3068 — Obesity Care Practices in the Veterans Health Administration

Noel PH (VERDICT/South Texas Veterans Health Care System), Copeland LA (VERDICT), Pugh MJ (VERDICT), DeLeon CA (VERDICT), Lee S (South Texas Veterans Health Care System), Kahwati LC (VHA National Center for Health Promotion and Disease Prevention), Tsevat J (Institute for the Study of Health, U. of Cincinnati ), Nelson KM (VA Puget Sound Health Care System), Arterburn D (Group Health Center for Health Studies ), Dundon M (VA Western New York Healthcare System at Buffalo)

Objectives:
To identify obesity care received by obese primary care (PC) patients served by the VHA and to assess factors predicting receipt of care.

Methods:
We selected veterans in PC in FY2002 who had heights and weights in the Corporate Data Warehouse FY2002-FY2006 from six geographically diverse VISNs (representing early or late adopters of the VHA’s MOVE! program for treating obesity). We defined obesity (BMI > 30) using modal height across study years and the median of median weights recorded each quarter for FY2002. We then determined the percentage of surviving obese patients seen in PC through FY2006 with an obesity diagnosis in the electronic medical record (EMR), and/or any of the following obesity treatments: 1) individual or group instruction on nutrition, exercise, or weight management; 2) anti-obesity medications; and 3) bariatric surgery. Logistic regression identified patient-level predictors of receipt of obesity care and included a VISN-level indicator of MOVE! program adoption.

Results:
In FY2002, 330,802 (35.5%) of 933,084 PC patients met criteria for obesity. Among the 264,671 obese survivors seen in PC through FY2006, only 141,536 (53.5%) received an obesity diagnosis; 90,240 (34.1%) received individual or group instruction on nutrition, exercise, or weight management; 1059 (0.4%) received anti-obesity medications; and 407 (0.2%) underwent bariatric surgery. Diagnosis of obesity was the strongest predictor of receiving any type of obesity care (OR = 4.183; 95% CI 4.10-4.26). Compared to Caucasians, African-Americans were more likely to receive obesity care (OR = 1.39). VISN-adoption of MOVE! (OR = 1.19), Charlson comorbidity score (OR = 1.11), medication count (OR = 1.11), and count of obesity-related comorbidities (OR = 1.16) had weak effects, although patients with a cancer diagnosis were less likely (OR = 0.76) to receive obesity care.

Implications:
Among the obese PC patients, only half had obesity diagnosed in the EMR, and only about one-third received any of the three obesity treatments. Patients who received an obesity diagnosis were 4.2 times more likely to receive obesity care. Future analyses will assess the impact of facility-level predictors of obesity care.

Impacts:
Under-diagnosis of obesity in the EMR has implications for health services research and may reflect level of interest in seeking obesity treatment. Obese veterans may benefit from increased access to the relatively new MOVE! program.