One of the central tenets of the patient safety movement is that modern medicine is a team sport. Unfortunately, its players – particularly physicians – were trained and socialized to be free-spirited individualists. We need the Celtics of the 80s; what we have is a collection of young John McEnroes.

While this theory has been generally accepted, there is less agreement regarding how to change things. When I speak about safety culture, many of the questions I’m asked focus on how we are going to train future generations of medical students and residents to be “different” (translation: not like the prima donnas I have to deal with in my daily practice). It’s as if people are fatalistic about the ability to transform the culture of today’s practitioners; perhaps the next crop of physicians will do, and be, better.

Those of us who aren’t resigned to a biological solution to this problem have been enthusiastic about teamwork training and crew resource management programs for years. These programs are modeled on similar programs introduced in aviation in the 1980s after it was discovered that several tragic crashes had their roots in remarkably poor teamwork and communication. The programs bring together multidisciplinary groups to learn habits of clear communication and teamwork, and to be trained in the use of tools to employ when the going gets tough – such as, say, when a flock of Canadian geese flies into your jet’s engines.

Five years ago, with funding from the Moore Foundation, we implemented such a program on the medical services at UCSF and at two nearby hospitals. Unfortunately, while the program’s participants believed that it made care safer, our relatively small numbers of patients and providers left us unable to show improvements on hard outcomes like mortality. Other studies have had similarly mixed results – enough to keep the candle burning for those of us who believe that culture is critical and that teamwork training is the likeliest way to improve it, but not enough to catalyze a national movement for more. And, because it is expensive (the outlays for the trainers are only a small fraction of the costs – the real costs are the lost productivity of scores of nurses and doctors taking a day away from their regular jobs), teamwork training has mostly remained a novelty, implemented by a few cutting edge institutions and true believers.

In a recent issue of JAMA, Neily and colleagues report the results of a teamwork training program implemented in 74 VA facilities. They found that the intervention (not just teamwork training, but more on this later) was associated with a 50% reduction in post-operative mortality, when compared to a contemporaneous control group of 34 facilities that had not yet implemented this training. While the raw mortality reduction is impressive, the finding of a dose-response curve (for each additional quarter-year of training, there were 0.5 fewer deaths for every 1000 procedures) makes the results all the more convincing.

I think pressure will now mount for healthcare organizations to institute teamwork training programs, particularly in their high risk, error-prone areas such as the OR and labor and delivery. As usual, the question will be, Where the money will come from? While our initial UCSF teamwork training program was grant-funded, the real test will be whether organizations will spend their own money for these programs. My guess is that most will not. This, of course, raises another question: Should regulators and accreditors now require teamwork training? Eventually, yes, but not just yet. I worry that an imposed solution will lead to half-hearted programs and resentful participants. This is the kind of complex, socio-cultural intervention that I’d love to see grow organically out of an organization’s recognition of its own needs. At the same time, it would be reasonable for the Joint Commission and other accreditors to mandate that every healthcare organization perform safety culture assessments, and that units with poor culture take active steps, such as a VA-like program, to address it. And, just as pilots are all required to undergo teamwork training to receive and maintain their license, such training should become a feature of every medical school and residency curriculum, and ultimately should be required by certifying boards.

We are still learning the best ways to teach teamwork: there is good team training and not-so-good team training. Some of the lessons we learned from our experience were: a) everybody needs to participate; nobody can be allowed to opt out (some ORs now require that surgeons and the other members of OR crews participate in teamwork training to maintain their staff privileges, and their hospitals shut down their ORs for a day to allow the training to be conducted with full participation); b) the training must be intensely multidisciplinary (i.e., you need to get docs, nurses, pharmacists, and others sitting together, working through clinical scenarios – teaching teamwork to separated cohorts of physicians or nurses is a perversion of the concept); c) the use of high-tech simulation can help by amping up the drama, but it isn’t crucial; d) leadership endorsement and support of the program is essential; and e) while there are many companies that can be brought in to help design and implement teamwork training programs (many of them staffed by former or active pilots), programs need to be localized and delivered, at least in part, by folks known and respected by the staff, not by outside consultants. Moreover, we learned that a single training program of 4-8 hours, while helpful, won’t have lasting impact unless it is followed by ongoing efforts to reinforce the lessons and the tools. The initial training is like a vaccination; the effect erodes if it isn’t followed by the appropriate boosters.

Another lesson, one emphasized by the JAMA authors and by Pronovost and Freischlag in their superb accompanying editorial, is that teamwork training is one of several interventions, and it is the combination that matters. Read the paper closely: while everyone is talking about the Crew Resource Management component, the actual program included pre-op briefings and post-op debriefings, the use of coaches, implementation of checklists, and several other tools. The risk here is that someone who hears about the VA study in passing might believe that implementing a one-and-done day of teamwork training will achieve the remarkable results described by Neily. This is analogous to the checklist experience, in which many people misinterpreted the Michigan ICU results as being simply the result of a checklist (how simple is that!). As Bosk and the Hopkins team that implemented the Michigan Keystone project later wrote in the Lancet,

“If we just tell the workers to use checklists, we will have solved the problem of catheter-related blood stream infections” is quite simply the wrong conclusion to draw from the Keystone study. The “simple checklist” stories in the press created excitement about progress in achieving patients’ safety and reassurance for the public and policy makers, but the real story of Keystone is messier and more complex. Although we all hope for the simple solution that with ease and no additional expense makes a stay in the ICU safer, there is some danger in mistaking hope for reality. The answer to the question of what a simple checklist can achieve is: on its own, not much.

And so it is with teamwork training: necessary but not sufficient.

The recent tenth anniversary of the safety field occasioned lots of chatter regarding how little progress has been made in preventing medical errors. While I share some of this disappointment, I also think that changing behavior and culture is not a sprint. It’s a marathon. This week’s JAMA study proves that culture can be modified with a feasible set of interventions, and that these efforts not only make providers feel warm and fuzzy, they also save lives.

If I were a patient in need of surgery, I’d want to know my surgeon’s outcomes and pedigree, to be sure. But I’d also want to know about the quality of the teamwork. Since there is no way for me to find this out, I’d be satisfied knowing that the members of my surgical team had participated in a robust team training program, along the lines of that reported by the VA. If they hadn’t, I’d look elsewhere. It’s that important.

This essay is cited in the KevinMD record on medical errors.

Bob Wachter is chair, American Board of Internal Medicine and professor of medicine, University of California, San Francisco. He coined the term “hospitalist” and is one of the nation’s leading experts in health care quality and patient safety. He is author of Understanding Patient Safety, Second Edition, and blogs at Wachter’s World, where this post originally appeared.