The Checklist Manifesto

Years ago I was responsible for creating the technology to support a clinical study designed to test the feasibility of improving patient outcomes for patients with diabetes. The resulting article was published in Diabetes Care Volume 24 Number 6 in June 2001. The title was “A Systematic Approach to Risk Stratification and Intervention Within a Managed Care Environment Improves Diabetes Outcomes and Patient Satisfaction.” Despite the exceedingly long title, the program was simple enough. We did some assessments. We fed the data into a program that, in turn, applied the appropriate risk stratification and spit out the collection of interventions that appeared in the standing orders of the system as a patient-specific checklist. We had a nurse work with the primary care provider to ensure that the provider approved the patient-specific checklist that the system created.

The greatest problem for patients with diabetes is that they’re largely seen by primary care providers – the same people who are treating the common cold, and every other kind of malady that you can imagine. They don’t have time to keep up on the latest advances in diabetes management and there’s too much to remember to provide every patient with the best practice care they deserve on every visit. The system made it easy for these providers to be successful. As a result of the study, more than 45% of the high risk patients saw a significant (>0.5%) positive change in the primary measure of long term glucose health – HbA_1c. I think about this study from time-to-time with great fondness. We actually made a difference to the care of numerous patients.

When my bride, Terri Bogue, mentioned The Checklist Manifesto during one of our conversations about how to improve health care, I knew I had to read it. I knew that I had to read more about how my love of flying might be connected to my desire to continue to make an impact in the lives of patients. You see flying uses checklists all over – and health care for the most part does not.

Too Much Plane for Man to Fly

I learned to fly twice. The first time was ground school at Delta College by my friend and mentor, Ben Gibson. I was in high school and didn’t have an abundance of money, so I never finished the practical component of my training. I later revisited the topic of flying with my friend Brian Proffitt. He and I both wanted to fly and so we learned together – and both got our license. That was about 10 years after my first training in flying.

When I finally got to the point of my first solo flight, I was in awe of how much there is coming at you. While it doesn’t seem as overwhelming now, I can remember the terror running through my brain as I approached the runway for landing, knowing that the instructor wasn’t sitting next to me to make sure that I didn’t make a mistake. I was more focused on the details of flying than I’ve ever been since. And despite that, each time I went around the pattern I used the checklist in front of me. I checked flap position, fuel mixture, and throttle settings according to the checklist.

I didn’t know that it had taken a colossal failure in 1935 to create the spark behind the checklist that I held in my hands. It was October 30^th and Boeing had the winning design for a government contract. The Model 299 crashed during a test flight that was expected to be the final step in a purchasing contract. One newspaper man wrote that it was just “too much airplane for one man to fly.” That contract was awarded to a competitor however after a checklist was developed to lighten the cognitive load of flying the Model 299 aircraft, the US government bought a huge number of this venerable bomber – given the designation of B17. (I’ve mentioned cognitive load in my review of Efficiency in Learning.)

When I started flying the second time, I even made my own checklists based on the ones that come with the Cessna 172 that I was flying. The ones provided by Cessna were in a rather cumbersome book, so I created some sheets, laminated them, and bound them. I sold a few of them for a few dollars – but largely, I used them to make my world easier. I was working hard to make sure I didn’t forget anything and that I did the same thing every time.

Our Problems Aren’t the Same

In aviation the checklist was well proven. However, in medicine it hadn’t shown its worth. The key objection was that the problems weren’t the same in medicine as they were in flying. After all, flying a plane is simple by comparison to treating a human. However, it seems like both processes might have some similarities. The author, Atul Gawande, quotes Brenda Zimmerman of York University and Sholom Glouberman of the University of Toronto in the assessment that there are three types of problems:

Simple – Like baking a cake from a mix
Complicated – Like sending a rocket to the moon; they’re capable of decomposition into many smaller tasks
Complex – Like raising a child; there is no one way to do it right and doing the same things to two different children may (and frequently does) result in wildly different results

Readers of this blog may notice that we’ve called Complex problems Wicked problems in our review of Dialogue Mapping. These are problems in which defining the problem can be just as hard as coming up with a possible solution – which will invariably change the situation and, therefore, the problem – for better or worse.

The easiest problems to solve are those that are susceptible to “forcing functions”, that is to say that you can change the behavior by modifying one aspect of the solution. For instance, if there is a contaminated well, you prevent people from consuming water from it. If there’s a disease for which there is a vaccine you supply the vaccine to everyone. However, many problems in medicine aren’t the same. So could a humble checklist make a real impact on healthcare?

Gawande quotes that the result of his surgical checklist developed with the World Health Organization (WHO) resulted in 36% fewer major complications and 47% fewer deaths after surgery – these results were measured across eight hospitals with vastly different situations. That’s a statistically significant change for something as humble as a checklist. His results here are certainly the most compelling, but not the only instances where checklists have proven their worth in medicine. It seems as though checklists, done correctly, could make a significant impact on improving the care of patients.

Making a Good Checklist

So acknowledging that the aviation industry has been working with checklists in a more pervasive and effective way than anyone else, it makes sense to take a few pages from the aviation playbook for how to create good checklists. That’s just what Gawande and his team did with their surgical checklist. They consulted Boeing’s Daniel Boorman and learned the secrets to creating a good checklist – which I’ve summarized here:

Decide between READ-DO and DO-CONFIRM – You’ll need to decide whether you want the checklist to be one where the person is supposed to read the step and then do it – or do the step then use the checklist to confirm it’s done right
Create / Find “Pause Points” – You need a psychological anchor point to start the checklist; in aviation that can be a warning light but when it’s not, it should be clear what the anchor or starting point is (For instance, in flying there’s a pause point before you run the engine up.)
No more than 90 seconds – If the checklist takes more than 90 seconds to do, it’s too long, shorten it
Checklist the mistakes – You don’t need to include the things that everyone will do, create checklists for the things people could miss; manage the conflict between brevity and effectiveness by skipping the things people always get right
Don’t checklist the obscure – If something rarely happens, don’t put it on the checklist; the checklist supports remembering the frequently missed things, not the obscure
Create Communication Steps – The most effective thing can be the functioning of a team, encourage everyone to introduce themselves and foster an air of openness and teamwork as a part of the checklist
Test, Test, Test – Test the checklist however possible – Use simulators, collect data on real-life usage, plan to constantly refine the checklist

Checklists are powerful things. They can guide behavior and keep people from making the simple cognitive slips that they’re likely to make. (See Thinking: Fast and Slow.)

Changing Behaviors

Kurt Lewin (who I’ve spoken about before in Nine Keys to SharePoint Success, Who Am I?, and Beyond Boundaries) said that behavior is a function of both person and environment. In other words, careful construction of the environment has the capability to have a profound impact on behavior. Some of the most innocuous changes can have a profound impact. (See Switch about how solution size and impact are not related and Diffusion of Innovations on how the impacts are not predictable.) Gawande relates one very powerful way that behavior was changed for positive effect.

Even in the most impoverished places, soap is generally available. People can – and do – wash their hands. However, in the impoverished areas there are factors – like cost – that impact how much people use soap. In a study of the ability for soap to reduce disease and suffering, he arranged for Safeguard soap – with and without an antibacterial additive – to be given out to poor families. The impact was that the incidence of diarrhea among children fell 52%. However, the crazy thing wasn’t that using soap improved health. The crazy thing was that the families already had soap. In fact, they used approximately 2 bars of soap per week – the distribution of Safeguard averaged only 3.3 bars of soap. It wasn’t the addition of soap that made the change, it was something else.

The something else, in Gawande’s assessment, was the instruction and systemization of the hand washing. It wasn’t that they didn’t have soap. It was that their use of soap wasn’t sufficient. By instructing families of when to use soap – and eliminating the economic barrier to using it when it was effective – they changed behavior. They changed the behavior of families so that they were washing with soap when they touched anything that was probably infected (such as human waste) – and again before they touched anything that would be substantially harmful if infected (such as food preparation.) The impact was substantial however not because the intervention was monumental – like a new plumbing and sewer system for the city. It was impactful because it changed the behaviors that were harmful.

Sometimes the key to change behavior is in deciding who is responsible for the change. For instance, in a two-pilot cockpit, the pilot not flying (which may or may not be the first officer) is the person walking through the checklist. They’re responsible for making sure that every line of the checklist is “checked off.” That’s because the cognitive load of the person flying the plane can be significant. Significant enough, in fact, that they’ll try to skip or skimp on the checks in the checklist. In these situations the “power” of the situation is shared between both pilots – the pilot flying is responsible to the “power” of the checklist. This is a subtle change in reassigning responsibilities to both pilots and aligning both of them to ensuring that the critical checks are completed.

Gawande discusses how checklists change the power in the operating room. A simple metal tent was placed over the scalpel in one set of operating rooms to serve as a physical reminder to ensure that the checklist was completed – and the nurse who hands the instruments to the surgeon was thereby empowered to own responsibility that the checklist was completed. This is a substantial shift in power where the surgeon is the most powerful figure in the room and the rest of the team are there to support him or her. Empowering a nurse to stop the surgery (by not providing the instruments) sends a strong message that the checklist must be followed.

Gawande does caution about forcing physicians to use the process. Some will refuse. He correctly states that forcing them to adhere to the checklist will build resistance and that it can be bad to the overall program. The participants will silently disengage and go through the motions and will quietly and subtly subvert the system. However, those that are willing to at least try a new approach will open up their surgical team and engage them – igniting the possibility for teamwork and for exceptional results. That’s what we expect out of professionals although we don’t always get it.

Professional Responsibilities

I think that an appropriate closing to the book, and to this review, is to talk about professional responsibility. The Checklist Manifesto ends with three components to a code of conduct that all professional occupations have – plus one more that was unique to aviation. Here are the three common ones:

Selflessness – We expect that professionals operate in their clients’ best interests – not their own
Skill – We expect that a professional has the necessary skills to do their work
Trustworthiness – We expect that professionals are responsible and worthy of the trust we place in them

The final characteristic – one that is admittedly very difficult – is discipline. That is, following the prudent procedure consistently. Discipline is hard on all levels. It’s hard to stick with something, even when we know we should do it, when it’s not easy.

That brings me to my last flying story. (If you’re from the FAA reading this, my brother was a certified flight instructor for instruments and we were on an instrument flight plan but I was the “pilot flying.”) As a private pilot, you don’t get much (if any) actual time in a cloud. I had quietly skipped a step on the checklist to turn on pitot heat. A pitot tube is used to drive the airspeed indicator. Because of the Bernoulli effect, sometimes ice can form in the tube and render it ineffective. I’d never had this happen and after a few minutes of flight – in the clouds – my airspeed dropped to zero with no change in thrust or angle of attack. I was noticeably concerned when my brother told me that I had missed turning on the pitot heat. (He later confessed that he had seen the miss earlier in the flight, but decided to let it go because it was educational for me to go through the problem.)

However, that wasn’t the last problem in the flight. In pilot training, you’re absolutely taught about vertigo and about trusting the instruments. On this day, I followed the right procedures and took off into the clouds, paying particular attention to airspeed, angle of attack, altitude, and track. That’s when I started feeling it. It felt like my artificial horizon was wrong. It showed level on the dials, but I felt like I was turning. I cross-checked with the turn-bank indicator which indicated level. I watched for a change in the directional gyro and even glanced up at the compass which was doing its bobble-head movement but it wasn’t turning.

Now I had a problem. I knew that we were straight (and climbing slightly) but I also knew that it didn’t feel right. I can’t exactly describe the feeling in a way that makes sense. I felt as apparently John Kennedy Jr. must have felt. I had vertigo and I had to transfer my trust to my training. I had to trust the instruments that I knew were fallible (although not generally all at once.) Luckily, my brother was sitting with me and he coached me to turn the strobe light off – which was reflecting off the clouds and destroying my night vision anyway. He also patiently reminded me that I knew what this was and I had to follow my training and do what the instruments (collectively) told me.

An eternity or a minute later, depending upon your point of view, my sense of balance was restored and we completed the flight without any incident.

The message, for me, was that I had to trust the checklist and my instruments – even when my gut was telling me something different. I was taught how to rely on things outside of myself to improve my effectiveness. I didn’t need to be a better pilot. I needed to have more discipline to use the tools I was given. I needed to really believe The Checklist Manifesto.