In alwinw/epocakir: Clinical Coding of Patients with Kidney Disease

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.height = 4,
  fig.width = 7
)

Introduction

The epocakir package makes clinical coding of patients with kidney disease using clinical practice guidelines easy. The guidelines used are the evidence-based KDIGO guidelines. This package covers acute kidney injury (AKI), anemia, and chronic liver disease(CKD).

Features

aki_staging(): Classification of AKI staging (aki_stages) with automatic selection of:
aki_bCr(): AKI based on baseline creatinine
aki_SCr(): AKI based on changes in serum creatinine
aki_UO(): AKI based on urine output
anemia(): Classification of anemia
Classification of albuminuria (Albuminuria_stages)
Albuminuria_staging_ACR(): Albuminuria based on Albumin excretion rate
Albuminuria_staging_AER(): Albuminuria based on Albumin-to-creatinine ratio
eGFR(): Estimation of glomerular filtration rate with automatic selection of:
eGFR_adult_SCr(): eGFR based on the 2009 CKD-EPI creatinine equation
eGFR_adult_SCysC(): eGFR based on the 2012 CKD-EPI cystatin C equation
eGFR_adult_SCr_SCysC(): eGFR based on the 2012 CKD-EPI creatinine-cystatin C equation
eGFR_child_SCr(): eGFR based on the pediatric creatinine-based equation
eGFR_child_SCr_BUN(): eGFR based on the pediatric creatinine-BUN equation
eGFR_child_SCysC(): eGFR based on the pediatric cystatin C-based equation
GFR_staging(): Staging of GFR (GFR_stages)
Multiple utility functions including:
conversion_factors: Conversion factors used throughout the KDIGO guidelines
as_metric(): Conversion of a measured value into metric units
dob2age(): Calculation of age from a date of birth
binary2factor(): Conversion of binary data into factors based on a column name
combine_date_time_cols(): Combining separate date and time columns into a single date and time column
combn_changes: Generating changes between measurements
Automatic conversion of units class objects
Tidy output allowing seamless integration with functions from the tidyverse
Tidyeval via programming with dplyr
Comprehensive tests and coverage

Examples

library(epocakir)
library(dplyr)
library(units)

Clinical Data

Often clinical data must be cleansed and tidied before analysis can begin. To assist in this, several utility functions have been included. To explore these, consider a sample clinical dataset clinical_obvs:

# Example workflow: clinical_obvs <- read.csv("cohort.csv")
glimpse(clinical_obvs)

tidy_obvs <- clinical_obvs %>%
  combine_date_time_cols() %>%
  mutate(
    Age = dob2age(`Date of Birth`),
    Height = as_metric(height = set_units(as.numeric(Height), "cm"))
  ) %>%
  binary2factor(Male, Surgery)

glimpse(tidy_obvs)

Make sure to use set_units() from the units package to convert all measurements into unit objects for automatic unit conversion in epocakir.

AKI Staging

Next consider the sample aki_pt_data dataset. It is possible to use aki_staging() to automatically classify the presence and staging of AKI. If a particular method is required, it is possible to classify AKI using aki_bCr(), aki_SCr() or aki_UO().

# Example workflow: aki_pt_data <- read.csv("aki.csv")
head(aki_pt_data)

aki_staging(aki_pt_data,
  SCr = "SCr_", bCr = "bCr_", UO = "UO_",
  dttm = "dttm_", pt_id = "pt_id_"
)

aki_pt_data %>%
  mutate(aki = aki_staging(
    SCr = SCr_, bCr = bCr_, UO = UO_,
    dttm = dttm_, pt_id = pt_id_
  )) %>%
  select(pt_id_, SCr_:dttm_, aki)

aki_pt_data %>%
  mutate(aki = aki_SCr(
    SCr = SCr_, dttm = dttm_, pt_id = pt_id_
  )) %>%
  select(pt_id_, SCr_:dttm_, aki)

Estimated Glomerular Filtration Rate

Similarly, eGFR() offers the ability to automatically select the appropriate formula to estimate the glomerular filtration rate. If a particular formula is required, then eGFR_adult_SCr, eGFR_adult_SCysC, eGFR_adult_SCr_SCysC, eGFR_child_SCr, eGFR_child_SCr_BUN, or eGFR_child_SCysC can be used.

# Example workflow: aki_pt_data <- read.csv("aki.csv")
head(eGFR_pt_data)

eGFR(eGFR_pt_data,
  SCr = "SCr_", SCysC = "SCysC_",
  Age = "Age_", height = "height_", BUN = "BUN_",
  male = "male_", black = "black_", pediatric = "pediatric_"
)

eGFR_pt_data %>%
  dplyr::mutate(eGFR = eGFR(
    SCr = SCr_, SCysC = SCysC_,
    Age = Age_, height = height_, BUN = BUN_,
    male = male_, black = black_, pediatric = pediatric_
  )) %>%
  select(SCr_:pediatric_, eGFR)

eGFR_pt_data %>%
  dplyr::mutate(eGFR = eGFR_adult_SCr(
    SCr = SCr_, Age = Age_, male = male_, black = black_
  )) %>%
  select(SCr_:pediatric_, eGFR)