Continuous Versus Episodic Diabetes Disease Management in an Underserved Population.

In public healthcare systems uninsured patients with diabetes are increasingly common. In order to improve outcomes, the Department of Health Services in Los Angeles County initiated a pilot diabetes disease management program in East Los Angeles, in an area where 90% of adults live below 185% of the federal poverty level. To control costs in the diabetes program patients are graduated back to their PCP after 6 months or when the A1C is <8%, whichever comes first. Patients am asked to return yearly for an annual assessment. This prospective study was designed to follow patients enrolled in the program and randomized to continuous follow-up or discharge at 6 months. Eighty-one patients were randomized to continuous (CONT) and 82 to episodic (EPI) follow-up. All patients were Latino, the majority from Mexico. Spanish was the most most common primary language (81% in CONT/88% in EPI). Seven years was the average time spent in school, and ∼25% had no formal education. The average age was 55 yrs, and duration of diabetes 11 yrs. Half were on SA's, 75% were on metformin, 23% on pioglitazone and 33% on insulin (patients were on one or more agent). The initial A1C was 8.9(2.0)(mean)(stdev)and was not different between groups. At 6 months, when both groups were receiving the same care, there was no difference in A1C's [7.9(1.8)CONT vs 7.7(1.6)EPI]. At one and two years the A1C's were also not significantly different in the two groups. At 6 months 65%(CONT) and 71%(EPI) patients acheived an A1C of <8% (the target for referral back to primary care). Approximately 60% of patients in both groups had A1C levels <8% at 12 and 24 months. Many fewer reached an A1C of <7% (CONT = 32% at 6 mos, 28% at 12 mos and 16% at 24 mos; EPI = 30% at 6 mos, 38% at 12 mos and 26% at 24 mos, p = NS). Therefore, improvements in A1C levels are possible in an underserved Latino population. Patients did as well when seen episodically as when seen continuously in the program. When analyzed based on A1C level achieved, rather than treatment group, those who reached an A1C of <7% had fewer ER/walk-in visits than did those with higher A1C levels (p=0.0319). Thus, increasingly cost-effective models for diabetes care can be developed, maximally utilizing limited resources and reducing costly urgent care visits. [ABSTRACT FROM AUTHOR]