MED.DRMod: Calculate MED for a DRMod object

Description

This function calculates the minimum effective dose (MED) for a DRMod object. The MED is the smallest dose that achieves a clinical relevant effect. See Bretz et al. (2005) for a detailed definition.

Usage

## S3 method for class 'DRMod':
MED(object, type = c("MED2", "MED1", "MED3"),
    clinRel, gamma = 0.05, old = FALSE,
    direction = c("increasing", "decreasing"),
    doseSeq = NULL, lenDose = 101, data = getData(object), uGrad, ...)

Arguments

object

An object of class DRMod

type

The type of MED estimator to be used. "MED1"-"MED3" are generalization of the estimators described in Bretz et al. (2005), for the case of no covariates. To obtain the estimators described in Bretz et al. (2005), set the argument old

clinRel

Numeric specifying the clinical threshold(s) to be used for MED estimation. Should be of the same length as gamma, or of length 1.

gamma

Numeric specifying the gamma parameter needed for the confidence interval used in the MED estimate. Should be in (0,0.5] as 1-2*gamma confidence intervals are used. gamma should be of the same length as clinRel, or of length 1.

old

Logical indicating, whether old versions of the MED estimates should be calculated.

direction

Character specifying the direction of the dose effect.

doseSeq

Sequence of values allowed for the dose estimator. Should include 0 (ie placebo) as a first value.

lenDose

If doseSeq is not specified an equally spaced grid of lenDose values between placebo and the maximal dose in the study is used.

data

Data frame containing the variables on which the DRMod object was fitted

uGrad

Function to return the gradient of a user defined model, see Examples of the fitDRModel function.

...

Additional arguments

Value

Returns a numeric containing the MED estimates, if the MED cannot be determined within the dose-range NA is returned.

References

Bornkamp B., Pinheiro J. C., and Bretz, F. (2009). MCPMod: An R Package for the Design and Analysis of Dose-Finding Studies, Journal of Statistical Software, 29(7), 1--23 Bretz, F., Pinheiro, J. C., and Branson, M. (2005), Combining multiple comparisons and modeling techniques in dose-response studies, Biometrics, 61, 738--748

Pinheiro, J. C., Bretz, F., and Branson, M. (2006). Analysis of dose-response studies - modeling approaches, in N. Ting (ed.). Dose Finding in Drug Development, Springer, New York, pp. 146--171

Examples

Run this code

data(biom)
fit <- fitDRModel(resp ~ dose, biom, "logistic")
## MED2 estimates MED as smallest dose where the dose-effect
## curve is larger than clinRel and the lower bound of the
## 1-2*gamma confidence interval for the mean is larger than 0
## select gamma 0.05, 0.1 and 0.5
## (for gamma=0.5 CI is irrelevant for definition of MED estimate)
MED(fit, "MED2", clinRel = 0.2, gamma = c(0.05, 0.1, 0.5))

## option "old" uses the definition given in Bretz et al. (2005)
## to calculate the MED based on the full dose-response function
## this is included for back-compatibility with the old MCPMod function.
## This option is not available if covariates are used
MED(fit, "MED2", clinRel = 0.2, gamma = c(0.05, 0.1, 0.5), old = TRUE)