable or out-of-date. Alternate measures are needed to complement and update estimates between censuses. This study demonstrates how public and private data sources that are commonly available for LMICs can be used to provide novel insight into the spatial distribution of poverty. We evaluate the relative value of modelling three traditional poverty measures using aggregate data from mobile operators and widely available geospatial data. Taken together, models combining these data sources provide the best predictive power (highest r2 = 0.78) and lowest error, but generally models employing mobile data only yield comparable results, offering the potential to measure poverty more frequently and at finer granularity. Stratifying models into urban and rural areas highlights the advantage of using mobile data in urban areas and different data in different contexts. The findings indicate the possibility to estimate and continually monitor poverty rates at high spatial resolution in countries with limited capacity to support traditional methods of data collection.

Census-derived migration data as a tool for informing malaria elimination policy.

BACKGROUND: Numerous countries around the world are approaching malaria elimination. Until global eradication is achieved, countries that successfully eliminate the disease will contend with parasite reintroduction through international movement of infected people. Human-mediated parasite mobility is also important within countries near elimination, as it drives parasite flows that affect disease transmission on a subnational scale. METHODS: Movement patterns exhibited in census-based migration data are compared with patterns exhibited in a mobile phone data set from Haiti to quantify how well migration data predict short-term movement patterns. Because short-term movement data were unavailable for Mesoamerica, a logistic regression model fit to migration data from three countries in Mesoamerica is used to predict flows of infected people between subnational administrative units throughout the region. RESULTS: Population flows predicted using census-based migration data correlated strongly with mobile phone-derived movements when used as a measure of relative connectivity. Relative population flows are therefore predicted using census data across Mesoamerica, informing the areas that are likely exporters and importers of infected people. Relative population flows are used to identify community structure, useful for coordinating interventions and elimination efforts to minimize importation risk. Finally, the ability of census microdata inform future intervention planning is discussed in a country-specific setting using Costa Rica as an example. CONCLUSIONS: These results show long-term migration data can effectively predict the relative flows of infected people to direct malaria elimination policy, a particularly relevant result because migration data are generally easier to obtain than short-term movement data such as mobile phone records. Further, predicted relative flows highlight policy-relevant population dynamics, such as major exporters across the region, and Nicaragua and Costa Rica's strong connection by movement of infected people, suggesting close coordination of their elimination efforts. Country-specific applications are discussed as well, such as predicting areas at relatively high risk of importation, which could inform surveillance and treatment strategies.

Rapid and Near Real-Time Assessments of Population Displacement Using Mobile Phone Data Following Disasters: The 2015 Nepal Earthquake.

INTRODUCTION: Sudden impact disasters often result in the displacement of large numbers of people. These movements can occur prior to events, due to early warning messages, or take place post-event due to damages to shelters and livelihoods as well as a result of long-term reconstruction efforts. Displaced populations are especially vulnerable and often in need of support. However, timely and accurate data on the numbers and destinations of displaced populations are extremely challenging to collect across temporal and spatial scales, especially in the aftermath of disasters. Mobile phone call detail records were shown to be a valid data source for estimates of population movements after the 2010 Haiti earthquake, but their potential to provide near real-time ongoing measurements of population displacements immediately after a natural disaster has not been demonstrated. METHODS: A computational architecture and analytical capacity were rapidly deployed within nine days of the Nepal earthquake of 25th April 2015, to provide spatiotemporally detailed estimates of population displacements from call detail records based on movements of 12 million de-identified mobile phones users. RESULTS: Analysis shows the evolution of population mobility patterns after the earthquake and the patterns of return to affected areas, at a high level of detail. Particularly notable is the movement of an estimated 390,000 people above normal from the Kathmandu valley after the earthquake, with most people moving to surrounding areas and the highly-populated areas in the central southern area of Nepal. DISCUSSION: This analysis provides an unprecedented level of information about human movement after a natural disaster, provided within a very short timeframe after the earthquake occurred. The patterns revealed using this method are almost impossible to find through other methods, and are of great interest to humanitarian agencies.

Detecting climate adaptation with mobile network data in Bangladesh: anomalies in communication, mobility and consumption patterns during cyclone Mahasen.

Large-scale data from digital infrastructure, like mobile phone networks, provides rich information on the behavior of millions of people in areas affected by climate stress. Using anonymized data on mobility and calling behavior from 5.1 million Grameenphone users in Barisal Division and Chittagong District, Bangladesh, we investigate the effect of Cyclone Mahasen, which struck Barisal and Chittagong in May 2013. We characterize spatiotemporal patterns and anomalies in calling frequency, mobile recharges, and population movements before, during and after the cyclone. While it was originally anticipated that the analysis might detect mass evacuations and displacement from coastal areas in the weeks following the storm, no evidence was found to suggest any permanent changes in population distributions. We detect anomalous patterns of mobility both around the time of early warning messages and the storm's landfall, showing where and when mobility occurred as well as its characteristics. We find that anomalous patterns of mobility and calling frequency correlate with rainfall intensity (r = .75, p < 0.05) and use calling frequency to construct a spatiotemporal distribution of cyclone impact as the storm moves across the affected region. Likewise, from mobile recharge purchases we show the spatiotemporal patterns in people's preparation for the storm in vulnerable areas. In addition to demonstrating how anomaly detection can be useful for modeling human adaptation to climate extremes, we also identify several promising avenues for future improvement of disaster planning and response activities.

Tracking social contact networks with online respondent-driven detection: who recruits whom?

BACKGROUND: Transmission of respiratory pathogens in a population depends on the contact network patterns of individuals. To accurately understand and explain epidemic behaviour information on contact networks is required, but only limited empirical data is available. Online respondent-driven detection can provide relevant epidemiological data on numbers of contact persons and dynamics of contacts between pairs of individuals. We aimed to analyse contact networks with respect to sociodemographic and geographical characteristics, vaccine-induced immunity and self-reported symptoms. METHODS: In 2014, volunteers from two large participatory surveillance panels in the Netherlands and Belgium were invited for a survey. Participants were asked to record numbers of contacts at different locations and self-reported influenza-like-illness symptoms, and to invite 4 individuals they had met face to face in the preceding 2 weeks. We calculated correlations between linked individuals to investigate mixing patterns. RESULTS: In total 1560 individuals completed the survey who reported in total 30591 contact persons; 488 recruiter-recruit pairs were analysed. Recruitment was assortative by age, education, household size, influenza vaccination status and sentiments, indicating that participants tended to recruit contact persons similar to themselves. We also found assortative recruitment by symptoms, reaffirming our objective of sampling contact persons whom a participant may infect or by whom a participant may get infected in case of an outbreak. Recruitment was random by sex and numbers of contact persons. Relationships between pairs were influenced by the spatial distribution of peer recruitment. CONCLUSIONS: Although complex mechanisms influence online peer recruitment, the observed statistical relationships reflected the observed contact network patterns in the general population relevant for the transmission of respiratory pathogens. This provides useful and innovative input for predictive epidemic models relying on network information.

Implementation of Web-Based Respondent Driven Sampling among Men Who Have Sex with Men in Sweden.

BACKGROUND: Respondent driven sampling (RDS) was designed to study 'hidden' populations, for which there are no available sampling frame. RDS has been shown to recruit far into social networks of the study population and achieve unbiased estimates when certain assumptions are fulfilled. Web-based respondent driven sampling (WebRDS) has been implemented among MSM in Vietnam and produced a sufficient sample of MSM. In order to see if WebRDS could work in a 'hidden' population in a high-income setting, we performed a WebRDS among MSM in Sweden to study a sensitive topic, sexual risk behaviour for HIV/STI and Internet use. METHODS: A cross-sectional survey was implemented between July 11, 2012 and January 21, 2013 by using a WebRDS software. Men, fifteen years old or above, who reported having ever had sex with another man were included. The web-survey explored sociodemographics, sexual risk behaviour for HIV/STI and Internet use. RESULTS: The WebRDS process created a sample of 123 eligible respondents. The mean age among participants was 32 years old. All respondents reported having had unprotected anal intercourse (UAI) with at least one regular and one casual sex partner during the last 12 months. On average participants reported having had UAI with three casual sexual partners and in total having had seven casual sex partners during the last 12 months. CONCLUSION: The WebRDS produced a sample of Internet-using MSM in Sweden who all reported sexual risk behaviour for HIV/STI during the last 12 months. It holds promise for future online studies among MSM and a possibility to reach MSM at risk for HIV/STI with interventions or information. Some challenges were found including short recruitment chains, and further research need to address how to optimize WebRDS online recruitment methods in high income settings.

Enhancing Syndromic Surveillance With Online Respondent-Driven Detection.

OBJECTIVES: We investigated the feasibility of combining an online chain recruitment method (respondent-driven detection) and participatory surveillance panels to collect previously undetected information on infectious diseases via social networks of participants. METHODS: In 2014, volunteers from 2 large panels in the Netherlands were invited to complete a survey focusing on symptoms of upper respiratory tract infections and to invite 4 individuals they had met in the preceding 2 weeks to take part in the study. We compared sociodemographic characteristics among panel participants, individuals who volunteered for our survey, and individuals recruited via respondent-driven detection. RESULTS: Starting from 1015 panel members, the survey spread through all provinces of the Netherlands and all age groups in 83 days. A total of 433 individuals completed the survey via peer recruitment. Participants who reported symptoms were 6.1% (95% confidence interval = 5.4, 6.9) more likely to invite contact persons than were participants who did not report symptoms. Participants with symptoms invited more symptomatic recruits to take part than did participants without symptoms. CONCLUSIONS: Our findings suggest that online respondent-driven detection can enhance identification of symptomatic patients by making use of individuals' local social networks.

Remotely measuring populations during a crisis by overlaying two data sources.

BACKGROUND: Societal instability and crises can cause rapid, large-scale movements. These movements are poorly understood and difficult to measure but strongly impact health. Data on these movements are important for planning response efforts. We retrospectively analyzed movement patterns surrounding a 2010 humanitarian crisis caused by internal political conflict in Côte d'Ivoire using two different methods. METHODS: We used two remote measures, nighttime lights satellite imagery and anonymized mobile phone call detail records, to assess average population sizes as well as dynamic population changes. These data sources detect movements across different spatial and temporal scales. RESULTS: The two data sources showed strong agreement in average measures of population sizes. Because the spatiotemporal resolution of the data sources differed, we were able to obtain measurements on long- and short-term dynamic elements of populations at different points throughout the crisis. CONCLUSIONS: Using complementary, remote data sources to measure movement shows promise for future use in humanitarian crises. We conclude with challenges of remotely measuring movement and provide suggestions for future research and methodological developments.

Using mobile phone data to predict the spatial spread of cholera.

Effective response to infectious disease epidemics requires focused control measures in areas predicted to be at high risk of new outbreaks. We aimed to test whether mobile operator data could predict the early spatial evolution of the 2010 Haiti cholera epidemic. Daily case data were analysed for 78 study areas from October 16 to December 16, 2010. Movements of 2.9 million anonymous mobile phone SIM cards were used to create a national mobility network. Two gravity models of population mobility were implemented for comparison. Both were optimized based on the complete retrospective epidemic data, available only after the end of the epidemic spread. Risk of an area experiencing an outbreak within seven days showed strong dose-response relationship with the mobile phone-based infectious pressure estimates. The mobile phone-based model performed better (AUC 0.79) than the retrospectively optimized gravity models (AUC 0.66 and 0.74, respectively). Infectious pressure at outbreak onset was significantly correlated with reported cholera cases during the first ten days of the epidemic (p < 0.05). Mobile operator data is a highly promising data source for improving preparedness and response efforts during cholera outbreaks. Findings may be particularly important for containment efforts of emerging infectious diseases, including high-mortality influenza strains.

Comparison of contact patterns relevant for transmission of respiratory pathogens in Thailand and The Netherlands using respondent-driven sampling.

Understanding infection dynamics of respiratory diseases requires the identification and quantification of behavioural, social and environmental factors that permit the transmission of these infections between humans. Little empirical information is available about contact patterns within real-world social networks, let alone on differences in these contact networks between populations that differ considerably on a socio-cultural level. Here we compared contact network data that were collected in The Netherlands and Thailand using a similar online respondent-driven method. By asking participants to recruit contact persons we studied network links relevant for the transmission of respiratory infections. We studied correlations between recruiter and recruited contacts to investigate mixing patterns in the observed social network components. In both countries, mixing patterns were assortative by demographic variables and random by total numbers of contacts. However, in Thailand participants reported overall more contacts which resulted in higher effective contact rates. Our findings provide new insights on numbers of contacts and mixing patterns in two different populations. These data could be used to improve parameterisation of mathematical models used to design control strategies. Although the spread of infections through populations depends on more factors, found similarities suggest that spread may be similar in The Netherlands and Thailand.

Strong propensity for HIV transmission among men who have sex with men in Vietnam: behavioural data and sexual network modelling.

OBJECTIVES: Survey data from men who have sex with men (MSM) in Asian cities indicate ongoing and drastic increases in HIV prevalence. It is unknown which behavioural factors are most important in driving these epidemics. We aimed to analyse detailed sexual behaviour data among MSM in Vietnam and to model HIV transmission using improved assumptions on sexual network structure. SETTING: Vietnam. PARTICIPANTS: Internet-using men who had ever had sex (any type) with a man, aged ≥18 years and living in Vietnam. The study was cross-sectional, population-based and performed in 2012, using online respondent-driven sampling. The Internet-based survey instrument was completed by 982 participants, of which 857 were eligible. Questions included sociodemography and retrospective sexual behaviour, including number of unprotected anal sex (UAS) acts per partner. PRIMARY AND SECONDARY OUTCOME MEASURES: Estimated basic reproductive number over 3 months as a function of transmission risk per UAS act; frequency distributions of number of UAS partners and UAS acts during last 3 months. RESULTS: 36% (CI 32% to 42%) reported UAS at least once during the last 3 months. 36% (CI 32% to 41%) had ever taken an HIV test and received the result. UAS partner numbers and number of UAS acts were both highly skewed and positively correlated. Using a weighted configuration model, taking into account partner numbers, frequency of UAS and their correlations, we estimated the basic reproductive number (R0) over 3 months. The results indicated rapid transmission over a wide range of values of per-act transmissibility. CONCLUSIONS: Men with multiple partners had unexpectedly high UAS frequency per partner, paired with low HIV testing rates. The study highlights the importance of collecting data on frequency of UAS acts and indicates the need to rapidly scale-up HIV prevention services and testing opportunities for MSM in Vietnam.