mStats (version 3.2.2)

diagTest: Statistics of Diagnostic Tests

Description

diagTest reports statistics of diagnostic tests

Usage

diagTest(x, y = NULL, p = NULL, rnd = 2, print.table = TRUE)
# S3 method for default
diagTest(...)
# S3 method for table
diagTest(x, y = NULL, p = NULL, rnd = 2, print.table = TRUE)
# S3 method for numeric
diagTest(x, y = NULL, p = NULL, rnd = 2, print.table = TRUE)

Arguments

x

a factor object or a table

y

an optional factor object

p

disease prevalence

rnd

specify rounding of numbers. See round.

print.table

logical value to display formatted outputs

...

optional arguments

Details

The screening tests are based on Bayes' Theorem. These tests help clinicians to correctly predict the presence or absence of a particular disease from the knowlege of test results (positive or negative) and/or the status of presenting symptoms (present or absent) or information regarding the likelihood of positive and negative test results and the likelihood of the presence or absence of a particular symptom in patients with and without a particular disease.

Statistics of diagnostic tests:

Sensitivity: True Positive (TP) rate among diseased (D+) = TP / D+

Specificity: True Negative (TN) rate among non-diseased (D-) = TN / D-

Positive predictive value (PPV): probability of D+ when all positives (AP) = D+ / AP

Negative predictive value (NPV): probability of D- when all negatives (AN) = D+ / AN

Likelihood ratio (LR) = To summarize information about a diagnostic test LR of positive result (LR+) = (Sensitivity) / (1 - Specificity) LR of negative result (LR-) = (1 - Sensitivity) / (Specificity)

Calculating confidence intervals:

95% Confidence intervals were calculated by using Wilson Score Interval as well as its continuity correction. The method was developed by Edwin Bidwell Wilson in 1927. This interval has good properties even for a small sample. Just lik Pearson's chi-squared test and Yates' continuity correction.

Several other methods have been developed such as Clopper<U+2013>Pearson interval and Agresti<U+2013>Coull interval. But Wilson score interval methods (with or without continuity correction) have been shown to be the most accurate and the most robust. For further details, see references.

Reference:

1. Biostatistics A Foundation for Analysis in the Health Sciences (10th Edition). Chapter 3.5 BAYES<U+2019> THEOREM, SCREENING TESTS, SENSITIVITY, SPECIFICITY

2. Avijit Hazra, Nithya Gogtay. Biostatistics Series Module 7: The Statistics of Diagnostic Tests. Indian J Dermatol. 2017 Jan-Feb; 62(1): 18-24. doi: 10.4103/0019-5154.198047

3. Brown, Lawrence D.; Cai, T. Tony; DasGupta, Anirban (2001). "Interval Estimation for a Binomial Proportion". Statistical Science. 16 (2): 101<U+2013>133.

4. Wallis, Sean A. (2013). "Binomial confidence intervals and contingency tests: mathematical fundamentals and the evaluation of alternative methods" .Journal of Quantitative Linguistics. 20 (3): 178<U+2013>208.

5. Newcombe, R. G. (1998). "Two-sided confidence intervals for the single proportion: comparison of seven methods". Statistics in Medicine. 17 (8): 857<U+2013>872. doi:10.1002/(SICI)1097-0258(19980430)17:8<857::AID-SIM777>3.0.CO;2-E. PMID 9595616.

Examples

# NOT RUN {

#### Biostatistics A Foundation for Analysis in the Health Sciences (10th Edition)
# numbers taken from Example 3.5.1

diagTest(as.table(matrix(c(436, 5, 14, 495), ncol = 2, byrow = TRUE)))
diagTest(as.table(matrix(c(436, 5, 14, 495), ncol = 2, byrow = TRUE)),
         p = .113)

#### Avijit Hazra, Nithya Gogtay. Biostatistics Series Module 7: The
#### Statistics of Diagnostic Tests. Indian J Dermatol. 2017 Jan-Feb; 62(1):
#### 18-24. doi: 10.4103/0019-5154.198047

diagTest(as.table(matrix(c(2, 18, 1, 182), ncol = 2, byrow = TRUE)))
diagTest(as.table(matrix(c(2, 18, 1, 182), ncol = 2, byrow = TRUE)),
         p = .10)
diagTest(as.table(matrix(c(30, 35, 23, 12), ncol = 2, byrow = TRUE)))
diagTest(as.table(matrix(c(30, 35, 23, 12), ncol = 2, byrow = TRUE)),
         p = .10)

#### Just an example to demonstrate
# diagTest(infert$case, infert$spontaneous)

# }