Bariatric Surgery for Severe Obesity: Determinants of Use and Economic Impact

Abstract

Objectives: Obesity in the U.S. population is a major public health problem with important clinical and economic implications. Bariatric surgery is currently the most effective long-term weight loss treatment for morbid obesity but less than 2% of the potentially eligible population has undergone the procedure. Our aims were to: 1) determine the patient characteristics associated with receiving a bariatric procedure among eligible patients; 2) identify patient characteristics that could be used as predictors of resource use and costs in severely obese patients with and without bariatric surgery; and 3) estimate and project the differences in long-term costs and outcomes across different clinically defined populations undergoing bariatric surgery compared to non-surgical approaches, refining a previously developed cost-effectiveness model. Methods: We used electronic medical records from members of Group Health Cooperative, based in Washington State to identify severely obese individuals, eligible for bariatric surgery, from 2004 to 2010. The probability of undergoing bariatric surgery was assessed using multivariate logistic regression adjusting for demographic and clinical characteristics, as well as patterns of practice. To determine the predictors of costs, we used generalized linear models for both groups with similar adjustment variables. We created a propensity score matched cohort based on specific characteristics previously identified at the date of surgery and the assigned index date for those who did not have surgery. Finally, we used these data to update a previously developed cost-effectiveness model to re-estimate the cost-effectiveness of bariatric surgery compared to non-surgical interventions. Results: A total of 48,166 subjects were identified as eligible for bariatric surgery. Only 1,129 had bariatric surgery. The characteristics associated with having a bariatric procedure were: having insurance coverage for the procedure (OR=5.61; 4.71-6.68), higher body mass index (BMI) (1.10; 1.09-1.11), and older age (1.00; 1.00-1.01). The presence of comorbidities was associated with higher odds of having surgery. Examining comorbidities individually, only coronary heart disease was not associated with the surgery (0.96; 0.74-1.26). These characteristics changed over time, showing time trends towards increasing numbers of older adults and subjects with lower BMI having the procedure in the recent years. For costs, total annual costs were higher prior to surgery in the bariatric surgery group but decreased more (in absolute and relative terms) after the surgery compared to the non-surgical group. The total health care costs post-surgery for both groups were primarily driven by inpatient costs. The presence of comorbid conditions was associated with greater annual total costs in both groups. Major cardiovascular risk factors, such as coronary heart disease, hypertension, diabetes, and a higher comorbidity index, were associated with the highest increase in mean annual total costs. Despite being more expensive, bariatric surgery is cost-effective due to the impact on weight loss, which is associated with improved life expectancy and quality of life after the procedure. The incremental cost-effectiveness ratio of the bariatric surgery is sensitive to the presence of multiple comorbidities, mainly due to the reduction in the differences in the lifetime costs between surgical and non-surgical interventions for patients with multiple comorbidities. Conclusions: There are demographic, clinical, and insurance differences among subjects eligible for bariatric surgery that receive and do not receive the surgery. The presence of specific comorbid conditions increases the probability of surgery and is associated with higher total costs in this population. The cost-effectiveness of these procedures also depends on the presence of specific comorbid conditions. Understanding these differences could better inform the decisions that patients, clinicians, payers and policy-makers face in addressing the problem of increasing population obesity.