raghava-thummapudi / miaocom

A R package for calculating comorbidity index for epidemiologists

MiaoCom

Overview

A R package for calculating comorbidity indexes (i.e. Charlson Comorbidity Index, Elixhauser Comorbidity Index, and C3 Index) for epidemiologists.

Installation

install.packages("devtools")
devtools::install_github("caimiao0714/MiaoCom@miao") 
# The master branch is just a initialized branch. The miao branch is the branch that I really work on.

Overview of the functions

library(MiaoCom)

• cci() is a function that calculates Charlson Comorbidity Index for patients. The ICD-10 coding algorithms for defining comorbidities refers to Hude Quan's paper in 2005. This function returns two Charlson Comorbidity Indexes:

  1. the 1987 version developed by Mary E. Charlson;
  2. the 2011 updated version by Hude Quan.

• eci() is a function that calculates Elixhauser Comorbidity Index for patients. The ICD-10 coding algorithms for defining comorbidities also refers to Hude Quan's paper in 2005.

• c3() is a function that calculates C3 Index specifically for cancer patients

Usage of functions

new_data <- cci(data, comorbidity, age)
new_data <- eci(data, comorbidity)
new_data <-  c3(data, comorbidity)

• data is a data.frame or alike objects from which you want to calculate comorbidity indexed from.
• comorbidity is a vector of the variable names of patients' comorbidity ICD-10 codings in the data. For example, c("comorbidity1", "comorbidity2","comorbidity3","comorbidity4","comorbidity5").
• age is the variable name of the age of the patients in your data, the name should be surrounded by single or double quotes (i.e. ', ").

A simple example

Step 1:　Construct a demo dataframe.

demo_data <- data.frame(
   comorbidityICD1 = c("I25.105", "I50.907", "I25.903", "I50.907", "I50.907", "I25.903",  
                       "I50.907", "I50.910", "I25.105", "I25.903"),
   comorbidityICD2 = c("I25.210", "I25.903", "I50.908", "J44.103", "I10.003", "I50.907", 
                       "K21.001", "J44.101", "I25.203", "I50.907"),
   comorbidityICD3 = c("I51.709", "K27.906", "E11.732", "I10.005", "E11.901", "E78.501", 
                       "Z98.8123", "E11.901", "I50.908", "E11.901"),
   comorbidityICD4 = c("I49.904", "I10.005", "J44.101", "K73.901", "I63.902", "J98.402", 
                       "B18.106", "E11.221", "E11.423", "E11.2211"),
   comorbidityICD5 = c("I45.101", "E11.901", "J96.903", "K70.301", "K76.811", "K72.901", 
                       "K74.602", "I15.002", "J44.003", "N18.916"),
   patient_age = c(72, 66, 81, 86, 75, 33, 63, 70, 71, 47)
)

Step 2: Apply cci() function to the demo data

newdata1 <- cci(data = demo_data, comorbidity = c("comorbidityICD1", "comorbidityICD2", 
                "comorbidityICD3", "comorbidityICD4", "comorbidityICD5"), age = "patient_age")

The generated newdata1 data.frame includes multiple new columns. The CCI_1987 and CCI_2011 columns deserve your special attention.

1. CCI_1987 is the 1987 version CCI developed by Mary E. Charlson;
2. CCI_2011 is the 2011 updated version CCI by Hude Quan.

Step 3: Apply eci() function to the demo data

newdata2 <- eci(data = demo_data, comorbidity = c("comorbidityICD1", "comorbidityICD2", 
                "comorbidityICD3", "comorbidityICD4", "comorbidityICD5"))

The generated newdata2 data.frame includes multiple new columns. The Elix_Index is your desired Elixhauser Comorbidity Index.

Step 4: Apply c3() function to the demo data

newdata3 <- c3(data = demo_data, comorbidity = c("comorbidityICD1", "comorbidityICD2", 
                "comorbidityICD3", "comorbidityICD4", "comorbidityICD5"))

The generated newdata3 data.frame includes multiple new columns. The C3 is your desired C3 Index. Please note that the C3 index is developed specifically for cancer patients. This is just a demonstration of the c3() function.

Note that CCI includes patient age as a variable for calculating the index. However, patient age is not included in calculating Elixhauser Comorbidity Index and C3 Index.