,
Marc Henrion [aut] ,
José Antonio Roldán Nofuentes [aut]
| Title: | Comparing Two Diagnostic Tests with Dichotomous Results using Paired Data |
|---|---|
| Description: | Provides a method for comparing the results of two binary diagnostic tests using paired data. Users can rapidly perform descriptive and inferential statistics in a single function call. Options permit users to select which parameters they are interested in comparing and methods for correction for multiple comparisons. Confidence intervals are calculated using the methods with the best coverage. Hypothesis tests use the methods with the best asymptotic performance. A summary of the methods is available in Roldán-Nofuentes (2020) <doi:10.1186/s12874-020-00988-y>. This package is targeted at clinical researchers who want to rapidly and effectively compare results from binary diagnostic tests. |
| Authors: | Kyle J. Wilson [cre, aut] ,
Marc Henrion [aut] ,
José Antonio Roldán Nofuentes [aut]
|
| Maintainer: | Kyle J. Wilson <[email protected]> |
| License: | GPL-3 |
| Version: | 1.1.1.9000 |
| Built: | 2024-11-20 14:02:01 UTC |
| Source: | https://github.com/kajlinko/testcomparer |
This data from the Coronary Artery Surgery Study evaluates two tests to determine the presence or absence of coronary artery disease by comparing to coronary angiography, the gold standard. Test 1 is an exercise stress test and Test 2 is a clinical history of chest pain.
casscass
A data frame with 871 rows and 3 columns:
Dichotomous result on exercise stress testing.
Presence of absence of chest pain based on medical history.
Dichotomous result on coronary angiography.
All three variables are dichotomous. 1 indicates a positive result; 0 indicates a negative result.
This data was originally presented in Weiner et al. (1979).
doi:10.1056/NEJM197908023010502
Weiner et al. (1979)) N Engl J Med. 1979;301(5):230-5 doi:10.1056/NEJM197908023010502
This data from the Cystic Fibrosis Foundation's Patient Registry (USA) evaluates risk factors for pulmonary exacerbation in patients with cystic fibrosis. The two risk factors evaluated are previous pulmonary exacerbation and previous colonisation with Pseudomonas aeruginosa. Each of the two risk factors was evaluated using data from 1995. If an instance occurred at any point in 1995 the 'test' was considered positive. If negative throughout 1995 the 'test' was considered negative. The gold standard was evidence of pulmonary exacerbation at any point in 1996.
cfprcfpr
A data frame with 11,960 rows and 3 columns:
Presence or absence of previous pulmonary exacerbation.
Presence or absence of Pseudomonas aeruginosa infection.
Presence or absence of severe infection (gold standard).
All three variables are dichotomous. 1 indicates presence; 0 indicates absence.
This data was originally presented in Moskowitz and Pepe (2006).
Data was sourced directly from the referenced paper. For up-to-date data requests contact: Cystic Fibrosis Foundation
Moskowitz and Pepe (2006) Clinical Trials. 2006;3(3):272-9. doi:10.1191/1740774506cn147oa
Calculates descriptive statistics and performs statistical inference on two binary diagnostic tests in a single function call. Handles multiple comparisons using methods in 'p.adjust()'.
compareR( df = NULL, test1 = NULL, test2 = NULL, gold = NULL, interpret = FALSE, alpha = 0.05, margins = FALSE, multi_corr = "holm", cc = TRUE, dp = 1, sesp = TRUE, ppvnpv = TRUE, plrnlr = TRUE, conf.int = "contemporary", test.names = c("Test 1", "Test 2"), ... )compareR( df = NULL, test1 = NULL, test2 = NULL, gold = NULL, interpret = FALSE, alpha = 0.05, margins = FALSE, multi_corr = "holm", cc = TRUE, dp = 1, sesp = TRUE, ppvnpv = TRUE, plrnlr = TRUE, conf.int = "contemporary", test.names = c("Test 1", "Test 2"), ... )
df |
A data frame or matrix with 3 columns (test1, test2, gold). Flexible coding of positive and negative results permitted. |
test1 |
Either a vector of values for Test 1 (if df is NULL) or a string or integer value to be used for subsetting df. |
test2 |
Either a vector of values for Test 2 (if df is NULL) or a string or integer value to be used for subsetting df. |
gold |
Either a vector of values for the gold standard test (if df is NULL) or a string or integer value to be used for subsetting df. |
interpret |
If TRUE provides a verbose plain English interpretation of the output by calling interpretR(). Defaults to FALSE. |
alpha |
An alpha value. Defaults to 0.05. |
margins |
A Boolean value indicating whether the contingency tables should have margins containing summed totals of rows and columns. |
multi_corr |
Method for multiple comparisons. Uses 'p.adjust.methods'. |
cc |
A Boolean value indicating whether McNemar's test should be applied with continuity correction. |
dp |
Number of decimal places of output in summary tables. Defaults to 1. |
sesp |
A Boolean value indicating whether output should include sensitivity and specificity. |
ppvnpv |
A Boolean value indicating whether output should include positive and negative predictive values. |
plrnlr |
A Boolean value indicating whether output should include positive and negative likelihood ratios. |
conf.int |
A character string, either "contemporary" or "classic". Indicates whether function should use contemporary or classic statistical methods to calculate confidence intervals. |
test.names |
A vector of length two giving the names of the two different binary diagnostic tests. This argument is not relevant when testing a single binary diagnostic test. |
... |
Rarely needs to be used. Allows additional arguments to be passed to internal functions. |
Confidence intervals for prevalence, diagnostic accuracies and predictive values are calculated using the interval for binomial proportions described by Yu et al. (2014) by default. Setting conf.int = "classic" uses the Clopper-Pearson method. Confidence intervals for likelihood ratios are calculated using the methods recommended by Martín-Andrés and Álvarez-Hernández (2014). Hypothesis testing for diagnostic accuracies uses different methods depending on disease prevalence and number of participants or samples as described by Roldán-Nofuentes and Sidaty-Regad (2019). Global hypothesis testing for predictive values uses a method described by Roldán-Nofuentes et al. (2012), with subsequent individual tests (where indicated) performed using methods described by Kosinksi (2012). The methods for hypothesis testing- for likelihood ratios are taken from Roldán-Nofuentes & Luna del Castillo (2007).
An excellent summary of these methods is provided by Roldán-Nofuentes (2020) along with an open-source program (compbdt) licensed under GPL-2. This R package can be considered an extension of this work and is therefore distributed under the same license. Please consider citing Roldán-Nofuentes (2020) when you are citing this package.
A list object summarising all calculated descriptive and inferential statistics.
Yu, Guo & Xu (2014) JSCS. 2014; 84:5,1022-1038 doi:10.1080/00949655.2012.738211
Clopper & Pearson (1934) Biometrika. 1934; 26,404-413 doi:10.2307/2331986
Martín Andrés & Álvarez Hernández (2014) Stat Comput. 2014; 24,65–75 doi:10.1007/s11222-012-9353-5
Roldán-Nofuentes & Sidaty-Regad (2019) JSCS. 2019; 89:14,2621-2644 doi:10.1080/00949655.2019.1628234
Roldán-Nofuentes, Luna del Castillo & Montero-Alonso (2012) Comput Stat Data Anal. 2012; 6,1161–1173. doi:10.1016/j.csda.2011.06.003
Kosinski (2012) Stat Med. 2012; 32,964-977 doi:10.1002/sim.5587
Roldán-Nofuentes, Luna del Castillo (2007) Stat Med. 2007; 26:4179–201. doi:10.1002/sim.2850
Roldán-Nofuentes (2020) BMC Med Res Methodol. 2020; 20,143 doi:10.1186/s12874-020-00988-y
# load data df <- cfpr # run compareR function compareR(df, margins = TRUE, multi_corr = "bonf", test.names = c("pulm.exac", "pseudomonas") )# load data df <- cfpr # run compareR function compareR(df, margins = TRUE, multi_corr = "bonf", test.names = c("pulm.exac", "pseudomonas") )
Produces a data frame which can be used by the compareR function using values commonly found in published literature. Useful for reviews and meta-analyses.
dataframeR(s11, s10, s01, s00, r11, r10, r01, r00)dataframeR(s11, s10, s01, s00, r11, r10, r01, r00)
s11 |
Number of cases where Test 1 is positive, Test 2 is positive and gold standard is positive. |
s10 |
Number of cases where Test 1 is positive, Test 2 is negative and gold standard is positive. |
s01 |
Number of cases where Test 1 is negative, Test 2 is positive and gold standard is positive. |
s00 |
Number of cases where Test 1 is negative, Test 2 is negative and gold standard is positive. |
r11 |
Number of cases where Test 1 is positive, Test 2 is positive and gold standard is negative. |
r10 |
Number of cases where Test 1 is positive, Test 2 is negative and gold standard is negative. |
r01 |
Number of cases where Test 1 is negative, Test 2 is positive and gold standard is negative. |
r00 |
Number of cases where Test 1 is negative, Test 2 is negative and gold standard is negative. |
Understanding the parameter names: s & r represent positive and negative results for the gold standard test, respectively. The first digit represents a positive (1) or negative (0) result for Test 1. The second digit represents a positive (1) or negative (0) result for Test 2.
A data frame populated with zeros and ones indicating positive or negative test results which can be passed to the compareR function.
# build data frame using numbers dataframeR(3, 3, 3, 3, 3, 3, 3, 3)# build data frame using numbers dataframeR(3, 3, 3, 3, 3, 3, 3, 3)
Provides a plain English readout of the results of the compareR function.
interpretR(result)interpretR(result)
result |
A list object with class 'compareR' output from the compareR function. |
A plain English summary of the findings produced by the compareR function.
# simulate data test1 <- c(rep(1, 300), rep(0, 100), rep(1, 55), rep(0, 145)) test2 <- c(rep(1, 280), rep(0, 120), rep(1, 45), rep(0, 155)) gold <- c(rep(1, 400), rep(0, 200)) dat <- data.frame(test1, test2, gold) # compare with compareR result <- compareR(dat) # provide a plain English readout with interpretR interpretR(result)# simulate data test1 <- c(rep(1, 300), rep(0, 100), rep(1, 55), rep(0, 145)) test2 <- c(rep(1, 280), rep(0, 120), rep(1, 45), rep(0, 155)) gold <- c(rep(1, 400), rep(0, 200)) dat <- data.frame(test1, test2, gold) # compare with compareR result <- compareR(dat) # provide a plain English readout with interpretR interpretR(result)
An S3 method to plot a simple visualisation of the results from the compareR function.
## S3 method for class 'compareR' plot(x, ...)## S3 method for class 'compareR' plot(x, ...)
x |
An object of class compareR. |
... |
Arguments such as graphical parameters. Not currently in use. |
Method to plot the most commonly used results of the compareR output.
A visualisation of the results for diagnostic accuracies and predictive values from the compareR output.
# generate result res <- compareR(cass, test1 = "exercise", test2 = "cp", gold = "angio", test.names = c("ExerciseStressTest", "ChestPain")) # run print method plot(res)# generate result res <- compareR(cass, test1 = "exercise", test2 = "cp", gold = "angio", test.names = c("ExerciseStressTest", "ChestPain")) # run print method plot(res)
An S3 method to print the results verbose from the compareR function.
## S3 method for class 'compareR' print(x, ...)## S3 method for class 'compareR' print(x, ...)
x |
An object of class compareR. |
... |
Further arguments passed to or from other methods. |
Method to print the pertinent results of the compareR output.
A printed results table of the compareR output.
# generate result res <- compareR(cass, test1 = "exercise", test2 = "cp", gold = "angio", test.names = c("ExerciseStressTest", "ChestPain")) # run print method print(res)# generate result res <- compareR(cass, test1 = "exercise", test2 = "cp", gold = "angio", test.names = c("ExerciseStressTest", "ChestPain")) # run print method print(res)
Summarises descriptive statistics associated with a single binary diagnostic test.
summariseR(df, dp = 1)summariseR(df, dp = 1)
df |
A data frame or matrix with 2 columns (test1, gold). Flexible coding of positive and negative results permitted. |
dp |
Number of decimal places of output in summary tables. Defaults to 1. Kappa defaults to 3 decimal places unless user selects more. |
Confidence intervals for prevalence, diagnostic accuracies and predictive values are calculated using the interval for binomial proportions described by Yu et al. (2014). Confidence intervals for likelihood ratios are calculated using the methods recommended by Martín-Andrés and Álvarez-Hernández (2014). Cohen's kappa is a value between -1 and 1 which describes the agreement of the two tests, taking account of random agreement. A score of zero or less indicates the agreement could be entirely due to chance.
A summary of the descriptive statistics of a binary diagnostic test, compared to a gold standard.
Yu, Guo & Xu (2014) JSCS. 2014; 84:5,1022-1038 doi:10.1080/00949655.2012.738211
Martín Andrés & Álvarez Hernández (2014) Stat Comput. 2014; 24,65–75 doi:10.1007/s11222-012-9353-5
Cohen (1960) Educ Psychol Meas. 1960; 20(1),37–46 doi:10.1177/001316446002000104
# simulate data test1 <- c(rep(1, 300), rep(0, 100), rep(1, 55), rep(0, 145)) gold <- c(rep(1, 400), rep(0, 200)) dat <- data.frame(test1, gold) # summarise descriptive statistics result <- summariseR(dat, dp = 4)# simulate data test1 <- c(rep(1, 300), rep(0, 100), rep(1, 55), rep(0, 145)) gold <- c(rep(1, 400), rep(0, 200)) dat <- data.frame(test1, gold) # summarise descriptive statistics result <- summariseR(dat, dp = 4)
An S3 method to summarise the rather verbose output of the compareR function.
## S3 method for class 'compareR' summary(object, ...)## S3 method for class 'compareR' summary(object, ...)
object |
An object of class compareR. |
... |
Additional arguments affecting the summary produced. |
Method to summarise the verbose compareR output.
A summary of the compareR output.
# generate result res <- compareR(cass, test1 = "exercise", test2 = "cp", gold = "angio", test.names = c("ExerciseStressTest", "ChestPain")) # run summary method summary(res)# generate result res <- compareR(cass, test1 = "exercise", test2 = "cp", gold = "angio", test.names = c("ExerciseStressTest", "ChestPain")) # run summary method summary(res)