relrisk: Regression Adjusted Relative Risks

In lbreg: Log-Binomial Regression with Constrained Optimization

Description

This function calculates the relative risks RR adjusted for covariates (acting on a previous log-binomial regression fit) and confidence intervals (by default 95 percent) for the estimated RR. The confidence interval is calculated from the log(RR) and backtransformed.

Usage

relrisk(object, alpha = 0.05, dispersion = FALSE)

Arguments

object	object of class 'lbreg'.
alpha	1 - desired confidence level.
dispersion	logical. TRUE if standard errors should be adjusted for dispersion estimate based on Pearson residuals.

Value

value

table with estimated relative risks, lower and upper bounds of condifidence intervals.

Author(s)

Bernardo B. Andrade

References

Andrade, BB; Andrade JML (2018) Some results for Maximum Likelihood Estimation of Adjusted Relative Risks. Communications in Statistics - Theory and Methods.

See Also

lbreg

Examples

require(lbreg)

# ungrouped data
# data preparation
data(PCS)
w <- PCS
w <- w[,-1]
w$race <- factor(w$race)
w$dpros <- factor(w$dpros)
w$dcaps <- factor(w$dcaps)

# log-binomial regression
fm <- lbreg(tumor ~ ., data=w)

# relative risks
relrisk(fm)
relrisk(fm, alpha=.10)

# grouped data 
require(lbreg)
data(Caesarian)
m1 <- lbreg( cbind(n1, n0) ~ RISK + NPLAN + ANTIB, data=Caesarian)
relrisk(m1)
relrisk(m1, dispersion=TRUE)

lbreg documentation built on Dec. 26, 2019, 5:06 p.m.