This function creates a summary table of ICD-10 diagnosis class distributions across exposure groups, excluding diagnoses matching a specified ICD-10 pattern (e.g., the exposure-defining diagnosis). It calculates patient percentages per diagnosis class and reshapes the result into a wide-format matrix suitable for plotting (e.g., radar charts or heatmaps).
Usage
classify_icd10_profile(
data,
diagnoses = diagnoses,
exposure_icd10 = "",
exposure_src = c("")
)Arguments
- data
A data frame containing the study population with at least columns `ID` and `exp.GROUP`, where `exp.GROUP` distinguishes between exposure and non-exposure groups. Can be created by classify_population() -function.
- diagnoses
A Original data frame of diagnoses that must include columns `ID`, `DGREG`, `DG`, `SRC`, and `ICD10_CLASS`.
- exposure_icd10
A regular expression string that defines the ICD-10 diagnosis used to define the exposure group. Diagnoses matching this pattern are excluded from the analysis.
- exposure_src
A character vector of source types (e.g., `"avohilmo"`, `"hilmo"`, etc.) to include in the filtering.
Value
A wide-format data frame with ICD-10 class percentages by exposure group. Includes artificial `"Max"` and `"Min"` rows (100 and 0) for visualization purposes.
Details
The function internally calculates: - Population sizes for each exposure group - Diagnosis counts and patient counts per ICD-10 class - Percentages of patients per group with each diagnosis class
If run inside a Shiny application, it will display progress bars using `withProgress()`.
Examples
if (FALSE) { # \dontrun{
classify_icd10_profile(
data = dpop,
diagnoses = diagnoses,
exposure_icd10 = "^E11",
exposure_src = c("avohilmo", "hilmo", "erko", "local")
)
} # }