Time Series Analysis of Malaria Trends — Burkina Faso (2016–2023)

R
Epidemiology
Malaria
DHS
climate
Time series
Quantify the trend in malaria incidence at health district level and identify the factors associated with malaria incidence in Burkina Faso from 2016-2023 using routine cases data
Author

Ousmane Diallo, MPH-PhD

Published

November 12, 2025

Project Role: Research Associate & Lead Author

Timeline: 2016-2023 Analysis Period | 2023-2024

Status: Oral presentation at ASTMH 2024 (New Orleans) | Manuscript in preparation

Overview

The aim of this study was to assess the malaria trends in Burkina Faso at the health district level to help understand the malaria transmission level between 2016 and 2023 and to determine the factors associated with malaria incidence to better understand the drivers of malaria in Burkina Faso. This analysis combined publicly available DHS/MIS and climate data with restricted HMIS surveillance data obtained from PNLP under data sharing agreement.

Role: Sole author responsible for data cleaning, quality assessment, analysis, modeling, mapping, and reporting.

Methodology Framework

Data sources & Integration

Variable Definition Source Temporal Period
Suspected cases Clinical malaria suspicion (fever) HMIS Monthly 2016–2023 (excl. 2019)
Tested cases Received RDT or microscopy HMIS Monthly 2016–2023
Confirmed cases Positive RDT or microscopy HMIS Monthly 2016–2023
Presumed cases Treated without positive test HMIS Monthly 2016–2023
Population District population HMIS Annual (Monthy est) 2016–2023
Treatment-seeking (U5) Public/private/no care DHS/MIS Every 3 years 2014, 2017–18, 2021

Data management

A standard data quality workflow before analysis was applied:

  • Standardized district and facility names across years.

  • Removed duplicates/inconsistencies; harmonized facility identifiers.

  • Classified facility reporting activity (active vs. inactive).

  • Detected and imputed outliers using mean ± 3 SD, Median absolute deviation (MAD), and Interquartile range (IQR) rules.

  • Managed missingness; validated reporting rates.

Quality Control Workflow
View code 
#| label: data-quality-workflow
#| eval: false
#| code-fold: true
#| code-summary: "View Data Quality Control Code"

# Classify HF activity (single implementation)

inactive_threshold <- 6

# Function to classify HF
classify_activity_status <- function(df, conf_col = "conf") {
  df %>%
    dplyr::group_by(UID) %>%
    complete(
      Date = seq(min(Date, na.rm = TRUE), max(Date, na.rm = TRUE), by = "month"),
      fill = setNames(list(NA), conf_col)
    ) %>%
    dplyr::filter(year(Date) != 2019) %>%       
    dplyr::arrange(Date, .by_group = TRUE) %>%
    dplyr::mutate(
      temp_status   = if_else(is.na(.data[[conf_col]]) | .data[[conf_col]] == 0, "inactive", "active"),
      active_status = zoo::na.locf(ifelse(temp_status == "active", temp_status, NA),
                                   na.rm = FALSE, fromLast = FALSE, default = "inactive"),
      rle_vals      = rle(temp_status),
      run_len       = rep(rle_vals$lengths, rle_vals$lengths),
      run_val       = rep(rle_vals$values,   rle_vals$lengths),
      HF_status     = case_when(
        all(temp_status == "inactive") ~ "inactive",
        run_val == "inactive" & run_len >= inactive_threshold ~ "inactive",
        TRUE ~ active_status
      )
    ) %>%
    dplyr::ungroup() %>%
    dplyr::select(-temp_status, -active_status, -rle_vals, -run_len, -run_val)
}

Incidence adjustments

Crude malaria incidence was calculated by dividing the number of reported confirmed cases for each health district and month by the district population and multiplying by 1000. Then, crude incidence was adjusted for each factor in accordance with the WHO framework.

Incidence Adjustment Framework (WHO Guidelines):

  • Crude: Raw confirmed cases / population × 1,000

  • Adjustment 1: Account for testing rate variations

  • Adjustment 2: Account for facility reporting completeness

  • Adjustment 3: Account for care-seeking behavior patterns

Statistical Analysis Methods

Time Series Decomposition

  • STL decomposition (LOESS) to separate seasonal, trend, residual components.
  • Sen’s slope for monotonic trend magnitude.
  • Mann–Kendall test for trend significance (α = 0.05).
View STL Decomposition Implementation 
# Normalization helper
getNormalized <- function(vec) {
  if (!is.numeric(vec) || all(is.na(vec))) {
    warning("Input vector is non-numeric or all NA; returning original vector")
    return(vec)
  }
  vec_mean <- mean(vec, na.rm = TRUE)
  vec_sd   <- sd(vec,  na.rm = TRUE)
  if (is.na(vec_sd) || vec_sd == 0) {
    warning("Standard deviation is 0 or NA; returning original vector")
    return(vec)
  }
  (vec - vec_mean) / vec_sd
}

monthly_DS_incidence <- monthly_DS_incide %>%
  dplyr::mutate(
    mal_cases_norm                            = getNormalized(`Incidence brute`),
    incidence_adj_presumed_cases_norm         = getNormalized(Adj1),
    incidence_adj_presumed_cases_RR_norm      = getNormalized(Adj2),
    incidence_adj_presumed_cases_RR_TSR_norm  = getNormalized(Adj3)
  )

Factors associated with incidence

  • Model Type: Generalized Additive Models (GAMs) for non-linear relationships

  • Covariates: Climate variables, ITN coverage, SMC implementation, stockout indicators

  • Outcome: Fully adjusted incidence (Adjustment 3)

Key Findings & Impact

Data Quality Insights

Fig. Monthly reporting rates per health district

Key Insight: Reporting completeness varied dramatically (50-100% across districts), with systematic gaps that biased crude incidence estimates. This finding led to policy recommendations for surveillance system strengthening.

Burden Estimation Results

Fig. Incidence estimates following WHO framework. A-D: Temporal trend for the four incidences (crude incidence, incidence adjusted by testing rate, incidence adjusted by testing rate and reporting rate, incidence adjusted by testing rate, reporting rate and care-seeking rate; E-D: Spatial analysis for the four incidences from 2023.

Key Insight:

  • Adjustments 2 & 3 revealed hidden burden: Poor reporting in northern/eastern districts inflated true incidence

  • Care-seeking adjustment critical: Added ~450 cases/1,000 in Gorom-Gorom, Gaoua, Kaya

  • Policy Impact: Demonstrated need for integrated care-seeking behavior in burden estimation

Temporal Trend Analysis

Fig. Sen’s slope coefficient for the trend of malaria incidence adjusted for testing rate, weighted reporting rate and care-seeking rate; Gray color: not significant.

Key Insight:

  • Increasing trends: Gorom-Gorom, Sapone, Kampti (climate/intervention factors)

  • Declining trends: 39 of 70 districts showed significant improvements

  • Spatial clustering: Northern border regions showed persistent high transmission

GAM mixed models analysis

Technical Innovation & Skills Demonstrated

Advanced R Programming

  • Time Series Analysis: STL decomposition, trend testing, forecasting

  • Spatial Statistics: District-level mapping, spatial autocorrelation analysis

  • Data Pipeline: Automated quality control, validation, and reporting workflows

  • Reproducible Research: Version-controlled analysis with comprehensive documentation

Statistical Methods Expertise

  • GAM Modeling: Non-linear relationship detection between climate and incidence

  • Robust Statistics: Multiple outlier detection methods (MAD, IQR, Z-score)

  • Survey Integration: Complex sampling design incorporation from DHS/MIS data

Public Health Applications

  • Surveillance System Evaluation: Quantified data quality impacts on burden estimates

  • Policy Translation: Converted complex statistical findings into actionable recommendations

  • Intervention Assessment: Evaluated ITN, SMC, and case management program effectiveness

Deliverables & Impact

Technical Outputs:

  • Comprehensive R analytical pipeline (fully documented)

  • Interactive dashboard for district-level monitoring

  • Policy brief for National Malaria Control Programme

  • Manuscript for peer-reviewed publication

Policy Impact:

  • Informed 2024-2028 National Malaria Strategic Plan development

  • Guided resource allocation for surveillance system strengthening

  • Provided evidence base for targeted intervention deployment

Methodological Contributions:

  • Established best practices for HMIS data quality assessment

  • Developed reproducible workflow for routine surveillance analysis

  • Created template for multi-country burden estimation studies

Repository & Reproducibility

  • Repository: Full code

  • Data availability:

    • Public: DHS/MIS surveys, and CHIRPS climate data are fully open-access.
    • Restricted: HMIS surveillance data (available under PNLP data sharing agreement).
    • Reproducibility: Complete analytical code provided for transparency

Collaboration & Leadership

Stakeholder Engagement:

  • Direct collaboration with Burkina Faso National Malaria Control Programme

  • World Health Organization

This project demonstrates expertise in epidemiological surveillance, advanced time series analysis, multi-source data integration, and evidence-based policy support using state-of-the-art 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