summary.sievePH {sievePH}

Summarizing Mark-Specific Proportional Hazards Model Fits

Description

summary method for an object of class sievePH.

Usage

## S3 method for class 'sievePH'
summary(
  object,
  markGrid,
  contrast = c("te", "hr", "loghr"),
  sieveAlternative = c("twoSided", "oneSided"),
  confLevel = 0.95,
  ...
)

## S3 method for class 'summary.sievePH'
print(x, digits = 4, ...)

Arguments

object

an object of class sievePH, usually a result of a call to sievePH

markGrid

a matrix specifying a grid of multivariate mark values, where rows correspond to different values on the (multivariate) grid and columns correspond to components of the mark. A numeric vector is allowed for univariate marks. The point and interval estimates of the contrast are calculated on this grid.

contrast

a character string specifying the treatment effect parameter of interest. The default value is "te" (treatment efficacy); other options are "hr" (hazard ratio) and "loghr" (log hazard ratio).

sieveAlternative

a character string specifying the alternative hypothesis for the sieve tests, which can be either "twoSided" (default) or, in case of a univariate mark, "oneSided". The one-sided option is unavailable for a multivariate mark.

confLevel

the confidence level (0.95 by default) of reported confidence intervals

...

further arguments passed to or from other methods

x

an object of class summary.sievePH, usually a result of a call to summary.sievePH

digits

the number of significant digits to use when printing (4 by default)

Details

print.summary.sievePH prints a formatted summary of results. Inference about coefficients in the mark-specific proportional hazards model is tabulated. Additionally, a summary is generated from the likelihood-ratio and Wald tests of two relevant null hypotheses: (1) {H_0: HR(v)=1 for all v}, and (2) {H_0: HR(v)=HR for all v}. For the tests of (2) and a univariate mark, sieveAlternative controls the choice of the alternative hypothesis.

Value

An object of class summary.sievePH, which is a list with the following components:

• coef: a data frame summarizing point and interval estimates of the density ratio model coefficients and the marginal log hazard ratio (the confidence level is specified by confLevel), and p-values from the two-sided Wald test of the null hypothesis that the parameter equals zero

• pLR.HRunity.2sided: a numeric vector with two named components: pLR.dRatio.2sided is a p-value from the two-sided profile likelihood-ratio test of the null hypothesis H_0: \beta=0, where \beta is the vector of mark coefficients in the mark density ratio model, and pLR.cox.2sided is a p-value from the two-sided partial likelihood-ratio test of the null hypothesis H_0: \gamma=0, where \gamma is the marginal log hazard ratio in the Cox model. The two p-values are intended for the use of the Simes (1986) procedure as described on page 4 in Juraska and Gilbert (2013).

• pWald.HRunity.2sided: a p-value from the two-sided Wald test of the null hypothesis {H_0: HR(v)=1 for all v}

• pWtWald.HRunity.1sided: a p-value from the one-sided weighted Wald test of the null hypothesis {H_0: HR(v)=1 for all v} against the alternative hypothesis {H_1: HR < 1 and HR(v) is increasing in each component of v}

• pLR.HRconstant.2sided: a p-value from the two-sided profile likelihood-ratio test of the null hypothesis {H_0: HR(v)=HR for all v}. This component is available if sieveAlternative="twoSided".

• pLR.HRconstant.1sided: a numeric vector with two named components: pLR.dRatio.2sided is a p-value from the two-sided profile likelihood-ratio test of the null hypothesis {H_0: HR(v)=HR for all v}, and estBeta is the point estimate of the univariate mark coefficient in the density ratio model. This component is available if the mark is univariate and sieveAlternative="oneSided".

• pWald.HRconstant.2sided: a p-value from the two-sided Wald test of the null hypothesis {H_0: HR(v)=HR for all v}. This component is available if sieveAlternative="twoSided".

• pWald.HRconstant.1sided: a p-value from the one-sided Wald test of the null hypothesis {H_0: HR(v)=HR for all v} against the alternative hypothesis {H_1: HR(v) is increasing in v}. This component is available if the mark is univariate and sieveAlternative="oneSided".

• te: a data frame summarizing point and interval estimates of the mark-specific treatment efficacy on the grid of mark values in markGrid (available if contrast="te"). The confidence level is specified by confLevel.

• hr: a data frame summarizing point and interval estimates of the mark-specific hazard ratio on the grid of mark values in markGrid (available if contrast="hr"). The confidence level is specified by confLevel.

• te: a data frame summarizing point and interval estimates of the mark-specific log hazard ratio on the grid of mark values in markGrid (available if contrast="loghr"). The confidence level is specified by confLevel.

References

Juraska, M. and Gilbert, P. B. (2013), Mark-specific hazard ratio model with multivariate continuous marks: an application to vaccine efficacy. Biometrics 69(2):328–337.

See Also

sievePH

Examples

n <- 500
tx <- rep(0:1, each=n/2)
tm <- c(rexp(n/2, 0.2), rexp(n/2, 0.2 * exp(-0.4)))
cens <- runif(n, 0, 15)
eventTime <- pmin(tm, cens, 3)
eventInd <- as.numeric(tm <= pmin(cens, 3))
mark1 <- ifelse(eventInd==1, c(rbeta(n/2, 2, 5), rbeta(n/2, 2, 2)), NA)
mark2 <- ifelse(eventInd==1, c(rbeta(n/2, 1, 3), rbeta(n/2, 5, 1)), NA)

# fit a model with a bivariate mark
fit <- sievePH(eventTime, eventInd, data.frame(mark1, mark2), tx)
sfit <- summary(fit, markGrid=matrix(c(0.3, 0.3, 0.6, 0.3, 0.3, 0.6, 0.6, 0.6),
                                     ncol=2, byrow=TRUE))
# print the formatted summary
sfit
# treatment efficacy estimates on the grid
sfit$te