,
Andy Kampfschulte [ctb] ,
Spectrum Health, Grand Rapids, MI [cph]| Title: | Power and Sample Size Calculation for the Cochran-Mantel-Haenszel Test |
|---|---|
| Description: | Calculates the power and sample size for Cochran-Mantel-Haenszel tests. There are also several helper functions for working with probability, odds, relative risk, and odds ratio values. |
| Authors: | Paul Egeler [aut, cre] ,
Andy Kampfschulte [ctb] ,
Spectrum Health, Grand Rapids, MI [cph] |
| Maintainer: | Paul Egeler <[email protected]> |
| License: | GPL-2 | GPL-3 |
| Version: | 0.0.5 |
| Built: | 2025-02-05 04:32:53 UTC |
| Source: | https://github.com/pegeler/samplesizecmh |
This data summarizes counts of a case-control study investigating the link between breast cancer rates and oral contraceptive use, stratified by age group. In toto, 10,890 subjects. See source for details.
data(contraceptives)data(contraceptives)
A 3-dimensional table.
OC Usage: Subject exposure to oral contraceptives.
Disease Status: Breast cancer present (case) or absent (control).
Age Group: Age group of the subject.
Hennekens, C. H., F. E. Speizer, R. J. Lipnick, B. Rosner, C. Bain, C. Belanger, M. J. Stampfer, W. Willett, and R. Peto. (1984). "A Case-Control Study of Oral Contraceptive Use and Breast Cancer." Journal of the National Cancer Institute 72 (1): 39–42. Table 1.
These functions will create either odds for a given probability, probability for a given odds, calculate the odds ratio between two probabilities, or calculate effect size (raise a probability by theta)
prop2odds(p) odds2prop(o) effect.size(p, theta) props2theta(p1, p2) rr2theta(rr, p1, p2) theta2rr(theta, p1, p2)prop2odds(p) odds2prop(o) effect.size(p, theta) props2theta(p1, p2) rr2theta(rr, p1, p2) theta2rr(theta, p1, p2)
p, p1, p2
|
Proportion vector. |
o |
Odds vector. |
theta |
Odds ratio vector. |
rr |
Relative risk vector ( |
A numeric vector.
Paul W. Egeler, M.S.
# Convert proportions of 0 through 1 to odds props <- seq(0,1,0.1) prop2odds(props) # Convert odds to proportions odds2prop(1:3) # Raise a proportion by an effect size theta effect.size(0.5, 2) # Find the odds ratio between two proportions props2theta(0.75, 0.5)# Convert proportions of 0 through 1 to odds props <- seq(0,1,0.1) prop2odds(props) # Convert odds to proportions odds2prop(1:3) # Raise a proportion by an effect size theta effect.size(0.5, 2) # Find the odds ratio between two proportions props2theta(0.75, 0.5)
Create an odds ratio estimate from a 2-by-2 table of frequencies or proportions
odds.ratio(x)odds.ratio(x)
x |
A two-dimensional matrix or table containing frequencies or proportions. |
A numeric vector.
Paul W. Egeler, M.S.
# Load in Titanic data from datasets package data(Titanic, package = "datasets") # Get marginal table of survival by sex marginal_table <- margin.table(Titanic, c(2,4)) marginal_table # Compute odds ratio of marginal table odds.ratio(marginal_table) # Get partial tables of survival by sex, stratified by class partial_tables <- margin.table(Titanic, c(2,4,1)) partial_tables # Compute odds ratio of each partial table apply(partial_tables, 3, odds.ratio)# Load in Titanic data from datasets package data(Titanic, package = "datasets") # Get marginal table of survival by sex marginal_table <- margin.table(Titanic, c(2,4)) marginal_table # Compute odds ratio of marginal table odds.ratio(marginal_table) # Get partial tables of survival by sex, stratified by class partial_tables <- margin.table(Titanic, c(2,4,1)) partial_tables # Compute odds ratio of each partial table apply(partial_tables, 3, odds.ratio)
Compute the post-hoc power or required number of subjects for the Cochran-Mantel-Haenszel test for association in J stratified 2 x 2 tables.
power.cmh.test( p1 = NULL, p2 = NULL, theta = NULL, N = NULL, sig.level = 0.05, power = 0.8, alternative = c("two.sided", "less", "greater"), s = 0.5, t = 1/J, correct = TRUE )power.cmh.test( p1 = NULL, p2 = NULL, theta = NULL, N = NULL, sig.level = 0.05, power = 0.8, alternative = c("two.sided", "less", "greater"), s = 0.5, t = 1/J, correct = TRUE )
p1 |
Vector of proportions of the J case groups. |
p2 |
Vector of proportions of the J control groups. |
theta |
Vector of odds ratios relating to the J 2 x 2 tables. |
N |
Total number of subjects. |
sig.level |
Significance level (Type I error probability). |
power |
Power of test (1 minus Type II error probability). |
alternative |
Two- or one-sided test. If one-sided, the direction of the association must be defined (less than 1 or greater than 1). Can be abbreviated. |
s |
Proportion (weight) of case versus control in J stratum. |
t |
Proportion (weight) of total number of cases of J stratum. |
correct |
Logical indicating whether to apply continuity correction. |
This sample size calculation is based on the derivations described in the Woolson et al. (1986). It is designed for case-control studies where one margin is fixed. The method is "based on the Cochran-Mantel-Haenszel statistic expressed as a weighted difference in binomial proportions."
Continuity corrected sample size is described in Nam's 1992 paper. This uses the weighted binomial sample size calculation described in Woolson et al. (1986) but is enhanced for use with the continuity corrected Cochran's test.
Power calculations are based on the writings of Wittes and Wallenstein (1987). They are functionally equivalent to the derivations of the sample size calculation described by Woolson and others and Nam, but have slightly added precision.
Terminology and symbolic conventions are borrowed from Woolson et al.
(1986). The p1 group is dubbed the Case group and p2
group is called the Control group.
An object of class "power.cmh": a list of the original arguments and
the calculated sample size or power. Also included are vectors of n's per
each group, an indicator or whether continuity correction was used, the
original function call, and N.effective.
The vectors of n's per each group, n1 and n2, are the
fractional n's required to achieve a final total N specified by the
calculation while satisfying the constraints of s and t.
However, the effective N, given the requirement of cell counts populated by
whole numbers is provided by N.effective. By default, the print method
is set to n.frac = FALSE, which will round each cell n up to the
nearest whole number.
To calculate power, the power parameter must be set to
NULL. To calculate sample size, the N parameter must
be set to NULL.
The J number of groups will be inferred by the maximum length of
p1, p2, or theta.
Effect size must be specified using one of the following combinations of arguments.
Both case and control proportion vectors, ex.,
p1 and p2 with theta = NULL.
One proportion vector and an effect size, ex.,
p1 and theta with p2 = NULL, or
p2 and theta with p1 = NULL.
Paul W. Egeler, M.S.
Gail, M. (1973). "The determination of sample sizes for trials involving several 2 x 2 tables." Journal of Chronic Disease 26: 669-673.
Munoz, A. and B. Rosner. (1984). "Power and sample size for a collection of 2 x 2 tables." Biometrics 40: 995-1004.
Nam, J. (1992). "Sample size determination for case-control studies and the comparison of stratified and unstratified analyses." Biometrics 48: 389-395.
Wittes, J. and S. Wallenstein. (1987). "The power of the Mantel-Haenszel test." Journal of the American Statistical Association 82: 1104-1109.
Woolson, R. F., Bean, J. A., and P. B. Rojas. (1986). "Sample size for case-control studies using Cochran's statistic." Biometrics 42: 927-932.
power.prop.test, mantelhaen.test, BreslowDayTest
# From "Sample size determination for case-control studies and the comparison # of stratified and unstratified analyses", (Nam 1992). See references. # Uncorrected sample size estimate first introduced # by Woolson and others in 1986 sample_size_uncorrected <- power.cmh.test( p2 = c(0.75,0.70,0.65,0.60), theta = 3, power = 0.9, t = c(0.10,0.40,0.35,0.15), alternative = "greater", correct = FALSE ) print(sample_size_uncorrected, detail = FALSE) # We see that the N is 171, the same as calculated by Nam sample_size_uncorrected$N # Continuity corrected sample size estimate added by Nam sample_size_corrected <- power.cmh.test( p2 = c(0.75,0.70,0.65,0.60), theta = 3, power = 0.9, t = c(0.10,0.40,0.35,0.15), alternative = "greater", correct = TRUE ) print(sample_size_corrected, n.frac = TRUE) # We see that the N is indeed equal to that which is reported in the paper sample_size_corrected$N# From "Sample size determination for case-control studies and the comparison # of stratified and unstratified analyses", (Nam 1992). See references. # Uncorrected sample size estimate first introduced # by Woolson and others in 1986 sample_size_uncorrected <- power.cmh.test( p2 = c(0.75,0.70,0.65,0.60), theta = 3, power = 0.9, t = c(0.10,0.40,0.35,0.15), alternative = "greater", correct = FALSE ) print(sample_size_uncorrected, detail = FALSE) # We see that the N is 171, the same as calculated by Nam sample_size_uncorrected$N # Continuity corrected sample size estimate added by Nam sample_size_corrected <- power.cmh.test( p2 = c(0.75,0.70,0.65,0.60), theta = 3, power = 0.9, t = c(0.10,0.40,0.35,0.15), alternative = "greater", correct = TRUE ) print(sample_size_corrected, n.frac = TRUE) # We see that the N is indeed equal to that which is reported in the paper sample_size_corrected$N
"power.cmh" objectThe S3 print method for the "power.cmh" object
## S3 method for class 'power.cmh' print(x, detail = TRUE, n.frac = FALSE, ...)## S3 method for class 'power.cmh' print(x, detail = TRUE, n.frac = FALSE, ...)
x |
A |
detail |
Logical to toggle detailed or simple output. |
n.frac |
Logical indicating whether sample n's should be rounded to the next whole number. |
... |
Ignored. |
Computes the relative risk of a specified column of a two-by-two table.
rel.risk(x, col.num = 1)rel.risk(x, col.num = 1)
x |
A table or matrix containing frequencies. |
col.num |
The column number upon which relative risk should be calculated. |
A numeric vector.
Paul W. Egeler, M.S.
This package provides functions relating to power and sample size calculation for the CMH test. There are also several helper functions for interconverting probability, odds, relative risk, and odds ratio values.
The Cochran-Mantel-Haenszel test (CMH) is an inferential test for the association between two binary variables, while controlling for a third confounding nominal variable. Two variables of interest, X and Y, are compared at each level of the confounder variable Z and the results are combined, creating a common odds ratio. Essentially, the CMH test examines the weighted association of X and Y. The CMH test is a common technique in the field of biostatistics, where it is often used for case-control studies.
Given a target power which the researcher would like to achieve, a
calculation can be performed in order to estimate the appropriate number of
subjects for a study. The power.cmh.test function calculates
the required number of subjects per group to achieve a specified power for a
Cochran-Mantel-Haenszel test.
Researchers interested in estimating the probability of detecting a true
positive result from an inferential test must perform a power calculation
using a known sample size, effect size, significance level, et cetera.
The power.cmh.test function can compute the power of a CMH test,
given parameters from the experiment.