By ROBERT WACHTER, MD

In this month’s Archives of Internal Medicine, my colleagues and I report the results of our early experience with hospitalist co-management of neurosurgery patients. We found stratospheric satisfaction among neurosurgeons and nurses, as well as impressive cost reductions ($1400/admission). At the same time, there was no impact on quality or safety, at least as judged by hard end-points such as mortality and readmission rates.

While these results might seem like a mixed bag, I believe that the overall impact of this service has been fantastic, for patients, surgeons, and our own hospitalists. Let me explain, beginning with a brief history of hospitalist co-management, folding in the history of our neurosurgery co-management effort (which we call the “Co-Management with Neurosurgery Service”, or CNS), and ending with some of the more subtle outcomes that lead me to feel that this is one of the most important things our hospitalist program has done since its inception in 1995.

A Brief History of Co-Management

When the hospitalist field took off in the mid-1990s, we projected that its growth would largely reflect the degree to which hospitalists assumed the care of inpatient internal medicine (and later, pediatrics) patients: those with pneumonia, heart failure, sepsis, GI bleed, and the like. Sure, I recognized that there would be increased opportunities for traditional medical consultation – we come when you call us – but I completely underestimated the siren call of co-management.

It turns out that once there are hospitalists in the house, the notion of having them actively co-manage surgical patients is hard to resist, for several reasons. First, many of the problems such patients experience before and after surgery are really medical, not surgical. Secondly, just as a hospitalist can provide on-site availability that the primary care physician can’t match for medical patients, he or she can do the same for surgical patients. (In this case, it’s not that the primary care doc is stuck in the office, but rather the surgeon is stuck in the OR.) Third, in an era of more widespread quality measurement and reporting, it seems likely that a hospitalist will improve quality measures such as DVT prophylaxis and evidence-based management of CHF more than a surgeon, flying solo, would be able to.

Finally, there are the economics. If, in helping to manage the post-operative care of a surgical patient, a hospitalist frees up the surgeon to do another case in the operating room, the economics for both the surgeon (payment for one more case) and the hospital (ditto) are highly favorable. If the surgeon earns $450,000/year and the hospitalist $200,000, it’s clearly more lucrative to have the latter take over some of the medical management if that frees up the former to scrub in for one more hip replacement or craniotomy.

But the economics are not simple. Precisely because hospitalists don’t perform procedures, their pro fee collections generally don’t come close to covering their salaries and benefits. In the case of medical patients, there’s no ambiguity regarding where to seek additional dollars to augment these collections: in more than 90% of hospitalist programs, it is the hospital that pays the difference between the collections and the costs, to the tune of about $100,000 per clinical FTE. The fact that the hospitalist field is the fastest-growing medical specialty in modern history tells you all you need to know about hospitals’ collective thinking about the wisdom of this investment.

But surgeons are generally paid a case-rate (a fixed payment per surgical case), which bundles in the compensation not just for the surgery but also the pre- and post-operative management. This raises a sticky question: if hospitalists require a support payment for co-management, shouldn’t some of the dollars come out of the surgeon’s pot o’cash? This very question held up our co-management efforts, and those at hundreds of hospitals around the country, for several years; more on this later.

In the late 1990s, several institutions began experimenting with hospitalist-surgical co-management. I collaborated with one of the most prominent: the Hospitalist-Orthopedics Team (HOT) at the Mayo Clinic. The Mayo folks demonstrated that the HOT resulted in patients being ready for discharge sooner, with a small decrease in minor complications, but no major improvements in hard end-points. Not exactly the Salk vaccine.

Despite these less-than-jawdropping results, interest in co-management grew – mostly, I suspect, because of the economic issues I described earlier rather than any ironclad evidence that this model improved quality or efficiency. This growth curve tilted upwards in 2003, occasioned by the ACGME’s first duty-hours limitations, which instantly removed tens of thousands of hours of surgical resident bandwidth from the system. If surgical housestaff previously worked 110 hours/week and now could only work 80 – and if surgical residents both want and need to scrub into as many cases as possible – something has to give, and that something is the amount of time residents are available on the wards to manage patients before and after surgery. Who could pick up the slack? Sure, some of the work could be assumed by NPs or PAs, but, for the more complex patient populations, hospitalist co-management seemed the best solution.

Our Early Experience With Co-Management

At UCSF, our hospitalist program dabbled in co-management with our orthopedic service about six years ago, and our experience demonstrated that these services could fail if both logistics and culture weren’t right. The rules of engagement were vague, the hospitalists felt that their time on the service was undercompensated, and our medical residents, who were helping the faculty co-manage the ortho patients, hated it (they felt they were substituting for the orthopedics housestaff, which, to a degree, was correct). For a variety of reasons, the sense of collegiality and mutual understanding vital to this kind of collaborative care was lacking: when there was a bad outcome on orthopedics, we felt that there was finger pointing (“where was the hospitalist?”) rather than an honest effort to learn from experience. After a fairly unpleasant year, we pulled the plug on the service, convinced that this was not the right time, right place or right people, and that, while co-management could work, both the logistics and the culture had to be just so.

Around this time – 2004, as I recall – my hospital’s Chief Medical Officer, Ernie Ring, approached me regarding the possibility of co-managing neurosurgery patients. The need was pretty obvious. This is one of the busiest and highest-profile services in the hospital. Ranked as one of the top five neurosurgery services in the nation, it receives referrals from all over the world and carries an average daily patient census of 50-60. When we analyzed the service, we realized that this large volume of patients, many of whom were quite ill, was being managed by one hapless second-year surgery resident who had drawn the short straw; the rest of the surgeons spent the day in the OR. It was not hard to find cases of poor outcomes because patients’ deterioration was not recognized quickly enough or addressed appropriately.

But we had this money problem: staffing the service with one hospitalist a day, 365 days a year, was a half-million-dollar proposition. For a couple of years, three-way meetings that included me, the CMO, and the neurosurgery department leadership ended with a philosophical agreement to begin co-management, but a snag over the source of the dollars. From the medical center’s perspective, the surgical department needed to ante up some of its own professional fee revenues to support co-management. The surgeons, like those everywhere, felt the hospital should be the source. Nobody blinked, and so the CNS remained a drawing board proposition.

Finally, in 2007, after 3 years of negotiation, the medical center leadership realized that it would have to bankroll the service if co-management was to become a reality. After a couple of bad outcomes that might have been prevented had there been a hospitalist on the service, the hospital agreed to provide all the necessary funding. The CNS was born.

The Co-Management With Neurosurgery Service: The Inside Story

Now that the funding was secure, we needed to sort out how the service would actually function. I asked Quinny Cheng, the wonderful director of our med consult service, to lead the CNS. Quinny and I met with the chair of neurosurgery, Mitch Berger, a world-class surgeon and former high school football academic
all-American. (Wooed to Alabama by Bear Bryant and Wake Forest by Arnold Palmer, he chose Harvard because “there’s life after football,” demonstrating uncommon maturity for an 18-year-old kid.) Mitch is a Just-Do-It kind of guy: larger than life, passionate about patient care, and a bit impatient. His initial view was that the co-managing hospitalist should round on every patient, every day. “Mitch, in a 12 hour shift, that’s about 10 minutes per patient,” we said. Instead, we advocated for having us actively co-manage the sickest 13-16 patients on the service (as determined by a triage rule described in the paper), being available to the nurses, physicians, and families to check in on the other 35-40 as needed.

Before we launched we were also determined to agree on clear rules of engagement; failure to achieve such an agreement, I was convinced, had been responsible for dooming our orthopedics co-management experiment. We would manage the medical problems, largely independently. The surgeons would retain ownership of the surgical issues (“We don’t do burr holes,” was our cheeky motto). Decisions around anticoagulation, so potentially hazardous in these patients, would be made jointly.

There was one other thing that needed to be hammered out, and I didn’t mince words. “If one of the surgeons yells at one of my hospitalists, we’ll give you three months notice and then leave the service.” Although the iconic image of the scalpel-throwing surgeon is somewhat hyperbolic, it does exist in nature, as does a physician hierarchy that has surgeons on top and general internists near the bottom. I needed to be clear that, at least as far as the hospitalists went, anything but a respectful partnership wouldn’t be tolerated.

The Outcomes of the CNS: Hard and Not-So-Hard

While the Archives paper describes the hard outcomes in our first 18 months of the service, it was some of the softer ones that convinced me that this was an extraordinarily positive experiment. For example, after having at least one or two cases each year of truly awful patient outcomes that could be pegged to insufficient floor oversight, I haven’t heard about such a case since we began co-management three years ago.

Moreover, I find the survey results compelling. We surveyed the non-nurse caregivers (mostly neurosurgeons with a smattering of NPs and pharmacists on the service). Their response to the statement: “Patients’ medical problems are promptly recognized and appropriately addressed” was 2.7 (on a 1-5 scale, where 5 is complete agreement) before the CNS began, and 4.4 afterwards, a highly significant uptick. Six months after the service began, this group’s response to the statement, “The presence of a hospitalist improves care for neurosurgery patients,” was 4.9 out of 5, representing near-uniform agreement.

Perhaps more impressively, 34 neurosurgery nurses answered the post-implementation survey. The average nurse had worked at UCSF for 11 years – they’d seen it all, including lots of failed quality improvement strategies implemented with great hype. Their score on the statement, “The presence of a hospitalist improves care for neurosurgery patients,” was 5.0, meaning all 34 gave it the highest rating. I have to believe that perceptions like these reflect real improvements in care.

Then there’s the cost data: $1,400 lower costs per surgical case, translating into a savings of more $1.5 million during the first 18 months of the program. This yields a return-on-investment of 2:1, independent of any revenue enhancements via greater surgeon availability to do cases, or any impacts on quality, malpractice liability, or nurse satisfaction and retention.

So, while the study does not prove that lives were saved, I find the evidence of benefit, and of the positive return-on-investment, to be pretty compelling.

Our Archives article was accompanied by an editorial written by Patrick O’Malley of the Uniformed Services University of the Health Sciences in Bethesda. In the editorial, subtitled, “Can We Afford to Do This?” the author states, “this well-performed study stops short of providing definitive evidence to support co-management or refute its value, from any perspective.” He then goes on to attack the economics of the co-management model:

I would argue that even the economic rationale for co-management is poor because it really involves shifting work to lower-paid workers (internists), allowing surgeons to spend more time in the operating room, where they get paid more by a dysfunctional reimbursement system that disproportionally rewards procedural care over more cognitive services.

Well, duh. As it happens, this is also the rationale for having NPs or nutritionists perform many primary care tasks in the Patient-Centered Medical Home, and for using Turbo Tax instead of an accountant for your taxes. In other words, this is what systems look like as they seek higher value: they shunt work down the income scale, allowing more highly paid specialists to focus on what they can uniquely do, and assigning some of their prior tasks to lower paid, less specialized individuals who can do the work well at lower unit costs. I’m still scratching my head trying to figure out why Dr. O’Malley finds this offensive, or demeaning. Sure, it is unfair that some surgeons make four times what a primary care internist makes. It’s also unfair that Alex Rodriguez makes $43,000 every time he gets up to bat. Tilting at this particular set of windmills doesn’t change the facts on the ground – particularly since, in this case, patients likely benefit from the new arrangement.

Perhaps Dr. O’Malley will be pleased to know that hospitalists benefit too, in ways that I could never have predicted.

Our negotiated agreement pays significantly more for a day’s work on neurosurgery than for a day on the medicine wards. (Note to my fellow hospitalist directors: I wouldn’t have agreed to co-management without this premium.) This has helped ease the job of recruiting our faculty to rotate on the CNS, as 6-8 of them do each year. And many of our faculty have really taken to the work. While hospitalists clearlyadd value on the medicine wards, the effect is attenuated by our army of very smart medicine housestaff who can function quite well without us. On CNS, our hospitalists have the gratifying feeling that the patients are getting far better care directly because of our presence.

My concern about the relationship between the surgeons and hospitalists has also proven to be a non-issue – quite the opposite, actually. In three years, I can recall only one incident in which a surgeon lost his temper with a hospitalist. One. Instead, the relationship has been characterized by tremendous mutual respect: we know that we couldn’t do what they do (extraordinarily well), and I am confident that they feel the same way about us. It feels like a real partnership.

This partnership has other value as well. When the medical center makes a bad decision that affects all services (cutting the availability of pharmacists, for example), it is no longer just our medical service advocating in isolation. We sit down with our neurosurgical colleagues and make our case together. And our collective voice is loud. Having a surgeon on your side – particularly a 6 foot, 4 inch former defensive tackle – is like having that big kid in class looking out for you in grade school.

Over the past two years, the fruits of this partnership have truly begun to flower. Quinny Cheng, the hospitalist who runs the CNS service, won a departmental teaching award last year… from theDepartment of Neurosurgery! And, several months ago, Mitch Berger called me for a meeting. “We want to be the leading neurosurgery department in the country in quality and safety,” he said. I was thrilled; I’ve been pushing this agenda for years throughout UCSF Medical Center. “I want to hire a physician to lead our efforts.” But where would he find a neurosurgeon with the skills, interests, and time to run a world-class quality and safety program, I wondered aloud. But Mitch was thinking differently. “I’d like to hire one of your hospitalists to run it.”

So, while not all co-management experiments will succeed, I’m convinced that our partnership with neurosurgery has not only improved quality, safety, and efficiency, but also resulted in exciting collaborations in education, research, and quality improvement. If you’re in the position to consider such a relationship, go in with your eyes open, make sure the conditions are right, and sweat the culture along with the logistics and the dollars. But if these stars align, I’d recommend you do it.

In healthcare, we talk a lot about unanticipated consequences, and when we do it’s usually not a happy tale. This is that rare case in which the unanticipated consequences have been uniformly positive. Thanks to all my colleagues – particularly Quinny Cheng, the other hospitalists who rotate on CNS, Andy Auerbach for taking the lead on the Archives paper, the medical center leaders who support the service, and our friends in neurosurgery, especially Mitch Berger – for making it so.

Robert Wachter, MD, is widely regarded as a leading figure in the modern patient safety movement. Together with Dr. Lee Goldman, he coined the term “hospitalist” in an influential 1996 essay in The New England Journal of Medicine. His most recent book, Understanding Patient Safety, (McGraw-Hill, 2008) examines the factors that have contributed to what is often described as “an epidemic” facing American hospitals. His posts appear semi-regularly on THCB and on his own blog, Wachter’s World.