"Dead by Mistake" Report Hits the News
Medical “goofs” probably isn't the best tone, from the St. Louis paper. This is serious stuff. A “goof” is getting cheese on your sandwich at the hospital cafeteria when you said “no cheese.”
It states the problem of medical errors (causing death and patient injury) is WORSE than 10 years ago when the landmark study “To Err is Human” was published. That study estimated that 44,000 to 98,000 patients died each year due to preventable errors (not including infections). Combine that with other published numbers that 99,000 patients die due to hospital-acquired infections, that's 200,000 deaths a year. I've compiled some related stats here.
It's unclear if there is any new real research here, I think they are reporting numbers from the previous research. The Hearst press release about the series had sort of given the impression that they had new numbers. Regardless of the exact numbers, many in healthcare know this is a huge problem.
Preventable mistakes made in medical care are the nation's leading cause of accidental death, a Hearst investigation has documented.
An analysis of key national research on the issue shows that the death toll from medical injury, including infections acquired during medical treatment, approaches 200,000 a year.
Ten years ago, a highly publicized federal report called the death toll shocking and challenged the medical community to cut it in half â€” within five years.
Instead, federal analysts say, the rate is increasing.
Here is what I wrote as a comment to that story (with hyperlinks added here).
This truly is appalling that the healthcare industry has not made more progress in the last 10 years.
Some will point to this report and say “see how broken WE are in the U.S.” when this is actually a problem with healthcare. This is a global problem. 72,000 a year die in the United Kingdom due to medical errors (source) and they are a much smaller country.
The good news that will, hopefully, come out in the news is that MANY hospitals are dramatically improving patient safety and quality. They are doing this through a number of practical methods, like methods adapted from Toyota, often called “Lean” that allow the employees of a hospital to improve quality AND reduce cost.
Look up online references to Virginia Mason Medical Center, ThedaCare, Seattle Children's Hospital, Cincinnati Children's Hospital and you'll be impressed with what they are doing.
Allegheny Medical Center in Pittsburgh reduced central line infection deaths to ZERO in 2007 through the use of Toyota methods. This can be done. It doesn't require a ton of technology. It requires common sense and leadership.
The Healthcare Value Leaders Network is a collaborative group of 16 hospitals from around the country. These hospitals are working together in their shared mission of improving healthcare quality through lean methods. Dr. John Toussaint, the CEO emeritus of ThedaCare, is doing great work in this space. I hope the media will talk to him to get his side of the story.
While I thought the Hearst report was complete in terms of the number of stories that were told, it was all “doom and gloom” negativity. At least for now, the series focused only on what's wrong, not the good things that are happening at SOME hospitals.
As our release says:
Many healthcare leaders recognize the problem, and that's why we are developing and implementing solutions to fix the problems. Lean healthcare eliminates the errors and mistakes that lead to death. It's not just about preventing mistakes; it's about preventing injuries and deaths. Lean healthcare is a defined methodology that works. It's all about system redesign coupled with a commitment to making daily improvements throughout the system.
For the past six years, ThedaCare, a community-owned healthcare system in Wisconsin, has used lean methods and tools to redesign the way patients are cared for in its hospital rooms and physician clinics. The results have been dramatic:
- Heart surgery mortality rates decreased from 4 percent in 2001 to 1.4 percent in 2008 and zero percent in 2009
- Medication reconciliations errors are at zero percent since February 2007, compared to one per patient stay prior to redesign
This story needs to be told — the good things that are happening. What do other hospitals learn? How do we get EVERY hospital to take action (whether they use “lean” or not, although my bias is clear that I think Lean methods and philosophies would really help).
The Hearst series focuses A LOT on the failings of government. We put out a report. Some money was spent to fund groups like the AHRQ. The series shows a bit of pro-government bias — their implication is that we need more government and more funding to fix this major patient safety and quality crisis?
The hospitals that have been real leaders have done it without federal funding. I view this as a hospital management issue, not a government issue.
The solutions, in my mind, are not just a matter of reporting and recording. We need action. This is not a political issue — medical errors should be a non-partisan problem that we can all rally around. Of course, a lot of the online reader discussion was reactionary and partisan. Some people assumed the said online the papers were writing about how bad healthcare quality is so that people would support the Democratic reform proposals.
One reader actually said 200,000 deaths is “not a staggering number.” What? I don't see how anyone can excuse poor quality in the name of stopping healthcare reform plans in DC. Federal government take over will NOT improve quality (deaths and patient harm are a huge problem in Canada and the UK, too). Whatever happens in D.C., we still have to solve these healthcare delivery problem.
Here is a link to case examples of errors and injuries. Sad cases, every one.
One story that jumped out at me was a revisiting of a famous medical error that occurred at Virginia Mason Medical Center in 2004, about two years into their Lean / VMPS journey actually. A patient died after being injected with a disinfectant solution instead of a contrast dye.
The radiology technologist who was involved in the error was blamed — he was fined by the state, he resigned his job, and he took personal responsibility. But this was arguably a preventable process error.
Was he set up for failure? Not Dorsey individually, but anybody.
Dorsey recalled mentioning this “setup” for error to a physician â€” not the one involved in the accident â€” months before the tragedy.
So if he said something, why wasn't the risk eliminated? Before hand, the old disinfectant was brown — hard to mistake for the clear dye. After the problem, they went to a swab instead of an injectable. It's a shame that a death had to prompt process improvement when they should have anticipated the problem. Why wasn't Dorsey listened to?
It all seems to come down to culture, management, and leadership.
In another case, a patient died after her feeding tube went into her lungs instead of her stomach.
A nurse twice notified the doctor that something was wrong, but she ordered the nurse to continue the feedings, Anderson said.
Here's another case of a broken culture. Why did the doctor not listen to the nurse?
This article did focus a bit on what hospitals are trying, but it focused (predictably) on technology — computer systems (Computer Physician Order Entry) and bar code scanning.
The claim is that CPOE improves safety:
But few hospitals use them. A survey of 3,049 hospitals published in March found that only 17 percent of U.S. hospitals used the electronic prescription pad, known as computerized provider order entry, or CPOE. Other surveys have found even fewer hospitals use bar coding.
Hospitals blame high cost. Here's a clear opportunity to eliminate waste in other areas and then use the financial savings to fund projects like this. Or, the federal government is jumping in as part of the stimulus package.
So if you invest in the systems, does it really help? Help anybody other than the technology vendors?
The Hospital Corporation of America, the largest private health care system in the country, uses bar coding in all of its 163 hospitals. It has CPOE, the more expensive technology, only in about 20. Metropolitan Methodist is one of those 20. But the CPOE system is still voluntary for physicians â€” making it mandatory would require agreement by the medical staff â€” and only about 5 percent of the hospital's doctors currently are using it. The rest continue to scribble on paper, putting patients at risk of misread prescriptions.
5% of 5% of their doctors are actually using CPOE. Goodness gracious. Is it so simple as to blame the doctors for being careless or lazy? I'd argue “no.” I've heard people complain first hand that computer systems are too hard to use in hospitals. One nurse said, “Every time we get new technology, it makes my job harder.” That's not right.
“When one looks historically at the software technologies that were available, they really weren't built with a professional friendliness,” said Perlin, who was named to head a federal advisory committee on health information technology standards.
Software and technology has to be easy to use — or better said, it needs to work in a way that supports user workflow. Toyota has this embedded in one of its 14 Toyota Way principles:
- Use only reliable, thoroughly tested technology that serves your people and processes.
A Harvard prof says that doctors are used to routines and:
“What electronic records do is disrupt that routine. And there's pretty good evidence that for about six months after you implement one of those things, you tend to be less efficient.”
After doctors become familiar with the systems, however, they become as efficient as before and rarely want to go back, Jha said.
How do you lead physicians and get them to adopt the systems? It's a very complicated “standardized work” management situation if the doctors are not employees. Toyota would always say “lead as if you have no authority” but you still have formal authority to fall back on (a Toyota manager can mandate that an employee wear safety equipment).
Are hospital leaders willing to anger doctors, pull their privileges, and risk losing them and their patients to another hospital?
Poorly designed software that users resist… but the software helps… right?
When Children's Hospital of Pittsburgh launched its CPOE system in 2001, the death rate actually rose for five months. Critics blamed poorly designed software.
“In the last several years, we've seen a literature emerge of medical errors caused by computer systems,” said Dr. Robert Wachter, professor of hospital medicine at the University of California, San Francisco. “The systems as they stand now are still fairly clunky and user unfriendly.”
So, in theory, these systems PREVENT errors. But done badly, they cause errors. So the impact of the federal stimulus money will be….. what, exactly?
The same piece focused on a hospital that used PROCESS improvement, not technology, to improve quality and cost (a very “lean” approach).
Catholic-affiliated Ascension Health is the nation's largest not-for-profit hospital chain, with 67 acute care hospitals in 20 states.
In 2002, three years after “To Err is Human” was published, Ascension officials laid out an ambitious goal to eliminate all preventable deaths in its hospitals. They estimated it would be about 900 a year, or three per day â€” roughly 15 percent of deaths involving patients who weren't hospitalized for end-of-life care.
That seems like the only acceptable goal — zero harm. Not 5% reductions, but ZERO. How did they do?
They underestimated. A year after the program began, the system saw a 21 percent drop in overall mortality â€” about 1,200 deaths. Not all of them were from errors, but they all were preventable, said Dr. David Pryor, Ascension's chief medical officer.
“We believe that our financial performance has been improved by the work we've done in safety,” Pryor said, adding that malpractice costs have declined 35 percent since 2005.
More evidence that cost and quality go hand in hand. Doing the right thing (protecting patients) means lower cost for the hospital. What more evidence do you need? What do you say to the other hospitals that resist improvement because “we can't afford it?”
Here's another “like lean” lesson:
“The way I look at it is to say automation and technology can be very important and helpful,” Pryor said. “But if what you do is automate a bad process, you've now got a very expensive bad process.”
Another academic chimes in:
“Some things have improved,” said Mary Stefl, dean of health care administration at Trinity University in San Antonio. “Some of the low-hanging fruit has been removed. Now, you have checklists prior to surgery; you mark the spot on which limb you were going to operate on. And afterwards, they count the surgical sponges and instruments so they presumably don't leave anything inside. But it still happens. Somebody assumes someone else did it, or a surgeon refused to go through it and it still happens.”
As I wrote about last year, the checklist is useless if it's not being used. It still happens. In cases like this in Rhode Island and others in Utah, problems occur when the time outs and checklists are not followed 100% of the time. This is a leadership problem, not a technology problem. We know what to do, it's just not always done. Check out the video I blogged about from the show “ER” that shows this dynamic of people not wanting to take the time to use the checklist.
Culture and leadership. Do you see a theme? How can the government mandate that hospitals change their culture and manage better?
One other example of broken culture:
“It's just been a culture thing that if an error happens, it's somebody's fault,” said Dr. Jan Patterson, chief of staff at Audie Murphy Veterans Hospital. “We really think that 85 percent of the time it is a system issue. That means that we need to make things foolproof, just like in the aviation industry and the nuclear industry, where there's all these different safeguards and checklists that you go through. We need to make it easier to do the right thing and make it harder to make an error.”
There's so much that's broken in healthcare. I am glad that Hearst brought up the issue and shared the stories — as much as those stories drive improvement and not hatred for hospitals and doctors.
Ironically, the Houston Chronicle web page on medical errors had a web error. When reading comments on the first article, clicking the “Dead by Mistake” logo takes you to a “404 not found” dead link. Human error, indeed. I'll admit, I make mistakes all the time…