Abstract

TOVAR CUEVAS, José Rafael. Bayesian inference and health research studies: an application case in clinical diagnosis. Revista médica Risaralda [online]. 2015, vol.21, n.1, pp.9-16. ISSN 0122-0667.

Introduction: the bayesian methods are a helpful and reliable alternative to analyze the data obtained in research Projects within the health sciences. One of the most common applications is given in the clinical diagnostic tests validation studies, in which the use of Bayes’ theorem to obtain predictive values of the test under evaluation is the natural choise. Material y methods: A method to estimate the sensitivity and specificity of magnetic resonance cholangiography as a screening test to classify individuals with choledocholithiasis was developed. The estimation process was developed using the Bayesian methodology. Results: The Bayesian estimates of the sensitivity and specificity had a greater degree of accuracy compared to the estimates obtained by the classical method of maximum likelihood. Conclusion: Bayesian methods are quite useful in situations such as discussed in this article that the sample sizes for the estimates can be considered very “small”. In those situations the results obtained using non Bayesian methods may be more inaccurate and unreliable.

Keywords : Bayesian inference; Clinical diagnosis; sensibility; specificity; Bayes theorem.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )