Is the source of domestic water associated with the risk of malaria infection? Spatial variability and a mixed-effects multilevel analysis

On flexible inverse probability of treatment and intensity weighting: Informative censoring, variable selection, and weight trimming

Many observational studies feature irregular longitudinal data, where the observation times are not common across individuals in the study. Furthermore, the observation times may be related to the longitudinal outcome. In this setting, failing to account for the informative observation process may result in biased causal estimates. This can be coupled with other sources of bias, including nonrandomized treatment assignments and informative censoring. This paper provides an overview of a flexible weighting method used to adjust for informative observation processes and nonrandomized treatment assignments. We investigate the sensitivity of the flexible weighting method to violations of the noninformative censoring assumption, examine variable selection for the observation process weighting model, known as inverse intensity weighting, and look at the impacts of weight trimming for the flexible weighting model. We show that the flexible weighting method is sensitive to violations of the noninformative censoring assumption and that a previously proposed extension fails under such violations. We also show that variables confounding the observation and outcome processes should always be included in the observation intensity model. Finally, we show scenarios where weight trimming should and should not be used, and highlight sensitivities of the flexible inverse probability of treatment and intensity weighting method to extreme weights. We conclude with an application of the methodology to a real data set to examine the impacts of household water sources on malaria diagnoses.

Factors affecting the place of death in patients with liver cancer in China, 2013-2020: A population-based study

Background

Despite the country's substantial liver cancer burden, there is limited research on the factors influencing the place of death (POD) of patients with liver cancer in China. This study aimed to delineate POD distribution among patients with liver cancer, identify the factors associated with hospital deaths, and offer valuable insights for the government to develop healthcare policies.

Methods

Data from 2013 to 2020 were obtained from the National Mortality Surveillance System (NMSS) of China. This analysis focused on the distribution of POD among individuals who succumbed to liver cancer. Variations in characteristic distributions across different categories were evaluated using a chi-squared test. We also applied a multilevel logistic regression analysis to identify the factors associated with hospital liver cancer deaths. The proportional change in variance was computed to evaluate the contributions of different factors in the model.

Results

From 2013 to 2020, the NMSS reported a total of 608,789 liver cancer-related deaths, of which 440,079 (72.29%) died at home, and 158,291 (26.00%) died in the hospital. Home remained the preferred POD among patients with liver cancer. The results demonstrated that female patients, aged between 0 and 14 years, of Han ethnicity, living in urban areas, unmarried, highly educated, and either employed in a professional, staff, or civil servant capacity, or retired patients tended to end their lives in the hospital.

Conclusions

In China, home continues to be the predominant POD for patients with liver cancer, with demographic and socioeconomic factors significantly influencing whether a hospital is their POD. Enhancing healthcare policymakers' understanding of the factors influencing the place of death for patients with liver cancer may assist in creating a more equitable distribution of healthcare resources and providing a variety of choices for minorities with distinct preferences for end-of-life care.

Combined application of metagenomics and FEAST to trace sources of microbial eukaryotic contamination in the Pasig-Marikina-San Juan (PAMARISAN) river system in Metro Manila, Philippines

Microbial eukaryotes are vital to global microbial diversity, but there is limited information about their composition and sources in contaminated surface waters. This study examined the pathogens and potential sources of microbial eukaryotic communities in polluted sink environments using the 18S rDNA amplicon sequencing combined with the fast expectation-maximization for microbial source tracking (FEAST) program. Six sampling sites were selected along the Pasig-Marikina-San Juan (PAMARISAN) River System, representing different locations within the waterway and classified as sinks (n = 12), whereas animal fecal samples collected from various farms were classified as sources (n = 29). Taxonomic composition revealed Stramenopila, Alveolata, Rhizaria (SAR), Archaeplastida, and Excavata in the rivers, accounting for 85.1%, 13.2%, and 0.36% mean abundance of microbial sink communities, respectively. Clinically relevant human pathogens were also observed in sink environments. The correlation test demonstrated that dissolved oxygen, total suspended solids, pH, temperature, fecal coliform count, and phosphates were important environmental factors driving community variations. Moreover, FEAST results indicated that sewage (19.6%) was the primary source of microbial eukaryotes, followed by duck (0.644%) and cow (0.566%) feces. Spatio-seasonal variations showed higher contributions at downstream stations and during the wet season, highlighting the role of rainfall in enhancing microbial dispersal. Results from community-based microbial source tracking can be used to explore factors shaping microbial eukaryotes in freshwater environments, assess potential pathogen-related hazards, and inform river conservation and management strategies. Furthermore, this also serves as preliminary data for microbial eukaryotic source tracking in the Philippines, laying groundwork for future research.

Post-war status of water supply, sanitation, hygiene and related reported diseases in Tigray, Ethiopia: A community-based cross-sectional study

Water, sanitation and hygiene (WASH) associated diseases remain a global public health issue and linked with Sustainable Development Goal 6. In November 2020, a war broke out in Tigray, Ethiopia, resulting in a negative health consequence. The post war status of WASH and its associated diseases are not documented. The aim of this study was to assess the status of drinking water, sanitation and hygiene practices and the prevalence of WASH-associated diseases in Tigray, Ethiopia following the war. A community-based cross-sectional study was conducted in 24 randomly selected accessible districts of Tigray, Ethiopia. A standardized questionnaire was used to collect data from households in the study. Data was collected from 2338 households. Descriptive statistics and binary logistic regression were used to analyze the data. The average age of respondents was 28.7 years (SD = 6.2). The majority of respondents 2030 (86.8%) were married and 1698 (72.6%) were rural residents. Nearly one third of the respondents were uneducated and around 40% have either radio or TV as means of communication. More than half (55.2%) of the respondents had a family size of over 5. A quarter (25%, 95% CI: 23.3, 26.8) of study participants had access to a basic water supply. Less than a tenth (7.7%, 95% CI: 6.6, 8.8) of households had access to basic sanitation. Basic hand washing was available in 2% of households. Malaria, diarrhoea, skin infection and eye infection were the common reported disease in the community. Marital status, family size, place of residence and liquid waste management were the most important predictors of reported diseases. Access to basic water, sanitation and hygiene services was low, and the prevalence of malaria, diarrhoea and skin infections was higher. There were differences in WASH services and reported diseases according to zone and place of residence (urban-rural). Post war, improved access to basic water, sanitation and hygiene services is recommended to prevent WASH-associated diseases in Tigray, Ethiopia. Furthermore, the prevention oriented policy of the country needs better implementation to reduce preventable diseases and ensure better health status in the community.

Assessment of malaria transmission in Kenya using multilevel logistic regression

Background

Kenya has a lower malaria incidence in comparison to other African malaria-endemic nations. Malaria is a significant public health concern in the country. The malaria indicator survey (MIS) data were analyzed using the logistic regression model. Nonetheless, independent data may be the cause of most MIS's hierarchical structure. This approach does not consider any association between data points within a cluster, as it assumes that the individual malaria statuses are independent of their causes. The approach may lead to biased analysis conclusions. The primary goal of this research is to determine the impact of sample enumeration areas (SEAs) and SEA features on individual malaria rapid diagnostic test (RDT) results. We are interested in identifying key factors influencing household members' malaria RDT findings or Kenya's malaria prevalence and assessing variation.

Methods

Our study utilized the robust 2020 Kenya National Malaria Indicator Surveys (KMIS) dataset, which is representative of the entire nation. This dataset, comprising 301 clusters (134 urban and 167 rural areas), was instrumental in applying several multilevel models, including random sample and sample Enumeration Area (SEA) effects. We also considered the weights used in the s survey design, which is used to adjust uneven probabilities of choice within clusters, further enhancing the reliability and relevance of our findings. The methods used in this study involved a rigorous analysis of the KMIS dataset, including applying multilevel models and considering survey design weights to ensure the robustness and strength of our results.

Results

This study's findings are significant and crucial in understanding the prevalence of malaria in Kenya. The findings reveal that factors such as region, place of residence, mosquito bed net use, water source location, wealth index, age, household size, and altitude are significantly associated with malaria's prevalence.After accounting for these variables, systematic changes across SEAs accounted for approximately 47.1 % of the remaining variability in malaria occurrence in the study locations. In contrast, the remaining 52.9 % was projected to be unmeasured differences between individuals or family units. These findings provide a detailed explanation of the various processes that influence malaria prevalence in Kenya.

Conclusions

The study's multilevel logistic regression model, which includes random effects, identified two SEA-level and eight individual/household risk factors for malaria infection. Thus, increasing the availability of insecticide-treated bed nets is one crucial element that public health policymakers should consider. Furthermore, health planners can organize spatially targeted initiatives to prevent malaria transmission with the help of spatial clustering data.

Malaria infection and predictor factors among Chadian nomads' children

BACKGROUND

In Chad, malaria remains a significant public health concern, particularly among nomadic populations. Geographical factors and the mobility of human populations have shown to be associated with the diversity of Plasmodium species. The study aims to describe the malaria prevalence among nomadic children and to investigate its associated factors.

METHODS

A cross-sectional study was conducted in February and October 2021 among nomadic communities in Chad. Blood sample were collected and tested from 187 Arab, Fulani and Dazagada nomadic children aged 3-59 months using malaria rapid diagnostic test (RDT). A structured electronic questionnaire was administered to their parents to collect information about the socio‑economic data. Malaria testing results were categorized according to the SD BIOLINE Malaria Ag Pf/Pan RDT procedures. Logistic regression analysis was used to determine key risk factors explaining the prevalence of malaria. STATA version IC 13 was used for statistical analysis.

RESULTS

The overall malaria prevalence in nomadic children was 24.60%, with 65.20% being Plasmodium falciparum species and 34.8% mixed species. Boys were twice as likely (COR = 1.83; 95% CI, 0.92-3.62; p = 0.083) to have malaria than girls. Children whose parents used to seek traditional drugs were five times more likely (AOR = 5.59; 95% CI, 1.40-22.30, p = 0.015) to have malaria than children whose parents used to seek health facilities. Children whose parents reported spending the last night under a mosquito net were one-fifth as likely (AOR = 0.17; 95% CI, 0.03-0.90, p = 0.037) to have malaria compared to children whose parents did not used a mosquito net. Furthermore, Daza children were seventeen times (1/0.06) less likely (AOR = 0.06; 95% CI, 0.01-0.70, p = 0.024) to have malaria than Fulani children and children from households piped water as the main source were seven times more likely (AOR = 7.05; 95% CI, 1.69-29.45; p = 0.007) to have malaria than those using surface water.

CONCLUSIONS

Malaria remains a significant public health issue in the nomadic communities of Chad. Community education and sensitization programs within nomad communities are recommended to raise awareness about malaria transmission and control methods, particularly among those living in remote rural areas. The National Malaria Control Program (NMCP) should increase both the coverage and use of long-lasting insecticidal nets (LLINs) and seasonal malaria chemoprevention (SMC) in addition to promoting treatment-seeking behaviors in nomadic communities.

Complete sources of cluster variation on the risk of under-five malaria in Uganda: a multilevel-weighted mixed effects logistic regression model approach

BACKGROUND

Malaria, a major cause of mortality worldwide is linked to a web of determinants ranging from individual to contextual factors. This calls for examining the magnitude of the effect of clustering within malaria data. Regrettably, researchers usually ignore cluster variation on the risk of malaria and also apply final survey weights in multilevel modelling instead of multilevel weights. This most likely produces biased estimates, misleads inference and lowers study power. The objective of this study was to determine the complete sources of cluster variation on the risk of under-five malaria and risk factors associated with under-five malaria in Uganda.

METHODS

This study applied a multilevel-weighted mixed effects logistic regression model to account for both individual and contextual factors.

RESULTS

Every additional year in a child's age was positively associated with malaria infection (AOR = 1.42; 95% CI 1.33-1.52). Children whose mothers had at least a secondary school education were less likely to suffer from malaria infection (AOR = 0.53; 95% CI 0.30-0.95) as well as those who dwelled in households in the two highest wealth quintiles (AOR = 0.42; 95% CI 0.27-0.64). An increase in altitude by 1 m was negatively associated with malaria infection (AOR = 0.98; 95% CI 0.97-0.99). About 77% of the total variation in the positive testing for malaria was attributable to differences between enumeration areas (ICC = 0.77; p < 0.001).

CONCLUSIONS

Interventions towards reducing the burden of under-five malaria should be prioritized to improve individual-level characteristics compared to household-level features. Enumeration area (EA) specific interventions may be more effective compared to household specific interventions.

Prevalence of malaria infection and the impact of mosquito bed net distribution among children aged 6-59 months in Ghana: Evidence from the Ghana demographic health and malarial indicator surveys

Objective

To assess the prevalence of malaria infection and further quantify the impact of mosquito bed net distribution on malaria infection among children aged 6-59 months in Ghana.

Methods

A cross-sectional study using Ghana Demographic Health (GDHS) and Malaria Indicator (GMIS) surveys (2014 GDHS, 2016 GMIS, and 2019 GMIS). The exposure and the main outcomes were mosquito bed net use (MBU) and malaria infection (MI). Relative percentage change (Δ) and prevalence ratio (PR) were estimated to assess the changes and the risk of MI by MBU respectively. The Propensity-score matching treatment effect model was employed to estimate the average treatment effect (ATE) of MBU on MI. All analyses were performed using Stata 16.1 and p-value<0.05 was deemed significant.

Results

The study involved 8781 children aged 6-59 months. MI ranged from 25.8%(22.3-29.7) in 2019 GMIS to 40.6%(37.0-44.2) in 2014 GDHS and the prevalence was significantly high among children who used mosquito bed net. The relative percentage change in MI prevalence showed a significant reduction rate and was high among non-MBU (p-value<0.05). In all, the adjusted PR of MI among children exposed to MBU was 1.21(1.08-1.35), 1.13(1.01-1.28), and 1.50(1.20-1.75) in 2014 GDHS, 2016 GMIS, and 2019 GMIS respectively. The average MI among participants who slept in mosquito bed net significantly increased by 8%(0.04 to 0.12), 4%(0.003 to 0.08), and 7%(0.03 to 0.11) in 2014 GDHS, 2016 GMIS, and 2019 GMIS respectively.

Conclusion

Even though malaria infection prevalence among children aged 6-59 months is decreasing, the reduction rate seems not to be directly linked with mosquito bed nets distribution and/or use in Ghana. For a continued distribution of mosquito bed nets, and for Ghana to achieve her Malaria Strategic Plan (NMSP) 2021-2025, program managers should ensure effective use of the distributed nets in addition to other preventive measures and nuanced consideration of community behaviours in Ghana. The effective use and care of bed nets should be emphasized as part of the distribution.

Multilevel logistic regression modelling to quantify variation in malaria prevalence in Ethiopia

Background

Ethiopia has low malaria prevalence compared to most other malaria-endemic countries in Africa. However, malaria is still a major public health problem in the country. The binary logistic regression model has been widely used to analyse malaria indicator survey (MIS) data. However, most MIS have a hierarchical structure which may result in dependent data. Since this model assumes that conditional on the covariates the malaria statuses of individuals are independent, it ignores potential intra-cluster correlation among observations within a cluster and may generate biased analysis results and conclusions. Therefore, the aim of this study was to quantify the variation in the prevalence of malaria between sample enumeration areas (SEAs) or clusters, the effects of cluster characteristics on the prevalence of malaria using the intra-class correlation coefficient as well as to identify significant factors that affect the prevalence of malaria using the multilevel logistic regression modelling in three major regions of Ethiopia, namely Amhara, Oromia and Southern Nations, Nationalities and Peoples’ (SNNP).

Methods

Dataset for three regional states extracted from the 2011 Ethiopian National Malaria Indicator Surveys (EMIS) national representative samples was used in this study. It contains 9272 sample individuals selected from these regions. Various multilevel models with random sample SEA effects were applied taking into account the survey design weights. These weights are scaled to address unequal probabilities of selection within clusters. The spatial clustering of malaria prevalence was assessed applying Getis-Ord statistic to best linear unbiased prediction values of model random effects.

Results

About 53.82 and 28.72 per cents of the sampled households in the study regions had no mosquito net and sprayed at least once within the last 12 months, respectively. The results of this study indicate that age, gender, household had mosquito nets, the dwelling has windows, source of drinking water, the two SEA-level variables, i.e. region and median altitude, were significantly related to the prevalence of malaria. After adjusting for these seven variables, about 45% of the residual variation in the prevalence of malaria in the study regions was due to systematic differences between SEAs, while the remaining 55% was due to unmeasured differences between persons or households. The estimated MOR, i.e. the unexplained SEA heterogeneity, was 4.784. This result suggests that there is high variation between SEAs in the prevalence of malaria. In addition, the 80% interval odds ratios (IORs) related to SEA-level variables contain one suggesting that the SEA variability is large in comparison with the effect of each of the variable.

Conclusions

The multilevel logistic regression with random effects model used in this paper identified five individual / household and two SEA-level risk factors of malaria infection. Therefore, the public health policy makers should pay attentions to those significant factors, such as improving the availability of pure drinking water. Further, the findings of spatial clustering provide information to health policymakers to plan geographically targeted interventions to control malaria transmission.