Table 7 Variables, data collection methods, analysis techniques, and quantification approaches used in the reviewed articles

Nixon et al. [9]

1. Location of households

Field observation

The distance between households and larval habitats

Hierarchical logistic regression models

Odds Ratio or Relative Risk

2. Location of larval habitats

Fornace et al. [13]

1. Elevation

Satellite images

Malaria reported cases

Multivariate logistic regression

Odds Ratio or Relative Risk

2. Distance from house to forest

Ahmad et al. [31]

1. Location of households

Field observation

The distance between households and larval habitats

The abundance of larva

Others: The percentage of the vectors

2. Location of larval habitats

Durnez et al. [28]

The characteristics of land covers in capture locations (forest plot, village)

Field observation

The density of mosquitoes (e.g. man biting rate (MBR))

Non-parametric classification and regression tree models

The relative importance (RI) score of discriminants that affect the mosquitoes' abundance

Ninphanomchai et al. [20]

Monthly meteorological data (rainfall, temperature, and humidity)

Field observation

Malaria reported cases

Poisson regression

Correlation, regression, or other coefficients

Xu et al. [25]

The characteristics of land covers around households (Hilly zone, Larval habitats within 100 m, Vegetation nearby)

Field observation

Malaria reported cases

Matched univariate and multivariate logistic regression analyses

Odds Ratio or Relative Risk

Sluydts et al. [14]

Location of malaria infection cases

Field observation

Malaria reported cases

Spatial clusters of malaria cases

Other: Prevalence of malaria infection from different villages

Lawpoolsri et al. [15]

Light forest coverage

Satellite images

Malaria reported cases

Multivariate generalized linear mixed models

Odds Ratio or Relative Risk

Zhao et al. [16]

1. Forest cover

Satellite images

Risk score of Multi-Criteria Decision Analysis (weighted linear combination)

Multi-criteria decision analysis

Correlation, regression, or other coefficients

2. Crop land

3. Water body

4. Elevation

5. Urbanization

6. Distance to road

7. Distance to health facilities

Fornace et al. [10]

Coverage of forest

Satellite images

Malaria reported cases

Bayesian hierarchical model

Prevalence of malaria

Jeffree et al. [11]

The presence of breed sites (stagnant water)

Field observation

Malaria reported cases

Multiple logistic regression

Odds Ratio or Relative Risk

Okami and Kohtake [21]

1. NDVI

Satellite images

Malaria reported cases

Generalized linear regression model

Correlation, regression, or other coefficients

2. NDWI

Inverse Distance Weight (for interpolation)

3. TWI

4. annual average temperature

Inthavong et al. [26]

The presence of breed sites (cattle near household)

Field observation

Malaria reported cases

Generalized linear regression model

Odds Ratio or Relative Risk

Grigg et al. [27]

The characteristics of land covers (tall grass, young forest, rice paddy field)

Field observation

Malaria reported cases

Logistic regression models

Odds Ratio or Relative Risk

Kaewpitoon et al. [24]

1. Land used (agriculture areas, number of houses, water reservoirs, forest areas)

Field observation

Malaria reported cases

Multiple regression

Correlation, regression, or other coefficients

2. Anopheles adult density in villages with reported cases

3. Average annual rainfall, average annual temperature, and annual relative humidity

Van Bortel et al. [29]

The characteristics of land covers where the mosquitoes were collected (forest, village, and a route between the forest and village)

Field observation

The density of mosquitoes

Negative binomial regression

The mean density of biting rate per night between of the village and forest areas

Zhang et al. [32]

Land covers: types of forest in villages

Satellite images

The density and diversity of mosquitoes

The abundance of mosquitoes and the indicator of species diversity

The diversity indices of mosquitos (Simpson’s diversity index and Shannon–Wiener’s index)

Sato et al. [17]

1. Wetland

Field observation

Malaria reported cases

Generalized linear mixture model

Prevalence of malaria infection using different land cover types

2. Monoculture palm oil plantation

3. Mosaic oil palm plantation

4. Monoculture rubber plantation

5. Dense forest

6. Degraded forest

7. Bush, cropland mosaic

Hasyim et al. [18]

1. Altitude

Field observation

Malaria reported cases

1. Ordinary least squares

Correlation, regression, or other coefficients

2. Distance from river

2. Geographically weighted regression

3. Distance from lake and pond

4. Distance from forests

5. Annual rainfall

Mercado et al. [12]

1. Land covers: forest coverage

Satellite images

Malaria reported cases

Pearson's correlation analysis

Correlation, regression, or other coefficients

2. Meteorological data: aggregate monthly temp, rainfall, humidity

Wangdi et al. [22]

Meteorological data (Maximum temperature without lag time)

Field observation

Malaria reported cases

Zero-inflated Poisson regression

Odds Ratio or Relative Risk

Fornace et al. [19]

The coverage of forest

Field observation

Malaria reported cases

General linearized mixed models with a negative binomial distribution

Odds Ratio or Relative Risk

Tangena et al. [33]

1. Immature rubber plants

Satellite images

The density of mosquitoes

Generalized estimating equations

Odds Ratio or Relative Risk

2. Mature rubber plants

3. Villages

4. Secondary forests

5. Season (wet and dry seasons)

Fornace et al. [30]

1. EVI

Satellite images

The density of mosquitoes

Bayesian model with Integrated Nested Laplace Approximation

Correlation, regression, or other coefficients

2. Distance to the forest

Yang et al. [23]

1. Annual average temperature

Field observation

Malaria reported cases

Geographically weighted regression

Correlation, regression, or other coefficients

2. Annual cumulative rainfall

3. Rice yield per square kilometer