Misclassification occurring in either outcome variables or categorical covariates or both is a common issue in medical science. It leads to biased results and distorted disease–exposure relationships. Moreover, it is often of clinical interest to obtain the estimates of sensitivity and specificity of some diagnostic methods even when neither gold standard nor prior knowledge about the parameters exists. We present a novel Bayesian approach in binomial regression when both the outcome variable and one binary covariate are subject to misclassification. Extensive simulation results under various scenarios and a real clinical example are given to illustrate the proposed approach. This approach is motivated and applied to a dataset from the Baylor Alzheimer's Disease and Memory Disorders Center.
CopelandKCheckowayHMcMichaelA. Bias due to misclassification in the estimation of relative risk. Am J Epidemiol1977; 105: 488–495.
3.
NeuhausJ. Bias and efficiency loss due to misclassified responses in binary regression. Biometrika1999; 86: 843–855.
4.
JohnsonWGastwirthJ. Bayesian inference for medical screening tests: approximations useful for the analysis of acquired immune deficiency syndrome. J R Stat Soc Ser B: Stat Methodol1991; 53: 427–439.
5.
TuXKowalskiJJiaG. Bayesian analysis of prevalence with covariates using simulation-based techniques: applications to HIV Screening. Stat Med1999; 18: 3059–3073.
6.
McInturffPJohnsonWCowlingD. Modelling risk when binary outcomes are subject to error. Stat Med2004; 23: 1095–1109.
7.
BarronB. The effects of misclassification on the estimation of relative risk. Biometrics1977; 33: 414–418.
8.
GreenlandS. Variance estimation for epidemiologic effect estimates under misclassification. Stat Med1988; 7: 745–757.
9.
MarshallR. Validation study methods for estimating exposure proportions and odds ratios with misclassified data. J Clin Epidemiol1990; 43: 941–947.
10.
MorrisseyMSpiegelmanD. Matrix methods for estimating odds ratios with misclassified exposure data: extensions and comparisons. Biometrics1999; 55: 338–344.
11.
SpiegelmanDRosnerBLoganR. Estimation and inference for logistic regression with covariate misclassification and measurement error in main study/validation study designs. J Am Stat Assoc2000; 95: 51–61.
12.
RichardsonSGilksW. A Bayesian approach to measurement error problems in epidemiology using conditional independence models. Am J Epidemiol1993; 138: 430–442.
13.
RichardsonSGilksW. Conditional independence models for epidemiological studies with covariate measurement error. Stat Med1993; 12: 1703–1722.
14.
DetryM. A logistic regression model with misclassification in the outcome variable and categorical covariate, Houston, TX: University of Texas School of Public Health, 2003.
15.
GreenlandS. The effect of misclassification in matched-pair case-control studies. Am J Epidemiol1982; 116: 402–406.
16.
GreenlandSEricsonCKleinbaumD. Correcting for misclassification in two-way tables and matched-pair studies. Int J Epidemiol1983; 12: 93–97.
17.
MagderLHughesJ. Logistic regression when the outcome is measured with uncertainty. Am J Epidemiol1997; 146: 195–203.
18.
RobinsJRotnitzkyAZhaoL. Estimation of regression coefficients when some regressors are not always observed. J Am Stat Assoc1994; 89: 846–866.
19.
PrescottGGarthwaiteP. A Bayesian approach to prospective binary outcome studies with misclassification in a binary risk factor. Stat Med2005; 24: 3463–3477.
20.
CarrollRRuppertDStefanskiL. Measurement error in nonlinear models: a modern perspective, Boca Raton, FL: Chapman and Hall/CRC, 2006.
21.
ShepsSSchechterM. The assessment of diagnostic tests. J Am Med Assoc1984; 17: 2418–2422.
22.
WacholderSArmstrongBHartgeP. Validation studies using an alloyed gold standard. Am J Epidemiol1993; 137: 1251–1258.
23.
AlbertPDoddL. On estimating diagnostic accuracy from studies with multiple raters and partial gold standard evaluation. J Am Stat Assoc2008; 103: 61–73.
24.
KosinskiAFlandersW. Evaluating the exposure and disease relationship with adjustment for different types of exposure misclassification: a regression approach. Stat Med1999; 18: 2795–2808.
25.
Mendoza-BlancoJTuXIyengarS. Bayesian inference on prevalence using a missing-data approach with simulation-based techniques: applications to HIV screening. Stat Med1996; 15: 2161–2176.
26.
RenDStoneR. A Bayesian adjustment for covariate misclassification with correlated binary outcome data. J Appl Stat2007; 34: 1019–1034.
27.
LunnDThomasABestN. WinBUGS-a Bayesian modelling framework: concepts, structure, and extensibility. Stat Comput2000; 10: 325–337.
28.
HuiSZhouX. Evaluation of diagnostic tests without gold standards. Stat Meth Med Res1998; 7: 354–370.
29.
RindskopfDRindskopfW. The value of latent class analysis in medical diagnosis. Stat Med1986; 5: 21–27.
30.
AlvordWDrummondJArthurL. A method for predicting individual HIV infection status in the absence of clinical information. AIDS Res Hum Retroviruses1988; 4: 295–304.
31.
FormannA. Measurement errors in caries diagnosis: some further latent class models. Biometrics1994; 50: 865–871.
32.
QuYTanMKutnerM. Random effects models in latent class analysis for evaluating accuracy of diagnostic tests. Biometrics1996; 52: 797–810.
33.
BlackMCraigB. Estimating disease prevalence in the absence of a gold standard. Stat Med2002; 21: 2653–2669.
34.
FujisawaHIzumiS. Inference about misclassification probabilities from repeated binary responses. Biometrics2000; 56(3): 706–711.
35.
GelmanACarlinJSternH. Bayesian data analysis, Boca Raton, FL: CRC Press, 2004.
36.
SpiegelhalterDBestNCarlinB. Bayesian measures of model complexity and fit. J R Stat Soc Ser B: Stat Methodol2002; 64(4): 583–639.
37.
CarlinBLouisT. Bayesian methods for data analysis, Boca Raton, FL: Chapman & Hall/CRC, 2009.
38.
KassRRafteryA. Bayes factors. J Am Stat Assoc1995; 90: 773–795.
39.
LewisSRafteryA. Estimating Bayes factors via posterior simulation with the Laplace-Metropolis estimator. J Am Stat Assoc1997; 92: 648–655.
40.
VacekP. The effect of conditional dependence on the evaluation of diagnostic tests. Biometrics1985; 41: 959–968.
41.
McKhannGDrachmanDFolsteinM. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's disease. Neurology1984; 34: 939–944.
42.
DoodyRPavlikVMassmanP. Changing patient characteristics and survival experience in an Alzheimer's center patient cohort. Dement Geriatr Cogn Disord2005; 20: 198–208.
43.
RountreeSChanWPavlikV. Persistent treatment with cholinesterase inhibitors and/or memantine slows clinical progression of Alzheimer disease. Alzheimers Res Ther2009; 7: 1–7.
44.
FolsteinMFolsteinSMcHughP. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res1975; 12: 189–198.
45.
RosenWMohsRDavisK. A new rating scale for Alzheimer's disease. Am J Psychiatry1984; 141: 1356–1364.
46.
HughesCBergLDanzigerW. A new clinical scale for the staging of dementia. Br J Psychiatry1982; 140: 566–572.
47.
MorrisJ. The clinical dementia rating (CDR): current version and scoring rules. Neurology1993; 43: 2412–2414.
48.
PerneczkyRWagenpfeilSKomossaK. Mapping scores onto stages: mini-mental state examination and clinical dementia rating. Am J Geriatr Psychiatry2006; 14: 139–144.
49.
O'BryantSWaringSCullumC. Staging dementia using clinical dementia rating scale sum of boxes scores – a Texas Alzheimer's research consortium study. Arch Neurol2008; 65: 1091–1095.
50.
DoodyRPavlikVMassmanP. Predicting progression of Alzheimer's disease. Alzheimers Res Ther2010; 2: 1–9.
51.
NagelkerkeNFidlerVBuwaldaM. Instrumental variables in the evaluation of diagnostic test procedures when the true disease state is unknown. Stat Med1988; 7: 739–744.
52.
PaulinoCSoaresPNeuhausJ. Binomial regression with misclassification. Biometrics2003; 59: 670–675.
53.
ChengDStameyJBranscumA. Bayesian approach to average power calculations for binary regression models with misclassified outcomes. Stat Med2009; 28: 848–863.
54.
ChengDBranscumAStameyJ. Accounting for response misclassification and covariate measurement error improves power and reduces bias in epidemiologic studies. Ann Epidemiol2010; 20: 562–567.
