comorbidityAnalysis: Comorbidity Analysis 'cAnalysis'

In aGutierrezSacristan/comorbidity: Analyze comorbidities from electronic health record data

Description

Given an object of type comoRbidity, a comorbidity analysis is perform, for the subset of population under specific conditions of age and sex. It generates a cAnalysis object.

Usage

comorbidityAnalysis(input, codesPth, databasePth, ageRange = c(0, 100),
  sex = "ALL", score, correctedPval = 1, correctionMethod = "fdr",
  oddsRatio, rr, phi, cores = 1, verbose = FALSE, warnings = TRUE)

Arguments

input	A comorbidity object, obtained with the query function.
codesPth	Determines the path where the file with the index diseases is located (indexDiseaseCode)
databasePth	Determines the path where the three required input files (patientData, diagnosisData, admissionData) are located.
ageRange	Determines what is the age range of interest for performing the comorbidity analysis. By default it is set from 0 to 100 years old.
sex	Determine what is the sex of interest for performing the comorbidity analysis. By default ALL. Change it to the sex of interest for your comorbidity analysis.
score	The comorbidity score is a measure based on the observed comorbidities and the expected ones, based on the occurrence of each disease.
correctedPval	By default 1.
correctionMethod	A Fisher exact test for each pair of diseases is performed to assess the null hypothesis of independence between the two diseases. The Benjamini-Hochberg false discovery rate method ("fdr") is applied to correct for multiple testing by default. However user can select the best correction method for the analysis. The adjustment methods include the Bonferroni correction ("bonferroni"), Holm correction ("holm"), Hochberg correction ("hochberg"), Hommel ("hommel") and Benjamini & Yekutieli ("BY").
oddsRatio	The odds ratio represents the increased chance that someone suffering disease X will have the comorbid disorder Y.
rr	The relative risk refers to the fraction between the number of patients diagnosed with both diseases and random expectation based on disease prevalence.
phi	The Pearsons correlation for binary variables (Phi) measures the robustness of the comorbidity association.
cores	By default 1. To run parallel computations on machines with multiple cores or CPUs, the cores argument can be changed.
verbose	By default FALSE. Change it to TRUE to get a on-time log from the function.
warnings	By default TRUE. Change it to FALSE to don't see the warnings.

Value

An object of class cAnalysis

Examples

load(system.file("extdata", "comorbidity.RData", package="comoRbidity"))
ex1 <- comorbidityAnalysis( 
              input              = comor_obj,
              databasePth      = system.file("extdata", package="comoRbidity"),
              codesPth         = system.file("extdata", package="comoRbidity"),
              ageRange         = c(0,50),
              sex              = "Female", 
              score            = 1,
              correctionMethod = "fdr", 
              correctedPval  = 1
              )

aGutierrezSacristan/comorbidity documentation built on April 10, 2020, 5:54 p.m.