|| Checking for direct PDF access through Ovid
To examine the relationship between hospital outcomes and expenditures in patients undergoing bariatric surgery in the United States.As one of the most common surgical procedures in the United States, bariatric surgery is a major focus of policy reforms aimed at reducing surgical costs. These policy mechanisms have made it imperative to understand the potential cost savings of quality-improvement initiatives.We performed a retrospective review of 38,374 Medicare beneficiaries undergoing bariatric surgery between 2011 and 2013. We ranked hospitals into quintiles by their risk and reliability-adjusted postoperative serious complications. We then examined the relationship between upper and lower outcome quintiles with risk-adjusted total episode payments. Additionally, we stratified patients by their risk (low, medium, high) of developing a complication to understand how this impacted payment.We found a strong correlation between hospital complication rates and episode payments. For example, hospitals in the lowest quintile of complication rates had average total episode payments that were $1321 per patient less than hospitals in the highest quintile ($11,112 vs $12,433; P < 0.005). Cost savings was more prominent amongst high-risk patients where the difference of total episode payments per patient between lowest and highest quintile hospitals was $2160 ($12,960 vs $15,120; P < 0.005). In addition to total episode payment savings, hospitals with the lowest complication rates also had decreased costs for index hospitalization, readmissions, physician services, and postdischarge ancillary care compared with hospitals with the highest complication rates.Medicare payments for bariatric surgery are significantly lower at hospitals with low complication rates. These findings suggest that efforts to improve bariatric surgical quality may ultimately help reduce costs. Additionally, these cost savings may be most prominent amongst the patients at the highest risk for complications.