This R package implements a pipeline to process clinical episode data, identify chronic pathologies, and calculate frailty and comorbidity scores based on patient diagnosis codes.
The pipeline performs the following steps:
Setup environment
Load and install required R packages and source supporting
scripts.
Data Preparation
Clean and format input episode data based on user-specified column
mappings.
Chronic Pathologies Identification
Apply algorithms to detect and propagate chronic conditions within
episodes.
Frailty Calculation
Calculate frailty indices from updated episode data.
Comorbidity and Frailty Summary
Combine frailty and comorbidity measures into final result
tables.
You can either clone the repository or download the ZIP file.
git clone https://github.com/bayaniazadeh/LabTNSCPSSPackage.git
cd LabTNSCPSSPackagePlace your input CSV in LABTNSCPSS_Data/, e.g.,
LABTNSCPSS_Data/testpackage.csv
Go to your directory that the LabTNSCPSSPackage folder exists and
run the file LabTNSCPSSPackage.Rproj, then open
Frailty_Comorbidity_Pipeline.R in your open R
studio.
In the Frailty_Comorbidity_Pipeline.R code edit these
information: - The input dataset should be a CSV file with episode-level
patient data. - Required columns (default mapping):
- Patient_id — patient ID
- ICD — ICD coding system diagnosis codes
- start_date — episode start date
- end_date — episode end date
- episode_id — unique episode identifier
You can customize these column names by modifying the
col_mapping list in the pipeline.
source("./LABTNSCPSS_Code/setup_package.R") # Load/install packages
source("./LABTNSCPSS_Code/source_scripts.R") # Load pipeline functions
coding_system <- get_coding_system()Here you should select the ICD vesion of according to your data: ICD-10-CA, ICD-10-CM, ICD-11, write it in the console part and press enter.
Run the rest of the code line by line.
Finally you can find the generated files at
LABTNSCPSS_Data/
In the folder data/ you can find all the mapping files
and categorizations in “.rda”” format. To be able to explore the
mappings in your R studio browser use this code, and replace “file_name”
with your desired data file :
df <- as.data.frame(file_name)
View(df)