Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

VA Health Systems Research

Go to the VA ORD website
Go to the QUERI website

IIR 05-261 – HSR Study

IIR 05-261
Longitudinal Analysis of Utilization and Cost by VA Patients with Diabetes
Yujing Shen, PhD
East Orange Campus of the VA New Jersey Health Care System, East Orange, NJ
East Orange, NJ
Funding Period: January 2008 - December 2011
Diabetes is a chronic disease that is the major cause of adult blindness, end stage renal disease, cardiovascular disease and amputations. Its direct medical costs in the US were estimated at $92 billion in 2002. However, no studies have evaluated utilization and costs for VHA patients with diabetes or pre-diabetes over time, or how health care and costs for VHA patients with diabetes are attributed to microvascular, macrovascular, metabolic, and all other medical conditions. No VA studies have assessed the non-VHA cost incurred by VHA patients with diabetes and cost sharing between VHA and Medicare.

Aim 1: Examine the longitudinal trends in utilization and its associated cost for VHA patients with diabetes, with pre-diabetes, and without diabetes, using combined VHA and Medicare data; assess the association between diabetes duration, obesity, mental health conditions and utilization/cost, controlling for other personal and system factors.
The primary hypotheses are:
H1-1: Utilization and cost (including pharmacy cost) associated with macrovascular conditions increases over the years preceding the diabetes diagnosis (pre-diabetes), compared to those without diabetes.
H1-2: Patients with obesity (BMI >=30) and diabetes experience higher utilization and cost than those with obesity but without diabetes; among patients with diabetes, morbid obesity (>35 BMI) significantly increases utilization and cost associated with macrovascular, microvascular and metabolic conditions.

Aim 2: Among VHA-Medicare dual enrollees with and without diabetes, evaluate the longitudinal trends in: 1) Medicare HMO/FFS enrollment; 2) VHA utilization/cost and Medicare Fee-for-Service (FFS) utilization/cost; and 3) reliance on VHA and Medicare.
The primary hypotheses are:
H2-1: The reliance on Medicare will be greater for patients with diabetes than those without diabetes.
H2-2: Incident "emergency inpatient admissions" in Medicare FFS are associated with increase in reliance on Medicare after controlling for personal and system level factors regardless of diabetes status.

This longitudinal study followed the study cohort (1999 Large Veterans Health Survey (LVHS) respondents with diabetes, pre-diabetes or non-diabetes) over the period of 1999-2005 by linking LVHS to VHA and Medicare administrative data. To accomplish the objectives of the study, we used multivariable and robust regression techniques to control for correlated data due to clustering and repeated observations.

1. We evaluated the association between BMI and expenditures among LVHS respondents with diabetes in fiscal year 1999, who were age 65 or older and enrolled in Medicare FFS (N = 79,934). VHA expenditures were from the Health Economics Resource Center average Cost database; Medicare expenditures were paid amounts recorded in the claims. BMI was grouped into: normal (18.5-24.9), overweight (25-29.9), obese (30-34.9), and morbidly obese (>=35). Generalized linear models (GLM) with the log-link function were used to examine the association between BMI categories and expenditures controlling for social-demographic factors, diabetes duration, health status, and health behavior. Almost half of the study sample was overweight (47.6%), followed by obese (22.6%), normal (20.7%), and morbidly obese (9.1%). Normal BMI patients had the highest average total expenditures ($10,470) followed by overweight ($7,526). Total expenditure was not significantly different between obese and morbidly obese BMI groups ($6,597 vs. $6,772). After controlling for all other variables, normal weight patients with diabetes had greater total, inpatient and outpatient expenditures.

2. We also evaluated whether the relationship between BMI categories and expenditures remains consistent over time, using longitudinal data on respondents of 1999 LVHS with newly diagnosed diabetes in 1999. GLM was fit to examine the association between 1999 BMI categories and annual expenditures over time controlling for factors similar to above. Overall, 12.8% of veterans with newly diagnosed diabetes were normal weight, 46% overweight, 25% obese and 13% morbid obese. The average expenditures were lowest ($7,006) for the obese group in 1999 and highest for morbidly obese ($12,052) in 2004. In both bivariate and multivariate analyses, the relationship between BMI categories and expenditures changed over time. For example, during the first year of incident diagnosis, compared to veterans with normal BMI, those with obesity had lower expenditures. However, in 2004, the relationship reversed: compared to veterans with normal BMI, those with obesity had greater expenditures. In a pooled model with year as one of the covariates, the interaction term between time and BMI categories was positive for all the categories.

3. We evaluated changes in utilization and reliance on VHA/Medicare of diabetes patients (with newly diagnosed diabetes in 1999 and with Medicare FFS coverage in FY2000-2004) in 4 years before and 4 years after the diagnosis. Reliance was measured by the percentages of using VHA only, private sector only, and both systems separately for hospitalization or outpatient visits. Utilization patterns were further examined by different categories of care: microvascular, macrovascular, metabolic related diabetes complications and non-diabetes related complications. There were 33,092 patients (97.8% male, 23% younger than 65). Overall, 32% of patients had hospitalization within the year of diagnosis compared to 23.7% per year preceding the diagnosis and 27.4% after diagnosis. Patients had 15.7 outpatient visits during the year of diagnosis, compared to 13 visits/year before and 14.3 visits/year after diagnosis. Non-diabetes related hospitalizations accounted for 75% to 80% of all hospitalizations while non-diabetes related visits arranged from 90% to 80%. Reliance on VHA for both inpatient and outpatient care was greatest during the year of diagnosis. However, there was an overall increase in patients using Medicare only for inpatient care and both systems for outpatient care after the diagnosis. Patients were more likely to rely on VHA if they had macrovascular/microvascular complications, and were more likely to be Medicare reliant for non-diabetes related complications.

Understanding the relationships between obesity and expenditures over time is critical to approaches that focus on pathways to reducing expenditures in veterans with chronic illnesses such as diabetes. Newly diagnosed diabetic veterans consumed more health care. Extra inpatient care triggered by diabetes diagnosis was most likely to be met by the private sector, while the extra outpatient care need was met by both VHA and the private sector. This fragmentation of care highlights the need for coordination of care across Federal systems in order to address continuity of care for this complex patient population, which presumably can result in improved quality and decreased costs.

External Links for this Project

Dimensions for VA

Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.

Learn more about Dimensions for VA.

VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address.
    Search Dimensions for this project


Journal Articles

  1. Shen Y, Sambamoorthi U, Rajan M, Miller D, Banerjea R, Pogach L. Obesity and expenditures among elderly Veterans Health Administration users with diabetes. Population health management. 2009 Oct 1; 12(5):255-64. [view]
Conference Presentations

  1. Shen Y, Sambamoorthi U, Soroka O, Rajan M, Pogach LM, Miller DR, Maney M. Medical Use for VHA-Medicare Dual Enrollees with New Onset Diabetes: Before and After Diagnosis. Poster session presented at: AcademyHealth Annual Research Meeting; 2012 Jun 25; Orlando, FL. [view]
  2. Shen Y, Sambamoorthi U, Soroka O, Rajan M, Pogach LM. Relationship between Body Mass Index Categories and Expenditures Over Time among Veteran Health Administration Users with Newly Diagnosed Diabetes. Paper presented at: AcademyHealth Annual Research Meeting; 2012 Jun 25; Orlando, FL. [view]

DRA: Health Systems, Diabetes and Other Endocrine Conditions
DRE: Epidemiology
Keywords: Diabetes, Utilization patterns
MeSH Terms: none

Questions about the HSR website? Email the Web Team

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.