The aim of matchedcc is to provide epidemiologists using R with Stata-like analysis of matched case-control data. This package has two functions, mcc() and mcci(), which are direct analogues of Stata’s own mcc and mcci commands.
You can install matchedcc from CRAN with:
install.packages("matchedcc")
You can also install the development version from GitHub with pak:
# install.packages("pak")
pak::pkg_install("simpar1471/matchedcc")
The functions in matchedcc are easy to use. To demonstrate their use, we will use the mccxmpl dataset, which is included in matchedcc. This dataset has two columns - cases and controls. In this dataset, cases had experienced a heart attack, and cases and controls were matched accordingly. Each column has only 1 or 0 values, which describe whether a case or control encountered our exposure - in this case, drinking >6 cups of coffee per day.
library(matchedcc) head(matchedcc::mccxmpl) #> case control #> 1 1 1 #> 2 1 0 #> 3 0 1 #> 4 0 0 #> 5 1 1 #> 6 1 1
The mcc() function will take vectors of 1s and 0s and use these to run a matched case control analysis:
mcc(cases = matchedcc::mccxmpl$case, controls = matchedcc::mccxmpl$control) #> $data #> Controls #> Cases Unexposed Exposed Total #> Unexposed 8 8 16 #> Exposed 3 8 11 #> Total 11 16 27 #> #> $mcnemar_chi2 #> #> McNemar's Chi-squared test #> #> data: mcc_table #> McNemar's chi-squared = 2.2727, df = 1, p-value = 0.1317 #> #> #> $mcnemar_exact_p #> Exact McNemar significance probability #> 0.2265625 #> #> $proportions #> Proportion with factor #> Cases Controls #> 0.5925926 0.4074074 #> #> $statistics #> estimate [95% CI] #> statistic estimate lower upper #> difference 0.1851852 -0.08225420 0.4526246 #> ratio 1.4545455 0.89110096 2.3742568 #> rel. diff. 0.3125000 -0.02436881 0.6493688 #> odds ratio 2.6666667 0.64003641 15.6064036
The mcc() function can also accept a 2x2 table with matched case-control data, provided it is in the following format:
mcc_table <- table(matchedcc::mccxmpl$control,
matchedcc::mccxmpl$case)
mcc(table = mcc_table)
#> $data
#> Controls
#> Cases Unexposed Exposed Total
#> Unexposed 8 8 16
#> Exposed 3 8 11
#> Total 11 16 27
#>
#> $mcnemar_chi2
#>
#> McNemar's Chi-squared test
#>
#> data: mcc_table
#> McNemar's chi-squared = 2.2727, df = 1, p-value = 0.1317
#>
#>
#> $mcnemar_exact_p
#> Exact McNemar significance probability
#> 0.2265625
#>
#> $proportions
#> Proportion with factor
#> Cases Controls
#> 0.5925926 0.4074074
#>
#> $statistics
#> estimate [95% CI]
#> statistic estimate lower upper
#> difference 0.1851852 -0.08225420 0.4526246
#> ratio 1.4545455 0.89110096 2.3742568
#> rel. diff. 0.3125000 -0.02436881 0.6493688
#> odds ratio 2.6666667 0.64003641 15.6064036
Last but not least, if you have individual cell counts from a 2x2 table, you can provide them to mcci():
mcci(a = 8, b = 8, c = 3, d = 8) #> $data #> Controls #> Cases Unexposed Exposed Total #> Unexposed 8 8 16 #> Exposed 3 8 11 #> Total 11 16 27 #> #> $mcnemar_chi2 #> #> McNemar's Chi-squared test #> #> data: mcc_table #> McNemar's chi-squared = 2.2727, df = 1, p-value = 0.1317 #> #> #> $mcnemar_exact_p #> Exact McNemar significance probability #> 0.2265625 #> #> $proportions #> Proportion with factor #> Cases Controls #> 0.5925926 0.4074074 #> #> $statistics #> estimate [95% CI] #> statistic estimate lower upper #> difference 0.1851852 -0.08225420 0.4526246 #> ratio 1.4545455 0.89110096 2.3742568 #> rel. diff. 0.3125000 -0.02436881 0.6493688 #> odds ratio 2.6666667 0.64003641 15.6064036
The package is validated against Stata’s own outputs, using 1000 randomly generated mcci runs from Stata. The code to generate these can be seen in /tests/testdata/run-stata-mcc.R.
Parker S (2024). matchedcc: Stata-like matched case-control analysis. https://github.com/simpar1471/matchedcc/.
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4