Skip to contents

Plot Distribution of Antibiotic Resistance Across Sample Types

Source: R/prep_eda_plots.R
plot_resistance_by_sample.Rd

Produces a horizontal 100% stacked bar chart showing the proportion of Resistant (R) and Susceptible (S) deduplicated AST results for the top n sample types. Each bar sums to 100% and is labelled with the R and S percentages.

Usage

plot_resistance_by_sample(
  data,
  n = 10,
  mode = c("faceted", "overall", "single"),
  center = NULL,
  patient_col = "PatientInformation_id",
  sample_col = "sample_type",
  organism_col = "organism_name",
  antibiotic_col = "antibiotic_name",
  value_col = "antibiotic_value",
  center_col = "center_name",
  colours = c(R = "#D73027", S = "#1A9850"),
  ncol = 2,
  base_size = 14,
  title = NULL,
  syndrome_col = NULL,
  syndrome_name = NULL
)

Arguments

data

Data frame. Long-format AMR dataset.

n

Integer. Number of top sample types to show (ranked by deduplicated test count). Default 10.

mode

Character. One of "faceted", "overall", or "single". Default "faceted".

center

Character. Required when mode = "single". Exact centre name.

patient_col

Character. Patient ID column. Used for deduplication. Default "PatientInformation_id".

sample_col

Character. Sample/specimen type column. Default "sample_type".

organism_col

Character. Organism name column. Used for deduplication. Default "organism_name".

antibiotic_col

Character. Antibiotic name column. Default "antibiotic_name".

value_col

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

center_col

Character. Centre/facility column. Default "center_name".

colours

Named character vector for R and S bars. Default c("R" = "#D73027", "S" = "#1A9850").

ncol

Integer. Facet columns (mode = "faceted" only). Default 2.

base_size

Numeric. Base font size. Default 14.

title

Character. Custom title. Auto-generated if NULL. Default NULL.

syndrome_col

Character or NULL. Column name containing syndrome/infection category labels (e.g. "infectious_syndrome"). If NULL (default), no syndrome filtering is applied.

syndrome_name

Character or NULL. The syndrome value to retain (e.g. "BSI", "VAP"). Requires syndrome_col to be set. If NULL (default), all syndromes are included.

Value

A ggplot object.

Details

Deduplication (worst-phenotype): before counting, rows are collapsed by patient x centre x sample type x organism x antibiotic. If the same antibiotic appears more than once for a patient-organism-sample combination (data entry error or repeat test), any single R result marks the episode as R. This prevents duplicate rows from inflating counts.

Counting unit: deduplicated AST test results – one record per patient x organism x antibiotic x sample type combination.

Top N ranking: sample types are ranked by total deduplicated test count, so the most-tested specimen types appear first.

Default sample column: "sample_type" – the raw column. Switch to "specimen_normalized" for standardised values.

Supports three display modes:

  • "faceted" (default) – top n sample types per centre independently, one panel per centre.

  • "overall" – all centres pooled; top n globally.

  • "single" – one specific centre.