Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/672808v1?rss=1

Authors: Wu, Z., Chen, I.

Abstract:

SO_SCPLOWUMMARYC_SCPLOWOptimal prevention and treatment strategies for a disease of multiple causes must be informed by the population distribution of causes among cases, or cause-specific case fraction (CSCFs) which may further depend on explanatory variables. However, the true causes are often not observed, motivating the use of non-gold-standard diagnostic tests that provide indirect etiologic evidence. Based on case-control multivariate binary data, this paper proposes a novel and unified modeling framework for estimating CSCF functions, closing the existing methodological gap in disease etiology research. With a novel shrinkage prior to encourage parsimonious approximation to a multivariate binary distribution given covariates, the model leverages critical control data for valid probabilistic cause assignment for cases. We derive an efficient Markov chain Monte Carlo algorithm for flexible posterior inference. We illustrate the inference of CSCF functions using extensive simulations and show that the proposed model produces less biased estimates and more valid inference of the overall CSCFs than an analysis that omits covariates. A regression analysis of pediatric pneumonia data reveals the dependence of CSCFs upon season, age, HIV status and disease severity.

Copy rights belong to original authors. Visit the link for more info