The beauty of checklists

I’ve just read The Checklist Manifesto, a lovely little book by Anul Gawande, a surgeon and New Yorker writer who wrote a notable piece last year about finding that more expensive healthcare wasn’t necessarily better healthcare.

The premise of the book is that a simple (but well devised) checklist is perhaps the best tool we have to make us consistently more successful at managing complex situations – whether that be operating on someone, choosing companies to invest in, landing a plane or building a skyscraper.

The airline industry in particular has learned to rely on pilot checklists – and a culture of always using those checklists – to maintain safety and their evolution makes for interesting reading. As Gawande notes, the reason no one died when an Airbus A320 was forced to land on the Hudson river in January 2009 was not only because of skill and professionalism heroism on the part of the pilot. It was because the pilot and crew all used the checklists provided for them for precisely that situation.

Gawande’s interest in checklists grew out of wanting to improve outcomes for surgical patients. He points to data showing that by 2004 surgeons around the world were performing 230 million major operations a year and estimates of complication rates ranged from 3 per cent to 17 per cent. “Worldwide, at least seven million people a year are left disabled and at least one million dead – a level of harm that approaches that of malaria, tuberculosis and other traditional public health concerns.”

The numbers were sufficiently alarming to prompt the World Health Organization (WHO) to ask him to spearhead a project to tackle the issue. The project group discussed all manner of interventions – guidelines, training, incentives – but ruled them out as too easily ignored, expensive or impractical.

Research showed that simple interventions were often the best in public health. Gawande cites as examples how removing the pump from an infected water well was enough to end a famous cholera outbreak in London in 1854, and how child mortality was reduced in poor districts of Karachi, Pakistan, by supplying free soap and instructions on how and when to wash hands – a kind of checklist.

A checklist had also proved useful in reducing infections at a US children’s hospital simply by ensuring antibiotics were administered at the right time.

So the group came up with a checklist and trialled it in eight countries, including here in New Zealand, at Auckland City Hospital. When I sat with my Dad at Auckland hospital recently, ahead of an exploratory surgery, the pre-op nurses worked through a set of forms and asked his name, date of birth and address, asked about allergies and breathlessness and so on – things that are on the WHO Surgical Safety Checklist.

The results of the trial were so positive WHO set about rolling the project out worldwide, and New Zealand is one of the countries to have signed up to encourage its use in every hospital. Not all have taken it up, but let’s hope they do soon. A ministerial review in NZ last year found that “44 thousand people admitted to hospital suffer an unintended injury caused in the management of their conditions, rather then the underlying disease – this is a similar rate to other countries. Although most of those people had relatively minor adverse events, about 15% resulted in permanent disability or death.”

In testing his theory about the checklist, Gawande visited building sites, the people who write the pilot checklists for Boeing, venture capitalists and top-flight restaurants. He found examples everywhere of simple checklists improving success rates, with the pilot checklists proving particularly instructive.

He learned, however, that not all checklists are created equal and he picked up tips along the way on writing a good one. Here are the main points:

  1. Identify pause points – the point in a process where the team needs to pause and use the checklist. In surgical checklists they came up with three: before the patient has anaesthesia, after anaesthesia but before incision, and after the operation but before the patient is wheeled out of the theatre.
  2. Decide whether it makes more sense to use a Do-Review or a Read-Do list. The first requires people to do their tasks then stop and review them before moving on. The other requires people to read the checklist and do each task in turn, like a recipe.
  3. Use simple, exact wording and use the language of the profession.
  4. Don’t make the checklist too long. If it’s too complex or takes too long people will give up and throw it away. Not everything has to be on the checklist. Just those things that are essential but that people can sometimes miss in the heat of the moment.
  5. Boeing found the best checklists have 6 to 9 points, are on one page that is sparing in colour and free of clutter, uses a mix of upper and lower case letters and a sans serif font. (Boeing use electronic checklists now too.)
  6. Find the right person in a given scenario to give the job of starting or overseeing the checklist. It may not be the most senior person. In some cases the checklist proved a tool for distributing responsiblity and power in a way that made people work as a team and unafraid to speak up.
  7. Buy-in is important and so is working as a team. As part of the surgical checklist operating staff are asked to introduce each other so they know one another’s names. Research had shown that people communicate better if they know each other’s names. Communicating well is as important as checking for potential infection agents.
  8. Checklists have to be road tested in the real world.
  9. Checklists are not static. They need to be adapted to individual situations and improved and updated over time.

While reading Checklist I started thinking about something that’s been irking me over at Whenever we push a set of changes to the website, I first preview them to check that everything’s working properly before we hit ‘deploy’.

But invariably over the following hours and days I will notice a slew of little problems – wrong or misspelled words, rogue spacing, things not uploading properly, bits of information not displaying right. Some of that’s to be expected, but it’s happened enough times now for me to think that I’m consistently missing things in the preview, which costs us time fiddling about later.

So I’m going to design myself a checklist for site updates and periodic maintenance checks. It’s got to be worth a shot.

I also started thinking about how checklists are or could be used in journalism. Craig Silverman of Regret the Error explores the subject here and offers a free Accuracy Checklist download here. I remember a journalist (woefully, not who) describing how he uses a spreadsheet to keep track of names, titles, quotes, facts and fact checking – essential for a big story covering multiple interviews over weeks or months.

I wonder if decision checklists might be useful too for making judgement calls in newsrooms. An example might be to help web editors decide on and execute value-added content such as video, slide shows and reader discussions. Another might be to help sniff out hoaxes. Food for thought.