R code for new CKD equation - with "if" statements-CodePudding

I'm new to coding in R and am having a difficult time coding the following equation

142 x (Scr/A)^B x 0.9938^age x (1.012 if female)

Female

Scr <=0.7; A = 0.7, B = -0.241
Scr >0.7; A = 0.7, B = -1.2

Male

Scr <=0.9; A = 0.9, B = -0.302
Scr >0.9; A = 0.9, B = -1.2

(There is an existing R package for the old version of this equation, but not this updated one).

Could someone help me with the code? Having a hard time figuring out how to create a new variable (eGFR) with values calculated from this equation. The components A and B of the equation depend on 2 categories of variable Scr (serum creatinine) and on gender (M/F). Thanks!

    wave1.data$baseline.egfr <- (if(wave1.data$Sex=="Male") {M1 <- wave1.data$Creatinine >= 0.9
      M2 <- wave1.data$Creatinine < 0.9
      eGFR <- (142*(wave1.data$Creatinine/A1)^-0.302*(0.9938^wave1.data$Age))
      eGFR <- (142*(wave1.data$Creatinine/A2)^-1.2*(0.9938^wave1.data$Age))}
    else if(wave1.data$Sex=="Female")
    {F1 <- wave1.data$Creatinine >= 0.7
    F2 <- wave1.data$Creatinine < 0.7
    eGFR <- (142*(wave1.data$Creatinine/A1)^-0.241*(0.9938^wave1.data$Age))
      eGFR <- (142*(wave1.data$Creatinine/A2)^-1.2*(0.9938^wave1.data$Age))})

CodePudding user response：

I hope this function well represents the formula for individual case:

ckdi <- function(sex_i, scr, age){
  female_scr <- c(0, 0.7)
  male_scr <- c(0, 0.9)
  female_A <- c(0.7, 0.7)
  female_B <- c(-0.241, -1.2)
  male_A <- c(0.9, 0.9)
  male_B <- c(-0.302, -1.2)
  
  if(sex_i == "F"){
    scrid <- findInterval(scr, female_scr, left.open = TRUE)
    A <- female_A[scrid]
    B <- female_B[scrid]
    k <- 1.012
  }
  if(sex_i == "M"){
    scrid <- findInterval(scr, male_scr, left.open = TRUE)
    A <- male_A[scrid]
    B <- male_B[scrid]
    k<- 1.000
  }
  result_i <- 142*(scr/A)^B * 0.9938^age * k
  return(result_i)
}

To apply to multiple data, mapply can be used. For example:

set.seed(1)
scr_all <- runif(3)
age_all <- runif(3, 30,50)
sex_all <- c("F", "M", "F")

mapply(ckdi, sex_all, scr_all, age_all)
#       F        M        F 
#134.5373 150.0365 111.9148

CodePudding user response：

You can do:

eGFR <- function(Scr, age, sex) {
  sex <- tolower(sex)
  A <- ifelse(sex == "female", 0.7, 0.9)
  B <- ifelse(Scr > A, -1.2,
              ifelse(sex == "female", -0.241, -0.302))
  mult <- ifelse(sex == "female", 1.012, 1)
  
 round(142 * (Scr/A)^B * 0.9938^age * mult)
}

eGFR(0.6, 35, "male")
#> [1] 129

eGFR(0.75, 35, "Male")
#> [1] 121

eGFR(0.8, 80, "female")
#> [1] 74

eGFR(Scr = c(0.6, 0.8), age = c(35, 80), sex = c("male", "female"))
#> [1] 129  74

^{Created on 2022-04-18 by the reprex package (v2.0.1)}