Saturday, April 28, 2012

The Checklist Manifesto, by Atul Gawande

Book cover

I think that The Checklist Manifesto is a book that every technical professional should read. It is simple to read, to the point and extremely useful. I first heard about it in a Scrum training and now, after reading it, I think it was the best thing that came out of it (and it was a pretty awesome training session). What is this book about, then? It is about a surgeon that researches the way a simple checklist can improve the daily routine in a multitude of domains, but mainly, of course, in surgery. And the results are astounding: a two fold reduction in operating room accidents and/or postoperatory infections and complications. Atul Gawande does not stop there, though, he uses examples from other fields to bring his point around, focusing a lot on the one that introduced the wide spread use of checklists: aviation.

There is a lot to learn from this book. I couldn't help always comparing what the author had to say about surgery with the job I am doing, software development, and with the Scrum system we are currently employing. I think that, given he would have heard of Scrum and the industrial management processes it evolved from, Gawande would have surely talked about it in the book. There is no technical field that could not benefit from this, including things like playing chess or one's daily routine. The main idea of the book is that checklists take care of the simple, dumb things that we have to do, in order to unclutter our brain for the complex and intuitive work. It enables self discipline and allows for unexpected increases in efficiency. I am certainly considering using in my own life some of the knowledge I gained, and not only at the workplace.

What I could skim from the book, things that I marked as worthy to remember:
  • Do not punish mistakes, instead give more chances to experience and learning - this is paramount to any analytical process. The purpose is not to kill the host, but to help it adapt to the disease. Own your mistakes, analyse them, learn from them.
  • Decentralize control - let professionals assume responsibility and handle their own jobs as they know best. Dictating every action from the top puts enormous pressure on few people that cannot possibly know everything and react with enough speed to the unpredictable
  • Communication is paramount in managing complex and unexpected situations, while things like checklists can take care of simple and necessary things - this is the main idea of the book, enabling creativity and intuition by checking off the routine stuff
  • A process can help by only changing behaviour - Gawande gives an example where soap was freely given to people, together with instructions on how and when to use it. It had significant beneficial effects on people, not because of the soap per sé, but because it changed behaviour. They were already buying and using soap, but the routine and discipline of soap use was the most important result
  • Team huddles - like in some American sports, when a team is trying to achieve a result, they need to communicate well. One of the important checks for all the lists in the book was a discussion between all team members describing what they are about to do. Equally important is communicating during the task, but also at the end, where conclusions can be drawn and outcomes discussed
  • Checklists can be bad - a good checklist is precise, to the point, easy to use. A long and verbose list can impede people from their task, rather than help them, while vague items in the lists cause more harm than good
  • A very important part of using a checklist system is to clearly define pause points - they are the moments at which people take the list and check things from it. An undefined or vaguely defined pause point is just as bad as useless checklist items
  • Checklists are of two flavours - READ-DO, like a food recipe, with clear actions that must be performed in order, and DO-CONFIRM, where people stop to see what was accomplished and what is left to do, like a shopping list
  • A good checklist should optimally have between five and nine items - the number of items the human brain can easily remember. This is not a strong rule, but it does help
  • Investigate failures - there is no other way to adapt
  • A checklist gotcha is the translation - people might make an effort to make a checklist do wonders in a certain context, only to find that translating it to other cultures is very difficult and prone to errors. A checklist is itself subject to failure investigation and adaptation
  • Lobbying and greed are hurting us - a particularly emotional bit of the book is a small rant in which the author describes how people would have jumped on a pill or an expensive surgical device that would have brought the same great results as checklists, only to observe that people are less interested in something easy to copy, distribute and that doesn't bring benefits to anyone except the patients. That was a painful lesson
  • The star test pilot is dead - there was a time when crazy brave test pilots would risk their lives to test airplanes. The checklist method has removed the need for unnecessary risks and slowly removed the danger and complexity in the test pilot work, thus destroying the mythos. That also reduced the number of useless deaths significantly.
  • The financial investors that behave most like airline captains are the most successful - they balance their own greed or need for excitement with carefully crafted checklists, enabling their "guts" with the certainty that small details were not missed or ignored for reasons of wishful thinking
  • The Hudson river hero(es) - an interesting point was made when describing the Hudson river airplane crash. Even if the crew worked perfectly with each other, keeping their calm in the face of both engines suddenly stopping, calming and preparing the passengers, carefully checking things off their lists and completing each other's tasks, the media pulled hard to make only the pilot a hero. Surely he denied it every time and said that it was a crew effort because he was modest. Clearly he had everything under control. That did not happen and it also explains why the checklist is so effective and yet so few people actually employ it. We dream of something else
  • We are not built for discipline - that is why discipline is something that enables itself. It takes a little discipline to become more disciplined. A checklist ensures a kind of formal discipline in cases previously analysed by yourself. It assumes control over the emotional need for risk and excitement.
  • Optimize the system, not the parts - it is always the best choice to look at something as a whole and improve it as a whole. The author mentions an experiment of building a car from the best parts, taken from different companies. The result was a junk car that was not very good. The way the parts interact with one another is often more important than individual performance

I am ending this review with the two YouTube videos on how to use and how not to use the WHO Surgical Safety Checklist that Atul Gawande created for surgical team all around the globe.


Carole said...

I was wondering whether you would be happy to put up a link in my brand new monthly series called “Books You Love”. The idea is for people to link up posts about a book they loved – it doesn’t have to be one they just posted about. It could be an old fave. I am hoping we will end up with a nice collection of books that can go on our reading lists. Here is the link Books You Loved May Edition

Siderite said...

Carole, I've used and abused the opportunity you gave me. Thank you for thinking of me.