Regular outpatient follow-up is important for all diabetes patients, with some needing frequent attention because their health is unstable, their treatment regimen is complex, or their social supports are inadequate. However, many patients live with access barriers that limit their use of outpatient services, fail to attend outpatient appointments, and experience worse outcomes than trials of aggressive management suggest is possible. Although labor-intensive, telephone care programs are one potential strategy for bringing diabetes management services into patients� homes and improving their glycemic control. Automated telephone disease management (ATDM) systems can augment telephone care by providing frequent monitoring and health education to large patient panels while focusing clinicians� attention on individuals who need it most. Although this technology has shown some promise, it has not been rigorously evaluated, particularly in VA.

Background:

Regular outpatient follow-up is important for all diabetes patients, with some needing frequent attention because their health is unstable, their treatment regimen is complex, or their social supports are inadequate. However, many patients live with access barriers that limit their use of outpatient services, fail to attend outpatient appointments, and experience worse outcomes than trials of aggressive management suggest is possible. Although labor-intensive, telephone care programs are one potential strategy for bringing diabetes management services into patients� homes and improving their glycemic control. Automated telephone disease management (ATDM) systems can augment telephone care by providing frequent monitoring and health education to large patient panels while focusing clinicians� attention on individuals who need it most. Although this technology has shown some promise, it has not been rigorously evaluated, particularly in VA.

Objectives:

This study evaluated Automated Telephone Disease Management (ATDM) calls with telephone nurse follow-up as a means of improving the quality of VA diabetes care. Specifically, we will determine whether this service improves patients' glucose control; improves other important outcomes such as their quality of life, satisfaction with care, and health service use; improves health behaviors such as self-monitoring of blood glucose, fat intake, and medication adherence; and has effects that vary across patient subgroups.

Methods:

Patients with diabetes mellitus using hypoglycemic medication were enrolled during outpatient visits to a university-affiliated VA health care system and randomized to usual care or bi-weekly ATDM assessment and self-care education calls with follow-up by a nurse educator. The intervention process was evaluated by examining patients� patterns of ATDM use and the reliability and validity of information they provided. Telephone surveys were used to measure intervention effects at 12-months on patients� self-care, symptoms, satisfaction with care, and perceived access barriers. The impact on VA utilization was evaluated using electronic utilization databases, and glycemic control was measured using laboratory tests. A total of 292 patients were randomized and 272 (93%) provided data at 12-months. Intervention patients completed ATDM assessments consistently throughout the observation period and the assessments identified groups of intervention patients with varying degrees of health risk at baseline. Compared to control patients, intervention patients at 12-months reported more frequent glucose self-monitoring, fewer access problems, and greater satisfaction with care (all p = 0.05). Intervention patients were more likely than controls to have been seen in podiatry clinics (53% versus 31%, p = 0.003) and diabetes specialty clinics (31% versus 17%, p = 0.03) during the study. The intervention did not influence mean endpoint HgA1c levels overall. However, among patients with baseline HgA1c = 8%, mean endpoint values among intervention and control patients were 8.7% and 9.2%, respectively (p = 0.05); intervention effects were even greater among patients with baseline HgA1c = 9%. Moreover, intervention patients at follow-up reported fewer symptoms of poor glycemic control than patients receiving usual care (3.6 versus 4.4, p = 0.03).

Status:

Completed.

Inclusion Criteria:

VA patients with diabetes taking hypoglycemic medications. Patients with serious mental disorders, no touch tone telephone, or a life expectancy of < 1 year were excluded.

Exclusion Criteria: