Using physician-led automated communications to improve patient health.

One common thread to health care reform in the United States is an emphasis on effectively managing the care of patients with chronic conditions. A landmark study by McGlynn et al demonstrated that patients receive about 55% of the treatment they need. While technological advances allow automated means for identifying and reaching out to patients in need of treatment, few studies have evaluated their impact. The purpose of this study is to measure how an automated outreach program can be used to improve the quality of care for patients with diabetes and hypertension. Billing and electronic medical records data from a large health system in Wisconsin were studied, identifying patients with a history of diabetes and hypertension but no visits recorded in billing data related to their condition in the past 6 months. The outcomes of interest were the occurrence of a chronic care-related visit and a necessary test within 6 months of the nonadherence date. Diabetes patients who were successfully contacted were significantly more likely to have both a chronic care-related visit and an HbA1c test (odds ratio [OR] = 4.61, 95% confidence interval [CI] 3.87-5.49) than their counterparts who were not contacted. As well, hypertension patients were significantly more likely to have both a chronic care-related visit and a systolic blood pressure reading recorded in an electronic medical record (OR = 3.18, 95% CI 2.90-3.48). An automated patient identification and outreach program can be an effective means to supplement existing practice patterns to ensure that patients with chronic conditions in need of care receive the necessary treatment.