Serviços Personalizados
Journal
Artigo
Indicadores
-
Citado por SciELO
-
Acessos
Links relacionados
-
Citado por Google
-
Similares em SciELO
-
Similares em Google
Compartilhar
Revista Colombiana de Estadística
versão impressa ISSN 0120-1751
Rev.Colomb.Estad. vol.35 no.3 Bogotá jul./dez. 2012
1Universidad del Rosario, Escuela de Medicina y Ciencias de la Salud, Centro de Investigaciones en Ciencias de la Salud (CISC), Bogotá, Colombia. Lecturer. Email: rtovar34@hotmail.com
2Universidade de São Paulo, Faculdade de Saúde, Departamento de Medicina Social FMRP, Riberão Preto, Brasil. Associate Professor. Email: achcar@fmrp.usp.br
In this paper, we introduce a Bayesian analysis to estimate the prevalence and performance test parameters of two diagnostic tests. We concentrated our interest in studies where the individuals with negative outcomes in both tests are not verified by a gold standard. Given that the screening tests are applied in the same individual we assume dependence between test results. Generally, to capture the possible existing dependence between test outcomes, it is assumed a binary covariance structure, but in this paper, as an alternative for this modeling, we consider the use of copula function structures. The posterior summaries of interest are obtained using standard MCMC (Markov Chain Monte Carlo) methods. We compare the results obtained with our approach with those obtained using binary covariance and assuming independence. We considerate two published medical data sets to illustrate the approach.
Key words: Bayes analysis, Copula, Dependence, Monte Carlo Simulation, Public health.
En este articulo introducimos un análisis Bayesiano para estimar la prevalencia y los parámetros de desempeño de pruebas para diagnóstico clínico, con datos obtenidos bajo estudios de tamizaje que incluyen el uso de dos pruebas diagnósticas en los cuales, los individuos con resultado negativo en las dos pruebas no son confirmados con una prueba patrón de oro. Dado que las pruebas de tamizaje son aplicadas al mismo indivíduo, nosotros asumimos dependencia entre los resultados de las pruebas. Generalmente, para capturar la posible dependencia existente entre los resultados de las pruebas diagnósticas, se asume una estrutura de covarianza binaria, pero en este artículo, nosotros consideramos el uso de estructuras que pueden ser modaladas usando funciones cópula, como una alternativa al modelamiento de la dependencia. Las estadísticas a posteriori de interés son obtenidas usando métodos MCMC. Los resultados obtenidos usando nuestra aproximación son comparados con los obtenidos usando modelos que asumen estructura binária y con los obtenidos usando modelos bajo el supuesto de independencia entre resultados de las pruebas para diagnóstico clínico. Para ilustrar la aplicación del método y para hacer las comparaciones se usaron los datos de dos estudios publicados en la literatura.
Palabras clave: análisis bayesiano, copula, dependencia, simulación Monte Carlo, salud pública.
Texto completo disponible en PDF
References
1. Ali, S., Moodambail, A., Hamrah, E., Bin-Nakhi, H. & Sadeq, S. (2007), 'Reliability of rapid dipstick test in detecting urinary tract infection in symptomatic children', Kuwait Medical Journal 39, 36-38. [ Links ]
2. Amblard, C. & Girard, S. (2002), 'Symmetry and dependence properties within a semiparametric family of bivariate copulas', Journal of Non-parametric Statistics 14, 715-727. [ Links ]
3. Amblard, C. & Girard, S. (2005), 'Estimation procedures for semiparametric family of bivariate copulas', Journal of Computational and Graphical Statistics 14, 363-377. [ Links ]
4. Amblard, C. & Girard, S. (2008), 'A new extension of bivariate FGM copulas', Metrika 70, 1-17. [ Links ]
5. Best, N., Cowles, M. & Vines, S. (1995), CODA: Convergence diagnosis and output analysis software for Gibbs sampling output, version 0.3;, MRC Biostatistics Unit, Cambridge, U.K. [ Links ].
6. Bohning, D. & Patilea, V. (2008), 'A capture-recapture approach for screening using two diagnostic tests with availability of disease status for the positives only', Journal of the American Statistical Association 103, 212-221. [ Links ]
7. Brenner, H. (1996), 'How independent are multiple diagnosis classifications?', Statistics in Medicine 15, 1377-1386. [ Links ]
8. Dendukuri, N. & Joseph, L. (2001), 'Bayesian approaches to modelling the conditional dependence between multiple diagnostic tests', Biometrics 57, 158-167. [ Links ]
9. Georgiadis, M., Johnson, W. & Gardner, I. (2003), 'Correlation adjusted estimation of sensitivity and specificity of two diagnostic tests', Journal of the Royal Statistical Society: Series C (Applied Statistics) 52, 63-76. [ Links ]
10. Gumbel, E. J. (1960), 'Bivariate exponential distributions', Journal of the American Statistical Association 55, 698-707. [ Links ]
11. Joseph, L., Gyorkos, T. W. & Coupal, L. (1995), 'Bayesian estimation of disease prevalence and the parameters of diagnostic tests in the absence of a gold standard', American Journal of Epidemiology 141, 263-272. [ Links ]
12. Martinez, E., Achcar, J. & Louzada, N. (2005), 'Bayesian estimation of diagnostic tests accuracy for semi-latent data with covariates', Journal of Biopharmaceutical Statistics 15, 809-821. [ Links ]
13. Nelsen, R. (1999), An Introduction to Copulas, Springer Verlag, New York. [ Links ]
14. Qu, Y. & Hadgu, A. (1998), 'A model for evaluating sensitivity and specificity for correlated diagnostic test in efficacy studies with an imperfect reference test', Journal of the American Statistical Association 93, 920-928. [ Links ]
15. Smith, D., Bullock, A. & Catalona, W. (1997), 'Racial differences in operating characteristics of prostate cancer screening tests', Journal of Urology 158, 1861-1865. [ Links ]
16. Spiegelhalter, D., Thomas, A., Best, N. & Lunn, D. (2003), Winbugs User Manual version 1.4, MRC Biostatistics Unit, Cambridge, U.K. [ Links ]
17. Thibodeau, L. (1981), 'Evaluating diagnostic tests', Biometrics 37, 801-804. [ Links ]
18. Torrance-Rynard, V. & Walter, S. (1997), 'Effects of dependent errors in the assessment of diagnostic tests performance', Statistics in Medicine 16, 2157-2175. [ Links ]
19. Tovar, J. R. (2012), 'Eliciting beta prior distributions for binomial sampling', Revista Brasileira de Biometría 30, 159-172. [ Links ]
20. Vacek, P. (1985), 'The effect of conditional dependence on the evaluation of diagnostic tests', Biometrics 41, 959-968. [ Links ]
21. Walter, S. & Irwig, L. (1988), 'Estimation of test error rates disease prevalence and relative risk from misclassified data: a review', Journal of Clinical Epidemiology 41, 923-937. [ Links ]
Este artículo se puede citar en LaTeX utilizando la siguiente referencia bibliográfica de BibTeX:
@ARTICLE{RCEv35n3a01,
AUTHOR = {Tovar, José Rafael and Achcar, Jorge Alberto},
TITLE = {{Two Dependent Diagnostic Tests: Use of Copula Functions in the Estimation of the Prevalence and Performance Test Parameters}},
JOURNAL = {Revista Colombiana de Estadística},
YEAR = {2012},
volume = {35},
number = {3},
pages = {331-347}
}