How accurate are WorldPop-Global-Unconstrained gridded population data at the cell-level?: A simulation analysis in urban Namibia.

Disaggregated population counts are needed to calculate health, economic, and development indicators in Low- and Middle-Income Countries (LMICs), especially in settings of rapid urbanisation. Censuses are often outdated and inaccurate in LMIC settings, and rarely disaggregated at fine geographic scale. Modelled gridded population datasets derived from census data have become widely used by development researchers and practitioners; however, accuracy in these datasets are evaluated at the spatial scale of model input data which is generally courser than the neighbourhood or cell-level scale of many applications. We simulate a realistic synthetic 2016 population in Khomas, Namibia, a majority urban region, and introduce several realistic levels of outdatedness (over 15 years) and inaccuracy in slum, non-slum, and rural areas. We aggregate the synthetic populations by census and administrative boundaries (to mimic census data), resulting in 32 gridded population datasets that are typical of LMIC settings using the WorldPop-Global-Unconstrained gridded population approach. We evaluate the cell-level accuracy of these gridded population datasets using the original synthetic population as a reference. In our simulation, we found large cell-level errors, particularly in slum cells. These were driven by the averaging of population densities in large areal units before model training. Age, accuracy, and aggregation of the input data also played a role in these errors. We suggest incorporating finer-scale training data into gridded population models generally, and WorldPop-Global-Unconstrained in particular (e.g., from routine household surveys or slum community population counts), and use of new building footprint datasets as a covariate to improve cell-level accuracy (as done in some new WorldPop-Global-Constrained datasets). It is important to measure accuracy of gridded population datasets at spatial scales more consistent with how the data are being applied, especially if they are to be used for monitoring key development indicators at neighbourhood scales within cities.

Population Estimates of GFR and Risk Factors for CKD in Guatemala.

INTRODUCTION: Chronic kidney disease (CKD) is an emerging public health priority in Central America. However, data on the prevalence of CKD in Guatemala, Central America's most populous country, are limited, especially for rural communities. METHODS: We conducted a population-representative survey of 2 rural agricultural municipalities in Guatemala. We collected anthropometric data, blood pressure, serum and urine creatinine, glycosylated hemoglobin, and urine albumin. Sociodemographic, health, and exposure data were self-reported. RESULTS: We enrolled 807 individuals (63% of all eligible, 35% male, mean age 39.5 years). An estimated 4.0% (95% confidence interval [CI] 2.4-6.6) had CKD, defined as an estimated glomerular filtration rate (eGFR) less than 60 ml/min per 1.73 m2. Most individuals with an eGFR below 60 ml/min per 1.73 m2 had diabetes or hypertension. In multivariable analysis, the important factors associated with risk for an eGFR less than 60 ml/min per 1.73 m2 included a history of diabetes or hypertension (adjusted odds ratio [aOR] 11.21; 95% CI 3.28-38.24), underweight (body mass index [BMI] <18.5) (aOR 21.09; 95% CI 2.05-217.0), and an interaction between sugar cane agriculture and poverty (aOR 1.10; 95% CI 1.01-1.19). CONCLUSIONS: In this population-based survey, most observed CKD was associated with diabetes and hypertension. These results emphasize the urgent public health need to address the emerging epidemic of diabetes, hypertension, and CKD in rural Guatemala. In addition, the association between CKD and sugar cane in individuals living in poverty provides some circumstantial evidence for existence of CKD of unknown etiology in the study communities, which requires further investigation.

Addressing Unintentional Exclusion of Vulnerable and Mobile Households in Traditional Surveys in Kathmandu, Dhaka, and Hanoi: a Mixed-Methods Feasibility Study.

The methods used in low- and middle-income countries' (LMICs) household surveys have not changed in four decades; however, LMIC societies have changed substantially and now face unprecedented rates of urbanization and urbanization of poverty. This mismatch may result in unintentional exclusion of vulnerable and mobile urban populations. We compare three survey method innovations with standard survey methods in Kathmandu, Dhaka, and Hanoi and summarize feasibility of our innovative methods in terms of time, cost, skill requirements, and experiences. We used descriptive statistics and regression techniques to compare respondent characteristics in samples drawn with innovative versus standard survey designs and household definitions, adjusting for sample probability weights and clustering. Feasibility of innovative methods was evaluated using a thematic framework analysis of focus group discussions with survey field staff, and via survey planner budgets. We found that a common household definition excluded single adults (46.9%) and migrant-headed households (6.7%), as well as non-married (8.5%), unemployed (10.5%), disabled (9.3%), and studying adults (14.3%). Further, standard two-stage sampling resulted in fewer single adult and non-family households than an innovative area-microcensus design; however, two-stage sampling resulted in more tent and shack dwellers. Our survey innovations provided good value for money, and field staff experiences were neutral or positive. Staff recommended streamlining field tools and pairing technical and survey content experts during fieldwork. This evidence of exclusion of vulnerable and mobile urban populations in LMIC household surveys is deeply concerning and underscores the need to modernize survey methods and practices.

Feasibility of satellite image and GIS sampling for population representative surveys: a case study from rural Guatemala.

BACKGROUND: Population-representative household survey methods require up-to-date sampling frames and sample designs that minimize time and cost of fieldwork especially in low- and middle-income countries. Traditional methods such as multi-stage cluster sampling, random-walk, or spatial sampling can be cumbersome, costly or inaccurate, leading to well-known biases. However, a new tool, Epicentre's Geo-Sampler program, allows simple random sampling of structures, which can eliminate some of these biases. We describe the study design process, experiences and lessons learned using Geo-Sampler for selection of a population representative sample for a kidney disease survey in two sites in Guatemala. RESULTS: We successfully used Epicentre's Geo-sampler tool to sample 650 structures in two semi-urban Guatemalan communities. Overall, 82% of sampled structures were residential and could be approached for recruitment. Sample selection could be conducted by one person after 30 min of training. The process from sample selection to creating field maps took approximately 40 h. CONCLUSION: In combination with our design protocols, the Epicentre Geo-Sampler tool provided a feasible, rapid and lower-cost alternative to select a representative population sample for a prevalence survey in our semi-urban Guatemalan setting. The tool may work less well in settings with heavy arboreal cover or densely populated urban settings with multiple living units per structure. Similarly, while the method is an efficient step forward for including non-traditional living arrangements (people residing permanently or temporarily in businesses, religious institutions or other structures), it does not account for some of the most marginalized and vulnerable people in a population-the unhoused, street dwellers or people living in vehicles.

Gridded population survey sampling: a systematic scoping review of the field and strategic research agenda.

INTRODUCTION: In low- and middle-income countries (LMICs), household survey data are a main source of information for planning, evaluation, and decision-making. Standard surveys are based on censuses, however, for many LMICs it has been more than 10 years since their last census and they face high urban growth rates. Over the last decade, survey designers have begun to use modelled gridded population estimates as sample frames. We summarize the state of the emerging field of gridded population survey sampling, focussing on LMICs. METHODS: We performed a systematic scoping review in Scopus of specific gridded population datasets and "population" or "household" "survey" reports, and solicited additional published and unpublished sources from colleagues. RESULTS: We identified 43 national and sub-national gridded population-based household surveys implemented across 29 LMICs. Gridded population surveys used automated and manual approaches to derive clusters from WorldPop and LandScan gridded population estimates. After sampling, some survey teams interviewed all households in each cluster or segment, and others sampled households from larger clusters. Tools to select gridded population survey clusters include the GridSample R package, Geo-sampling tool, and GridSample.org. In the field, gridded population surveys generally relied on geographically accurate maps based on satellite imagery or OpenStreetMap, and a tablet or GPS technology for navigation. CONCLUSIONS: For gridded population survey sampling to be adopted more widely, several strategic questions need answering regarding cell-level accuracy and uncertainty of gridded population estimates, the methods used to group/split cells into sample frame units, design effects of new sample designs, and feasibility of tools and methods to implement surveys across diverse settings.

A grid-based sample design framework for household surveys.

Traditional sample designs for household surveys are contingent upon the availability of a representative primary sampling frame. This is defined using enumeration units and population counts retrieved from decennial national censuses that can become rapidly inaccurate in highly dynamic demographic settings. To tackle the need for representative sampling frames, we propose an original grid-based sample design framework introducing essential concepts of spatial sampling in household surveys. In this framework, the sampling frame is defined based on gridded population estimates and formalized as a bi-dimensional random field, characterized by spatial trends, spatial autocorrelation, and stratification. The sampling design reflects the characteristics of the random field by combining contextual stratification and proportional to population size sampling. A nonparametric estimator is applied to evaluate the sampling design and inform sample size estimation. We demonstrate an application of the proposed framework through a case study developed in two provinces located in the western part of the Democratic Republic of the Congo. We define a sampling frame consisting of settled cells with associated population estimates. We then perform a contextual stratification by applying a principal component analysis (PCA) and k-means clustering to a set of gridded geospatial covariates, and sample settled cells proportionally to population size. Lastly, we evaluate the sampling design by contrasting the empirical cumulative distribution function for the entire population of interest and its weighted counterpart across different sample sizes and identify an adequate sample size using the Kolmogorov-Smirnov distance between the two functions. The results of the case study underscore the strengths and limitations of the proposed grid-based sample design framework and foster further research into the application of spatial sampling concepts in household surveys.

Correction to: Extending Data for Urban Health Decision-Making: a Menu of New and Potential Neighborhood-Level Health Determinants Datasets in LMICs.

Readers should note an additional Acknowledgment for this article: Dana Thomson is funded by the Economic and Social Research Council grant number ES/5500161/1.

Extending Data for Urban Health Decision-Making: a Menu of New and Potential Neighborhood-Level Health Determinants Datasets in LMICs.

Area-level indicators of the determinants of health are vital to plan and monitor progress toward targets such as the Sustainable Development Goals (SDGs). Tools such as the Urban Health Equity Assessment and Response Tool (Urban HEART) and UN-Habitat Urban Inequities Surveys identify dozens of area-level health determinant indicators that decision-makers can use to track and attempt to address population health burdens and inequalities. However, questions remain as to how such indicators can be measured in a cost-effective way. Area-level health determinants reflect the physical, ecological, and social environments that influence health outcomes at community and societal levels, and include, among others, access to quality health facilities, safe parks, and other urban services, traffic density, level of informality, level of air pollution, degree of social exclusion, and extent of social networks. The identification and disaggregation of indicators is necessarily constrained by which datasets are available. Typically, these include household- and individual-level survey, census, administrative, and health system data. However, continued advancements in earth observation (EO), geographical information system (GIS), and mobile technologies mean that new sources of area-level health determinant indicators derived from satellite imagery, aggregated anonymized mobile phone data, and other sources are also becoming available at granular geographic scale. Not only can these data be used to directly calculate neighborhood- and city-level indicators, they can be combined with survey, census, administrative and health system data to model household- and individual-level outcomes (e.g., population density, household wealth) with tremendous detail and accuracy. WorldPop and the Demographic and Health Surveys (DHS) have already modeled dozens of household survey indicators at country or continental scales at resolutions of 1 × 1 km or even smaller. This paper aims to broaden perceptions about which types of datasets are available for health and development decision-making. For data scientists, we flag area-level indicators at city and sub-city scales identified by health decision-makers in the SDGs, Urban HEART, and other initiatives. For local health decision-makers, we summarize a menu of new datasets that can be feasibly generated from EO, mobile phone, and other spatial data-ideally to be made free and publicly available-and offer lay descriptions of some of the difficulties in generating such data products.

Using routinely collected laboratory data to identify high rifampicin-resistant tuberculosis burden communities in the Western Cape Province, South Africa: A retrospective spatiotemporal analysis.

BACKGROUND: South Africa has the highest tuberculosis incidence globally (781/100,000), with an estimated 4.3% of cases being rifampicin resistant (RR). Control and elimination strategies will require detailed spatial information to understand where drug-resistant tuberculosis exists and why it persists in those communities. We demonstrate a method to enable drug-resistant tuberculosis monitoring by identifying high-burden communities in the Western Cape Province using routinely collected laboratory data. METHODS AND FINDINGS: We retrospectively identified cases of microbiologically confirmed tuberculosis and RR-tuberculosis from all biological samples submitted for tuberculosis testing (n = 2,219,891) to the Western Cape National Health Laboratory Services (NHLS) between January 1, 2008, and June 30, 2013. Because the NHLS database lacks unique patient identifiers, we performed a series of record-linking processes to match specimen records to individual patients. We counted an individual as having a single disease episode if their positive samples came from within two years of each other. Cases were aggregated by clinic location (n = 302) to estimate the percentage of tuberculosis cases with rifampicin resistance per clinic. We used inverse distance weighting (IDW) to produce heatmaps of the RR-tuberculosis percentage across the province. Regression was used to estimate annual changes in the RR-tuberculosis percentage by clinic, and estimated average size and direction of change was mapped. We identified 799,779 individuals who had specimens submitted from mappable clinics for testing, of whom 222,735 (27.8%) had microbiologically confirmed tuberculosis. The study population was 43% female, the median age was 36 years (IQR 27-44), and 10,255 (4.6%, 95% CI: 4.6-4.7) cases had documented rifampicin resistance. Among individuals with microbiologically confirmed tuberculosis, 8,947 (4.0%) had more than one disease episode during the study period. The percentage of tuberculosis cases with rifampicin resistance documented among these individuals was 11.4% (95% CI: 10.7-12.0). Overall, the percentage of tuberculosis cases that were RR-tuberculosis was spatially heterogeneous, ranging from 0% to 25% across the province. Our maps reveal significant yearly fluctuations in RR-tuberculosis percentages at several locations. Additionally, the directions of change over time in RR-tuberculosis percentage were not uniform. The main limitation of this study is the lack of unique patient identifiers in the NHLS database, rendering findings to be estimates reliant on the accuracy of the person-matching algorithm. CONCLUSIONS: Our maps reveal striking spatial and temporal heterogeneity in RR-tuberculosis percentages across this province. We demonstrate the potential to monitor RR-tuberculosis spatially and temporally with routinely collected laboratory data, enabling improved resource targeting and more rapid locally appropriate interventions.

Early changes in intervention coverage and mortality rates following the implementation of an integrated health system intervention in Madagascar.

INTRODUCTION: The Sustainable Development Goals framed an unprecedented commitment to achieve global convergence in child and maternal mortality rates through 2030. To meet those targets, essential health services must be scaled via integration with strengthened health systems. This is especially urgent in Madagascar, the country with the lowest level of financing for health in the world. Here, we present an interim evaluation of the first 2 years of a district-level health system strengthening (HSS) initiative in rural Madagascar, using estimates of intervention coverage and mortality rates from a district-wide longitudinal cohort. METHODS: We carried out a district representative household survey at baseline of the HSS intervention in over 1500 households in Ifanadiana district. The first follow-up was after the first 2 years of the initiative. For each survey, we estimated maternal, newborn and child health (MNCH) coverage, healthcare inequalities and child mortality rates both in the initial intervention catchment area and in the rest of the district. We evaluated changes between the two areas through difference-in-differences analyses. We estimated annual changes in health centre per capita utilisation from 2013 to 2016. RESULTS: The intervention was associated with 19.1% and 36.4% decreases in under-five and neonatal mortality, respectively, although these were not statistically significant. The composite coverage index (a summary measure of MNCH coverage) increased by 30.1%, with a notable 63% increase in deliveries in health facilities. Improvements in coverage were substantially larger in the HSS catchment area and led to an overall reduction in healthcare inequalities. Health centre utilisation rates in the catchment tripled for most types of care during the study period. CONCLUSION: At the earliest stages of an HSS intervention, the rapid improvements observed for Ifanadiana add to preliminary evidence supporting the untapped and poorly understood potential of integrated HSS interventions on population health.

Impact of a health system strengthening intervention on maternal and child health outputs and outcomes in rural Rwanda 2005-2010.

INTRODUCTION: Although Rwanda's health system underwent major reforms and improvements after the 1994 Genocide, the health system and population health in the southeast lagged behind other areas. In 2005, Partners In Health and the Rwandan Ministry of Health began a health system strengthening intervention in this region. We evaluate potential impacts of the intervention on maternal and child health indicators. METHODS: Combining results from the 2005 and 2010 Demographic and Health Surveys with those from a supplemental 2010 survey, we compared changes in health system output indicators and population health outcomes between 2005 and 2010 as reported by women living in the intervention area with those reported by the pooled population of women from all other rural areas of the country, controlling for potential confounding by economic and demographic variables. RESULTS: Overall health system coverage improved similarly in the comparison groups between 2005 and 2010, with an indicator of composite coverage of child health interventions increasing from 57.9% to 75.0% in the intervention area and from 58.7% to 73.8% in the other rural areas. Under-five mortality declined by an annual rate of 12.8% in the intervention area, from 229.8 to 83.2 deaths per 1000 live births, and by 8.9% in other rural areas, from 157.7 to 75.8 deaths per 1000 live births. Improvements were most marked among the poorest households. CONCLUSION: We observed dramatic improvements in population health outcomes including under-five mortality between 2005 and 2010 in rural Rwanda generally and in the intervention area specifically.

Next wave of interventions to reduce under-five mortality in Rwanda: a cross-sectional analysis of demographic and health survey data.

BACKGROUND: Sustained investments in Rwanda's health system have led to historic reductions in under five (U5) mortality. Although Rwanda achieved an estimated 68% decrease in the national under U5 mortality rate between 2002 and 2012, according to the national census, 5.8% of children still do not reach their fifth birthday, requiring the next wave of child mortality prevention strategies. METHODS: This is a cross-sectional study of 9002 births to 6328 women age 15-49 in the 2010 Rwanda Demographic and Health Survey. We tested bivariate associations between 29 covariates and U5 mortality, retaining covariates with an odds ratio p < 0.1 for model building. We used manual backward stepwise logistic regression to identify correlates of U5 mortality in all children U5, 0-11 months, and 12-59 months. Analyses were performed in Stata v12, adjusting for complex sample design. RESULTS: Of 14 covariates associated with U5 mortality in bivariate analysis, the following remained associated with U5 mortality in multivariate analysis: household being among the poorest of the poor (OR = 1.98), child being a twin (OR = 2.40), mother having 3-4 births in the past 5 years (OR = 3.97) compared to 1-2 births, mother being HIV positive (OR = 2.27), and mother not using contraceptives (OR = 1.37) compared to using a modern method (p < 0.05 for all). Mother experiencing physical or sexual violence in the last 12 months was associated with U5 mortality in children ages 1-4 years (OR = 1.48, p < 0.05). U5 survival was associated with a preceding birth interval 25-50 months (OR = 0.67) compared to 9-24 months, and having a mosquito net (OR = 0.46) (p < 0.05 for both). CONCLUSIONS: In the past decade, Rwanda rolled out integrated management of childhood illness, near universal coverage of childhood vaccinations, a national community health worker program, and a universal health insurance scheme. Identifying factors that continue to be associated with childhood mortality supports determination of which interventions to strengthen to reduce it further. This study suggests that Rwanda's next wave of U5 mortality reduction should target programs in improving neonatal outcomes, poverty reduction, family planning, HIV services, malaria prevention, and prevention of intimate partner violence.

Exogenous factors matter when interpreting the results of an impact evaluation: a case study of rainfall and child health programme intervention in Rwanda.

OBJECTIVE: Public health interventions are often implemented at large scale, and their evaluation seems to be difficult because they are usually multiple and their pathways to effect are complex and subject to modification by contextual factors. We assessed whether controlling for rainfall-related variables altered estimates of the efficacy of a health programme in rural Rwanda and have a quantifiable effect on an intervention evaluation outcomes. METHODS: We conducted a retrospective quasi-experimental study using previously collected cross-sectional data from the 2005 and 2010 Rwanda Demographic and Health Surveys (DHS), 2010 DHS oversampled data, monthly rainfall data collected from meteorological stations over the same period, and modelled output of long-term rainfall averages, soil moisture, and rain water run-off. Difference-in-difference models were used. RESULTS: Rainfall factors confounded the PIH intervention impact evaluation. When we adjusted our estimates of programme effect by controlling for a variety of rainfall variables, several effectiveness estimates changed by 10% or more. The analyses that did not adjust for rainfall-related variables underestimated the intervention effect on the prevalence of ARI by 14.3%, fever by 52.4% and stunting by 10.2%. Conversely, the unadjusted analysis overestimated the intervention's effect on diarrhoea by 56.5% and wasting by 80%. CONCLUSION: Rainfall-related patterns have a quantifiable effect on programme evaluation results and highlighted the importance and complexity of controlling for contextual factors in quasi-experimental design evaluations.

GridSample: an R package to generate household survey primary sampling units (PSUs) from gridded population data.

BACKGROUND: Household survey data are collected by governments, international organizations, and companies to prioritize policies and allocate billions of dollars. Surveys are typically selected from recent census data; however, census data are often outdated or inaccurate. This paper describes how gridded population data might instead be used as a sample frame, and introduces the R GridSample algorithm for selecting primary sampling units (PSU) for complex household surveys with gridded population data. With a gridded population dataset and geographic boundary of the study area, GridSample allows a two-step process to sample "seed" cells with probability proportionate to estimated population size, then "grows" PSUs until a minimum population is achieved in each PSU. The algorithm permits stratification and oversampling of urban or rural areas. The approximately uniform size and shape of grid cells allows for spatial oversampling, not possible in typical surveys, possibly improving small area estimates with survey results. RESULTS: We replicated the 2010 Rwanda Demographic and Health Survey (DHS) in GridSample by sampling the WorldPop 2010 UN-adjusted 100 m × 100 m gridded population dataset, stratifying by Rwanda's 30 districts, and oversampling in urban areas. The 2010 Rwanda DHS had 79 urban PSUs, 413 rural PSUs, with an average PSU population of 610 people. An equivalent sample in GridSample had 75 urban PSUs, 405 rural PSUs, and a median PSU population of 612 people. The number of PSUs differed because DHS added urban PSUs from specific districts while GridSample reallocated rural-to-urban PSUs across all districts. CONCLUSIONS: Gridded population sampling is a promising alternative to typical census-based sampling when census data are moderately outdated or inaccurate. Four approaches to implementation have been tried: (1) using gridded PSU boundaries produced by GridSample, (2) manually segmenting gridded PSU using satellite imagery, (3) non-probability sampling (e.g. random-walk, "spin-the-pen"), and random sampling of households. Gridded population sampling is in its infancy, and further research is needed to assess the accuracy and feasibility of gridded population sampling. The GridSample R algorithm can be used to forward this research agenda.

Fertility and HIV following universal access to ART in Rwanda: a cross-sectional analysis of Demographic and Health Survey data.

BACKGROUND: HIV infection is linked to decreased fertility and fertility desires in sub-Saharan Africa due to biological and social factors. We investigate the relationship between HIV infection and fertility or fertility desires in the context of universal access to antiretroviral therapy introduced in 2004 in Rwanda. METHODS: We used data from 3532 and 4527 women aged 20-49 from the 2005 and 2010 Rwandan Demographic and Health Surveys (RDHS), respectively. The RDHSs included blood-tests for HIV, as well as detailed interviews about fertility, demographic and behavioral outcomes. In both years, multiple logistic regression was used to assess the association between HIV and fertility outcomes within three age categories (20-29, 30-39 and 40-49 years), controlling for confounders and compensating for the complex survey design. RESULTS: In 2010, we did not find a difference in the odds of pregnancy in the last 5 years between HIV-seropositive and HIV-seronegative women after controlling for potential biological and social confounders. Controlling for the same confounders, we found that HIV-seropositive women under age 40 were less likely to desire more children compared to HIV-seronegative women (20-29 years adjusted odds ratio (AOR) = 0.31, 95% CI: 0.17, 0.58; 30-39 years AOR = 0.24, 95% CI: 0.14, 0.43), but no difference was found among women aged 40 or older. No associations between HIV and fertility or fertility desire were found in 2005. CONCLUSIONS: These findings suggest no difference in births or current pregnancy among HIV-seropositive and HIV-seronegative women. That in 2010 HIV-seropositive women in their earlier childbearing years desired fewer children than HIV-seronegative women could suggest more women with HIV survived; and stigma, fear of transmitting HIV, or realism about living with HIV and prematurely dying from HIV may affect their desire to have children. These findings emphasize the importance of delivering appropriate information about pregnancy and childbearing to HIV-infected women, enabling women living with HIV to make informed decisions about their reproductive life.

Applied statistical training to strengthen analysis and health research capacity in Rwanda.

BACKGROUND: To guide efficient investment of limited health resources in sub-Saharan Africa, local researchers need to be involved in, and guide, health system and policy research. While extensive survey and census data are available to health researchers and program officers in resource-limited countries, local involvement and leadership in research is limited due to inadequate experience, lack of dedicated research time and weak interagency connections, among other challenges. Many research-strengthening initiatives host prolonged fellowships out-of-country, yet their approaches have not been evaluated for effectiveness in involvement and development of local leadership in research. METHODS: We developed, implemented and evaluated a multi-month, deliverable-driven, survey analysis training based in Rwanda to strengthen skills of five local research leaders, 15 statisticians, and a PhD candidate. Research leaders applied with a specific research question relevant to country challenges and committed to leading an analysis to publication. Statisticians with prerequisite statistical training and experience with a statistical software applied to participate in class-based trainings and complete an assigned analysis. Both statisticians and research leaders were provided ongoing in-country mentoring for analysis and manuscript writing. RESULTS: Participants reported a high level of skill, knowledge and collaborator development from class-based trainings and out-of-class mentorship that were sustained 1 year later. Five of six manuscripts were authored by multi-institution teams and submitted to international peer-reviewed scientific journals, and three-quarters of the participants mentored others in survey data analysis or conducted an additional survey analysis in the year following the training. CONCLUSIONS: Our model was effective in utilizing existing survey data and strengthening skills among full-time working professionals without disrupting ongoing work commitments and using few resources. Critical to our success were a transparent, robust application process and time limited training supplemented by ongoing, in-country mentoring toward manuscript deliverables that were led by Rwanda's health research leaders.

Facility-Based Delivery during the Ebola Virus Disease Epidemic in Rural Liberia: Analysis from a Cross-Sectional, Population-Based Household Survey.

BACKGROUND: The Ebola virus disease (EVD) epidemic has threatened access to basic health services through facility closures, resource diversion, and decreased demand due to community fear and distrust. While modeling studies have attempted to estimate the impact of these disruptions, no studies have yet utilized population-based survey data. METHODS AND FINDINGS: We conducted a two-stage, cluster-sample household survey in Rivercess County, Liberia, in March-April 2015, which included a maternal and reproductive health module. We constructed a retrospective cohort of births beginning 4 y before the first day of survey administration (beginning March 24, 2011). We then fit logistic regression models to estimate associations between our primary outcome, facility-based delivery (FBD), and time period, defined as the pre-EVD period (March 24, 2011-June 14, 2014) or EVD period (June 15, 2014-April 13, 2015). We fit both univariable and multivariable models, adjusted for known predictors of facility delivery, accounting for clustering using linearized standard errors. To strengthen causal inference, we also conducted stratified analyses to assess changes in FBD by whether respondents believed that health facility attendance was an EVD risk factor. A total of 1,298 women from 941 households completed the survey. Median age at the time of survey was 29 y, and over 80% had a primary education or less. There were 686 births reported in the pre-EVD period and 212 in the EVD period. The unadjusted odds ratio of facility-based delivery in the EVD period was 0.66 (95% confidence interval [CI] 0.48-0.90, p-value = 0.010). Adjustment for potential confounders did not change the observed association, either in the principal model (adjusted odds ratio [AOR] = 0.70, 95%CI 0.50-0.98, p = 0.037) or a fully adjusted model (AOR = 0.69, 95%CI 0.50-0.97, p = 0.033). The association was robust in sensitivity analyses. The reduction in FBD during the EVD period was observed among those reporting a belief that health facilities are or may be a source of Ebola transmission (AOR = 0.59, 95%CI 0.36-0.97, p = 0.038), but not those without such a belief (AOR = 0.90, 95%CI 0.59-1.37, p = 0.612). Limitations include the possibility of FBD secular trends coincident with the EVD period, recall errors, and social desirability bias. CONCLUSIONS: We detected a 30% decreased odds of FBD after the start of EVD in a rural Liberian county with relatively few cases. Because health facilities never closed in Rivercess County, this estimate may under-approximate the effect seen in the most heavily affected areas. These are the first population-based survey data to show collateral disruptions to facility-based delivery caused by the West African EVD epidemic, and they reinforce the need to consider the full spectrum of implications caused by public health emergencies.

Factors associated with postnatal care utilisation in Rwanda: A secondary analysis of 2010 Demographic and Health Survey data.

BACKGROUND: Postnatal care (PNC) in the first seven days is important for preventing morbidity and mortality in mothers and new-borns. Sub-Saharan African countries, which account for 62 % of maternal deaths globally, have made major efforts to increase PNC utilisation, but utilisation rates remains low even in countries like Rwanda where PNC services are universally available for free. This study identifies key socio-economic and demographic factors associated with PNC utilisation in Rwanda to inform improved PNC policies and programs. METHODS: This is a secondary analysis of the 2010 Demographic and Health Survey, a national multi-stage, cross-sectional survey. In bivariate analysis, we used chi-square tests to identify demographic and socio-economic factors associated with PNC utilisation at α = 0.1. Pearson's R statistic (r > 0.5) was used to identify collinear covariates, and to choose which covariate was more strongly associated with PNC utilisation. Manual backward stepwise logistic regression was performed on the remaining covariates to identify key factors associated with PNC utilisation at α = 0.05. All analyses were performed in Stata 13 adjusting for sampling weights, clustering, and stratification. RESULTS: Of the 2,748 women with a live birth in the last two years who answered question about PNC utilisation, 353 (12.8 %) returned for PNC services within seven days after birth. Three factors were positively associated with PNC use: delivering at a health facility (OR: 2.97; 95 % CI: 2.28-3.87), being married but not involved with one's own health care decision-making (OR: 1.69; 95 % CI: 1.17, 2.44) compared to being married and involved; and being in the second (OR: 1.46; 95 % CI: 1.01-2.09) or richest wealth quintile (OR: 2.04; 95 % CI: 1.27-3.29) compared to the poorest. Mother's older age at delivery was negatively associated with PNC use (20-29 - OR: 0.51, 95 % CI: 0.29-0.87; 30-39 - OR: 0.47, 95 % CI: 0.27-0.83; 40-49 - OR: 0.32, 95 % CI: 0.16-0.64). CONCLUSIONS: Low PNC utilisation in Rwanda appears to be a universal problem though older age and poverty are further barriers to PNC utilisation. A recent change in the provision of BCG vaccination to new-borns might promote widespread PNC utilisation. We further recommend targeted campaigns to older mothers and poorest mothers, focusing on perceptions of health system quality, cultural beliefs, and pregnancy risks.

Social, economic and environmental risk factors for acute lower respiratory infections among children under five years of age in Rwanda.

BACKGROUND: In low and middle-income countries, acute lower respiratory illness is responsible for roughly 1 in every 5 child deaths. Rwanda has made major health system improvements including its community health worker systems, and it is one of the few countries in Africa to meet the 2015 Millennium Development Goals, although prevalence of acute lower respiratory infections (4 %) is similar to other countries in sub-Saharan Africa. This study aims to assess social, economic, and environmental factors associated with acute lower respiratory infections among children under five to inform potential further improvements in the health system. METHODS: This is a cross-sectional study using data collected from women interviewed in the 2010 DHS about 8,484 surviving children under five. Based on a literature review, we defined 19 health, social, economic, and environmental potential risk factors, tested bivariate associations with acute lower respiratory infections, and advanced variables significant at the 0.1 confidence level to logistic regression modelling. We used manual backward stepwise regression to arrive at a final model. All analyses were performed in Stata v13 and adjusted for complex sample design. RESULTS: The following factors were independently associated with acute lower respiratory infections: child's age, anemia level, and receipt of Vitamin A; household toilet type and residence, and season of interview. In multivariate regression, being in the bottom ten percent of households (OR: 1.27, 95 % CI: 0.85-1.87) or being interviewed during the rainy season (OR: 1.61, 95 % CI: 1.24-2.09) was positively associated with acute lower respiratory infections, while urban residence (OR: 0.58, 95 % CI: 0.38-0.88) and being age 24-59 months versus 0-11 months (OR: 0.53, 95 % CI: 0.40-0.69) was negatively associated with acute lower respiratory infections. CONCLUSION: Potential areas for intervention including community campaigns about acute lower respiratory infections symptoms and treatment, and continued poverty reduction through rural electrification and modern stove distribution which may reduce use of dirty cooking fuel, improve living conditions, and reduce barriers to health care.