Checking research on checklists
17 September 2019Checklists have become a routine part of modern life. In simple form, as to-do lists and shopping lists, they help us all keep track of the overlapping and numerous tasks of our day.
They have far more significant impacts in surgical wards and on construction job sites around the world. In the lifting industry, they are often promoted by safety experts as a key way to make sure everyone leaves the jobsite as healthy as when they started the day.
But it's in the medical field that checklists have undergone the most rigorous, academic checking. A recent study exposes some of the problems checklists cannot solve on their own.
A pioneer in the scientific study of checklists is Atul Gawande. Author of 'The Checklist Manifesto', published in 2009, Gawande conducted research in 2007 that saw him develop a checklist of steps to ensure that surgical teams didn't miss important steps: from handwashing through to ensuring that tools are removed from patients' bodies. Before key stages in surgery, checklists make sure everyone on the team is in the right place, with the right tools. When these lists were tested at eight hospitals around the world, the results were stark: mortality rates were nearly halved, falling from 1.5% to 0.8%. In other words, for every 1,000 surgeries at a hospital using a checklist, seven extra people would live.
Part of the argument for checklists is that they impose best practice on everyone: the bowtied surgical genius and the hardhatted veteran crane operator, the junior nurse and the signaller on their first job. Whether their challenge is a lack of experience, or the over-confidence that can come from too much experience, everyone who follows the checklist, follows the process.
But Gawande's later work, and that of others, shows this is not always the case. In a 2014 study in India, his team conducted trials at 60 Indian maternity facilities, involving 160,000 women. While birth attendants who were coached to use the checklists performed the correct steps far more than those who weren't, mortality levels remained the same.
A study published this summer in JAMA Network Open, by James Suliburk, Quentin Buck, and Chris Pirko, suggests some of the issues with checklists, and a possible solution. The research looked into ways to categorise 'human performance deficiencies'. They found that more than half of adverse outcomes in hospitals they studied were due to human error, and categorised more than half of these as 'cognitive error'. Discussing their results, they note the importance of errors of recognition: failing to spot a risk. Describing growing research into 'checklist burnout', they suggest that the use of systembased approaches should be accompanied by cognitive training. Reflecting practice in the aviation industry, they recommend 'simulated playback' of incidents. Perhaps a similar approach, working through the steps of an incident, would be a useful tool in the construction sector, to highlight how risks can be missed.