In a previous study, the investigators delivered graphs to family physicians that outlined the proportion of patients with a history of diabetes or heart disease achieving evidence-based quality targets derived from guideline recommendations. A qualitative evaluation found that participating family physicians did not act upon the feedback for two main reasons. First, they felt that targets recommended in guidelines often did not apply for particular patients. Second, they complained that had difficulty using the feedback reports that only provided aggregate level data for clinical action. In this cluster-randomized trial, the investigators test two approaches to conducting audit and feedback that aims to address these issues. The investigators hypothesize that feedback identifying a small number of patients at high-risk for cardiovascular events requiring action will more effectively lead to changes in clinical behavior than feedback identifying all patients not reaching optimal care targets.

For the last two years, all physicians contributing data to the Electronic Medical Record Administrative data Linked Database (EMRALD) have received two feedback reports. The first focused on diabetes and the second on patients with heart disease. The reports provide aggregate information regarding the proportion of the family physician's patients meeting quality targets, but no patient-specific information. Currently the feedback is sent to physicians by courier from the EMRALD team every six months.

The intervention arms in this trial are as follows:

• Arm 1 - Standard, aggregate-level feedback reports focusing on the proportion of patients with hypertension and/or diabetes and/or ischemic heart disease meeting targets sent via courier every six months and available on a password protected website. Family physicians in this arm will also have access through this website to patient-level data to identify patients not achieving optimal quality of care targets.
• Arm 2 - Aggregate-level feedback reports focusing on the proportion of patients with hypertension and/or diabetes and/or ischemic heart disease meeting criteria for high-risk sent via courier every six months and available on a password protected website. Family physicians in this arm will also have access through this website to a list of chart numbers identifying those patients at highest risk.

Both arms will have the opportunity to receive continuing medical education credits by completing worksheets that prompt them to reflect upon the data. Family physicians in Arm 1 are asked to complete a worksheet that follows continuous quality improvement principles, including setting an aim statement, engagement with team members in the clinic, testing change concepts at first on a small scale, and then scaling up in a effort to spread best practices. Family physicians in Arm 2 are asked to complete a worksheet that also includes goal setting and action planning, but focuses on reducing the number of patients with high-risk criteria and offers some suggested practice-based approaches.

Inclusion Criteria:

• Family physicians belonging to and sharing data with the Electronic Medical Record Administrative Linked Database in Ontario
• Patients rostered to these family physicians with diabetes or hypertension or ischemic heart disease

Exclusion Criteria:

• Family physicians without at least two years of Electronic Medical Record data in EMRALD
• Family physicians without at least 100 rostered, active patients