Developing Composite Quality Measures for EHR-Enabled Primary Care Practices in New York City.

We describe the process of developing composite quality measures (CQMs). During the initial consultative process, we grouped quality measures based on the associated clinical workflow and difficulty, and then confirmed the groupings with factor analysis. The CQMs are estimated as the mean of the measures for each group. We used analysis of variance followed by a post hoc analysis to assess: (1) performance among the different CQMs each year; and (2) the performance trend for each of the composite measures from 2009 to 2011. The four CQMs were Control-BP, Control-Other, Assessment, and Screening. Performance was highest at baseline for Control-BP (58%, SD 15.07), followed by Control-Other (48.04%, SD 22.75), Screening (46.49%, SD 20.21), and Assessment (42.15%, SD 19.08). Performance on the CQMs increased significantly with time, whereas the gap between the CQMs decreased significantly over time. The CQMs reflect the clinical care domains, and practice performance is influenced by electronic health record functionality, clinician workflow, and clinical difficulty.