Practice Transformation Support and Patient Engagement to Improve Cardiovascular Care

From EvidenceNOW Southwest (ENSW)

W. Perry Dickinson, MD; Donald E. Nease, Jr., MD; Robert L. Rhyne, MD; Kyle E. Knierim, MD; Douglas H. Fernald, MA; Dionisia R. de la Cerda, MPA; L. Miriam Dickinson, PhD


J Am Board Fam Med. 2020;33(5):675-686. 

In This Article


Practice transformation support using practice facilitation, clinical HIT support, and collaborative learning sessions resulted in primary care practices improving quality of care by implementing elements of the Building Blocks of High-Performing Primary Care and evidence-based interventions to reduce cardiovascular risk. Adding support for implementing patient and family advisory councils and other forms of patient engagement in quality improvement efforts increased practice implementation of various Building Blocks but was not associated with differential improvement on the ABCS CQMs. The overall intervention yielded improvements in important CVD risk reduction metrics, especially when compared with the external comparison group.

Patient engagement in practice improvement and redesign projects has been encouraged as a method for making care more patient centered. This project provides evidence that practices respond to support for patient engagement activities and that the resulting patient involvement can improve practice implementation of elements of high-performing primary care. Although there were not significant differences between the study groups in CQM improvements, the 9-month intervention provided a limited time for practices to implement a fully functional patient and family advisory council and then to see the impact of the council on outcomes. This is particularly true for CQMs, such as blood pressure control, that generally have a lag between implementation of changes in care and improvements in the measures, and this could be partially responsible for the lack of significant differences between the 2 groups on ABCS measure improvement. It also is possible that the patient and family advisory councils focused practice attention on improvements not directly tied to the ABCS CQMs. It should be noted that the study design did not exclude practices that had pre-existing patient and family advisory councils from the standard intervention group, and this may have attenuated differences in outcomes between the 2 groups.

Mediator analyses showed that increased implementation of the first 6 Building Blocks of High-Performing Primary Care served as mediators for improvements in blood pressure control and, to a lesser extent, aspirin use CQMs. This is particularly interesting in that the blood pressure control measure is the only 1 of the ABCS measures directly tied to clinical outcomes, with the other 3 measure tied to process of care. Process of care measures are often easier to change than clinical outcomes, in part because (1) improvements in data entry and capture can result in major improvements and (2) changes in process are more immediately reflected in the measure. This can been seen in our results, in which blood pressure did not improve as impressively as the process of care measures. However, for a clinical outcome such as blood pressure control that is difficult to change, the implementation of the Building Blocks might be particularly important in producing improvements, whereas the Building Blocks may be less important to produce changes in process of care measures. This will be interesting to explore in future projects.

This study adds to the growing evidence for the effectiveness of practice facilitation as a method for improving implementation of evidence-based interventions. ENSW results are consistent with other research demonstrating that practice facilitation can assist practices in improvements in cardiovascular care.[16] In addition, this study shows that practice facilitation can induce practices to take up important patient engagement activities. The intervention period for this project was relatively short for the level of practice transformation work being targeted. The study was further complicated by major changes in the cholesterol treatment guidelines just as the project was starting. Difficulties and delays with obtaining clean CQM data as reported elsewhere[57,58] made quality improvement efforts focused on specific measures problematic as well. Despite those barriers and limitations, practices improved all 4 of the ABCS measures significantly when compared with an external comparison group. Furthermore, they showed significant improvement in their implementation of many advanced primary care activities.

There are other potential limitations beyond the limited intervention and follow-up periods. Although a broad range of practice types across Colorado and New Mexico were engaged, they may not be representative of all primary care practices. Similarly, the external comparison group, drawn from a pool of practices involved in activities with research, safety, and quality activities, differed somewhat from ENSW practices and may not provide a representative comparison population. CQMs from comparison practices, which were obtained via data extraction from electronic health records, were substantially lower than ENSW practices for all 4 CQMs. The lower rates for external comparison practices may be due to lower ascertainment, lack of onsite HIT support for practices, and/or absence of focus specifically on ABCS measures. However, although some measures are quite low in comparison practices, the trajectories should be a reasonable reflection of change over time because all patient-level electronic health record data from the entire time frame (from which these measures were generated) were obtained and processed from a single data extraction in each comparison practice and processed using the same methods. Randomization of geographic regions (rather than practices) was necessary before recruiting practices to conduct regional boot camp translation activities, and although this was not ideal, it was necessary to conduct the study, as discussed elsewhere.[38] Because multiple PTOs provided the practice facilitation support, there could be underlying relationships between practice demographic characteristics and individual PTOs, so that differences seen for practice types could be due to PTO performance rather than the practice characteristics. However, results of sensitivity analyses adjusted for PTO were very similar.

With multiple ongoing and upcoming efforts underway to expand primary care capacity, investment in supporting practices through practice facilitation, clinical HIT support, and other forms of practice transformation support should be considered. Patient engagement in practice transformation efforts seems effective for improving key process of care outcomes, although further studies are needed.