Table 2 Definition of variables
From: Evaluation of household coverage with long-lasting insecticidal nets in central Côte d’Ivoire
Variable | Description |
|---|---|
Age (in years) | Data was categorized into three groups: <5, 5–15 and >15 years |
Altitude (in meters) | Was categorized into 2 groups around the median: 148–209 (very low altitude) and 210–258 (low altitude) |
Household LLIN access (based on 1 LLIN per 2 persons) | A new variable was generated which divided the total number of nets reported (regardless if used or not) in a household by total number of household inhabitants and re-categorized into a binary variable at a value of ≥0.5 (sufficient nets) |
Household LLIN access (based on 1 net per sleeping unit) | A new variable was created by dividing the number of nets in a household by the total number of sleeping units in the household, and a binary variable created with households with a ratio of ≥1 tagged as having sufficient nets to cover each sleeping unit |
Household LLIN ownership | The discrete variable, number of nets (whether used or not), was re-categorized as a binary variable (no net reported and ≥1 LLIN reported) |
Household size | Was categorized into 3 equal groups with the following cut-offs: 1–3, 4–7 and 8–19 members |
Nearest distance to a health facility (in km) | Was categorized into 3 equal groups with the following cut-offs: 0.4–3.6, 3.7–8.0 and 8.1–13.5 |
Nearest distance to a lake (in km) | Was categorized into 4 equal groups with the following cut-offs: < 4.2, 4.3–7.7, 7.8–11.6 and 11.7–36.0 |
Normalized difference vegetation index | Was categorized into 4 groups with the following cut-offs: low (71–5700), low-moderate (5701–7526), high-moderate (7527–8003) and high (8004- 9078) vegetation cover |
Population access (based on 1 LLIN per 2 persons) | The number of potential net users was calculated for each household by multiplying the number of nets by 2 and dividing it by the household size (with a maximum value of 1) and the mean obtained across the study population |
Population access (based on 1 LLIN per sleeping unit) | A new variable was generated which divided the total number of nets by the number of sleeping units in a household (with a maximum value of this ratio being 1) and then multiplied by household size to obtain how many persons would have been potentially covered within the sleeping spaces. The sum of these potential users was obtained and divided by the sum of household sizes |
Population density (in persons/km2) | Was categorized into 2 groups around the median: 36–339 (very low) and 340–505 (low) |