Risk Factors of ITN Ownership and Use – Guinea DHS 2018

R
Epidemiology
Malaria
DHS
Factors associated with the ownership and use of insecticide-treated nets in Guinea: an analysis of the 2018 Demographic and Health Survey
Author

Ousmane DIALLO

Published

August 30, 2025

Project Role: Research Assistant Professor & Lead Author

Status: Published in malaria journal

Overview

Analysis of the 2018 Guinea Demographic and Health Survey (DHS) to identify determinants of ownership and use of insecticide-treated nets (ITNs) at both household and individual levels. Understanding the factors influencing household ITN ownership and usage among those with access will enable the Guinea National Malaria Control Programme to design targeted interventions aimed at increasing bed net coverage and utilization.

Research Leadership & Study Design

Independent Research Initiative:

Conceived, designed, and executed comprehensive epidemiological analysis of ITN utilization patterns in Guinea using complex survey methodology. Led complete research lifecycle from study conceptualization through manuscript publication and policy dissemination.

Research Innovation:

Developed analytical framework addressing both household-level ownership and individual-level usage patterns, incorporating complex survey design features often overlooked in DHS analyses. Published methodology serves as template for similar national-level studies.

Methodology

Study Design

  • Data Source: Guinea DHS 2018 (nationally representative)
  • Study Population: Households and individuals across Guinea’s administrative regions
  • Analytical Approach: Multi-level logistic regression with survey weights

Key Variables Analyzed

Outcomes:

  • ITN ownership at household level

  • ITN usage the previous night (among those with access) at individual-level

Predictors:

  • Socioeconomic: Wealth quintiles, education level, household size

  • Demographic: Age and sex of household head, marital status, pregnancy status

  • Geographic: Urban/rural residence, administrative region

  • Structural: Number of rooms, presence of children under 5

Statistical Methods

Complex Survey Design Adjustment:

  • Survey weights applied for population representativeness

  • Stratification and clustering accounted for in variance estimation

  • Design effects calculated for precision assessment

Analytical Framework:

View code 
## Install and charge the survey design library

library(survey)

## 
# Survey design specification

design_sample <- svydesign(
  ids = ~hv021, 
  strata = ~hv022, 
  weights = ~wt,
  num_p=1,
  nest = T,
  data = hh_ex
)

## Bivariate analysis

explanatory_vars <- c("wealth", "sex", "urb", "Num_childre", "region", "rooms", "hh_size", "Edu", "head_age", "Marital")

for(i in 1:length(explanatory_vars)){
  col = explanatory_vars[i]
  tbl <- survey::svytable(~HH_at_least_one + col, design_sample_hh)
  t = summary(tbl, statistic="Chisq")
  t = plyr::ldply(t)
}


## Univariate analysis of different risk factors for ITN ownership at the country level 

models <- explanatory_vars %>%      
  stringr::str_c("HH_at_least_one ~ ", .) %>%      
  
  # iterate through each univariate formula
  purrr::map(                               
    .f = ~survey::svyglm(                      
      formula = as.formula(.x),      
      family = "binomial",           
      design = design_sample_hh)) %>%       
  

  purrr::map(
    .f = ~broom::tidy(
      .x, 
      exponentiate = TRUE,           
      conf.int = TRUE)) %>%         
  
 
  dplyr::bind_rows() %>% 
  
  dplyr::mutate(across(where(is.numeric), round, digits = 2))

Key Findings

Visual Highlights

Regional variation in Guinea of A household ITN ownership; B proportion of the population with access to an ITN; C ITN usage; and D ITN usage among those with access.

Two-dimensional histogram showing the number of people who could use nets owned by the household if all nets were in use, stratified by household size, in the 2018 Guinea DHS.

ITN Ownership Patterns

Geographic Disparities:

  • Significant regional variation in ownership rates

  • Urban-rural differentials identified

  • Hotspot mapping revealed priority intervention areas

Socioeconomic Determinants:

  • Wealth quintile emerged as strongest predictor

  • Education level positively associated with ownership

  • Household composition effects documented

Usage Among Those with Access

Behavioral Factors:

  • Gap identified between access and utilization

  • Individual-level characteristics influence usage patterns

  • Opportunity for targeted behavior change interventions

Technical Implementation

Tools & Packages Used

  • R Programming: survey, srvyr, dplyr, ggplot2

  • Geospatial Analysis: tmap, sf for mapping

  • Reproducible Reporting: Quarto/R Markdown workflow

Data Management

  • Complex survey data cleaning and validation
  • Multiple imputation for missing data
  • Quality control procedures implemented

Visualization Approach

View code 
# Example: Regional variation mapping
library(tmap)
guinea_map <- tm_shape(guinea_regions) +
  tm_polygons("ITN_ownership", 
              style = "quantile",
              palette = "Blues",
              title = "ITN Ownership %") +
  tm_layout(title = "Regional ITN Ownership Patterns")

Impact & Applications

Policy Implications

  • Targeted Distribution: Evidence-based identification of priority populations

  • Resource Allocation: Geographic prioritization for intervention funding

  • Program Design: Tailored approaches for different demographic groups

Public Health Impact

  • Informed national malaria control strategy development
  • Evidence for behavior change communication messaging
  • Baseline data for intervention impact evaluation

Methodological Contributions

  • Demonstrated complex survey analysis best practices
  • Reproducible analytical workflow established
  • Template for similar cross-sectional studies

Deliverables

Technical Outputs:

  • Comprehensive statistical analysis report

  • Interactive geospatial visualizations

  • Reproducible R analysis code

  • Policy brief with actionable recommendations

Skills Demonstrated:

  • Survey Methodology: Complex sampling design analysis

  • Statistical Modeling: Multi-level logistic regression

  • Geospatial Analysis: Regional disparity mapping

  • Reproducible Research: Version-controlled analytical workflows

  • Policy Translation: Research-to-practice communication

Code Repository

Reproducibility: Full analytical code and documentation available on https://github.com/Ousmanerabi/Risk_factors_ITN_Guinea_DHS_2018

Data Access: Analysis conducted on publicly available DHS data with appropriate use agreements

Research Impact & Publications

Scientific Contribution: This analysis was published as a peer-reviewed research article, contributing to the evidence base for malaria prevention strategies in West Africa. The methodology developed in this study has been referenced by subsequent DHS analyses in the region.

This project demonstrates expertise in epidemiological research, complex survey analysis, and evidence-based public health recommendations using industry-standard statistical methods and reproducible research practices.*

Ousmane Diallo, MPH-PhD – Biostatistician, Data Scientist & Epidemiologist based in Chicago, Illinois, USA. Specializing in SAS programming, CDISC standards, and real-world evidence for clinical research.

Back to top