Create Event IDs from Patient-Level Data

Converts patient-level isolate data to event-level data by grouping repeat cultures from the same infection episode.

Usage

create_event_ids(
  data,
  patient_col = "patient_id",
  date_col = "date_of_culture",
  organism_col = "organism_normalized",
  specimen_col = "specimen_type",
  antibiotic_col = "antibiotic_name",
  value_col = "antibiotic_value",
  gap_days = 14
)

Arguments

data

Data frame (long format – one row per antibiotic test).

patient_col

Patient ID column. Default "patient_id".

date_col

Culture date column. Default "date_of_culture".

organism_col

Organism column. Default "organism_normalized".

specimen_col

Specimen type column. Default "specimen_type".

antibiotic_col

Antibiotic name column. Default "antibiotic_name".

value_col

Susceptibility result column (S/I/R). Default "antibiotic_value".

gap_days

Days threshold: gap > gap_days triggers a new event. Default 14.

Value

Original data frame with event_id column added.

Details

Event classification rules (in order of precedence):

| Scenario | Result | |—————————————————|—————-| | Different body sites (any date) | Separate events| | Same site, same day, same organism, same ABG | One event | | Same site, same day, same organism, diff ABG | Separate events| | Same site, same organism, within gap_days, same ABG | Same event | | Same site, same organism, within gap_days, ABG changed | New event | | Same site, same organism, after > gap_days | New event | | Same site, same day, different organisms | Separate events|

Event IDs are numbered GLOBALLY per patient in chronological order, so a patient's events across all organisms/sites form one consistent sequence.