2006 HSR&D National Meeting Abstract
3031 — The Effectiveness of a Care Coordination/Home-Telehealth Program for Veterans with Diabetes
Barnett TE (VA HSR&D/RR&D Rehabilitation Outcomes Research Center)
Chumbler NR (VA HSR&D/RR&D Rehabilitation Outcomes Research Center)
Vogel WB (VA HSR&D/RR&D Rehabilitation Outcomes Research Center)
Beyth RJ (VA HSR&D/RR&D Rehabilitation Outcomes Research Center)
Qin H (VA HSR&D/RR&D Rehabilitation Outcomes Research Center)
To examine the effectiveness of a VA care coordination home telehealth (CC/HT) program for veterans with diabetes.
We analyzed the differences in health care utilization (hospital admissions, days of hospitalization, emergency room visits, unscheduled primary care visits, podiatry visits, and ophthalmology visits) of 400 enrolled veterans in a CC/HT program (from 02/01/00 – 05/31/02) with diabetes and a matched comparison group of 400 veterans with diabetes who did not receive the CC/HT intervention during the same time period. We used baseline and 24-month post intervention utilization for both the treatment and comparison groups. Propensity scores were employed to improve the balance between the treatment and comparison cohorts. A difference-in-differences (DiD) method was employed in the multivariate models to assess the treatment effect for veterans in the CC/HT program.
There was a significant difference in hospital admissions 24 months after the program began (p < .05), with the treatment group decreasing in hospital admissions from 37.31% at baseline to 28.63% while the control group increased from 29.80% at baseline to 31.64% at follow-up. Over 24-months, both the treatment and comparison groups experienced significant decreases in emergency room visits (both all-cause and diabetes-related); however, the comparison group decreased at a faster rate than the treatment group (p<.001 all-cause, diabetes-related). There were no differences found for days of hospitalization, unscheduled primary care visits, podiatry, or ophthalmology visits.
The lower hospital admission rate for the treatment group after 24 months suggests program effectiveness for enabling the patient to better care for himself at home through education and monitoring. Future research should explore, in-depth, why the emergency department visits decreased at a higher rate in the comparison group than the treatment group.
We employed a sophisticated DiD design that avoided methodological shortcomings that have been representative of many previous disease management studies. This rigor strengthened the finding that the VA CC/HT program was effective in decreasing hospitalizations for veterans with diabetes. Following the intervention longer than one year is imperative to determine effectiveness of the program in controlling health care utilization over time.