This book was referenced by so many of the other books I've read. Atul Gawande is a surgeon and prolific author.
The simple introduction of a humble checklist (and system to ensure its use) into John's Hopkins Medical Center was enough to save dozens of lives and millions of dollars over the course of the first year it was introduced. The post surgery Infection rate went from a few percent to zero.
There are essentially two types of failures: those that come about through ignorance (where we don't know the right course of action to take), and those that come about due to ineptitude (where we know what to do, we just failed to do it). As the collective knowledge of humanity expands, the proportion of failures caused by ineptitude is quickly dwarfing failures caused by ignorance.
As humanity's collective knowledge grows, it is no longer feasible to be an expert in all arenas. Doctors used to just be doctors. They would diagnose the sick, do surgeries, and perform therapies all from their one little office building. Now there aren't really many “general practitioners” in hospitals - and general practitioners often just serve as the gateway to get the patient to the appropriate specialists. What used to be considered “specialized medicine” isn't even that specialized any more. You're not just a surgeon, you're an oncologist, and not just an oncologist, but an oncologist who specialized into pediatric oncology. We have gone from generalists and specialists to specialists and ultra specialists. Not just in medicine, either. It used to be possible to be a Renaissance man and know literally everything there was to know about math, physics, astronomy, chemistry, and engineering. Now our chemical engineers have specialized into petroleum engineering versus biomedical engineering versus materials engineering.
When the military went to see the Boeing B52 for its test flight, the test flight ended in a crash that killed several crewman. The problem was that the pilot forgot a step in the the new, bigger, more complicated airplane's more complicated start up sequence. Rather than insist in more reasoning, they fixed the problem by coming up with the concept of a preflight checklist.
When constructing skyscrapers, the companies that organize and carry out their build utilize massively detailed schedules, which are effectively just checklist of things that need to happen in the right sequence.... this makes sure all the little, obvious things go right... but they also utilize communications checklists that specify when certain conversations between certain groups with specialized scope and expertise should talk about specific topics. This type of checklist ensures the right conversations are happening. It is at a higher level than the “get the little steps right” checklist, but every bit as important. Most failures in endeavors of more than one person are from poor communication.
Have clear streaking points, even the checklist should be used.
Are either “do confirm” checklist or “read do” checklist (like a recipe)
Keep it to 5-9 items. Don't be too strict on this.
Keep the checklist to 60 to 90 seconds to run through. Focus on the Killer items, those vital steps that sometimes get missed
Wording should be exact and use the language of the profession. No I'm should be vague, but it shouldn't be overly prescriptive.
A checklist should fit on one page and not be cluttered with extra imagery or colors.
Checklists need to be tested, and need to be revised as they are run and things ate learned.