National Alliance for Physician Competence Discovery Workshop


Complexity in HealthCare

Curt Lindberg

Plexus Institute works with people and organizations to understand the science of complexity and how it applies to the work they are doing. We are working with the healthcare community to improve the use of technology (like EMR) to improve patient outcomes. They have asked three firms to provide them with some guidance – two traditional consulting firms and us. They wanted very traditional guidance around governance, measurement, etc. This aligns well with the classical view of organizations – that they are predictable and controllable.

The challenge that this organization is facing, however, is fundamentally unknowable – the development of the technologies, the acceptance by the healthcare community, and the impact on patient outcomes. This is a complex system – it cannot be predicted and it cannot be controlled. But it can be influenced.

Organizations are complex adaptive systems. They are comprised of individual agents who are diverse and who learn. What is most interesting about these systems is the interactions among agents – the focus is local, but the patterns of behavior and output are nonlinear and unpredictable. In these nonlinear systems, it is often very small inputs or interactions that create massive effects throughout the system. Effective strategies focus, therefore, on these interactions and their focus is very small. Relationships are critical.

Effective strategies focus, therefore, on these interactions and their focus is very small. Relationships are critical.

If you give up the notion that the future is predictable and controllable, what are you supposed to do? In complex systems, you should pay close attention to what is going on and try to make sense of it. Draw on a diverse group of people with a wide variety of perspectives. The primary role of management is to learn and to make sense of the organization and the world.

This is a very creative art – very much like a jazz band. This is very improvisational. Use the building blocks that you already have in place to move your organization in the direction that you would like. All organizations are unique. It can be very damaging if you do not respect the history and norms of an organization. Try some experiments and see how they go, and then try something else.

There is a growing body of research about complexity in health care. Ruth Anderson at Duke is looking at the quality of care in nursing homes. When the families, nurses and physicians interact better, there are fewer falls, fewer pressure ulcers, etc. The outcomes are better when these groups work together to make sense of the patient’s condition.

What is the nature of networks in your organization and your field? We mapped the network of a hospital to understand how they are working together to prevent the transmission of drug-resistance microbes. By using a process called positive deviance, we have massively increased the connectivity within the network, and we have helped the network connect to more outside expertise. When you think about these organizations, look at the structure of the networks.

In addition to the structure, look carefully at the processes that are used to help the networks interact with each other. Help them have different kinds of conversations. You can use techniques like open space, positive deviance, appreciative inquiry, improvisation, conversation café, and small-group discussions.

The client asked us to present all of our findings in a 15-minute presentation with five minutes for questions. We explained to them that the outcomes would be far richer with greater interactions among the consultants, staff, decision-makers, and other stakeholders. The structure of their meetings is not conducive to learning, interaction and sense-making. As you think about structure, focus too on the processes of interaction.

An open space meeting would be convened around a particular topic. The meeting begins with the group standing in a circle. The entire group is invited to recommend topics for the group to work on. Each person who proposes an idea takes responsibility for setting up a small working group and facilitating the creation of a deliverable. This is a very self-organizing process that is very effective at engaging people.

The better strategy is to try a lot of different, little experiments. Make some sense of those experiments, and then try some more.

If you cannot predict the future, how do you begin? Traditionally, we spend a lot of time developing a single, big vision, and then we spend a lot of energy getting everyone to buy into the plan. If you cannot predict the future, complexity science tells us that this is a very risky strategy. The better strategy is to try a lot of different, little experiments. Make some sense of those experiments, and then try some more. The direction of the movement emerges from the results of these experiments, not from a big plan of some sort.

The traditional kind of planning process was not working at a big hospital. The nursing staff tried a different approach. They began to have a lot of different conversations. They agreed that nurses could spend one day every two weeks working on something that they felt passionate about to improve the health of the community. They were also allocated budget. Within two months, there were thirty projects underway. Many of them failed, but several of them were wildly successful. Make connections. Start small. Experiment. See what works and build on it.

Dr. Crabtree in New Jersey has been working to improve outcomes for very common problems. They have used some very non-traditional techniques – they have focused on improving the interactions among the providers within the practices. Better interactions yield better patient outcomes.

I will share with you a bibliography of a lot of the research that is going on related to health care.

Your focus is on physician competence. It is important to recognize that physicians work within a variety of other complex adaptive systems. Nurses are probably doing more than any other medical discipline in terms of incorporating complexity science in their training.


Q: We have tried two different approaches to large scale change. We invited programs to develop different approaches to teaching the competencies. 20% of the programs became very engaged, but now most of them want to be told what to do. We have shifted from interactive to prescriptive.

With a limitation of duty hours, we went the opposite way. We dictated that duty hours be limited, no matter what the outcomes were. Now we are starting to work with several of these programs to develop some small experiments to see how it might work.

A: Look very carefully at the complexity of the challenge. Simple challenges can be addressed with simple, dictated solutions. More complex challenges require more interactive approaches. Some organizations resist taking on the additional burden of experimentation, but by trying to impose a one-size-fits-all solution, they may actually increase their burden even further to deal with the unintended consequences.

Q: It seems as though many populations that we deal with are not prepared to approach projects in this way. Most people seem very content with the status quo.

A: Look for the early adopters. Some people will be more disposed to making changes like this. Find a few early adopters, and launch a few small experiments. Build consensus more slowly by building on success. It may be helpful to present this idea of change as a low-risk proposition – a small number of experiments and a lot of them will fail. We need order for this approach to work, but we also need tension with greater innovation and change. The attitude should be very playful.

Stories drive legislation. We must be very careful to craft the story around this experimentation – a failed experiment for us could result in dead bodies.

Top of Page