The most popular article in last week’s New England Journal of Medicine did not tout the discovery of a novel gene, nor describe a cardiology clinical trial with a clever acronym as its title. Rather, it was the report of a case in which a surgeon at the Massachusetts General Hospital performed the wrong operation on a 65-year-old woman.
This was a breakthrough for the Journal – the first time in its storied 86-year history that the Case
Records of the MGH published such a report. But it was not the first opportunity the NEJM had to publish such a piece… that occurred a decade earlier. The story of the path from then to now reflects the evolution of the patient safety movement. It’s a story I know well since it involved one of the lowest points in my professional life.
Before I share the back story, a word on last week’s article. David Ring, a prominent Harvard hand specialist, performed a carpal tunnel release on a patient who actually needed a trigger finger release – an entirely different operation. Showing great courage, Ring described his own error, with safety expert Gregg Meyer providing the color commentary. As always, the pathophysiology of this misfire was a combination of active (i.e., somebody did something wrong) and latent (the system was a setup for failure) errors that jibed entirely with Jim Reason’s famous “Swiss cheese model” of “organizational accidents.” The errors included:
- The patient spoke only Spanish, so that the surgeon, who spoke the language, ended up serving as the interpreter, distracting him from his primary task.
- None of the other members of the OR crew spoke Spanish, so they mistook the surgeon’s pre-op conversation with the patient for a Time Out. The result was that no formal Time Out was performed.
- A nurse, rather than the operating surgeon, marked the site pre-operatively.
- The nurse’s pre-operative site marking was washed away when the site was prepped and cleaned.
- The OR’s slick new computer screens blocked the nurses’ view of the patient.
- It was “one of those days” and everyone was stressed: the OR was running well behind schedule, nurses were being pulled off their previously assigned cases and sent to staff others (including the nurse who performed the pre-op site marking), and Dr. Ring had just completed another case on a very anxious patient. Talking her down just added to his level of distraction.
You get the picture.
In other words, it was a case of the usual stuff conspiring to cause a good person and institution to commit a breathtaking error.
OK, now the back story. In 1999, Dr. Kaveh Shojania arrived at UCSF to become the nation’s first hospitalist fellow. Kaveh had just completed his internal medicine residency at Harvard’s Brigham & Women’s Hospital, where he was introduced to patient safety by several legends of the fledgling field, including David Bates, Lucian Leape, and Atul Gawande. He came to UCSF to learn to be an academic hospitalist, and to continue his work on patient safety. With me.
There was one little problem: I didn’t know the first thing about patient safety. I wondered what kind of project we might do together.
One day early in his fellowship, Kaveh was in my office and picked up a copy of the New England Journal of Medicine he saw lying on my table. “Here’s the Case Records of the MGH,” he said, sounding slightly irritated. “Every case highlights some genius Harvard professor making an obscure diagnosis. I’ve never heard of half these diagnoses.” (An amazing statement, since Kaveh knows more clinical medicine than virtually anyone except Gregory House.) “Why don’t they discuss the stuff we actually see – like cases of medical mistakes?”
I immediately knew Kaveh was on to something. After spending a little time honing the idea and enlisting our colleagues Sanjay Saint and Amy Markowitz to help, we described the concept to my friend Mark Smith, the CEO of the California HealthCare Foundation. Mark loved it and agreed to fund us to produce a series of case-based articles on medical errors if we could find a journal to publish it. On July 7, 2000, only a few months after the publication of the IOM report on medical errors, I pitched the series to the New England Journal with this letter:
As with traditional CPC [Clinicopathologic Conference] presentations, we will select “interesting cases,” both in terms of clinical content and the quality issues they illuminate. However, the rounds will focus on identifying and analyzing the confluence of operator factors with systems failures often associated with severe medical injuries. After the expert’s discussion of the case, the manuscript will go on to review what became of the institution’s efforts to resolve the quality problem identified by the case. The expert will offer concluding remarks… to highlight the lessons gleaned from peering into the “sausage factory” of institutions as they attempt to rectify important problems identified by their practitioners…. Through the publication of this bimonthly series, the New England Journal will be taking a leadership role in highlighting medical errors and other quality problems, educating physicians and other providers about them, and participating in the search for solutions.
The Journal replied swiftly and positively, going so far as to permit us to recruit authors by telling them that the series had “tentatively been accepted for publication in the NEJM pending peer review.” Before giving it a final green light, though, they asked to see a couple of examples of cases and commentaries. Fair enough.
For our first manuscript, we had the perfect case in mind: an elderly woman admitted to a teaching hospital for a neurovascular procedure mistakenly underwent a cardiac electrophysiology procedure intended for another patient with a similar last name. Dr. Mark Chassin, then chair of health policy at Mt. Sinai (and now CEO of the Joint Commission), agreed to write the discussion with a junior colleague. And what a discussion it was: their detailed review of all of the errors and the lessons from this case was thorough, well written, and provocative. I thought the paper, entitled “The Wrong Patient,” was destined to be a classic. I sent it, along with a terrific case/commentary on “cascade iatrogenesis” authored by Tim Hofer and Rod Hayward of the University of Michigan, to the NEJM. In the meantime, we recruited a who’s who of safety and quality leaders to write subsequent articles in the series: David Bates, Julie Gerberding, Troy Brennan, Peter Pronovost, Lee Goldman, and others of similar stature.
?I awaited a “thick envelope” from the NEJM, but the journal’s response was anything but: they didn’t like the manuscripts and were inclined to reject them. They weren’t abandoning the idea of the series, but it was clear that the articles we’d submitted were not what they were looking for. The editors asked if I could fly to Boston to meet with them. From the tone of the rejection, I wondered whether the trip was futile, but this was the New England Journal of Medicine, so it was worth a shot.
Walking into the offices of the NEJM is intimidating: you enter Harvard Medical School’s Countway Library, take an elevator to the top floor, and then you’re there. The feeling is like someone has lifted up the curtain and you’re about to see the Wizard. In this case, the Wizard was Dr. Jeff Drazen, the pulmonologist who had taken the helm of the journal earlier that year. Drazen’s office was busy but neat; its most distinguishing characteristic was that it was filled with homemade grandfather clocks, which sounded off every 15 minutes or so. Mark Smith joined us by phone from California. We quickly got to the matter at hand.
“We love the concept,” Drazen said, “but we need the articles to focus on things that doctors can do to fix the problem of medical mistakes.” Referring specifically to the Chassin article, which had identified 17 distinct errors (not unlike last week’s wrong surgery case), Drazen continued. “It’s just too messy, there are simply too many problems.”
I retorted that the analysis reflected our new understanding of the patient safety field, frantically trying to win him over with a brief tutorial on systems thinking, the Swiss cheese model, active vs. latent errors. It was an area physicians knew nothing about, but one that we needed to learn if we were going to effectively tackle the problem of medical mistakes, I pleaded.
Even as I spoke, a famous scene from the movie Amadeus kept playing in the back of my mind, the one in which the precocious Mozart presents his masterwork to Emperor Joseph II. The Emperor’s face looks like he’s just sucked on a lemon. After a few painful moments of hemming and hawing, the monarch delivers his verdict: “Too many notes,” he concludes. “Just cut a few and it’ll be perfect.”
“Which few did you have in mind, Majesty?” blurts an outraged Mozart.
And now our manuscripts had received the same fatal diagnosis: they had too many notes! I left the meeting with Drazen crestfallen and called Mark Smith, who had heard the whole exchange. “You look up, ‘Just don’t get it’ in the dictionary, and you see his picture,” he said. “I don’t think we can get them to understand.” And he was right – while the editors invited us to revise and resubmit the manuscripts, it was clear that we could never meet their expectations without violating our understanding of the field and dishonoring the work of Chassin and our other authors. Quality Grand Rounds, at least as a series in the NEJM, was dead.
On Feb 19, 2001, I wrote this painful letter to Jeff Drazen:
We have decided not to resubmit the papers to the Journal. In the end, we came to believe that our vision for the series – to focus on the systems issues that we believe are crucial to improving patient safety – was incompatible with your desire to focus on the actions of individual physicians in manuscripts that were… less complex.
All’s well that ends well. The Annals of Internal Medicine became our partner in the series, ultimately publishing 13 highly regarded articles from 2002 to 2006. They were a joy to work with, and the series had a significant influence on the patient safety field.
And, despite taking a pass on Quality Grand Rounds, the New England Journal did a terrific job covering the safety story over the next decade, publishing many of the field’s seminal articles, including the central line infection and surgical checklist studies and several important thought pieces (including my “no blame” vs. accountability piece, written with Peter Pronovost). It was clear that the NEJM editors – like thoughtful physicians everywhere – ultimately came to embrace this new paradigm of systems thinking, and made key contributions to advancing public and professional understanding of it.
And now, exactly 10 years after my disheartening experience, the New England Journal has published a case of a serious medical error and analyzed the active and latent errors that allowed it to occur. It’s a great paper, partly because it contains just the right number of notes.
Bob,
I can not get over the eloquence of your descriptions and metaphors.
However, I note a disparity between your report and the detail described in the NEJM report. Additionally, the NEJM report itself is internally inconsistent, stating that “the poor placement of computer monitors and the consequent diversion of the operating room nurses from the task at hand…”
and also: “and — as we learned during our later investigation — inattention by the nurses that was fostered by placement of clinical computer monitors in a way that diverted the nurses’ gaze away from the patient.
Which is it, and by their opaque language, are they defending the EHR devices that have been subject to recent doubt and safety concerns as reported in JAMIA?
You state: “The OR’s slick new computer screens blocked the nurses’ view of the patient.”
Bob, the problem with the computer “screens” appears to have nothing to do with “blocking” the view of the nurses.
Rather, the patient appears to have suffered the consequences of the typical and well described cognitive distraction that the user unfriendly layouts of the jabberwock seen on the computer screens (regardless of where they are located) engenders to the users of these instruments.
Thus, this is but another example of unintended consequences of HIT facilitated error, even though it was not a “pure” error from the HIT devices. The HIT devices sure are not the cure all and did not prevent the error, here!
Best regards,
Menoalittle
Thats really appreciating! Nice thoughts!
I agree with Menoalittle that the unintended consequences of Health Information Technology are contributing to errors in medical care. HIT helps in many ways but it creates new patient safety concerns that I don’t believe have been adequately studied in our current health care environment. HIT is here to stay but it is not the ” be all and end all” that it has been touted to be. As a profession we need to continue to study medical errors as they truly harm our patients. HIT causes medical errors and we need to challenge the mantra that HIT will solve many of the problems facing healthcare, including medical errors. The medical profession needs to engage and challenge the for profit motivated vendors in HIT to develop systems that will truly benefit our patients, be user friendly and have sufficient redundancy so that HIT medical errors become rare. From the trenches.
After reading the article I was most taken with Dr. Drazen’s statement :
“…but we need the articles to focus on things that doctors can do to fix the problem of medical mistakes.” I can imagine your dismay, to be speaking to a colleague at a venerated institution concerning a topic of vital importance, only to be met with naivety about the true nature of medical mistakes. Our tendency as a society is to want a simple, pat answer to tragedies or unexpected events – which implies there is a simple solution or culprit. Rarely, is this the case when something goes wrong within a complex system. In his excellent book of essays, “What the Dog Saw”, author Malcolm Gladwell described the ultimate findings of the committee who studied the spaceship Challenger tragedy. The final conclusion: many small things, and a few not so small things, led to the tragedy. There was no single “smoking gun”, no single Machiavellian entity who was responsible. His conclusion was that this will happen again because the system is complex and there are an infinite number of things that can go wrong. We in the medical profession, if we are honest, know that this is true. Our task should be to highlight system errors, human or otherwise, and try to make meaningful changes. This is not an easy task, but I would argue that it is the most important task we can tackle in medicine. Kudos to all who paved the way for real discussions of safety issues.
Nice article