Audits distributions per measurement/analyte across sites or within a dataset to detect unit discordance (e.g. Creatinine mg/dL vs umol/L x88.42, Glucose mg/dL vs mmol/L x18.016).
Usage
detect_unit_discordance(
con_or_df,
measurement_concept_id = NULL,
measurement_concept_ids = NULL,
concept_id = NULL,
target_unit = NULL,
site_col = "site",
measurement_table = "measurement",
value_col = "value_as_number",
concept_col = "measurement_concept_id",
tolerance = 0.35
)Arguments
- con_or_df
Active DuckDB connection or data frame.
- measurement_concept_id
Single measurement concept ID.
- measurement_concept_ids
Vector of measurement concept IDs.
- concept_id
Alias for measurement_concept_id.
- target_unit
Canonical unit to compare against.
- site_col
Hospital site column name (default 'site').
- measurement_table
Measurement table name (default 'measurement').
- value_col
Value column name (default 'value_as_number').
- concept_col
Measurement concept column name (default 'measurement_concept_id').
- tolerance
Ratio deviation tolerance (default 0.35).