Monthly malaria surveillance data for 47 counties in Kenya, 2015–2025

Overview:This dataset contains monthly panel data on malaria incidence and potential environmental, climatic, and intervention‑related drivers for 47 counties in Kenya from January 2015 to December 2025. It was compiled to support the development and validation of machine learning models for the multi‑state classification of malaria transmission intensity. The data are aggregated at the county level and include temporally lagged features to enable time‑aware predictive modelling.

Geographic coverage:47 counties of Kenya, representing the full administrative county structure as defined by the Independent Electoral and Boundaries Commission (IEBC).

Temporal coverage:January 2015 – December 2025 (132 months per county, totalling 6,204 observations).

Variable name
Description

county
Name of the county (categorical)

year, month, date
Temporal identifiers

temperature
Average monthly temperature (°C)

precipitation
Total monthly precipitation (mm)

NDVI
Normalised Difference Vegetation Index (proxy for vegetation greenness)

ITN_coverage
Estimated proportion of the population with access to an insecticide‑treated net (0–1)

elevation
Mean elevation of the county (metres)

population_density
Estimated population per square kilometre

incidence_per_1000
Monthly malaria incidence (confirmed + probable cases per 1,000 population)

transmission_state
Target variable – categorical transmission state derived from incidence thresholds: 0 = no/low, 1 = moderate, 2 = high, 3 = epidemic (thresholds detailed in the data dictionary)

*_lag1, *_lag2
One‑month and two‑month lagged versions of temperature, precipitation, NDVI, ITN coverage, and transmission state

Data sources:

Malaria case data: Kenya Health Information System (KHIS) and National Malaria Control Programme (NMCP) routine surveillance (aggregated, de‑identified).

Climatic and environmental data: ERA5 (temperature), CHIRPS (precipitation), MODIS (NDVI).

ITN coverage: NMCP household survey estimates and routine distribution data, disaggregated to county level.

Elevation: SRTM digital elevation model.

Population density: WorldPop and Kenya National Bureau of Statistics (KNBS) projections.