Bayesian Bivariate Diagnostic Meta-analysis via R-INLA
Bivariate generalized mixed modeling is currently recommended for joint meta-analysis of diagnostic test sensitivity and specificity. Estimation is commonly performed using frequentist likelihood-based techniques assuming bivariate normally distributed, correlated logit transformations of sensitivity and specificity. These estimation techniques are fraught with non-convergence and invalid confidence intervals and correlation parameters especially with sparse data. Bayesian approaches, though likely to surmount these and other problems, have not previously been applied. Recently, integrated nested laplacian approximation has been developed (INLA)as a computationally fast, deterministic alternative to Markov chain Monte Carlo (MCMC)-based bayesian modeling and an R interface to the C-based inla program has been applied to diagnostic Meta-analysis. This presentation shows how to facilely interface R-INLA estimation with data pre- and post-processing within Stata. A user-written ado-file allows user-friendly application of INLA by Stata users.
Year of publication: |
2010-07-20
|
---|---|
Authors: | Dwamena, Ben Adarkwa ; Rue, Harvard |
Institutions: | Stata User Group |
Saved in:
Saved in favorites
Similar items by person
-
A midas retouch regarding diagnostic meta-analysis
Dwamena, Ben Adarkwa, (2014)
-
Meta-analytic depiction of ordered categorical diagnostic test accuracy in ROC space
Dwamena, Ben Adarkwa, (2009)
-
Nonstandard Deviation: Making the Global Local
Pagano, Marcello, (2014)
- More ...