Skip to contents

Filter Profiles to Classes with Actual RR Estimates

Source: R/burden.R, R/yll.R
filter_profiles_to_rr_classes.Rd

After assign_rr_to_profiles(), some profiles may be resistant only to antibiotic classes that have no RR estimate in the 7-centres LOS data (e.g. a class present in the susceptibility dataset but not tested in the LOS cohort). Those profiles receive fallback_rr = 1 and contribute zero to the PAF numerator while their probability mass still enters the denominator, silently distorting the estimate.

Drops profiles where every resistant class is absent from the RR table (they would receive fallback_rr = 1 and silently distort the PAF), then re-normalises the surviving profile probabilities to sum to 1.

Usage

filter_profiles_to_rr_classes(
  profiles_with_rr,
  rr_table,
  pathogen_col = "pathogen",
  class_col = "antibiotic_class",
  probability_col = "probability",
  fallback_rr = 1
)

filter_profiles_to_rr_classes(
  profiles_with_rr,
  rr_table,
  pathogen_col = "pathogen",
  class_col = "antibiotic_class",
  probability_col = "probability",
  fallback_rr = 1
)

Arguments

profiles_with_rr

Named list from assign_rr_to_profiles().

rr_table

Data frame. Must have pathogen_col and class_col.

pathogen_col

Character. Default "pathogen".

class_col

Character. Default "antibiotic_class".

probability_col

Character. Default "probability".

fallback_rr

Numeric. Default 1.

Value

Named list with the same structure as profiles_with_rr but with unmatched profiles removed and probabilities re-normalised.

Named list with unmatched profiles removed and probabilities re-normalised.

Details

This function drops any profile where every resistant class (\(\delta_c = 1\)) is absent from the per-pathogen RR table, then re-normalises the surviving profile probabilities to sum to 1.