Bayesian methods for medical test accuracy are presented, beginning with the basic measures for tests with binary scores: true positive fraction, false positive fraction, positive predictive values, and negative predictive value. The Bayesian approach is taken because of its efficient use of prior information, and the analysis is executed with a Bayesian software package WinBUGS？. The ROC (receiver operating characteristic) curve gives the intrinsic accuracy of medical tests that have ordinal or continuous scores, and the Bayesian approach is illustrated with many examples from cancer and other diseases. Medical tests include X-ray, mammography, ultrasound, computed tomography, magnetic resonance imaging, nuclear medicine and tests based on biomarkers, such as blood glucose values for diabetes. The presentation continues with more specialized methods suitable for measuring the accuracies of clinical studies that have verification bias, and medical tests without a gold standard. Lastly, the review is concluded with Bayesian methods for measuring the accuracy of the combination of two or more tests.
References
[1]
Broemeling, L.D. Bayesian Biostatistics and Diagnostic Medicine; Taylor & Francis: Boca Raton, FL, USA, 2007.
[2]
Pepe, M.S. The Statistical Evaluation of Medical Tests for Classification and Prediction; Oxford University of Press: Oxford, UK, 2003.
[3]
Zhou, X.H.; Obuchowski, N.A.; McClish, D.K. Statistical Methods in Diagnostic Medicine; John Wiley & Sons: New York, NY, USA, 2002.
[4]
Wiener, D.A.; Ryan, T.J.; McCabe, C.H.; Kennedy, J.W.; Schloss, M.; Tristani, F.; Chaitman, B.R.; Fisher, L.D. Correlations among history of angina, ST-segmented response and prevalence of coronary artery disease. N. Engl. J. Med. 1979, 301, 230.
[5]
Woodworth, G.G. Biostatistics, A Bayesian Introduction; John Wiley & Sons: Hoboken, NJ, USA, 2004.
[6]
O'Malley, J.A.; Zou, K.H.; Fielding, J.R.; Tampany, C.M.C. Bayesian regression methodology for estimating a receiver operating characteristic curve with two radiologic applications: Prostate biopsy and spiral CT of ureteral stones. Acad. Radiol. 2001, 8, 713–725.
[7]
Egan, J.P. Signal Detection Theory and ROC Analysis; Academic Press: New York, NY, USA, 1975.
[8]
Greenes, R.; Begg, C. Assessment of diagnostic technologies: Mmethodology for unbiased estimation from samples of selected verified ppatients. Invest. Radiol. 1985, 20, 751–756.
[9]
Bates, A.S.; Margolis, P.A.; Evans, A.T. Verification bias in pediatric studies evaluating diagnostic tests. J. Pediatr. 1993, 122, 585–590.
[10]
Philbrick, J.T.; Horwitz, R.I.; Feinstein, A.R. Methodologic problems of exercise testing for coronary artery disease. Am. J. Cardiol. 1980, 46, 807–812.
[11]
Reid, M.C.; Lachs, M.S.; Feinstein, A.R. Use of methodologic standards in diagnostic test research. Getting better but still not good. J. Am. Med. Assoc. 1995, 274, 645–651.
[12]
Drum, D.; Christacopoulos, J. Hepatic scintigraphy in clinical decision making. J. Nucl. Med. 1969, 13, 908–915.
[13]
Joseph, L.; Gyorkos, T.W.; Coupal, L. Bayesian estimation of disease prevalence and the parameters of diagnostic tests in the absence of a gold standard. Am. J. Epidemiol. 1995, 3, 263–272.
[14]
Dendukuri, N.; Joseph, L. Bayesian approaches to modeling the conditional dependence between multiple diagnostics tests. Biometrics 2001, 57, 158–167.
[15]
Buscombe, J.R.; Cwikla, J.B.; Holloway, B.; Hilson, A.J.W. Prediction of the usefulness of combined mammography and scintimammography in suspected primary breast cancer using ROC curves. J. Nucl. Med. 2001, 42, 3–8.
[16]
Berg, W.A.; Gutierrez, L.; NessAlver, M.S.; Carter, W.B.; Bhargavan, M.; Lewis, R.S.; Loffe, O.B. Diagnostic accuracy of mmammography, clinical examination, US and MR, imaging in preoperative assessment of breast cancer. Radiology 2004, 233, 830–849.
[17]
Van Overhagen, H.; Brakel, K.; Heijenbrok, M.W.; van Kasteren, J.H.L.M.; van de Moosdijk, C.N.F.; Roldaan, A.C.; van Gils, A.P.; Hansen, B.E. Metastases in supraclavicular lymph nodes in lung cancer: Assessment with palpation, US, and CT. Radiology 2004, 232, 75–80.
[18]
Pauleit, D.; Zimmerman, A.; Stoffels, G.; Bauer, D.; Risse, J.; Fluss, M.O.; Hamacher, K.; Coenene, H.H.; Langen, K.J. 18F-FET PET compared with 18F-FDG PET and CT in patients with head and neck cancer. J. Nucl. Med. 2006, 47, 256–261.
[19]
Schaffler, G.J.; Wolf, G.W.; Schoellnast, H.; Groell, R.; Maier, A.; Smolle-Juttner, F.M.; Woltsche, M.; Fasching, G.; Nicolletti, R.; Aigner, R.M. Non-small cell lung cancer: Evaluation of pleural abnormalities on CT scans with 18F-FET PET. Radiology 2004, 231, 858–865.
[20]
Gerber, B.L.; Coche, E.; Pasquet, A.; Ketelslegers, E.; Vancraeynest, D.; Grandin, C.; van Beers, B.E.; Vanocerschelde, J.L.J. Coronary artery stenosis: Direct comparison of four-section multi-detector row CT and 3D navigator MR imaging for detection—Initial results. Radiology 2005, 234, 98–108.
[21]
Gierada, M.; Pilgrim, T.K.; Ford, M. Lung cancer interobserver agreement on interpretation of pulmonary findings at low-dose CT screening. Radiology 2008, 246, 265–272.
[22]
Wieand, S.; Gail, M.H.; James, B.R. A family of nonparamtric statistics for comparing diagnostic markers with paired or unpaired data. Biometrics 1989, 76, 585.
[23]
Broemeling, L.D. Bayesian Methods for Measures of Agreement; Taylor & Francis Group: Boca Raton, FL, USA, 2009.
[24]
Erkanli, A.; Sung, M.; Costello, E.J.; Angold, A. Bayesian semi-parametric ROC analysis. Stat. Med. 2006, 25, 3905–3928.
[25]
Hanson, T.E.; Kottas, A.; Branscum, A.J. Modelling stochastic order in the analysis of receiver operating characteristic data. Appl. Stat. 2008, 57, 207–225.
[26]
Dendukuri, N.; Rahme, E.; Belisle, I.; Joseph, L. Bayesian sample size determination for prevalence and diagnostic test studies in the absence of a gold standard test. Biometrics 2004, 60, 378–387.
[27]
Cheng, D.; Branscum, A.J.; Stamey, J.E. A Bayesian approach to sample size determination for studies designed to evaluate continuous medical tests. Comp. Stat. Data Analy. 2010, 54, 298–307.
