The Checklist: If something so simple can transform intensive care, what else can it do?
newyorker.com
newyorker.com
It sounds simple to institute a checklist, but the key issue isn't the checklist, it's about allowing nurses to hold doctors accountable and the hospital backing the nurses up in those disputes.
Say you're a night shift ICU nurse. Your patient craps out at 3am, and your intensivist crawls out of bed to come and "line up" your patient and put them on life support. Do you think that Dr. Intensivist with 15 years of school wants to hear at 3am is a nurse with an associates degree tell him to stop what he's doing and start over because he didn't wash his hands properly before starting. I've had charts thrown at me for stuff like that.
This issue is accountability and incentives. This plan places accountability on the doctors, but it puts a lot of strain on the doctor/nurse relationship when that nurse is the one blowing the whistle on the doctor.
What we need is an incentive system for doctors. You need both a carrot and a stick. Doctors only get paid when the patient is sick. And, if the patient gets sicker, that doctor spends more time with the patient and bills more. There is no incentive to keeping the patient healthy, only payment for taking care of as many sick patients as possible. Doctors don't like it they generally want to take good care of patients. They just have to pay of $200,000 of student loans, and the only way to do that is to rush and see as many patients as possible.
If the doctors received a large bonus every year if their lines did not get infected, or if their patients did not get a pneumonia when the patient was on life support, you better believe that complication rates would go down.
If nurses got a small bonus every year based on our infection control practices, we'd all wash our hands a lot more often. But we don't. We're all paid by the hour on a patient assembly line.
The doctors or nurses haven't set up this system, the insurance system has. It's just the unintended consequences of our third party payer system.
Although, there are winds of change. Medicare is starting to threaten hospitals by not paying them if a patient gets an infection from the hospital.
No, I don't think I'm cynical. I'm speaking from over 15 years experience in the Critical Care business. I've worked at over 35 hospitals and medical facilities in 6 different states.
And, it's a plain fact of nature. People do what you pay them to do. Just because someone is in health care doesn't change the fact that they want to get paid. And, we should give people financial incentives to give good care. Pay them to do a good job.
A founder gets paid if their company has a successful exit. So, we bust our buts to make a successful exit for our company. Stock options reward employees for increasing the stock price. So, employees will work to increase that stock price. If you pay people per hour, they work a lot of hours. If you salary people without regards to performance, performance drops. If you pay a person per widget made, they make a lot of widgets.
I'm speaking from experience. The best ER's that I've worked in were ER's where the doctors are paid a base rate and then a bonus for performance based standards. That ensures excellent, efficient practice of medicine. Those are the ER's where the doctors push the nurses to line up more patients to be seen. The ER that I work in now is one of those. I'm working with some of the best doctors I've ever worked with.
I've worked in ER's like that where the group of doctors got paid a flat rate to staff the ER per year. So, if they understaffed the ER, they got paid more. Needless to say the patient wait times in that ER were often over 16 hours. I saw patients waiting a full 24 hours just to see a doctor for a broken arm or a gallbladder attack.
Another, potentially profitable, solution to the problem would be relatively easy using something similar to those new Philips ICU monitors that modularly record pulseox, blood pressure, respiration, etc... When the doctor is about to insert a central line, the nurse can click "central line" and the computer will say, audibly, "wash hands." Then the doctor says "next" and it says "create full-body sterile field. All non-necessary personnel, leave the room now." Then, just before the needle insertion, it asks very quickly "Have you washed your hands, created a sterile field, cleansed with chlorahexadine, and removed all non-necessary personnel?" Again an auditory "yes" will make the machine turn green or something, signifying that the doctor can insert the central line. It will add maybe 10 or 15 seconds to the process, and can have these kinds of flows for every routine procedure.
Many doctors would probably consider a machine doing that insulting. But if it saves "thousands" of lives, it's definitely worth it for someone to create a company and build them.
I was visiting an ICU for a couple of weeks last month, and I noticed that the nurses almost always do things by memory-- one of them forgot to check the dressing of a central line, and when the doctor asked why, she said she forgot. It says on the piece of paper at her desk that she was supposed to check it, but she doesn't get to sit at her desk very often.
There are lots of possibilities for technology to help improve the chances of critical care patients surviving. Like the auditory checklist above, why aren't scheduled things like dressing check/changes, IV fluid refills, and body movement/cleaning automated on some sort of central ICU computer which communicates with nurses using wireless PDAs? The airline Virgin America makes their flight attendants use wireless UMPCs so they are instantly notified when someone orders something from the in-flight entertainment system.
When a a patient's IV bag runs out, the nurses PDA can vibrate and the task can be added to a list. Why is this so difficult? It's complicated, expensive, and tedious. But if it saves lives and reduces infection rates, it might pay for itself very quickly.
My point is, even though the checklist procedure seems very stupid and simple, it should be integrated with sophisticated technology so it is never accidentally forgotten. I think that's where the future of healthcare is going, especially in critical care. It's too bad there aren't many (any) startups pursuing this field; it's dominated by Siemens and Phillips, and I think that's a reason there's so little innovation in hospital-based systems for patient care. So much money goes into research, as the article points out, but very little goes into implementing the results into practical patient care. That's where technology should come in, not just paper checklists.
RE: monitors talking to me... My monitors already "talk to me" too much. There are many many beeps and buzzes that I ignore constantly. There are one or two kinds of beeps that I watch very closely, however. When I hear those, I go running to the monitor. Everything else, I ignore. A voice telling me to wash my hands would get ignored like the others.
Re: computerizing tasks. It's a decent thought, but resistance to change is extremely high in health care, especially technological change. I work at a hospital now that transitioned to a new computer system hospital wide two months before I started. The entire hospital was pretty much in full-on revolt against the new system, and even now, 18 months later, there's still tons of animosity against the software and the IT people. It was a nightmare. So, there are many, many cultural issues that prevent adoption of high tech practices.
Battery time. I work 12 hour shifts, and in many hospitals I've worked at, I haven't even gotten a lunch break. A number of places I've worked, I don't get time to go to the bathroom for 12 hours. A PDA battery is not going to last a 12 hour shift, and nurses don't have time to change the batteries. If that's the case, they won't use the technology.
RE: startups in the health care field... The liability involved is astronomical. Getting a new medical device or new product through the FDA hoops is a nightmare and extremely costly. And, what happens if you have an obscure bug, and your IV pump that contains a life support drug runs dry, and the nurse isn't notified. In a number of cases, the patient could die if the drug is stopped for more than a few minutes.
What happens when the patient dies from a random complications, and the family is upset about the care received? They sue everyone involved. Even though your software/PDA product wasn't involved in the patient's death, you will get named in the lawsuit simply because you are considered a "deep pocket" and your product was involved in the care of the patient. You end up hiring a very extensive legal department when you have any sort of medical device. You have to.
I think beeping should be one of those sacred controlled sounds that are specifically designed, like fighter jet codes, to mean a single thing and only a single thing. They should be regulated extremely, and only used when absolutely necessary. One of the IV pumps I've seen played a little song when the bag ran out; it was similar to Jingle Bells. What the fuck?
Just because your monitors talk to you too much doesn't mean that having them talk to you more is a bad thing. The problem is that they beep and talk to you too much for stupid things. It needs to be better-regulated. If the low-resp alarm starts going off, no one even flinches until it's going on for more than fifteen seconds, because the machines beep all day long for false alarms. How hard would it be for someone to fix the software so that it waits 3-5 seconds to check for errors before it starts beeping?
I think that voice-controlled computers for checklisting would be a great idea that would save lives. And if you think that's too annoying, maybe there should be a financial incentive for you to use them. If you don't use the voice checklisting, you should lose money. There's pavlovian psychology that causes instant gratification.
Battery time is an interesting issue that I hadn't thought of. But to discount the use of PDAs because of batteries is kind of strange, especially if you think the idea is good. If it saves lives or helps nurses do their jobs, then the technology can be adjusted to fit the situation. There are batteries out there that can probably last 12 or 15 hours; and if they don't work, why is it so hard for a nurse to simply exchange a PDA in in the nurses station (taking 5 seconds)?
Regarding the adoption of technology in hospitals: The hospital I visited also had just a year earlier introduced a hospital-wide computer system for patient tracking. The nurses hated it because it is "a pain to use" and the computer locks the medicine drawers until it thinks the patient needs the medicine. Even the crash cart required a 10-character password before it would dispense any medication; how fucking stupid is that? My point is that the resistance to adoption is probably due to the poor design of the software. In web software, good user interfaces and ease of use are important to get business. In hospitals, it's imperative because it saves lives. And yet startups and innovative companies with talented software designers are afraid to enter the healthcare market because of the huge costs involved, like you said.
I think it's a shame. A well-designed hospital-wide system that contains all of the hospital operations (BP monitors, IV pumps, pulseox, CT scans, etc, etc) can prevent deaths because of standardization and ease of use.
Can you be a little less technical?
Empirical evidence is given about IV line infections, a major problem in ICUs. A 5 step checklist lowers the 10-day line infection rate from 11% to 0% at Johns Hopkins, an almost unbelievable result. It is confirmed over 15 months.
These results are again confirmed at an understaffed inner city hospital in Michigan with similarly dramatic results. The study from the Michigan hospital is published in the December 2006 publication of The New England Journal of Medicine. Despite these dramatic results, there is resistance against the increased use of checklists.
The checklist study author says he can introduce ICU checklists across America for 3 million dollars at the most. Because of the resistance in the USA, the country of Spain will be the first country to use checklists nationwide. The study author hopes that the USA will not be the last country to start using ICU checklists.
If anyone here has money to burn, lobbying for the increased use of checklists in intensive care units would get more net local social returns (in lives saved) than building Stanford a new emergency room and it would cost less money.
(Case in point at a client of mine: When the parking lot was scheduled for paving, email was sent to everyone. For whatever reason, about half didn't read it or know about it. The next time, paper notices were posted on both sides of every door. Not one person parked in the lot. Who knew?)