Skip to contents

Compute Per-Profile Fatal Prevalence of Resistance R_Kdelta* (Eq. 13)

Source: R/yll.R
compute_fatal_resistance_prevalence.Rd

For each **resistant** profile delta of pathogen K, computes the fatal prevalence of resistance:

Usage

compute_fatal_resistance_prevalence(
  profiles_with_rr,
  probability_col = "probability",
  rr_profile_col = "RR_LOS_profile",
  dominant_class_col = "dominant_class",
  facility_col = NULL,
  facility_name = NULL
)

Arguments

profiles_with_rr

Named list from assign_rr_to_profiles(rr_col = "RR_death"), one element per pathogen. Each element must contain probability_col, rr_profile_col, and dominant_class_col.

probability_col

Character. Profile prevalence column R'_Kdelta. Default "probability".

rr_profile_col

Character. Dominant-class converted RR. Default "RR_LOS_profile".

dominant_class_col

Character. Column identifying the dominant class (or "all_susceptible"). Default "dominant_class".

facility_col

Character or NULL. Facility identifier column. Default NULL.

facility_name

Character or NULL. If provided, stored in the output for provenance tracking. Default NULL.

Value

Named list (one per pathogen). Each element:

per_profile

Data frame of resistant profiles augmented with R_star (R*_Kdelta per profile) and numerator_delta.

R_K_star

Scalar: \(\sum_\delta R^*_{K\delta}\) = total fatal resistance prevalence for pathogen K.

sum_r_prime

Scalar: \(\sum_\delta R'_{K\delta}\) (resistant profiles only).

susceptible_fraction

Scalar: \(1 - \sum_\delta R'_{K\delta}\).

denominator

Scalar: shared denominator for all profiles.

n_resistant_profiles

Integer: number of resistant profiles.

Details

$$ R^*_{K\delta} = \frac{R'_{K\delta} \cdot RR_{Kd^*}} {\left(1 - \textstyle\sum_\delta R'_{K\delta}\right) + \textstyle\sum_\delta R'_{K\delta} \cdot RR_{Kd^*}} $$

where the sums \(\sum_\delta\) run over resistant profiles only (profiles with at least one resistant antibiotic class, i.e. dominant_class != "all_susceptible"). The term \((1 - \sum_\delta R'_{K\delta})\) is the susceptible fraction and must be > 0 for the formula to give R*_Kdelta < 1.

Important: pass the output of assign_rr_to_profiles() directly – do not call filter_profiles_to_rr_classes() first, because renormalisation would set \(\sum R'_{K\delta} = 1\) and collapse the denominator, producing R*_Kdelta = 1 for every pathogen.

References

Bhaswati Ganguli. DALY Methodology for AMR (YLD notes). March 2026. Eq. 13.