Combining information from multiple data sources to create multivariable risk models: Illustration and preliminary assessment of a new method

ASU Author/Contributor (non-ASU co-authors, if there are any, appear on document)
Martin Root Ph.D, Associate Professor (Creator)
Appalachian State University (ASU )
Web Site:

Abstract: A common practice of metanalysis is combining the results of numerous studies onthe effects of a risk factor on a disease outcome. If several of these composite relativerisks are estimated from the medical literature for a specific disease, they cannot becombined in a multivariate risk model, as is often done in individual studies, becausemethods are not available to overcome the issues of risk factor colinearity andheterogeneity of the different cohorts. We propose a solution to these problems forgeneral linear regression of continuous outcomes using a simple example ofcombining two independent variables from two sources in estimating a joint outcome.We demonstrate that when explicitly modifying the underlying data characteristics(correlation coefficients, standard deviations, and univariate betas) over a wide range,the predicted outcomes remain reasonable estimates of empirically derived outcomes(gold standard). This method shows the most promise in situations where the primaryinterest is in generating predicted values as when identifying a high-risk group ofindividuals. The resulting partial regression coefficients are less robust than thepredicted values.

Additional Information

Greg Samsa, Guizhou Hu, and Martin Root(2005) Combining Information From Multiple Data Sources to Create Multivariable Risk Models: Illustration and Preliminary Assessment of a New Method. Journal of Biomedicine and Biotechnology
Language: English
Date: 2005
, multivariable, preliminary, assessment, new, method,

Email this document to