Sunday, July 20, 2008

Did you do this on purpose?

My post below and a similar one of the Wall Street Journal Health Blog have engendered a lot of comments about punishment after medical errors. The discussion is important and is not yet complete. Let's expand on the topic here.

Thanks to Don Berwick from IHI who referred me to a recent article by Dr. Charles Denham, entitled "May I have the envelope please." (Journal of Patient Safety. 2008 Jun;4(2):119-123.) Chuck relates the marvelous approach to error used by Jeannette Ives-Erickson at the Massachusetts General Hospital. When there is a screw-up in nursing, she calls the involved nurse into her office and asks one question: “Did you do this on purpose?” If the nurse answers, “No,” then Jeannette says, “Well then it is my fault.... Errors stem from systems flaws.... I am responsible for creating safe systems."

As Tom Botts mentions below and as Chuck reinforces in his article, "When we push the envelope in health care, senior leaders and many clinician often never know about the adverse events because these events are often hidden and masked by the complexity and fragmentation of care.... We automatically fall in a name-blame-shame cycle citing violated policies and ignore the laws of human performance and our responsibility as leaders."

Turning back to Ives-Erickson, Chuck notes, "In a few short moments with a caregiver after an accident, the leader declares ownership of the systems envelope, and the performance envelope of her caregivers, and creates a healing constructive opportunity to prevent a repeat occurrence."

Recognizing that the comments made on this blog and the WSJ blog may or may not be representative of the general public, I was nonetheless impressed by the degree to which people felt that punishment was an essential part of process improvement. It also occurred to me that the easy path for a hospital administrator in this kind of environment would be to punish the wrong-doer, bolt on a new process, protocol, procedure, or requirement, and declare the problem solved. After all, that shows decisive and timely leadership.

There's only one problem. That doesn't work. Or if does, only for a short time or until a new glitch is uncovered.

Many of the comments show to me the level of dissatisfaction with and anger about the health care system in general, and perhaps also individuals' experience with certain "god-like" physicians. But, if those admittedly understandable emotional reactions guide our approach to process improvement, we will not make the kind of progress we need.

Lee Carter, chairman of the board at Cincinnati Children's Hospital -- a national leader in the quality and safety movement -- put it in elegant, all-American Midwest terms: Transparency depends on that one can report an error without getting whacked. I absolutely agree with your blog in both the lack of punishment for this event and reserving the right to punish for events in the future. If punishment were to be meted out, it should be spread to everyone in the OR who didn't call for a time-out. The point is that it wasn't only the surgeon's responsibility. This is what we are working very hard to spread throughout Cincinnati Children's and we are making slow progress.

Think about it. One of the national leaders says that his place is making slow progress. Let's learn from that. Let's not let our own impatience with the errors that occur cause us to leap to a type of solution that appears easy and direct but that is fundamentally flawed.


Medical Quack said...

I agree and also posted my views with comparing a recent incident in Worcester and the somewhat lack of transparency by comparison and someone else making an airline industry comparison on the same, and they make mistakes and deal with the same issues of punishment, there's no general rule that can be applied by any means.

Anonymous said...

This issue has kind of grabbed me today. Wow, that WSJ blog commentary was a revelation! I think it's because it seems like such a stupid error; the average person wouldn't understand how easy, in fact, it is to have this happen.
Within the restriction of the understandable need for confidentiality, I for one would be very interested in the results of the root cause analysis before trying to think about punishment. I read in one place that the correct side had been prepared, and in another that the wrong side had been prepared. Which? Was the correct side marked and then the wrong side draped? How did it come about that the wrong side was thought of in the first place - did the surgeon "remember" which side, was a CT put up backwards, was the white board in the OR wrong, what?
The reason I ask is that we are assuming that the time out would have caught the error if performed. Is that true in this case?
I have to say although I favor the no blame approach, I am kind of unable to get past the fact that an established policy WAS apparently ignored/forgotten by the entire team. That is hard to get around; more data would help.


Anonymous said...

Why can't there be both punishment and a solution?
You can implement change without a bolt-on improvement and still make an example out of someone who ignores a policy. It was an ignorant mistake. Is there really room at BIDMC for ignorant practitioners? It was the luck of the draw that you got a group of providers who didn't think of doing a time-out. All it takes is one person to initiate a time-out. If for example there are 5 people in the room, that was 0/5 who remembered to do a time-out. Perhaps in most cases 5/5 people would to a time out. Maybe it's only 1/5 normally. Again, all it takes is that one person. I bet you'll get closer to 5/5 if you make an example out of those ignorant ones who don't do it.

This error was in a class of its own. Perhaps people shouldn't be afraid of reporting minor to moderate errors - I'm sure the culture can be created even with punishment being handed out in this case. In big cases like this, obviously they'll be discovered even if the entire team in the OR didn't say anything. Others will not be any more or less afraid to report errors because you discipline someone for a major mistake like this one. I love the comment on the WSJ blog about the airline pilot who crashes because he ignored a fundamental rule. He doesn't get a free ride just because "it was human error" and just because he can turn on the water-works after the event because he feels bad and is scared for his livelihood. Of course the surgeon, anesthesiologist, residents, nurses, scrub techs etc felt bad after the event and had a convincing emotional reaction afterwards. That doesn't change the fact that they were NEGLIGENT in the first place. They were emotional afterwards, which you have said has guided your reaction to the event - did you consider that they were emotional because they're scared half to death of the consequences? I remember getting in trouble as a child and knowing that if I cried after and said I didn't mean to do it that I'd get off easier, and it seems glaringly obvious here that this team did the same thing, and that administration fell for it.
One unrelated question for you - is it inevitable that the public is going to find out more specifics of this case? Are they going to be drawn out in a year or so through the courts?

Anonymous said...


The time out presumably would have caught the error. The correct side was marked. The wrong side was operated on. No one in the room asked for the timeout.

But I disagree a little with you. Those details are not what matters, as I discuss below.

Anon 8:31,

Ditto. More specifics don't tell you much. The key point is this: No one in the room asked for the timeout. If any one of the people had, the result probably would have been different.

Why they didn't do this might depend on particular circumstances at the moment. But you don't design solutions for particular circumstances. You design solutions that have deep and lasting applicability even in the face of unanticipated circumstances.

Your use of the word "ignorant" is misplaced. These are not ignorant people. They -- individually and as a group -- made mistake.

By the way, this error was not in a class of its own. Wrong-side surgeries occur throughout the US even when protocols and check lists are in place. In fact, there was another in MA recently even after a timeout took place.

I think the experts in the field are telling us that they would disagree with you on this point: "Others will not be any more or less afraid to report errors because you discipline someone for a major mistake like this one".

Finally to this point: "They were emotional afterwards, which you have said has guided your reaction to the event - did you consider that they were emotional because they're scared half to death of the consequences?" Read my post again for what guided my reaction. But also, please consider that they were emotional mainly because these very well intentioned people had already SEEN the consequences, a patient that they had harmed.

Hey, You said...

This is one of the most thought-provoking blogs around. Kudos to you, Paul Levy, for not being afraid to take on the difficult arguments.

This seems so obvious to me, but it also is obvious to me that it is not obvious to lots of other people: extrinsic motivations don't work. Punishment is an extrinsic motivation.

My doctor wants me to eat right and exercise. If I'm doing it to please him, eventually I'm going to be tempted to lie to him. That will become the only way to get his approval, if I haven't been eating right and exercising. And if what I want is his approval, then telling him what he wants to hear gets me what I want.

Isn't it better, though, if rather than his approval, I choose as my goal improving my health because that's what *I* want. In that case, I eat right and exercise because it's what I want to do. And if he fails to voice approval at my next appointment, it doesn't exactly set me back.

If you are a parent, you will instantly understand that if you change your child's heart attitude, you'll get far better results than if you just rely on rewarding good behavior and punishing bad behavior. But changing heart attitudes is a lot harder, so most parents rely mostly on the behavioral thing.

So how do you get people to "do the right thing" if you don't punish them for doing the wrong thing? Well, I think it's pretty tough. In an ideal world, you'd carefully hire people who are intrinsically motivated. I think you also have to nurture a culture that values intrinsic motivation, and where top performers challenge everyone else to match or exceed their performance. A culture of excellence relies on peer encouragement and tough standards for praise.

Punishment should be reserved for people you expect to fire. If they have an attitude change, maybe you won't fire them. But punishment should be the shot across the bow.

Anonymous said...

Do you think that it's time to deal with the incident with the patient and the involved parties privatly now that you've received feedback? After reading some of those WSJ comments, things are getting out of hand and I don't want you to regret being honest. Transparency shouldn't trump respect.

Anonymous said...

Many thanks, anon 6:24,

Part of that comes with the world of blogging. Some people say things in ways they never would if they were dealing with you face to face or if they had to sign their names. Most of my readers, though, are quite respectful and value the interchange.

Rest assured that we are dealing with the participants (patient and staff) in a private and appropriate manner. I think they understand why I have posted these questions for the purpose of debate. The medical field needs some of this to be more out in the open, even if some commenters feel a need to be harsh.

Anonymous said...

Hello Paul,
Thanks for another great post in the world’s best live workshop on patient safety. I know that you know you have a global audience – myself included – who appreciates your openness and candor as we all work through this thing about punishment as a pathway to clinical (or organizational) improvement.

When I was about half the age I am now, I served as the first chair of hospital-wide quality assurance in a small (150 beds) community hospital. In those early days, the culture of punishment and shame was the rule rather than the exception. Happily, we are recovering as Lee Carter says.

There are a couple of points worth making about Dr. Denham’s description of Ives-Erickson’s approach. He is right on the money in saying the goal is to create a “healing constructive opportunity” – for an upgrade in the clinical process/workflow. The trick is to sustain the healing and constructive climate long enough for the team to do the work of discovering that upgrade. Punishment – and its organizational antecedents: prosecutorial investigation, trial and sentencing – serve only to kill off all opportunity to see the true cause of the error as all involved are forced into a defensive posture.

The problem with having the manager take sole responsibility – “… errors stem from system flaws …” – is that it takes the nurse off the hook and puts the manager firmly on it. This is not the most powerful way to investigate the sources of what really happened. None of us are off the hook in an such a critical incident. Rather, we are co-responsible for both the error and the upgrade.

From a process/workflow perspective, mistakes are the doorway in to the inner workings of a process which ordinarily are hidden or glossed over by the kind of language seen in most procedure manuals, for example. So holding the door open long enough is critical to being able to see the true source of the error. This idea is embodied in “Ask why 5 times” as practiced in the discipline of Lean.

“Think about it” – you say. Wow. It’s having me think about safety, quality and continuous improvement in new ways I have never done before. Thanks.
/Dr. Pete

Anonymous said...

Excellent blog.

Medical care is at its best when there is a partnership between patient, providers, and payers. The belt and suspenders approach is to have the patient "double check" and draw attention to safety.I understand in this case, though, the patient was marked.

One corollary is lab work. Often, the patient's default assumption is that no news is good news. But, that default assumption can be changed. Post a sign in the doctor's office- If you do not receive your results in one week, for example, you will receive your money back. The patient now understands that no news is a failure of the system; a failure that can be corrected by a phone call.Incidentally, this is exactly how it works when you buy a bottle of water at the airport. there is a sign that says if you do not receive a receipt, your purchase is free. The rationale: the owner as co-opted my support as a purchaser to guard the register. This prevents the vendor from pocketing the money.

Jeff Segal, MD

Anonymous said...


The details matter to me in this case because, as you put it, "it depends" on the circumstances of each case, when trying to assess the balance between a no-blame approach and accountability. Most medical errors result from more than one small error adding up to disaster. If the correct side was marked, then was the correct side draped? Does the surgeon participate in the draping? I am trying to assess the role of each team member in the sequence leading up to the lack of time out, rather than just fixing my attention on the one error of no timeout. Each error is equally important.

I am still stuck on this issue of not following policy, as the BIDMC
nurse alluded to (more harshly than I). She is right that "forgetting" to follow policy would result in harsh punishment elsewhere (I am thinking of how we treated phlebotomists who failed to follow policy for drawing the proper patient) - you just can't make a system that is absolutely error proof if people don't follow the policies in the system. Humans have to touch the patient, after all! This is an unresolved issue in my mind.


Anonymous said...

While I am not convinced that punishment is needed, I still have trouble accepting this as a systems failure. The system put in place is fail proof. The problem is that the culture put in place allows/accepts the system to be bypassed. Imagine if pilots one day decided to selectively complete checklists? No one in the OR spoke up to say, "hey, we need to take a time out." Clearly, this wasn't the first time a time out was skipped. Change the culture not the system.

Anonymous said...

Oh, seriously! If someone is asked right off the bat, "did you do this on purpose," do you really think they'll honestly say yes? More likely, their first instinct will be to defend or justify their action, and right away the interview will spiral off into emotion rather than fact.

The first question should be: "Tell me what happened." And don't be naive enough to think you'll get the whole story through an interview with just one employee.

Facts always need to come first.

Anonymous said...

I’m struck at the conversation ongoing on the blogs around accountability and responsibility for error. The approach taken by BIDMC is certainly consistent with everything I have learned. It is also a piece of a journey. Staff, no matter how expert, suffers from being human. As such, they will make mistakes. We as leaders, in conjunction with our staff, patients and families, must put in place systems that support safe practices and mitigate the chances of error getting to patients and causing harm. As leaders, teams, and individuals, we have both shared and individual accountability and responsibility. That isn’t something that you define as you go along. Instead, it is captured by an organizational expectation of a fair and just culture that is memorialized in an informed policy/practice, approved at all levels of the organization including Governance, widely shared and practiced. When an event occurs, accountability and responsibility is judged against this fair and just culture. Everyone knows, in advance of an event, how accountability will be considered. Julie Morath and Minn. Children’s has done great work in this area as has Maureen Connor and Dana farber Cancer Institute. David Marx and Allan Frankel are national content experts and the overall area of accountability has been a priority focus of the Mass Coalition for the Prevention of Medical Errors.

There is no question. It is often “easier” to say “we fired the staff, we rewrote the policies, it won’t happen again.” It’s just rarely if ever effective. Thirteen years ago at DFCI we learned we could have fired everyone and done nothing to improve the medication system or reduced the chances of a preventable death from happening again. Our systems must support safe practice and the culture must be fair and just. Organizations and individuals deliver high quality safe care in this context.

Jim Conway
Senior Vice President,
Institute for Healthcare Improvement

Unknown said...


Excellent approach to tackling this issue head-on. Kudo's to you and your senior leadership staff for developing a culture where this could be reported and a root cause analysis done to determine the true cause of the error.

Also, your comments regarding punishment, protocol and policy change, etc. were also dead on. It is very easy for those outside health care to attempt to look inside, apply rules from a coal mine and think we can operate our hospitals in the same manner.

Finally, I think you were very accurate in saying that the involved staff were emotional not because they were caught, but because they saw the face of their error and had to look deep within themselves. As a rule we do have good, caring people in health care and I know from experience how these events can affect physicians, nurses, techs, even administrators.

Excellent blog, I will check back!

Russ Gardner, MBA, CMPE

Anonymous said...

Paul, when you asked the surgeon why he didn't call a time-out, what did he say?

Anonymous said...

He was distracted by thinking through the procedure, i.e., mental "rehearsing". It was a procedure that he felt required careful consideration, and he was therefore focused on the actual execution of it.

Anonymous said...

Dear Concerned in Boston,

No, I am not writing about that topic. I would be happy to look into the issues you mention if you want to send a summary to my office.


Anonymous said...

My question is why nobody in the room seemed to notice that the surgery was not needed as an incision was made into the wrong side? Why did the surgery proceed if there was nothing to be fixed by doing it?

Have we hit a point of "auto-pilot" in the operating room that nobody can say "Hey, I don't see anything wrong here that requires this procedure?"

If we are going to address errors, then figuring out what allows sensible people to remove healthy tissue would seem to be part of the examination.

Anonymous said...

anon 6:13:

We are not told what kind of case it was, but not all abnormalities are externally visible. If the case were a knee arthroscopy, for instance, the surgeon would not know that the knee was normal till he got the scope in there.