# risk.ratio.confidence: Confidence intervals for the Risk Ratio Integral In pifpaf: Potential Impact Fraction and Population Attributable Fraction for Cross-Sectional Data

## Description

Function that calculates confidence interval for the integral

\int RR(x;θ)f(x)dx

where f(x) is the density function of the exposure X, RR(x;θ) the relative risk of the exposure with associated parameter θ.

## Usage

 1 2 3 risk.ratio.confidence(X, thetahat, rr, thetavar, weights = rep(1/nrow(as.matrix(X)), nrow(as.matrix(X))), nsim = 1000, confidence = 95, check_thetas = TRUE, force.min = FALSE) 

## Arguments

 X Random sample (data.frame) which includes exposure and covariates. thetahat Estimator (vector) of theta for the Relative Risk function. rr function for Relative Risk which uses parameter theta. The order of the parameters shound be rr(X, theta). **Optional** thetavar Estimator of variance of thetahat weights Normalized survey weights for the sample X. nsim Number of simulations. confidence Confidence level % (default: 95). check_thetas Checks that theta parameters are correctly inputed. force.min Boolean indicating whether to force the rr to have a minimum value of 1 instead of 0 (not recommended).

## Note

The force.min option forces the relative risk rr to have a minimum of 1 and thus an rr < 1 is NOT possible. This is only for when absolute certainty is had that rr > 1 and should be used under careful consideration. The confidence interval to acheive such an rr is based on the paper by Do Le Minh and Y. .s. Sherif

## Author(s)

Rodrigo Zepeda-Tello [email protected]

Dalia Camacho-Garc<c3><ad>a-Forment<c3><ad> [email protected]

## References

Sherif, Y. .s. (1989). The lower confidence limit for the mean of positive random variables. Microelectronics Reliability, 29(2), 151-152.

## See Also

risk.ratio.approximate.confidence for a method when only mean(X) and var(X) are known.

## Examples

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 ## Not run: #Example 1: Exponential Relative Risk #-------------------------------------------- set.seed(18427) X <- data.frame(rnorm(100)) thetahat <- 0.1 thetavar <- 0.2 rr <- function(X, theta){exp(theta*X)} risk.ratio.confidence(X, thetahat, rr, thetavar) #We can force RR'.s CI to be >= 1. #This should be done with extra methodological (philosophical) care as #RR>= 1 should only be assumed with absolute mathematical certainty risk.ratio.confidence(X, thetahat, rr, thetavar, force.min = TRUE) #Example 2: Multivariate Relative Risk #-------------------------------------------- set.seed(18427) X1 <- rnorm(1000) X2 <- runif(1000) X <- data.frame(cbind(X1,X2)) thetahat <- c(0.02, 0.01) thetavar <- matrix(c(0.1, 0, 0, 0.4), byrow = TRUE, nrow = 2) rr <- function(X, theta){exp(theta[1]*X[,1] + theta[2]*X[,2])} risk.ratio.confidence(X, thetahat, rr, thetavar) #Example 3: Categorical Relative Risk & Exposure #-------------------------------------------- set.seed(18427) X <- as.data.frame(sample(c("Normal","Overweight","Obese"), 100, replace = TRUE, prob = c(0.4, 0.1, 0.5))) thetahat <- c(1, 1.2, 1.5) thetavar <- diag(c(0.1, 0.2, 0.4)) #Categorical relative risk function rr <- function(X, theta){ #Create return vector with default risk of 1 r_risk <- rep(1, length(X)) #Assign categorical relative risk r_risk[which(X == "Normal")] <- thetahat[1] r_risk[which(X == "Overweight")] <- thetahat[2] r_risk[which(X == "Obese")] <- thetahat[3] return(r_risk) } risk.ratio.confidence(data.frame(X), thetahat, rr, thetavar) #Example 4: Categorical Relative Risk & continuous exposure #---------------------------------------------------------- set.seed(18427) BMI <- data.frame(rlnorm(100, 3.1, sdlog = 0.1)) thetahat <- c(Malnourished = 2.2, Normal = 1, Overweight = 1.8, Obese = 2.5) thetavar <- diag(c(0.5, 0.1, 0.1, 0.2)) rr <- function(X, theta){ #Create return vector with default risk of 1 r_risk <- rep(1, length(X)) #Assign categorical relative risk r_risk[which(X < 20)] <- theta[1] #Malnourished r_risk[intersect(which(X >= 20), which(X < 25))] <- theta[2] #Normal r_risk[intersect(which(X >= 25), which(X < 30))] <- theta[3] #Overweight r_risk[which(X >= 30)] <- theta[4] #Obese return(r_risk) } risk.ratio.confidence(BMI, thetahat, rr, thetavar) #Example 5: Bivariate exposure and rr ("classical rr") #------------------------------------------------------------------ set.seed(18427) X <- sample(c("Exposed","Unexposed"), 1000, replace = TRUE, prob = c(0.1, 0.9)) thetahat <- c("Exposed" = 2.5, "Unexposed" = 1.2) thetavar <- matrix(c(0.1, 0.2, 0.2, 0.4), ncol = 2) rr <- function(X, theta){ #Create relative risk function r_risk <- rep(1, length(X)) #Assign values of relative risk r_risk[which(X == "Unexposed")] <- theta["Unexposed"] r_risk[which(X == "Exposed")] <- theta["Exposed"] return(r_risk) } risk.ratio.confidence(data.frame(X), thetahat, rr, thetavar) ## End(Not run) 

pifpaf documentation built on Sept. 29, 2017, 1:03 a.m.