Getting all your “ducks in a row” with standardized order sets and procedures may help reduce costs and improve care.
Courtesy of Thinkstock


Published 11/1/2012
Terry Stanton

Applying a Factory Model to Knee Surgery

Can standardizing knee replacement treatment improve outcomes?

What can medicine learn from a national restaurant chain? In a recent article in The New Yorker, Atul Gawan-de, MD, author of The Checklist Manifesto, made the case that the standardized operating methods used in food service might be good for healthcare services as well.

Dr. Gawande’s article examined the approach taken by The Cheesecake Factory restaurants to providing a quality product on a consistent level with little waste. He contrasted that with medical services, which he noted are greatly fragmented, inconsistent, and inefficient. And he looked for examples of where Cheesecake Factory–style standardization has been implemented with positive results. He found one at his own institution, Brigham and Women’s Hospital, in the orthopaedics department.

R. John Wright, MD, is an associate orthopaedic surgeon at the hospital and assistant clinical professor of orthopaedic surgery at Harvard Medical School. He performed a revision total knee replacement on Dr. Gawande’s mother. And, Dr. Gawande explained, “He has led what is now a decade-long experiment in standardizing joint-replacement surgery.”

According to Dr. Gawande, Dr. Wright brought together all participants involved in the care of knee-replacement patients—from surgeons to nurses, physical therapists, and anesthesiologists—“to formulate a single default way of doing knee replacements.” Together, “they studied what the best people were doing, figured out how to standardize it, and then tried to get everyone to follow suit.”

In implementing a standardized approach, Dr. Wright hoped to achieve new efficiency and cost savings, but his paramount goal was to improve patient outcomes. In an interview with AAOS Now, Dr. Wright recalled how he implemented the program and explained why, as he said to Dr. Gawande, “customization should be 5 percent, not 95 percent, of what we do.”

AAOS Now: What made you undertake this effort in the first place?

Dr. Wright: Our total knee replacement patients were not doing as well as they could, and the experience was not as good as it ought to be. They were staying in hospital too long. So we teamed up with nursing, physical therapy, care coordination, pharmacy, and anesthesiology to see how we could improve this situation. First, how can we address their pain better than we are? How can we address their understanding and expectation of what is going to happen? How can we rehabilitate them faster so their milestones will be achieved more reliably and earlier, so they can get back to function or to work? The average age of our patients is 63, so many of them are still working.

We process-mapped what we were doing and looked at ways we might be able to improve. We made efficiency and lower costs our guiding principles, but not necessarily our priorities. The priority was improving patient outcomes.

AAOS Now: Where did you start?

Dr. Wright: First, we looked at dealing with pain. For a long time we had been using epidural anesthesia for 48 hours and then mobilization. But that meant that patients were in bed for a couple of days. So we reduced the epidural to one day, started using long-acting instead of short-acting narcotics, and tried to reduce the amount of breakthrough medication, so that we could get them up sooner. That required a commitment by the nurses and the therapists to really mobilize these patients.

After we moved the epidural cessation up a day and got the patients moving sooner, they were able to leave the hospital a day earlier. Pain levels on the second day were the same as they had been on the third day, and their walking distance was better.

AAOS Now: What do you do in terms of informing patients and managing their expectations?

Dr. Wright: We worked with the care coordinators who set up the patient services for rehabilitation or home care, so the services were in place in advance. We negotiated with insurers for pre-approval on patients going to rehabilitation.

Because we found that patients were concerned about leaving the hospital just a day or two after joint replacement surgery, we started an educational process, to tell them what to expect, and why it is better and safe. We encouraged patients to attend the class with a family member. To make it easier, we scheduled class for when the patient was coming in to see the anesthesiologist. We had a standardized message.

In survey interviews, we found we were doing a good job of telling patients what would happen in the hospital, but we were not so good at telling them what would happen down the road, and that was causing them some anxiety. We tried to focus on expectations, provide reinforcing messages, and encourage them to call us if they are not making their milestones.

AAOS Now: What adjustments did you make as you adopted the new approach?

Dr. Wright: After a year or two, we reassessed. We identified some cost and time concerns involving epidurals, so we worked with the anesthesiologists and settled on a femoral nerve block and a spinal anesthetic or a short general. Because the pain relief would not be quite as good as an epidural, we established a perioperative pain protocol with a standard set of medications that patients would receive before surgery, such as a long-acting narcotic and an anti-inflammatory, to be administered with the antibiotics.

AAOS Now: What other changes did you make?

Dr. Wright: We redesigned the whole postoperative course of care. We developed postoperative order sets to include every aspect of care and set them up in ways to drive behaviors. The changes in the pain protocol enabled patients to get up faster. We found that patients were coming out of the operation with lower pain scores.

To see if we could mobilize patients more quickly, we looked at the continuous passive motion (CPM) machines. They are widely used, but the data on their use after knee replacement show that they don’t really make a difference. I found out what we were spending on CPM and suggested that we get rid of the machines and hire two more full-time physical therapy aides, so it was a wash in terms of cost. But cost wasn’t the driver; mobilizing the patients for better outcomes was the driver.

Everyone, including nurses, is responsible for mobilizing patients. We want patients to get up and out of bed six times a day at least and to have all their meals out of bed. We want them to be rapidly mobilized—not waiting around for the physical therapist. The intention was to have patients walking the day after surgery and ready to do stairs the next day on the way home.

AAOS Now: What were the results?

Dr. Wright: On the first protocol, the length of stay was 3 to 5 days. Our goal was to get patients who were going home or to rehabilitation out in 2 days, and Medicare patients going to rehabilitation out in 3 days. That meant that patients needed to share that expectation before their surgery.

When we compared outcomes for patients under the new protocols with outcomes for patients under the previous protocols, we saw a big difference. With the new protocols, patients had significantly less pain; pain scores dropped from 7 to 5 on the first day, and continued to be lower on the second day. Using the CPM, patients had a range of motion of just 72 degrees on day 3, but under the new protocol, without CPM but with everyone focused on mobilization, the average range of motion was 91 degrees on day 2. Walking distance was better. The ratio of patients going into rehabilitation instead of home flipped from 60/40 to about 40/60. The majority going to rehabilitation were those who lived alone.

AAOS Now: What are the keys to successfully implementing standard treatment protocols for knee replacement surgery?

Dr. Wright: You need a multidisciplinary buy-in—including the surgeons. But because so many of the protocols are perioperative—rather than strictly surgical—you can establish pathways that don’t really affect them. All the defaults in the order sets have the same standardized pathway, which makes it difficult for a surgeon to do things his or her own way.

If we achieve an efficiency but the patients aren’t doing as well, that’s not a win. It’s true that not all patients are the same, but patients are more similar than they are different. By standardizing order sets and training, you ensure that most of what happens happens reliably every time. People don’t forget to give the antibiotic or get the patients up for breakfast, because it is done for everyone.

That enables you to concentrate on the issues that should be customized for the patient to give him or her an excellent experience. You can focus all your cognitive attention on what your patient needs. The rest is handled by your subconscious, so you can focus on the important things. That’s the guiding principle.

Disclosure information: Dr. Wright—DePuy, A Johnson & Johnson Company

Terry Stanton is senior science writer at AAOS Now. He can be reached at

Bottom Line

  • Care improvement should be the priority in any attempts to standardize treatment.
  • Establishing standard protocols for most perioperative issues enables the surgeon and hospital staff to focus on customizing issues important to the individual patient to ensure an excellent experience.
  • Multimodal pain protocols, early mobilization, and managing expectations through education and family involvement can help reduce hospitals stays and improve outcomes after total knee arthroplasty.