Checklists are hot. Surgeon/writer Atul Gawande, MD, has promoted them in The New Yorker and a new book, and many less famous quality improvement advocates are excited about the concept, too.
“A lot of the buzz at SHM [the Society of Hospital Medicine's annual meeting] among people in my peer group and people really passionate about [quality improvement] was around this idea,” said Michelle Mourad, ACP Member, director of quality for the division of hospital medicine at the University of California, San Francisco Medical Center.
But it isn't just trendiness driving hospitalists' interest in checklists. According to Dr. Mourad and other inpatient physicians who have developed and implemented checklists, they can be an effective and efficient way to improve care, if the process is handled correctly.
“If you've decided that there's an optimal way to discharge a patient…, how do you make that a reality? For me, that's where a checklist comes in,” said Dr. Mourad.
Checklists can also help physicians cover the ever-growing number of tasks and best practices involved in the course of a hospitalization. “There are always more important things to do than the mundane things,” said Craig Coopersmith, MD, associate director of the Center for Critical Care at Emory University in Atlanta.
“What checklists do to some degree is they force you to think about things that you wouldn't think about in the day-to-day. It's never going to hit number one on your priority list, but it's no less fatal if you forget it or mess it up,” he said.
The discharge process is a popular target of hospitalists' checklists, and a checklist developed by Peter Pronovost, MD, to prevent central line infections is well known, but the concept can be applied to a wide range of evidence-based practices, as Dr. Coopersmith and colleagues did in the intensive care unit. Their checklist, a study of which was published in Critical Care Medicine in October 2009, included items ranging from clot prophylaxis to bed elevation.
“Improved processes using checklists could be described for a lot of the clinical activities that we do,” said Brian Jack, MD, associate professor of family medicine at Boston University School of Medicine/Boston Medical Center and developer of the Project RED (Re-engineered Discharge) checklist.
In some cases, the areas in need of improvement may be already obvious. If not, the first step to building a checklist is to ask around. “We had key people from around the hospital get together, and bought them pizza and said, ‘Here's a process map [of discharge]. If you were going to re-engineer it, what would you do?’” described Dr. Jack.
At the least, the group who develops the checklist should include a representative of every provider who is going to have to implement it. “You don't want to just suddenly spring on your pharmacist, ‘Hey, we've got this checklist, so you're now responsible for doing medication teaching on all patients with these high-risk medications,’” said Dr. Mourad. “Have a pharmacist in your group developing the checklist to make sure that your expectations for the items on your checklist are reasonable.”
Depending on the nature of your checklist, input from outside the hospital may be required too, said Jeffrey L. Schnipper, ACP Member, assistant professor of medicine at Harvard University and a hospitalist working on discharge checklists. “My co-chair went to a lot of the rehab hospitals that we did business with and asked them for stories of lousy discharge summaries, and asked them what they would have wanted,” Dr. Schnipper said.
Since these checklist pioneers have already done a lot of investigating, you may be able to skip some of the research. “There is a wealth of literature out there already,” said Dr. Coopersmith. “One doesn't have to start from scratch.”
How precisely one program can replicate another's checklist is a matter of more debate. “I think there really are principles that ought to be done, not unlike the pilots taking off in the plane. That checklist applies to flights everywhere,” said Dr. Jack.
On the other hand, Charles Bosk, PhD, a sociologist who has studied checklists, warns that some decision-making should be local. “Checklists have worked in some cases and not in others. There's heterogeneity between groups in using checklists successfully,” he said.
The chances of success may be increased by allowing those who will use your checklist to customize it. “Places gain ownership by modifying something and making it their own,” said Dr. Schnipper.
Customization can also accomplish the intent of a checklist while bringing it in line with a hospital's standard procedures, for example, on how to prevent blood clots. “There's more than one strategy that is effective, and the question is not whether or not we try to prevent blood clots but how,” said Dr. Coopersmith.
Not too short, not too long
A checklist may also need to be modified so it can be implemented effectively. When Dr. Schnipper's colleague surveyed outside clinicians about their wants for a discharge summary, he came back with 100 items. The checklist that the Harvard researchers more recently developed has 11 steps, which Dr. Schnipper has concluded might be too many.
“I really have boiled it down more to [steps] 9, 10, and 11. We can generally get though [those] in about three minutes. Even that might be too long. I think checklists ideally are designed to be very short,” he said.
But don't make them too short to be effective, cautioned Dr. Mourad. “There isn't a perfect length to a checklist. They likely vary by the complexity of the task. For example, you wouldn't want the pre-NASA-shuttle-takeoff checklist to only be five items long,” she said.
Checklist developers should also keep an eye on the feasibility of accomplishing all the items on the list, and if there's a problem, either remove the item or make it possible. Dr. Mourad gave an example from Dr. Pronovost's project.
“They realized chlorhexidine prep was nowhere to be found in the hospital, so that wasn't an easily achievable item on their central venous catheter infection prevention checklist. They made sure that chlorhexidine prep was immediately available everywhere central venous catheters were being placed,” she said.
It may sound obvious, but checklists should also be based on solid evidence. “You're not saying, ‘This is my way. I want things done my way,’” said Dr. Mourad. Instead, collect data and present it to the potential users, advised Dr. Coopersmith. “We were able to say, ‘Here's our problem. We think we're doing a great job, but these are the statistics.’”
Evidence is necessary, but patient anecdotes can provide the additional push to get everyone to believe in a checklist. “A heartfelt story about an adverse event when things didn't go perfectly, and a checklist might have improved things, goes a long way toward convincing people that things need to change,” said Dr. Mourad.
Like every effort at system change, checklists also need champions—on the medical staff and in the administration. Those champions may be able to help with the publicity efforts required to get a checklist into use.
“Every Friday during bagel rounds, we would go down there and sort of exhort them to use it. We would ask the residents whether they had received it or not in front of their attendings,” said Dr. Schnipper.
Still, some hospitalists—and other clinicians—are probably going to insist that they're too busy to take on a checklist. Dr. Schnipper's project found that only about half of hospitalists in the program had adopted the discharge checklist. “If I were to do it over again, I'd probably make it half as long and do a little more exhortation to get people to use it. Maybe we would do some role-playing,” he said.
Check up on your checklist
But if you've exhorted until you're hoarse, there is one other way to assure widespread implementation of checklists: Make their use a requirement. “They are better than nothing when they are optional, but they need to be mandatory,” said Dr. Coopersmith. “If you really want to get it done, it has to be mandated for every provider, every day, on every patient.” Continuing the air traffic analogy, he noted that no one would ever consider making a pre-flight checklist optional.
Even a mandatory checklist will not achieve perfect compliance with every measure, however. “People will automatically say yes even if they didn't do something, because they want to move on to the next patient,” said Dr. Coopersmith, giving the example of head-of-bed elevation. “The answer from every resident in history is yes.”
To make sure that the beds actually were being placed at the proper elevation, the ICU team put a dot on each bed that was visible only when the head was at the recommended angle. “You need to figure out a way to audit it intermittently to see whether you're doing what you say you're doing,” said Dr. Coopersmith. If a checklist calls for prophylactic medication, for example, you could audit medication records to see how regularly it's being prescribed.
Monitoring of process measures like these is important, but you should also examine the checklist's effect on outcomes, advised Dr. Jack. “Outcome measures are [issues like] satisfaction; how many come back to the emergency department, the hospital; what's the cost to the system?”
Based on the results of auditing, it may be necessary to make changes. “We've always viewed the checklist as a living, breathing document. Every six months to a year, we would assess the utility of what was on it,” said Dr. Coopersmith.
Some items may be performed so consistently that they don't need to be on the checklist anymore, while new things will turn up that should be added. The danger is that the general trend will likely be toward expanding the checklist. “If you add five things, maybe one thing is detracted. Over time, they get longer and longer and people pay less and less attention to them,” said Dr. Coopersmith.
There are other potential risks in checklist implementation. “In a lot of teaching institutions at least, they tend to get relegated to the most junior person, who may be the least competent person to deal with patient safety issues,” Dr. Coopersmith said.
Whether the person in charge of making sure the checklist is complete (and there should be one person who's officially accountable, the experts noted) is a nurse, a discharge coordinator or a resident, the new task will take them away from other duties. “Whoever it is is busy and they already have a job,” said Dr. Jack.
That points to another downside of checklists. “What are you going to not do if you spend 10 minutes doing this? Are you going to leave something out that is equally important?” said Dr. Schnipper. “It's always a trade-off.”
The potential risks of checklists were discussed in a lecture at the American Thoracic Society's annual meeting in May by Damon Scales, MD, an ICU physician at the University of Toronto.
“They can make us better but they should not allow us to become lazier,” Dr. Scales said. Checklists could potentially cause clinicians to treat patients based on general rules, instead of thinking about each individual's specific needs, he warned. There's also the risk that the evidence on a topic could change, while the checklist stays the same. “It's imperative that we continue to re-evaluate them to make sure that they are still evidence-based and appropriate given the most recent evidence,” Dr. Scales said.
Checklists can even be an outlet for that creativity. At Dr. Mourad's hospital, the interns complete a rotation on which they spend two weeks learning about quality improvement, including checklists, and two weeks with the hospitalist procedure service. A group of interns on this rotation found that when they took a call to the procedure service, they were often forgetting to ask important questions, so they created a checklist of items to ask the referring physician.
“It was really a beautiful moment, when checklists started to creep into the procedure service rotation through the innovation of the interns,” said Dr. Mourad. “They wanted to make their process more efficient and more optimal and seized on the idea of a checklist as a way to make that change.”