Health Services Research & Development

Veterans Crisis Line Badge
Go to the ORD website
Go to the QUERI website

2006 HSR&D National Meeting Abstract


3011 — Obesity and VHA Cost of Treating Veterans with Diabetes

Author List:
Shen Y (Center for Healthcare Knowledge Management)
Sambamoorthi U (Center for Healthcare Knowledge Management)
Rajan M (Center for Healthcare Knowledge Management)
Miller D (CHQOER)
Banerjea R (Center for Healthcare Knowledge Management)
Pogach L (Center for Healthcare Knowledge Management)

Objectives:
Research shows that excess medical cost associated with obesity is largely attributable to the increased prevalence of diabetes. However, little is known about the association of obesity upon medical care costs among patients with established diabetes. We examined the association among VHA patients with diabetes.

Methods:
We identified survey respondents of the 1999 Large Veterans Health Survey with diabetes by linking the survey with the 1999 VHA Diabetes Epidemiology Cohort database (N=161,398). We obtained FY1999 VHA costs from the HERC Average Cost database. We assessed the inpatient cost, outpatient cost, and their summation as total cost. In addition, we examined the diabetes-related cost which was the sum of inpatient and outpatient costs of treating microvascular, macrovascular metabolic conditions. We used primary diagnosis codes to identify these conditions. Based on the self-report on weight and height, we calculated BMI (body-mass-index) and grouped diabetes patients into four exclusive categories: underweight (<18.5), normal (18.5-25), overweight (25-29), and obese (>=30). We used multivariate analysis to examine how BMI affected the total cost as well as other cost components, controlling for other personal characteristics including diabetes duration, health status, self-reported physical and mental co-morbidities, smoking behavior, etc. To deal with outliers and skewness of the cost items, we took the log-transformation of the costs as dependent variables.

Results:
A majority of the population was either overweight (41%) or obese (37%). About 6.3% were under weight, 16.4% were normal weight. Underweight patients had the highest total cost, followed by obese patients. The total cost was similar between normal weight and overweight groups. For both outpatient and diabetes-related costs, obese patients cost the most; normal weight patients cost the least; while the other two groups had similar costs. However, we found a negative association between obesity and inpatient cost.

Implications:
The association between obesity and costs was complex and varied by type of service and complications.

Impacts:
A clearer understanding of resource needs for obese patients with established diabetes will help to improve treatment for this specific cohort. Further research is needed to explore the impact of obesity on inpatient cost.