Skip to contents

Combine WHO and GTFCC Vaccination Data

Source: R/combine_vaccination_data.R
combine_vaccination_data.Rd

This function intelligently combines vaccination data from WHO and GTFCC sources, prioritizing GTFCC data while identifying and including WHO campaigns that are missing from GTFCC. It uses multiple matching criteria to determine if campaigns are duplicates, accounting for date discrepancies and dose variations.

Usage

combine_vaccination_data(PATHS, date_tolerance = 60, dose_tolerance = 0.2)

Arguments

PATHS

A list containing file paths for input and output data. The list should include:

  • MODEL_INPUT: Path to the directory containing processed vaccination data files and where combined data will be saved.

date_tolerance

Number of days tolerance for matching campaign dates between sources (default: 60 days)

dose_tolerance

Proportion tolerance for matching doses between sources (default: 0.2 = 20%)

Value

The function saves the combined vaccination data to a CSV file in the directory specified by PATHS$MODEL_INPUT. It also returns the combined data as a data frame for further use in R.

Details

The function performs intelligent campaign matching using the following approach:

  1. Load Processed Data: - Reads WHO and GTFCC processed vaccination data files

  2. Intelligent Matching: - Matches campaigns by country, approximate date (within tolerance), and dose similarity - Uses multiple passes with different tolerance levels to maximize accurate matching - Identifies truly unique WHO campaigns not present in GTFCC

  3. Data Combination: - Prioritizes GTFCC data (marked as source = "GTFCC") - Adds unique WHO campaigns (marked as source = "WHO_only") - Includes campaigns present in both (marked as source = "GTFCC_WHO_matched") - match_confidence is "high" (Step 1: +/-7 days, +/-5% doses), "medium" (Step 2: within the tolerances), "low" (Step 3: +/-30 days, any dose, or Step 4 only), or "GTFCC_only" / "WHO_only" - In Steps 1-3 each GTFCC campaign absorbs at most one WHO campaign, so a genuinely separate second campaign within 30 days is kept - When GTFCC records a campaign under the WHO row's ICG request number, only that request's campaign(s) are candidates in every step - Step 4 (same ICG request): GTFCC rows are per-request totals (all deliveries summed) while WHO rows are per shipment, so a still-unmatched WHO shipment is absorbed by the GTFCC campaign with the same ICG request number (WHO 20174 = GTFCC 201704) as long as the WHO doses assigned to that campaign stay within (1 + dose_tolerance) of its GTFCC doses (e.g. MOZ 2017-I04: 709.1K GTFCC vs 329.6K + 354.6K WHO); a WHO row whose request number GTFCC does not use is matched on the ICG decision date (+/-7 days) within the dose tolerance instead - Step 5 (repeated WHO-only rows): WHO-only rows sharing country, ICG request and decision date are kept in campaign-date order only while their summed doses stay within (1 + dose_tolerance) of the request's approved total; further rows are dropped as duplicate listings (MWI 20182: two 500,600-dose rows against 500,600 approved) - The GTFCC request columns (req_id, delivery_schedule, round_sequence, round_basis; see process_GTFCC_vaccination_data) are carried on every GTFCC row, matched or not. WHO-only rows have no delivery schedule and no round information (round_basis = "unknown"), so est_vaccination_rate releases their doses from the campaign date and counts them as first doses.

  4. Quality Assurance: - Validates data structure matches downstream requirements - Ensures all required columns are present - Maintains date and ID formatting standards

Examples

if (FALSE) { # \dontrun{
# Example usage (requires set_root_directory() + the MOSAIC-data repo)
PATHS <- get_paths()
combined_data <- combine_vaccination_data(PATHS)

# With custom tolerance settings
combined_data <- combine_vaccination_data(PATHS, date_tolerance = 30, dose_tolerance = 0.1)
} # }