Shortening the Team Learning Curve: A Study

In the following interview, Cath Lab Digest talks with Amy Edmondson, PhD, about the results of a study by Harvard Business School researchers. The study found that medical teams who embrace the more complex, active meaning of team have a learning advantage over groups that represent the more superficial definition. Active teams can learn new procedures faster and more efficiently than teams whose definition lies on the superficial end of the spectrum. In other words, We can do it together, works more quickly than I will do it (with an anonymous, interchangeable group there to assist). To study teams and their learning processes, Harvard Business School researchers (Gary Pisano, Richard Bohmer, and Amy Edmondson) followed 16 cardiac surgical teams who adopted a minimally invasive procedure.¹ This procedure (required by the manufacturer of the technology to remain unnamed) required a whole new level of involvement beyond what team members (physicians and staff) had been accustomed to in the past. Researchers’ hypothesis was that teams who emphasized the learning process and teamwork would learn much faster than those with equivalent levels of experience. Researchers also found that [a]n organization with less cumulative experience than its rivals can still achieve a performance advantage if it more thoroughly exploits its opportunities for learning. ¹ Study conclusions were based on the measurement of two dimensions of learning: reduction in the time required to perform the new procedure and overall implementation success, or degree to which the departments incorporated the new technique into ongoing practice. All participating teams were chosen because they were regarded as excellent in their field. Researcher Amy Edmondson was kind enough to talk with Cath Lab Digest about some of the universally applicable insights on team learning that came out of this study. What does organizational learning mean? Organizational learning is a notion that organizations, as entities, need to adapt to keep up with changes in the external environment. There’s a lot of very disparate kinds of research that pertains to that general issue. Where would you place yourself? I think the field divides broadly into macro and micro. Macro are people who think of the firm, organization, or hospital as the unit of analysis and develop theory and sometimes empirical research on large data sets from economic sources about factors that lead organizations to adapt or not to adapt to external change. In contrast, I tend to look from a micro perspective, investigating the actual behaviors that individuals and small groups engage in as part of the adaptation process. These behaviors either help them or block them from making effective changes in response to external pressures. Are there differences in the way individuals and teams learn? There are important differences. To begin with, individuals, at least as children, are programmed to learn and are more or less hard-wired to be curious. As we grow older, we have less of the child’s irrepressible drive to learn. But we’re still pretty good at learning when we put our minds to it. On the other hand, groups and organizations are almost hard-wired not to learn. Groups tend to get extremely facile with current routines. This creates a barrier to change, because the groups get really good at what they do over and over, and know who does what. This can create a powerful disincentive to change. Another barrier is the interpersonal challenges inherent in speaking up to question what others are doing. People are often reluctant to say there’s a mistake or a problem. No one wants to be the skunk at the picnic. So there are all sorts of social dynamics that can cause people to be reluctant to engage in the very behavior that they need to engage in if their groups are going to learn and change. Organizations are a much more complex version of a group, in the sense that there’s both interpersonal pressures and managerial and organizational system complexities that also favor stability over change. Why would a business school researcher choose to study medical teams instead of a more traditional business example? The medical field has real world importance and significance. This is not to say that business organizations do not, but everybody cares about human life, and many people recognize that health care is going through enormous change, making team and organizational learning especially important in this context. Also, my colleague, Gary Pisano, already had done considerable research in pharmaceutical manufacturing. Together, we had an interest in the healthcare sector in general, but wanted to look more carefully at the delivery side. Gary was curious to see if there would be differences in the learning curves when different organizations took on the same challenge at roughly the same time. I was interested in teams, and what helps teams learn. This was a situation where teams had to learn something. I’ve done previous research on adverse drug events and incidents of errors across different hospital units, and I found that differences in the social climate really made a difference in how comfortable people were in catching each other’s errors. Richard Bohmer, another colleague, is a physician, so he had a direct interest in contributing to knowledge about process improvement in healthcare. We all three came together in those partially overlapping areas of interest. The procedure that the teams were learning was a minimally invasive cardiac procedure. It was CABG, mitral valve repair, and other surgery done through a small incision, with a particular product that we agreed not to reveal at the outset of doing this study. The teams attended a course to learn how to operate this product? Right. They attended as teams, with the anesthesiologist, the nurses and usually two perfusionists. What was the interaction between the researchers and the surgical team? Were the teams aware of the reason they were being studied? Yes, they were. We went with one team to training. While we were there, we met a couple of other teams that were in training at the same time, and we ended up studying one of those teams as well. We shadowed one group for three days. We learned a lot from this experience about the technology and the challenge that these teams faced. Everyone knew why we were there when we interviewed them. We also had access to clinical data that the company collected from the hospitals. It held procedure times, patient co-morbidities and other controls that might contribute to procedure times, allowing a detailed analysis of the learning curves at each hospital. We were basically interested in: The slope of the learning curve; Whether there would be differences; How quickly the teams came down the learning curve. How did the reputation for excellence of the various institutions affect results? It did not. The sites were all well-respected cardiac centers, although some were truly famous and some were less so. Reputation did not have an effect, especially on the slope of procedure time reduction. I think that’s accounted for in part, but not all, by the fact that some centers were not as interested in procedure time, although everyone is interested in it to some extent for reasons of patient safety and cost. At least one place was training residents, which could contribute to longer procedure times. There were also differences in how quickly the teams brought new people in, which could make a difference. Similarly, how did the status of the physician within the institution itself (someone with clout versus perhaps a junior physician, for example) affect team performance? It did not. The distinction that made a difference was not status but rather whether and how the surgeon invited other team members to participate in a new kind of teamwork in the operating room. Your study reinforced the idea that cumulative experience is a significant predictor of learning but there is something important about the way the experience itself is structured. As conventional wisdom and previous research has suggested, the more you do, the better you get. (Similarly, the more you do, the faster you get.) But that’s old news. What’s new is that it was not the same slope for everybody. Some places learned faster than others. Those places that focused more on teamwork, that deliberately put together a team, learned faster. The composition of the team might look exactly the same across centers, but the communication of the composition was different. The surgeons told other team members (and I assumed, believed it) that they were picked for a reason say they’re good at training others, they’re very experienced, they have a good trusting relationship with them and the surgeon was putting together a special team for this effort. When the surgeons also took the time to do a careful dry run in advance of the first patient following training, that also made a difference. Teamwork wasn’t just teamwork it was fostered by some of these practices. Was the presence of the teamwork-fostering factors more dependent on the individual surgeon or on the culture at the various institutions? I think it’s a little of both. I think the individual surgeons count for probably 50% of the variance and the culture of the institution perhaps for the other half. Did you find that those physicians who were more adept at the process of learning through teamwork tended to be located at similarly-minded facilities? It was a little bit random. One of the best sites was a rural community hospital that had excellent cardiac service, but certainly would not be on the public’s radar screen like some of the more famous institutions in this study. There were surprising pockets of wonderful leadership, of people who knew how to lead a team or figured it out in the learning context as opposed to the routine delivery context. If you’re a staff member and your physician is not a leadership figure, what can you do? In the specific context of surgical teams, we did not see much in the way of evidence that individuals could take action, speak up or influence the powerful tide of leader direction. In theory, it’s possible to make attempts to speak up, suggest debrief sessions, ask more questions or challenge things and seek feedback, but we just didn’t see that happen. I think the status differences between surgeons, nurses and others is so great in this particular environment that it’s hard for people to do that. Is there any other research that you’re aware of that looks at learning in the medical field? There’s a fair amount of research that looks at learning. Certainly there’s research on the volume learning relationship. I wrote a paper several years ago on the results of a drug error study. Learning from mistakes is pretty challenging in the medical environment. Given how much is at stake, it’s still hard for people to speak up, ask for help, and challenge each other. Yet in some climates, in some units, they really manage to do that quite effectively, and I think a lot of it had to do with the leadership of the nurse managers. Is there any advice you would offer facilities whose teams are constantly having to learn something new new technologies, new procedures, etc.? Sure. I think a little bit of thought up front as to how to best launch a learning project is important. Take some time to talk about what the objectives are and what the goals are. Ensure that those with higher status who are involved in the project are open to hearing what others are thinking. The role and actions of the team leader are critical. A little bit of coaching and a little bit of thought go a long way, because people are so eager to do the right thing in the medical setting. It’s just that people are also eager to avoid violating status norms, so you really have to make it clear, over and over, that the invitation for a comments and feedback is genuine. In an article we wrote for the Harvard Business Review, we offer three guidelines for creating a learning team: 1. Design your team to learn.Choose the best possible individuals and don’t assume that any member can be interchanged with another simply because they have the same responsibilities or credentials. Recognize that different people have different strengths, and use them to your advantage. 2. Frame the learning experience as a challenge for the entire team. Acknowledge that the new procedure may be difficult and cause frustration for everyone. Emphasize the importance of each individual’s contribution in achieving success. Team leaders should also serve as a model, acknowledging their own frustrations and mistakes. 3. Foster an environment supportive of team member suggestions. Teams may run into a greater number of dead ends than individuals as they attempt to implement something new. All teams tended to emphasize the importance of experimentation for finding fast and efficient methods of working. In order for that to fully occur, team members must function in an environment where back-and-forth discussion is encouraged and they are not made to feel embarrassed or fearful of speaking out.

1. Pisano G, Bohmer R, Edmondson A. Organizational differences in rates of learning: Evidence from the adoption of minimally invasive cardiac surgery. Management Science, an INFORMS publication 2001;47(6):752-768.

2. Edmondson A, Bohmer R, Pisano G. Speeding Up Team Learning. Harvard Business Review 2001;79(9):125-132.

3. Gawane A. The Learning Curve. The New Yorker, January 28, 2002:52-61.