No: The data do not show that outcomes are better in COEs.
The Centers of Excellence has evolved into the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP), which is an outcomes benchmarking and quality improvement program. So we are, in many ways, looking in the rear-view mirror when we talk about whether or not the historical accreditation process improved outcomes.
I’m going to make two arguments against the historical way it was done. The first argument is that limiting care to Centers of Excellence (COEs) as they were defined had very little benefit on outcomes and may have caused some harm.
Dr. Justin B. Dimick
A lot of the data out there to evaluate the COE program are a little misleading. They looked at Medicare data for outcomes before and after the Medicare COE mandate was put in place requiring that bariatric surgery be performed in a COE to qualify for coverage. The problem with that is the outcomes are improving so quickly that you can’t just do a before and after study. You need to adjust for that time frame to make sure you’re not getting a spurious finding.
We did a study comparing improvement trends for any complications, serious complications, and reoperations before and after Medicare implemented its coverage decision. Surgeons were already getting better before that. There was no change in the trajectory of improvements when the policy was implemented. We concluded that the policy that selectively referred patients to Centers of Excellence didn’t have any benefit (JAMA 2013;309:792-9). We originally included mortality in our analysis but deaths were so rare that it didn’t meet power specifications set by the journal.
In that paper, we did a head-to-head comparison of COEs vs. non-COEs in 12 large states. It’s hard to do a randomized controlled trial for this, so we did an instrumental variable analysis, a sophisticated analysis that takes out a lot of the unmeasured confounding in selection bias. We found that there was no benefit to having bariatric surgery in a COE.
We noticed that many hospitals that traditionally served minorities did not achieve accreditation. One can assume that’s because they probably didn’t have the resources to participate in the program. After the policy was implemented, the rate of bariatric surgery in minorities dropped substantially, whereas it went up in non-minority patients (JAMA 2013;310:1399-1400). So, there’s potential evidence of harm with these policies.
Medicare eliminated their facility accreditation requirement for bariatric surgery last September.
My second argument is that we could be doing something so much better than accreditation to improve bariatric surgery outcomes. Michigan’s approach is a large quality improvement collaborative funded by Blue Cross Blue Shield. We’re using lessons learned in Michigan to implement change on a broader scale without the involvement of payers. We have 70 surgeons and 32 bariatric surgery programs benchmarking outcomes and using those in strategic ways to learn from each other, hopefully improving everybody’s outcomes.
We standardized venous thromboembolism (VTE) prophylaxis across the state and implemented decision support. VTE rates fell by half. We saw a 90% improvement in mortality and a 24% drop in overall complications. Yes, bariatric surgery centers everywhere were improving at that same time, but when we compared Michigan hospitals participating in our collaborative to non-Michigan hospitals, we saw steeper declines in adverse outcomes in Michigan. That’s very robust evidence that this approach is effective for improving perioperative care.
We have launched a technical skill initiative in which surgeons participating in the collaborative give us a video of a typical procedure that gets rated in a blinded fashion by their peers. We just got funded to use these videos in a coaching program to help surgeons in Michigan improve our technique.
Doing this kind of collaborative work doesn’t take a ton of money. It does take time, strategy, and social capital – developing networks of surgeons that work together and learn together. The historical Centers of Excellence program wasn’t a bad idea, but it wasn’t correlated with perioperative outcomes, and it did have harms.
Dr. Dimick is chief of minimally invasive surgery at the University of Michigan Health System, Ann Arbor. He disclosed being a stockholder in ArborMetrix.