power.mpcc: Power Calculation for Matched-Pairs Case-Control Odds Ratio...

In clayford/bme: Biostatistical Methods in Epidemiology

Description

Compute the power of the Matched-Pairs Case-Control Odds Ratio Test, or determine parameters to obtain a target power.

Usage

power.mpcc(n = NULL, OR, p2, v, sig.level = 0.05, power = NULL, M = 1,
  alternative = c("two.sided", "one.sided"), tol = .Machine$double.eps^0.25)

Arguments

n	number of matched pairs needed for study
OR	Odds Ratio (ie, the "ratio worth detecting")
p2	probability a control has a history of exposure
v	measure of concordance (of exposure) in the case and control samples. v = 1 if and only if the matching variables are not associated with exposure in the case and control samples.
sig.level	significance level (Type I error probability)
power	power of test (1 minus Type II error probability)
M	number of controls per case (default = 1)
alternative	one- or two-sided test. Can be abbreviated.
tol	numerical tolerance used in root finding, the default providing (at least) four significant digits. Root finding refers to uniroot, which is used to find sample size given power.

Details

Exactly one of the parameters n and power must be passed as NULL, and that parameter is determined from the others.

When pre-study information on concordance is limited, Fleiss and Levin (1988) recommend using a relatively large value of v, such as 2.5.

Sample size estimates for M >=1 can be overestimated when the probability is small that a sample control has a history of exposure.

Value

Object of class "power.htest", a list of the arguments (including the computed one) augmented with method and note elements.

References

Newman (2001), pages 288 - 291

Dupont, WD. (1988) Power calculations for matched case-control studies. Biometrics 44, 1157-1168.

Fleiss, JL and Levin B. (1988) Sample size determination in studies with matched pairs. Journal of Clinical Epidemiology 41, 727-730.

Examples

## Example 14.6
power.mpcc(power = 0.80, OR = 3, p2 = 0.05, v = 4.82)

clayford/bme documentation built on May 13, 2019, 7:37 p.m.