Preliminary Analyses and Data Summaries. All analyses will begin by examining numerical and graphical summaries of patient and practice characteristics, and outcomes (e.g., HTN control and BP at both the patient and practice levels). Smoothed bivariate plots will be used to examine unadjusted pairwise associations between variables and variability among stages of the trial and over time throughout the trial’s duration. Particular attention will be paid to assessing whether distributions appear to be approximately normal within each study group and whether there appear to be any secular trends, conspicuously nonlinear relationships, or outlying observations. We will use analysis of variance (ANOVA) (potentially with rank-transformation if non-parametric tests are needed) and distribution-appropriate bivariate tests for pairwise and overall differences between and among study groups in baseline characteristics. Preliminary analyses will be shared with the DSMB as required by the sponsor.
