A SAS Program Combining R Functionalities to Implement Pattern-Mixture Models

Pattern-mixture models have gained considerable interest in recent years. Pattern mixture modeling allows the analysis of incomplete longitudinal outcomes under a variety of missingness mechanisms. In this manuscript, we describe a SAS program which combines R functionalities to fit pattern-mixture models, considering the cases that missingness mechanisms are at random and not at random. Patterns are defined based on missingness at every time point and parameter estimation is based on a full group-by time interaction. The program implements a multiple imputation method under so-called identifying restrictions. The code is illustrated using data from a placebo-controlled clinical trial. This manuscript and the program are directed to SAS users with minimal knowledge of the R language.

Publication year:2015

Keywords:MAR, MNAR, pattern-mixture model, identifying restriction, multiple imputation

BOF-keylabel:yes

IOF-keylabel:yes

BOF-publication weight:10

CSS-citation score:1

Authors:International

Authors from:Higher Education

Accessibility:Open