Table 2.
Description of variables and parameters in ZERO-G method.
| Variable | Description | Source |
|---|---|---|
| The reported incidence in zone i at time t | Data | |
| Binomial variable representing if there was zero reported incidence | Data | |
| Number of non-focal disease consultations | Data | |
| Population count of zone | Data | |
| Longitude of zone i | Data | |
| Latitude of zone i | Data | |
| Distance between zone i and health facility j. Can be calculated using Euclidean distance or based on actual routing | Data | |
| Value of health service s provided at health facility j at time t | Data | |
| Month of year (e.g. Jan-Dec) at time t | Data | |
| The ZERO-G estimated incidence after correcting for under-ascertainment | Equation (1) | |
| FCA-based healthcare access | Equation (2, 3) | |
| Services provided by health facility j at time t | Equation (4) | |
| Coefficient for health service s | Equation (4) | |
| Distance decay coefficient | Equation (5) | |
| Competition at health facility j at time t | Equation (6) | |
| Scaling coefficient for competition at health facilities | Equation (6) | |
| Function describing temporal trend in . Specific function can be adjusted based on need | Equation (7) | |
| Probability of zero reported incidence | Equation (8, 9) | |
| Coefficients used to estimate probability of zero reported incidence | Equation (9) | |
| Zero-adjusted reported incidence | Equation (10) | |
| Function describing random forest algorithm used in data imputation | Equation (10) | |
| The sampling intensity in zone i at time t | Equation (11) | |
| Scaling coefficients for SI | Equation (12) |
Parameters estimated via the ZERO-G method have the corresponding equations listed as their source.
Note Subscript refers to zone and subscript refers to month .