Publish (version 2023.01.17)

publish.coxph: Tabulize hazard ratios with confidence intervals and p-values.

Description

Tabulize the part of the result of a Cox regression analysis which is commonly shown in publications.

Usage

# S3 method for coxph
publish(
  object,
  confint.method,
  pvalue.method,
  print = TRUE,
  factor.reference = "extraline",
  units = NULL,
  probindex = FALSE,
  ...
)

Value

Table with hazard ratios, confidence intervals and p-values.

Arguments

object

A coxph object.

confint.method

See regressionTable

pvalue.method

See regressionTable

print

If FALSE do not print results.

factor.reference

See regressionTable

units

See regressionTable

probindex

Logical. If TRUE show coefficients on probabilistic index scale instead of hazard ratio scale.

...

passed to summary.regressionTable and also to labelUnits.

Author

Thomas Alexander Gerds

Details

Transforms the log hazard ratios to hazard ratios and returns them with confidence limits and p-values. If explanatory variables are log transformed or log2 transformed, a scaling factor is multiplied to both the log-hazard ratio and its standard-error.

Examples

library(survival)
data(pbc)
pbc$edema <- factor(pbc$edema,
             levels=c("0","0.5","1"), labels=c("0","0.5","1"))
fit = coxph(Surv(time,status!=0)~age+sex+edema+log(bili)+log(albumin),
            data=na.omit(pbc))
publish(fit)
## forest plot
plot(publish(fit),cex=1.3)

publish(fit,ci.digits=2,pvalue.eps=0.01,pvalue.digits=2,pvalue.stars=TRUE)
publish(fit,ci.digits=2,ci.handler="prettyNum",pvalue.eps=0.01,
        pvalue.digits=2,pvalue.stars=TRUE)
publish(fit, ci.digits=2, ci.handler="sprintf", pvalue.eps=0.01,
        pvalue.digits=2,pvalue.stars=TRUE, ci.trim=FALSE)

fit2 = coxph(Surv(time,status!=0)~age+sex+edema+log(bili,base=2)+log(albumin)+log(protime),
    data=na.omit(pbc))
publish(fit2)

# with cluster variable
fit3 = coxph(Surv(time,status!=0)~age+cluster(sex)+edema+log(bili,base=2)
                                    +log(albumin)+log(protime),
    data=na.omit(pbc))
publish(fit3)

# with strata and cluster variable
fit4 = coxph(Surv(time,status!=0)~age+cluster(sex)+strata(edema)+log(bili,base=2)
                 +log(albumin)+log(protime),
    data=pbc)
publish(fit4)