The researchers used data from clinical systems to identify patients who received most of their primary care from a specific doctor or from different doctors in a specific practice. Other patients were considered not connected to either a physician or a practice; usually, these patients were transitioning out of the health care network. The researchers then looked for a relationship between “connectedness” and several measures of quality of care: cancer screening in eligible patients, diabetes monitoring and control in patients with diabetes, and cholesterol monitoring and control in patients with diabetes and heart disease.