First, Gawande describes the disparate organizational approaches to the Katrina crisis in New Orleans. He concludes that organizations are prone to break down in a crisis unless their management cedes control to the most far-flung workers. Rigid federal and state authorities froze up, as did many local authorities, but police and firefighters accepted and coordinated the help of hundreds of small boat owners to conduct unorthodox rescues of tens of thousands of residents. Many private businesses flailed in the complex aftermath of the disastrous storm, but WalMart shone. Its CEO announced:
This company will respond to the level of this disaster. A lot of you are going to have to make decisions above your level. Make the best decision that you can with the information that's available to you at the time, and, above all, do the right thing.Wal-Mart had to close 126 stores; 20,000 employees and family members were displaced. But within 48 hours, more than half of the stores were open again and able to consider, "Oh, my God, what can we do to help these people?" On their own authority, managers began distributing diapers, water, baby formula, and ice to storm victims. They improvised crude paper-slip credit systems for first responders at a time when FEMA was still paralyzed. One assistant manager went through her badly damaged store with a bulldozer, salvaging what she could and giving it away in the parking lot. She also broke into the store's pharmacy in response to a local hospital's call for help. Senior Wal-Mart officials, rather than micromanaging or second-guessing these efforts, concentrated on coordinating with line employees and state agencies to meet needs as they arose. What has this got to do with checklists? I'm not sure. Gawande is struggling toward an approach to unpredictable complexity that avoids stultifying central control without accepting anarchy as its alternative. He approves of "a seemingly contradictory mix of freedom and expectation -- expectation to coordinate, for example, and also to measure progress toward common goals." He thinks this approach can be codified into simple checklists.
The second anecdote concerned a CDC worker in Karachi who tried an inexpensive, low-tech solution to infectious disease outbreaks. Procter + Gamble wanted to prove the value of a new antibacterial soap. The CDC worker got a grant to study three groups: one supplied with ordinary P+G soap, one with the new antibacterial soap, and one left to its own devices for soap. Test subjects were encouraged to use the soap in six specific situations, such as just before feeding an infant. After a year, various infectious diseases dropped by between 35% and 52% in both groups using the P+G soap. The study was a failure in P+G's eyes, because the new antibacterial soap conferred no noticeable advantage, but it showed the CDC how a simple routine could combat persistent public health problems. The soap, Gawande says, was a "behavior-change delivery vehicle." It came with instructions for its use in six separate situations. The households already were using soap at the rate of two bars per week on average, but the study apparently caused them to use it more systematically and, because it was pleasant-smelling and well-lathering, more enthusiastically. "Global multinational corporations," as the CDC worker noted, "are really focused on having a good consumer experience, which sometimes public health people are not." What's more, the people enjoyed receiving the soap: "The public health field-workers were bringing them a gift rather than wagging a finger."
Getting back somewhat more convincingly to the "checklist" theme, Gawande interviewed a successful investor who put his colleagues into six categories according to how they evaluated the entrepreneurs who were seeking their venture capital. "Art Critics" assess entrepreneurs at a glance on the basis of intuition and long experience. "Sponges" gather information exhaustively, then go with their gut. "Prosecutors" challenge entrepreneurs with hard questions about how they would handle hypothetical situations. "Suitors" woo entrepreneurs rather than evaluate them. "Terminators" make a superficial initial choice and plan to fire and replace incompetents ruthlessly later. Finally, "Airline Captains" take a methodical, checklist-driven approach, consciously overriding their intuition. Gawande reports that the final category outperformed the rest dramatically, though they made up a small minority of the whole, which was dominated by Art Critics and Sponges. Why were the Air Captains so rare? Gawande muses that something in us makes checklists seem like buzzkills, like an abandonment of romantic ideals of competence. When his own experimental surgical checklist project for the World Health Organization showed impressive gains in reducing complications from infection and errors, he nevertheless felt a personal reluctance to implement them for his own surgical team: "[I]f I told the truth -- did I think the checklist would make much of a difference in my cases? No. In my cases? Please." It did, though, and it strengthened his conviction that modern humans are engaging in complex cooperative tasks that require a new approach to discipline and focus than comes naturally to us.
I have used the checklist technique several times in my career, specifically when I knew I was in over my head. I kept a steno notebook, and wrote down my notes about what I needed to do. I have never found a shortage of people ready and willing to tell me what I needed to do to accomplish any task I set. I did this when I got married, when I had a new job, and during each move. I also used it when I had a construction project on my house.
ReplyDeleteI would take notes every time somebody told me what I needed to do, and then I would re-read my old notes, crossing out the things that I'd taken care of, and repeating the ones that were left over.
It keeps you from forgetting the obvious, and it works like a charm.
Valerie
Senior Wal-Mart officials, rather than micromanaging or second-guessing these efforts, concentrated on coordinating with line employees and state agencies to meet needs as they arose. What has this got to do with checklists? I'm not sure.
ReplyDeleteI would say that to me, this says that while checklists are completely reasonable and useful in day to day operations, for emergency situations, it is better to let folks get about the business of emergency management rather than getting bogged down in procedure. Perhaps that's not the author's point, but that's what I got out of that anecdote.
The second example was more in line of what his actual point seemed to be. That following a checklist system (in this case, use the soap before doing these things, like feeding the baby) can change behaviors for the better, even if that wasn't the original goal of P&G. It's a cute story from the standpoint of "Multinational Corp does good without actually planning on it", and I did like the concept that by giving the soap as a "gift" the study participants were more likely to use it than if they were being scolded to have better hygiene. And it's another good point that the pleasant smelling soap made them more likely to use it has very powerful implications for health initiatives across the globe. People are sensory based animals, and we like things that are pleasant. That it should come as a surprise to anyone is actually surprising.
I agree that the Walmart example stands athwart the main point that checklists improve how we handle things, but everything else you mentioned makes sense to me, at least.
"He concludes that organizations are prone to break down in a crisis unless their management cedes control to the most far-flung workers."
ReplyDeleteSee Dietrich Doerner's interesting book The Logic of Failure, which I reviewed here:
http://photonplaza.blogspot.com/2003_12_28_photonplaza_archive.html#107275628844222087
MikeD, I agree that it's hard to fit the WalMart story into his thesis. I enjoy his writing, but it typically suffers from this fault. He's got great stories to tell, but is struggling to fit them into a larger conceptual framework and can't quite get there. He ended the WalMart story with his ideas about the struggle to reconcile freedom and anarchy in an organization struggling with a complex emergency. He told other stories about how checklists improved function in complex emergencies, but they didn't really address that tension between freedom and anarchy. In the sections from which I didn't try to quote, he did play around with the notion that checklists had the unexpected tendency to foster teamwork and communication. At one point he even seems to be arguing that the necessary tension between autonomy and communication among "cogs in the machine" can be boiled down in a checklist, but this is the least satisfying part of the book.
ReplyDeleteHe makes a great case for checklists' working, but not for why.
David, thank you, I'm going to order "The Logic of Failure" right away. Gawande talks a bit about how easy it is for checklists to degenerate into the kind of failure you wrote about in your review. He's perceiving a way for experts in a complex system to boil down some of the most routine, repeated errors into a short list to be reviewed at critical moments. The pre-op checklist can remind doctors, for instance, that prophylactic antibiotics must be given between one hour and one minute before the first incision in order to be effective, but patients often have been left waiting in the hall while an earlier surgical emergency has been finished up. It was remarkable how often a pre-incision checklist led even very experienced and diligent teams to realize they needed to give a new antibiotic injection and wait one more minute before starting the surgery. The trick was to boil down the common errors to the half-dozen or so that were mostly likely to be overlooked and the most devastating if not corrected. For whatever reason, aircraft pilots got a headstart on this checklist approach, and others had to learn from them.
ReplyDeletePS to David -- I hope you're willing to take full responsibility for the flood of books I've just downloaded! Your old website "Photon Plaza" is full of interesting book reviews. Luckily I didn't have anything planned for the next week or two.
ReplyDelete...hope you enjoy 'em all, Tex!
ReplyDeleteChecklists are essential, but they can rarely provide full solutions to a problem. Consider Capt Sullenberger's landing on the Hudson. His aircraft certainly had checklists for engine failure and for water landing, but when it came to the decision whether to land on the river or try to make Teterboro, he had to consider numerous factors, including altitude, wind, aircraft gliding capability, river conditions and water traffic, aircraft ability to float long enough to make passenger rescue likely, ATC's ability to clear a runway at Teterboro quickly enough that he could have landed without hitting someone else, and the historical success rate of those who have attempted to stretch glides in marginal conditions (which isn't great.)
In such situations, checklists cannot substitute for trained and knowledgeable intuition, coupled with advanced thinking about things that could go wrong and their remedies.
I wondered, too, how Gawande really thought the Sullenberger episode fit in with his concept of checklists. All he really seemed to be saying was that the whole aircraft team functioned well as a team, and that checklists sometimes foster teamwork. Another example of a great anecdote about success that didn't quite fit into his overall theme.
ReplyDeleteSAC was really big on checklists--they allowed rapid training, and prompt and error free execution of complex and/or lengthy tasks that had to be done quickly.
ReplyDeleteBut the USAF also had a saying to the effect that when the emergency arose, it was too late to consult the checklist. Roughly 95% of the time in our simulators was devoted to practicing emergency procedures. The point was to reduce those particular procedures to eyeball and muscle memory. Sullenberger and his crew benefited from that kind of training. They also benefited from executing the preflight, before takeoff, takeoff, before landing, approach, and landing checklists. Those make sure that important steps are all executed in their proper order, and where appropriate, things are set up in a standard way--which if nothing else, provides a known start condition for emergencies.
They work that way in business, too, only they're called (in AT&T, anyway) M&P--Methods and Procedures. Walmart's performance after Katrina, though, was more a matter of leadership trusting their staff than execution of checklists or well-trained emergency procedures. Though I suspect the local managers had already worked out, at least sub rosa, what they would do after a hurricane.
Eric Hines