Yale-New Haven Hospital's Mary I. O'Connor, MD, describes an effort to use a bundled payment program to bring orthopedic surgeons into better alignment with the organization's goals.

When orthopedic surgeon Mary I. O’Connor, MD, took the helm of the new Center for Musculoskeletal Care at Yale School of Medicine and Yale-New Haven Hospital in May 2015, she immediately saw an opportunity to reduce the cost of the system’s implant purchasing program. Less obvious was the opportunity to bring the orthopedic surgeons into better alignment with the organization by using a specific component of the Medicare Bundled Payments for Care Improvement (BPCI) program. It would take close collaboration with a top strategic and financial-planning executive to pull that off.

O’Connor discussed the initiative with Physician Business Adviser in advance of a presentation at The National Symposium on Value Innovation at Yale, April 27-29.

What were your goals for supply chain initiatives coming into the job of director of the center?

My overriding goal and that of our entire team was to drive value for our patients, providers, and payers. We all wanted to save money and improve the quality of our BPCI bundle for hip and knee arthroplasty, and we recognized an opportunity to improve patient outcomes with better prevention of postoperative blood clots through adoption of more standardized care pathways. We also knew there was an opportunity to decrease implant costs. While the bundle includes all care related to the surgery for 90 days, and there are a number of pathways to generating savings, such as decreasing discharges to skilled nursing facilities and rehabilitation units, reducing the cost of the implants seemed like a good initiative, particularly given the volume of joint replacements across our three-hospital health system.

I knew from years of experience with vendor negotiations at Mayo Clinic [where O’Connor was chair of orthopedic surgery for eight years] that vendors look for an increase in volume to offset a decrease in price. So limiting the number of vendors typically creates the opportunity for price reduction. There was pushback from my surgical colleagues who did not want to have their choice limited.

Was this just a matter of the natural resistance to change?

Not entirely. Product preference is not just a whim in orthopedic surgery: The instrumentation differs from one implant system to the next, and people who have training and extensive experience with implant A don’t necessarily want to switch to an unfamiliar implant B. However, for the implants we were considering that there is no clinical data to support the superiority of one product over another, which I was able to demonstrate to my colleagues.

In the end, after going through a lengthy RFP process with the help of Keith Murphy, the executive director of our corporate supply chain, and his team, we selected two vendors who now provide 90 percent of our primary joints. There are always outlier cases where a different implant is truly needed, so you have to leave a little wiggle room. We have a simple form the surgeon completes to document the reason for the exemption.

But I wanted to find a “win” in this change for the surgeons. The health system was saving money, the vendors were increasing volume and profit. So I said to Steve Allegretto, vice president of strategic analytics and financial planning, “Wouldn’t it be great if we could share some of these savings with the surgeons and link that sharing with a quality improvement effort?”

And it was Steve who came up with the brilliant idea of using the internal cost savings (ICS) component of the BPCI program to share part of what we saved on the implants with the people who use them.  Despite my prior knowledge of bundles, I had been unaware of this ICS portion of the bundle.

How does the ICS work?

There are parameters; there’s only so much we can share, and the savings have to be linked to a quality improvement process, which was perfect for us. In looking for a relevant quality metric, I found that there was a Yale task force already working on DVT [deep vein thrombosis] prophylaxis protocols—that is, how to reduce blood clots in our orthopedic patients, who are at risk for this complication after surgery. We thought our DVT numbers could be better.

So we were able to leverage that work. We brought the task force in and said, “OK, you folks have already done a lot of work on this; let’s refine it, let’s get a protocol we all agree on and make it part of our quality improvement process.” That way, we have the conceptual framework for a program that drives quality by linking physician performance related to a specific quality metric (no blood clot) to the potential for shared savings.  

So now, when we do a hip or knee replacement on a Medicare patient, it works like this: If the surgeon follows the DVT prophylaxis protocol, gives the right preventive medication at the right time, and the patient doesn’t get a blood clot prior to discharge, and if there are savings on the implant costs (which there almost always are, given the discounted pricing with only two vendors), we will share with the surgeon half of what we save on the case up to the maximum CMS [the Centers for Medicare & Medicaid Services] allows, which is half of the surgeon’s professional fee. If the surgeon’s fee is $1,200 dollars and we have saved $600 on the implant costs, we share savings of $300 with the surgeon (the maximum amount we would be able to share would be $600).

The reason we are able to share savings with our surgeons through the ICS program is because of the robust cost-accounting infrastructure that Yale-New Haven Health system has invested in, which allows Steve and his team to dig down and look at the expense per surgeon per patient in a very detailed way.

So all the surgeons who do hip and knee replacements are on board with the program?

We lost only one high-volume surgeon who moved the majority of his practice to a non-system hospital so he can use the implants he preferred. And while there are some surgeons who did not sign up for the shared savings program, everyone has to comply with our two-vendor program.

What kind of results have you seen so far?

The program is designed to run for six quarters. It just went into effect in the last quarter of 2015, and we’re close to making our first payments to the surgeons. It’s taking a bit longer because we want to be very careful with our numbers and it’s our first initiative like this.

The incidence of DVT remains low, and we’re doing very well in terms of people getting the right medication. Of course, our hope is that they won’t develop a blood clot at any time during the entire 90-day period included in the bundle. But because we wanted to provide a solid clinical endpoint for CMS, and because we can’t always control what patients do or where they go for care following discharge, we chose to use the absence of clots just during the initial hospitalization as our metric. I think we’ll eventually see a decrease in the number of DVTs. It is just too early for data.

What factors do you think have contributed to the success of the initiative?

Structuring our program on an individual patient basis was important. We did not calculate shared savings to the surgeon by aggregating the savings on cases with lower implant costs minus expenses on cases with higher costs. I want my colleagues to put the needs of their patients first and make implant decisions based on those needs. For example, if a patient needs a special implant that is more expensive, that higher cost will not negate savings realized on other patients.

Clinical leadership was another factor in successfully engaging the surgeons and driving the process. It’s one thing for a supply chain executive to try to get surgeons to use just two vendors and another for me to say to my colleagues, “Show me the data proving that the knee or hip implant you want to use is better than either of these two knees or hips”.

But supply chain leadership needs to be involved, right?

Absolutely. You cannot be successful with this kind of program unless you involve leaders from surgery, finance, and supply chain—as I learned when Steve Allegretto introduced me to the ICS component of BPCI and made the whole thing possible. 

I would add that it’s also very helpful to engage OR operations staff, because the vendor decision isn’t based on cost alone. You want to know if a company’s reps are any good, what the service is like. For example, we don’t want to select a vendor that doesn’t bring the surgical trays in the day before so we can process them in a timely manner. We want good partners. 

What else would you recommend to other organizations interested in developing a similar program?

Communicate, communicate, communicate. When a surgeon says he or she doesn’t know about the protocol, about the cost sharing, about the implant options, you want to be able to say, here are all the methods we used to try to get this information to you.

Finally, it’s important to build in some transition time, time to work out bumps so that when you finally go live, everything is in place. We are figuring this out as we go along—there are only a handful of organizations that have implemented this nationwide, and we don’t know another hospital that’s done it quite the way that we have. 

Lauren Phillips is president of Phillips Medical Writers, Ltd., Bellingham, Wash.

Interviewed for this article:  Mary I. O’Connor, MD, director, Musculoskeletal Center, Yale School of Medicine and Yale-New Haven Hospital.


a. In the spring of 2017, CMS may require Yale to replace the BPCI with the Comprehensive Care for Joint Replacement program, a new bundle it has mandated across 67 geographic regions.

Publication Date: Monday, April 11, 2016