During my recent stint covering the Neuro ICU I noticed for the first time a checklist posted above each patient bed. The checklist covered the steps to undergo whenever performing an invasive procedure on the patient. I was glad to see that the checklist phenomenon had penetrated my hospital, although the implementation of safety checklists is far from complete.
A recent study published in the BMJ offers support for the efficacy of using checklists to reduce complications and improve patient outcomes. This is a retrospective study looking at mortality and length of stay in Michigan area ICUs, comparing those that had implemented the Michigan Keystone ICU project (including a safety checklist for the placement of central lines) with local ICUs that had not implemented the project. They found a 10% decrease in overall mortality, but the results were not significant for length of stay. Because this was a retrospective study it was not designed to prove cause and effect, but it is highly suggestive of the efficacy of implementing such checklists.
The checklist trend represents a culture change within medicine – and a good one. This change received its greatest boost with the publication of The Checklist Manifesto by Dr. Atul Gawande. He presents a compelling case for the need and efficacy of using checklists in order to minimize error.
He argues that historically medicine has had a culture of quality control through individual excellence and training. This culture still pervades medicine. Each year, for example, I have to go through a long list of safety and other training – the standard response of the powers that be is to institute a new training and certification program for each new regulation or identified safety issue. Training is good, but increasingly there is recognition that it is not adequate.
The problem, Dr. Gawande points out, is that there are areas of our complex civilization that are too complex for mere humans to adequately master. Or you can look at it from the perspective of minimizing error. Training to deal with a complex system can only minimize error to a certain degree. There are inherent human limitations of memory, attention, and consistency that mean that error will be inevitable. In situations where minor mental errors can have catastrophic consequences (like flying planes or performing major surgery) relying on training alone is folly. In such situations the implementation of a simple checklist can significantly reduce error far below what training alone can. It is a lot easier to remember to follow the checklist than to remember each item on the checklist.
There is no question that medicine is a high stakes and complex game. While I am a strong advocate of science-based medicine, we have to recognize its limitations. The opportunities for catastrophic error in medicine are enormous – from prescribing the wrong medication or dose, to forgetting important steps in a complex procedure, to removing the wrong limb. Even minor errors or oversights can have extreme consequences.
In medicine the overarching consideration of any intervention is risk vs benefit. We only use interventions that have potential benefit that is greater than the potential risk (while also understanding that our information is probability-based and imperfect). Often our knowledge is based upon clinical trials which are highly controlled, and therefore do not have the same risk of error that is likely to exist when implemented in the “real” world outside of a clinical trial. In any case minimizing error is key to minimizing risk and optimizing the risk-benefit ratio of medical interventions.
It seems that we have pushed the limits of training. Medicine has become highly technical, specialized, and complex. While extensive training is necessary, it is no longer sufficient to minimize risk. We are now entering the age of the checklist. This is a simple procedure that can significantly improve human performance. The latest study is further evidence in support of this. A 10% reduction in mortality is highly significant.
A related phenomenon, in my opinion, is the movement toward a team approach to patient care, especially in highly complex cases. There is increasing recognition that group intelligence can vastly outperform individual intelligence, and that a group can be smarter than even its smartest member. Complex or high risk cases can benefit from a team of experts, especially with a variety of specialties, collaborating on care. This is nothing new in medicine – tumor boards and multi-disciplinary clinics have been around for decades. But there is movement toward greater reliance on teams than on individual experts.
This is related to the checklist phenomenon in that both trends represent a movement away from over-reliance on the individual and training to minimize error and maximize performance. Both recognize the crushing complexity of modern medicine, and the need to be humble before this complexity.
To broaden the context further, I think these phenomena represent increasing recognition that we need to pay attention in medicine to how our knowledge is implemented, not just to the acquisition of greater knowledge. Pushing the limits of medical knowledge is, of course, incredibly important. But we also have to pay attention to how that knowledge is disseminated, how it is received by the public, how it affects regulation, and how it is implemented by systems and by individuals. We also need better understanding of these processes – we need increased medical meta-knowledge. We need to learn how to deal with this vast body of scientific information we are rapidly accumulating.