This study evaluated the impact of a virtual version of the Diabetes Prevention Program (DPP) on healthcare utilization and costs in a Medicare Advantage population. The program was offered during 2015 to a random sample of 9,497 individuals who had metabolic syndrome or prediabetes. Program enrollees (n=501) received a 12-month virtual DPP including a wireless scale, pedometer, nutrition tracker, educational lessons, health coaching, and peer group support through an online platform. Participants with available administrative claims data during the 12 months before and 24 months following program start (n=495, mean age=69 years; 64% female; 85% white) were propensity-score matched on demographic, behavioral, and clinical factors in a ratio of 1:1 to a comparison group (n=495, selected from 6,490; mean age=69 years; 58% female; 87% white) who did not receive the DPP. In the 24 months following DPP enrollment, participants averaged 0.2 inpatient admissions, 0.3 emergency department visits, and 12.2 physician visits, compared to 0.2, 0.4, and 12.0 for controls. While the utilization values do not show a large change, a difference-in-differences regression analysis of total medical and pharmacy cost for 24 months following program start showed cost savings. The adjusted difference-in-differences effect on average cumulative cost difference was $1,110 per participant, or $46.25/month. Medical and pharmacy costs were consistently lower for participants in the post-program year. Pharmacy savings were statistically significant during the last 6 months, with a 24-month cumulative adjusted savings of $408. These cost estimates do not include the cost of the program. Results suggest that a virtual DPP can change the pattern of utilization and reduce costs in a Medicare population.


V. Chiguluri: None. D. Barthold: None. R. Gumpina: None. C. Castro Sweet: Employee; Self; Omada Health, Inc.. J. Pieratt: None. T.A. Cordier: Employee; Self; Humana Inc.. Stock/Shareholder; Self; Humana Inc.. R. Matanich: None. A. Renda: None. T.G. Prewitt: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at