PS2-15: Coding for Obesity in a Health Plan Claims Database

BACKGROUND: and Aims: The Centers for Disease Control estimated the obesity rate in New Mexico for 2008 to be 25.2%. Sources estimate the following associations between obesity and type 2 diabetes (80%); cardiovascular disease (70%); hypertension (26 %). Yet obesity is infrequently coded as a secondary diagnosis among providers submitting claims. This study examines the frequency with which obesity is documented on claims forms, the relationship between age, gender, and obesity coding, and the relationship between obesity coding and healthcare utilization. METHODS: Lovelace Health Plan (LHP) claims for calendar year 2008 were queried to identify a diagnosis of obesity documented with an ICD-9 code. Of 4,559,975 claims, 559,672 (12.3%), were for individuals who had a secondary diagnosis of obesity. Outpatient, inpatient, emergency, and total claims per patient were compared for patients with and without a diagnosis of obesity by age and gender. A comparison of major diagnostic categories taken from primary diagnoses for patients with and without coded obesity was also made. RESULTS: Mean annual claims for patients coded for obesity equaled $10,983, compared to $5,924 for patients not coded for obesity. For males coded for obesity, the mean annual claims paid were $12,165, compared to $10,409 for females. The figures increased as the age of the patient increased for both patients coded and not coded for obesity. For three major diagnostic categories Endocrine, Metabolic Diseases, (OR=2.5), Skin Diseases (2.0), and Circulatory Diseases (2.0), the odds of having a claim submitted for patients coded for obesity were at least double. CONCLUSIONS: The prevalence of patients coded for obesity in LHP claims is far lower than the estimated prevalence in New Mexico. Obesity is associated with greater utilization of health care. The odds of patients being coded for obesity are at least double for endocrine, skin, and circulatory diseases. This study describes an association that is assumed to be causative, since obesity is preventable and reversible. However, further studies need to be conducted to determine accuracy of coding. Perhaps incentives for providers to code for obesity should be considered.