Supporting Practices to Adopt Registry-Based Care

This project is designed to test the effectiveness and cost-effectiveness of a multi-faceted and largely self-directed organizational change intervention for implementation and use of diabetes registry in primary care practice. Participating practices will use a diabetes care registry to target evidence-based processes of care with the goal of improving patient attainment of recommended outcomes of care targets for HbA1c, blood pressure control and lipid control. Following recruitment and randomization of 30 primary care practices into intervention and control arms, clinician champions will be identified by practice leaders and attend a series of three meetings over the course of the intervention. At these meetings, champions will receive brief instruction in population health concepts, effective team leadership, organizational self-assessment, work process redesign and use of a computerized registry system. In addition, champions will share successes and challenges faced in the implementation process to encourage sharing successful techniques. At each participating practice, the clinician champion will be responsible for the intervention and will lead a process of organizational self-assessment and work process redesign to support implementation and use of a diabetes care registry. During the eleven-month implementation and initial registry use period an outside Peer Mentor will make quarterly visits to each practice and schedule monthly phone consultations with local clinician champions. During their visits, the Peer Mentors will meet with practice champions and advise them on methods for overcoming implementation barriers related to work process redesign. Peer Mentors will consult with a work process redesign expert as needed. For technical or computer-based barriers the Peer Mentor will coordinate access to a technical support staff member of the project.

Practice champion meetings The intervention will begin with a half-day training meeting for representatives from each of the 15 intervention practices who will then serve as "champions" for the intervention in their organizations. Prior to this stage, practice champions will be selected by practice leaders from among the clinicians in their practice using the following criteria: the clinician 1) has their principal employment in the practice, 2) is authorized to lead a practice improvement process and 3) regularly provides care to patients with diabetes. Practices may send more than one practice member to the training session (for example, a member of the nursing or clinical support staff could also attend the meeting). However, one clinician champion will be identified for each practice to ensure the maintenance of the relatively low cost of the intervention. In addition, practice managers will be provided with a detailed description of the intervention so that they can assist the champion in managing needed work process changes. Our experience in the pilot study suggests that a half-day session delivered in the evening is acceptable to primary care clinicians and is sufficient for delivery of the proposed training content. This meeting will be used to introduce practice champions to 1) the objectives of the project (including the focus on using the registry to target evidence-based processes of care improvements59 and cost-effective preventive care60), 2) population health concepts and how these relate to diabetes care in primary care practices, 3) principles of effective team leadership, 4) principles of patient registry construction and use, 5) an introduction to a free stand-alone registry software solution, 6) an organizational self-assessment process that focuses on planning work process changes related to using a registry in their practice, and 7) instruction in the use of study materials to coordinate changes in practice work processes with practice managers. Follow up meetings will be held at three and six months after the initial meeting of practice champions to assess progress with registry implementation, reinforce the learning from the initial meeting, and to share successes and challenges faced in the implementation process. The goal is that champions foster cross-practice learning by sharing successful techniques with each other. Peer Mentors will also attend these meetings.

Practice self-assessment and work process redesign Following the initial meeting of practice champions, practices will initiate a process of organizational self-assessment and work process redesign to ensure effective implementation and use of a diabetes registry. Clinician champions will use a self-assessment checklist designed to ensure that as part of this process champions focus on learning 1) current practice systems and procedures relating to diabetes care and the management and control of diabetes, 2) how these systems and procedures may be affected by registry use, and 3) which work processes will need to be changed or adapted to accommodate new registry-related work (including data entry, patient follow up, and visit preparation). Champions will then use these tools to focus on redesigning work processes to support these new tasks. These changes will necessarily be specific to each participating practice and will be tailored to the skills of the available staff members.

Peer Mentor activities The Peer Mentors will focus on assisting practice champions with addressing implementation and use hurdles specific to their practice. This assistance will be limited to an advisory role and will focus on reinforcing the need for practice champions to anticipate and plan for work process changes, encourage the development and use of effective team leadership techniques, and assist with addressing technical issues with regard to installation and use of the registry program. Peer Mentors will visit participating practices on a quarterly basis and have monthly phone conferences with practice champions throughout the 12-month intervention period.

Delayed intervention Based on initial findings from the evaluation of the intervention as delivered at baseline, we will adjust the intervention to focus on those activities from the initial intervention found to be most helpful for successful implementation and use of a diabetes registry. This modified intervention will be delivered, using a crossover design, to the initial control practices. Since the project timeline will not allow for long-term follow up of these practices we will compare the effectiveness of the initial and delayed interventions by comparing effects on patient outcomes observed in the first post-intervention assessment of initial intervention practices to effects on patient outcomes collected in the final observation period in delayed intervention practices.

Medical record review To assess the effect of the intervention on the quality of diabetes care delivered in participating practices, we will audit the medical records of 100 randomly selected patients with diabetes in each practice. The quality of diabetes care will be assessed for 100 patients from each practice at baseline, in year 3 following the initial intervention and then again in year 5 following the delivery of the delayed intervention to control practices for a total of 3000 medical records at each assessment period.