Q&A: How Direct to Business Coverage Can Help Infection Care

September 18, 2020

Brent James, MD: “If you had an inpatient who developed an inpatient infection, it was like he got a console from ID in about three or four seconds. You just called up the program. You had to tell it the sites of infection you’re interested in, but then it did an epidemiologic evaluation of that patient.”

In his long and distinguished career, the Brent James, MD, an internationally known health care expert, worked on both the insurance and provider sides. James is currently pushing a system called direct to business, which allows large employers and large hospital systems in a region to contract directly, cutting out the health insurer middleman. James recently sat down with Infection Control Today® and provided examples of how such a system might bolster infection control and prevention. “Time after time, I’ve seen it dramatically improve clinical outcomes and reduce costs,” says James.

Infection Control Today®: Direct to business arrangements. What are they? How they come about?

Brent James, MD: Why did they set up in the first place? You know, the best way I know to introduce it: Some years ago, I was asked to chair a strategic planning committee at Intermountain. Eventually it produced our population health initiative. In the committee, I had senior executives from across the system, including the three senior leaders from our own health plan, Intermountain Select Health, the largest commercial insurer in the region. Well as we were working our way through, I asked a question. I said: “Exactly how does an insurance company add value?” We came up with a list of 10 major items, there are lots of smaller ones just to give you a flavor for it. They do something called transaction processing, insurance claims processing—surprisingly difficult, by the way, real expertise around that. Actuarial risk analysis. Part of that is reinsurance. When you reinsure a patient. It’s for really high-cost patients. They did a certain amount of benefit design. They did a certain amount of customer support. Oh, a big one. They did network formation, where they found a sufficient number of hospitals, clinics, physicians to supply all health needs for the uninsured population. The list goes on. We did the analysis. The thing that we found is, number one, almost all of them were available as commodities in the general market. You could purchase them, outsource them, the claims process and the actuarial risk analysis, for example. You could purchase them at very good prices competently done in the external market. The only element that they couldn't outsource was network formation. But we realized that the system was large enough that it could form its own network. They had a sufficient blend of physicians and facilities that they can handle that on their own. The second piece, we came back next and said, “All right, given that this is the value add, how much should it cost?” We express it as percentage of the premium dollar. The answer we got was between 4% and 8% of the premium dollar. Seemed low. But I checked in our community. We had a number of nonprofit insurers. Desert Mutual Benefit Administrators, the DMBA, that’s the LDS Church’s self-insurance plan. PEHP, that’s the public employees health plan. It’s a consortium of public employee groups from around the state [Utah]. Educators Mutual, that’s for schoolteachers. They were all running at about 4%. That’s roughly where Medicare runs too, just in passing. It turns out the 4% to 8% estimate was pretty reasonable. Oh, the trick. Major for-profit insurers will pull somewhere between about 15% and 20% of the premium dollar to put on to their bottom line in the stock market. And it was that difference, let’s say between 6% and 18%. Twelve percent of the premium dollar is a lot of money. And that was the economic argument that if a large integrated system could go direct to self-insured employers, ERISA plans, that they could then through negotiation, decide how to divide that pie of 12% of the premium dollar. Now, frankly, given that it’s the health system that puts it together, and the health system that has to do the care management within it, I’m prejudice. I think it ought to swing a little bit more to the care delivery system side. But that’s what made it so very attractive. Now let’s shift gears and go to the other side, the care management side. A lot of insurance companies claim that that’s how they added value. That’s how they justified the additional spend. Fact is, we’ve proven that care management works much better at the level of the care delivery system than at the level of insurance. And we had our data showing that makes sense. It happens better if it’s closer to the point of action, where a patient interacts with the system, and the insurance companies are kind of two giant steps away. We knew that, “Oh, if you start to manage care, that’s where you get into the whole population health thing.” We estimated that we could drop the cost of care by 30% to 50%. So, you’ve got, let’s say, 14% here, but then you’ve got over here, you’ve got waste elimination to better care management. The cool thing about the waste elimination, the way you get it is to improve your clinical outcomes. The money aligns to your mission, better clinical outcomes, lower costs. And you’ve got an even bigger pot of money to share.

ICT®: How does it standardize care and prevent defensive medicine?

James: Well, a classic. It was an infection control case. The patient went in for open heart surgery. He developed an infection, a post-op infection from a staph epidermidis, a common bacterium that lives on everyone’s skin and it’s usually completely innocuous. This was a strange variant. The patient had a very fast, bad course and died. Shortly after he died, his estate sued the hospital and the physicians because we didn’t prophylax that staph epidermidis. Turns out it was right in the protocol. They had a list of organisms that we did prophylax. They had a list that they purposefully did not. It turns out that we had the evidence based best standard. It doesn’t say you have to do everything. He just says that you deliver good care, and by consensus in our infectious disease staff, and the consensus among our physicians and nurses, that was best care. And what we did is we pulled out the protocol and then showed it to the plaintiff’s attorney. Once again, another lawsuit dropped fairly quickly before you go through all the trouble of preparation. That’s where the real money goes, by the way, in discovery and preparation. We had a long list of those cases. Now imagine the opposite happens. Imagine you have the standard, and you need to verify. All right, well, two things. We had built into every protocol, it was just boilerplate our attorneys have helped us with, it said this protocol is designed for the general use of most patients, but may need to be adapted to specific use of a particular patient by the patient’s caregiver. All right, it was right there in the protocol. It was honest, it’s exactly how we used it. It also became number two. That’s when you need the note. When you need that six-word note is when you choose to vary. You just need a clinical explanation, but it became a marker: “Oh, this is where I need the note to protect me.” You see, I have to tell you honestly on long experience, it reduced our malpractice risk because it established a legitimate written standard of care. And it told us in the document so on that our malpractice risk went down.

ICT®: What about defensive medicine? That’s a big problem.

James: There are two elements to that too. The first was this. What turned out what was happening…. We investigated this a bit. So, you measure variation in care. one guy’s doing it one way, another person is doing it a different way. Third person’s doing it differently still, and we go back and ask him why, why did you do that? And nearly always they come back and say defensive medicine. It was a really common response. What we discovered, though, is it probably wasn’t defensive medicine that was just plain old garden variation. I was doing it in my opinion with very little support. Right. So, it was a marker for variation, and it probably wasn’t a legitimate answer. To be honest. It’s just a handy hook to hang your hat on when somebody challenges about why you’re doing this in this crazy way. You see, well, that kind of went away, but you have to give that pressure relief valve. On average, you know, we did this for over 125 protocols. You vary about 5% to 15% typically. I used to say it this way to my colleagues just to make it real easy: “Look, guys, we’re tracking protocol brands. You will get as much scrutiny for complying with protocols as for complying too little. If you’re different from your colleagues, we want to take a look. The first time that really played out was in infectious disease interestingly enough. John Burke, the head of infectious disease, had a system. If you had an inpatient who developed an inpatient infection, it was like he got a console from ID in about three or four seconds. You just called up the program. You had to tell it the sites of infection you’re interested in, but then it did an epidemiologic evaluation of that patient. The first screen listed: Here are the most likely causative pathogens in our experience at this hospital. It gives you a differential of causative pathogens. There’s a 72% chance it’s this one. Oh, with this pattern of sensitivities, there’s a 59% chance it’s this one. There’s a 31% chance with this one. With each sensitivity in probability order. If you went one screen ahead, it said, here’s the best antibiotics to treat this set of infections. It actually gives you several choices. It was in priority order by efficacy number one, and then if it had several similar efficacies, it prioritized them in cost order. And if you saw one you like, give me a number two. It ordered the drug, it alerted the nurses. If you had level testing, it automatically did the level testing. Now initially, people were really kind of watchful around this thing. Did it deliver good care, as you would expect from any good group of clinicians. But it turned out to work and it worked pretty well. The real problem is, oh, two months later, they came to us and said, “You know, we don’t really need to see the differential. Just show us a list of antibiotics. It will shorten our time.” And then they came back and said, “Don’t show us the whole list of antibiotics. We pretty much always want the one on the top. Just show us the top one.” It reduced it to a single click, single return. And it took the time that when I was practicing surgery, it used to take me 10 to 15 minutes. It probably cut it by half. You see, that was really popular. But then we saw the downside. The downside was John Burke, using the same system, started to pick up cases where they should have buried, and didn't. It made it so dang easy to follow the protocol that, you know, people aren’t thinking as hard as they should. But it turns out the same system that detected it…. At the time at LDS hospital, you might get a console from ID even if you didn’t order one. Because John Burke was monitoring the whole hospital and he’d come by on cases that looked like they needed a little expert oversight and just show up. They learned how to interact with their colleagues in a positive way. What we saw—by the way, this was published in New England Journal of Medicine—was a rather dramatic drop in…. Well, it was infection related mortality was what they were tracking in that study. And side-by-side with it, a drop in costs. I love the way that one of our ID docs said it. He said it was more targeted therapies where you were getting the right therapy to the right patient at the right time.

ICT®: Anything that you want to add, Doctor?

James: You know, I’ve been looking for my whole career for things that make this happen faster and better. Time after time, I’ve seen it dramatically improve clinical outcomes and reduce costs. Sometimes it’s a real uphill fight. And frankly, on my side, it mostly has to do with administration worried about the revenues. When you move into these straight to business relationships, technically, it’s a form of shared savings, and it solves the financial side so that medically we can do what’s right. And I’m pretty sanguine about it. I like it a lot. I think it’s a good way to practice medicine that will make us better doctors, help us learn faster, and help us deliver dramatically better care to our patients. I hope it takes over the world. I think it will make us better doctors and a better practice in general.

ICT®: Better infection preventionists?

James: Oh, yes. Oh, yes. The examples I just used were all infection, weren’t they?

This interview has been edited for clarity and length.