Editor's note: Albert Wu, MD, MPH, is Professor of Health Policy and Management at the Johns Hopkins School of Public Health. A leading expert on disclosure and the psychological impact of medical errors on both patients and caregivers, he may be best known for coining the term "second victim" in a 2000 British Medical Journal article. We discussed the second victim phenomenon with him, including what is known about efforts to ameliorate the toll that serious medical errors take on providers.

Dr. Robert Wachter, Editor, AHRQ WebM&M: In 2000, you coined the term the second victim. Tell us what you meant by that.

Dr. Albert Wu: The term came from the recognition that when a patient is seriously injured by health care, there are almost always two victims and one of those victims is obviously the patient. A second victim is the health care provider involved in the incident who feels in some way responsible and is emotionally traumatized by what happened.

RW: I noticed your article came out in March of 2000, a couple months after the Institute of Medicine (IOM) report on medical errors. How much linkage was there in your thinking about this concept of the second victim and the main theme of the IOM report, which was that most errors are system problems rather than people problems?

AW: I had been thinking about this issue for a few years. In fact, I began thinking about it when you and I were RWJ Fellows in San Francisco and I did a study asking interns and residents to describe their worst mistake. It was supposed to be anonymous, but one came up to me and apologized. I said, "Don't apologize to me." And she said, "No, I'm really sorry. I wanted to fill in your survey, but I just couldn't." I realized pretty quickly what she was saying. She meant that she had an incident that she thought she should tell me about but couldn't bring herself to do it. Already I had realized that we did not handle people that well. Over the years, we gradually began to realize that most things are at least partly the responsibility of the system: things are built into the system that allow other things to go wrong. But even though individuals are often not responsible at all for things that go wrong, they still feel responsible.

RW: So did the shift in the patient safety field, emphasizing the role of systems, make it easier for individuals to feel a little bit less responsible? Does it change the nature of the second victim?

AW: I think it does. If people really absorb that message, if they truly internalize that they're not responsible for everything that goes right with patients, and as a corollary, they're not always directly responsible when things go wrong with patients either. That can be comforting, but I think that individual attitudes and beliefs have lagged a good deal behind that thinking. We still like to think that we can do positive things for our patients, and we still believe that when things go wrong that it's really our fault. So if people would learn that lesson more thoroughly, I think they would take comfort from it.

RW: Tell us about findings from your research on second victims and any other research over the last decade or so. What do we know about what happens to providers after errors?

AW: I think that there are two ways that health care workers are traumatized. The first are things we do to ourselves; these are self-inflicted wounds that result from internalized judgments. When we do things wrong, we feel bad: not surprisingly, we feel anxious, depressed, and demoralized. On the other hand, some health care workers are actively victimized by elements of the system or by other organizations. Those are people who work in places that still believe in shame-and-blame or crime-and-punishment, and they're sometimes official bodies like nursing boards and so forth. They actually punish people and, I think, add insult to those self-inflicted injuries. There has been some survey research, and an unsurprising finding is that after these things happen, people feel bad for a while, days or even months. What's surprising is the extent to which it happens.

One survey at the University of Missouri found that 30% of people surveyed reported feeling persistently bad and at least somewhat impaired during the year after an incident in which they consider themselves a second victim. This is a very prevalent problem. Some people have persistent problems—the incidents go on to cause lasting harm. Some people are so depressed that they need to be hospitalized; some people are suicidal. People quit medicine or nursing because of what they did. I hear lots of stories about these things. I was told recently about a surgeon who was one of the highest volume laparoscopists in his region who was doing a pretty simple procedure. The patient began to bleed, and bled out in the operating room. The surgeon cleaned up, left the hospital, and never came back. He left and stopped doing surgery. In our hospital, we had a very unfortunate incident in our emergency department. A boy with head trauma came in and was badly injured. He needed to be intubated, could not be intubated and could not be trached, and died in the emergency room. It was just a horrific incident and everyone felt unimaginably bad. But even though the emergency physician got some support, others were a little critical of his actions, and he quit clinical medicine within the year.

RW: What do we know about risk factors for people being very traumatized from these events?

AW: Not surprisingly, people who were already under a lot of stress or had preexisting conditions that predisposed them to be anxious or depressed are much more likely to have subsequent problems. What is more interesting and perhaps not so self-evident is that people who felt supported—who were able to have discussions with clinical peers and even in one study people who talked to the patient about the incident—reported having better long-term psychological outcomes. But we're pretty short on longer term data. Even where there are starting to be interventions, we really just need more evaluations.

RW: So let's shift to those interventions. I assume there's some piece of this that relates to the overall culture of the institution and how it handles errors. If an organization wanted to start a program, what would that program look like? Are there examples of innovations that you think are very positive?

AW: First of all, I think that interventions are really needed. Every survey suggests that even though people need support, they usually don't get it. Not only do they not get it, there's a general lack of recognition that they might need it. As a result there's limited self-learning, and there's also pretty limited organizational learning after this sort of incident. Most people look for internal support first from peers, and probably more people than any talk to family members, but they really want and feel like they would benefit from having a detailed clinical discussion with someone who understands the nature of the work. They can explore how to get some meaning from the incident. A few programs are underway, I would say the two furthest along are something that is called MITSS, or the Medically Induced Trauma Support Services, which was started up by Linda Kenney, a patient who was a primary victim of medical care, and Rick Van Pelt, a physician at Brigham and Women's Hospital. They together set up this system recognizing that both patients and providers need help after these incidents. What they basically do is provide either one-to-one support, or group sessions if it's a major enough incident, for both patients and physicians. I think they recognize that in many ways the responses of patients, families, and health care workers are similar. People feel badly, and we have a limited repertoire of responses. We really feel bad when bad things happen. We all want to behave like human beings.

RW: Are these counselors primarily doing the work? Or are they peers?

AW: Some of them are counselors, but a lot of them are peers trained to provide emotional support, as well as to provide clinical input at a time when the physician or nurse really needs help interpreting what happened. There is a well-developed system at the University of Missouri, led by Sue Scott. They call it the ForYOU Team, and it's a 50-plus member rapid response team on-call 24 hours a day to help providers when there are bad outcomes. Those people undergo a detailed training program in patient safety, counseling, the idea of the second victim, and they also receive information about resources that exist, because they realize that some people are going to be fine with peer support, some need more expert support, and some need intensive professional services. It's very important to triage and make sure that people get referred when they need to.

RW: Some of these models blend peer support and counseling with clinical feedback, and you can envision a time when those are in conflict. The clinical feedback is, in fact, you did make a terrible error. This is not a systems problem, you blew it in some way, and that may be obviously the last thing that someone needs to hear when the job is to support them through it. How do you balance that?

AW: Well, they may conflict, that's the truth. I think understanding what happened and what could have been done, or the lessons that might be learned, can be sobering and perhaps not what you want to hear. But in most cases it does provide meaning. When something goes badly wrong for any of us, we spend lots of time ruminating and turning the thing over and over in our minds and wondering; part of the distress is from the uncertainty. What people would really like to do is be able to talk to a trusted mentor, someone who is not judging them and someone who is not their boss. Someone who is not evaluating them, but who can provide an objective sophisticated assist to help them think through what happened. We did a survey recently and asked people what was comforting to them when they experienced something of this nature, and several of them said that the person helped them to understand what happened and what steps might be taken in the future. It's really interesting that both patients and providers very early on are interested in knowing what can be done to prevent future incidents. I think it's because we have such a disbelief that these things happen that when something really goes wrong, it's an existential crisis—we go on a mad search for meaning and understanding, so that learning that things can be done and will be done is comforting by itself.

RW: The literature has emphasized how patients value apologies and that it doesn't appear to increase the risk of losses. What is the role of the apology for the provider?

AW: Nancy Berlinger and I wrote a piece about apology and how, in Western culture, apology may be the first step—a kind of penance that allows one to be forgiven. But she emphasizes very strongly that you shouldn't apologize as a way to seek grace, or to escape personal responsibility or liability. You're doing it because it's the right thing to do, and there should be no, what she calls "cheap grace." I suppose that phrase might be more meaningful to someone who is Catholic, but the apology is for the patient. On the other hand, it can be a relief to unburden yourself of something, to say or do something that you know that you should be doing. But I think that is a secondary gain for our second victims. Disclosing is not as risky as in the bad-old days, and as hospital lawyers and malpractice companies used to assert. One can worry, is there a tension between disclosing and protecting oneself or institutions from medical malpractice? I think the bottom line on this is that the initial disclosure discussion winds up being something of a malpractice-free zone. Rick Kidwell, our former hospital counsel, used to advise us to tell the patient what happened, be honest, and that this was an expected part of a good relationship between doctor and patient. Incidentally, an apology has never been entered as evidence in a lawsuit. If anything, plaintiffs' attorneys hate it when doctors apologize because they can seem sympathetic. He also said that in a really bad outcome, anything that's really important will come out in deposition. He said, "Don't worry about it. You're having that first conversation, it's not being recorded, it's you letting the patient know what happened and that you feel bad, and it's what's humanly expected."

RW: I reread your 2000 BMJ "second victim" piece. In the first paragraph you talk about a house officer who committed an error. Then you write, "The news spread rapidly. The case tried repeatedly before an incredulous jury of peers who returned a summary judgment of incompetence." It struck me that in our zeal these days to try to disseminate messages from errors more broadly, to try to prevent them from happening again, we sometimes do have conference after conference discussing our errors where either the person who is the middle of it is the presenter or has to sit through all of this. How do we balance this tension between an organizational imperative to try to disseminate messages from errors and what must feel like this bad movie being shown over and over again to the second victim?

AW: That's a really interesting question. On one hand, in the case that I described from my house officer days, it was really the anonymous mob who gossiped for several days about this particular incident. It's a terrible feeling to be on the receiving end of that, be the subject of all that gossip, but I think today we are getting better at handling incidents, especially with patients and families. It's possible to have a good disclosure but poor follow-through, or a good disclosure but to not handle the rest of the case particularly well. In many ways you could say that we're doing better, but these days that we're feeling worse. We're handling patients well, but even at our own institution, providers will sometimes say that they feel like they've been thrown under the bus. In a way they suffer at the expense of the rest of the organization's learning about systems of care. I think part of the reason for that is that we are not fully evolved. We've moved from thinking about patient safety in the first place to understanding we need to disclose and we need to report to reporting systems. But we haven't uniformly reached the next point yet, to understand that there's always a second victim in a really bad adverse outcome and that we also need to take that into account. We need to deal much more comprehensively with these incidents so that we take care of ourselves too. We really cannot afford at this point to lose the kind of people who actually feel bad when things go wrong.

RW: And we probably on the other hand have seen people for whom the act of presenting the case in another forum was an act of altruism for which they felt quite good. In some ways it was redemptive.

AW: I think so. It can be cathartic. I feel this myself in talking about my clinical errors over my career. Other people too, have said that they feel that others are benefitting even if they are rubbing a little bit of salt in their own wound. They feel relieved that there could be some benefit from these incidents.

RW: You mentioned that you thought the no blame approach was somewhat helpful in terms of mitigating the harm to the second victim. Is there a risk that things will get worse again if we shift back toward a more accountability-focused framework?

AW: Well, I don't think that being open about these incidents with patients and families, and ourselves, should be a way to escape personal responsibility, or even to escape liability. There is and will always be a very important role for personal responsibility. There are many cases that are more purely caused by the system, and there are cases where people are truly negligent or violated rules. We do need to accept our own responsibility as parts of the system, and since we are often closest to the patient at the sharp end of care, we need to perform as well as we possibly can. So a little bit of pendulum swing back toward personal responsibility is fine. On the other hand, I don't think we can afford the crime-and-punishment approach, and to be purely selfish, we need information so that we can manage risk and improve safety. We need health care workers who can function as well as they can, and if they are impaired because their needs have been ignored, then our own institutions are going to suffer. But we do need to hang on to people who really care about their patients' welfare.

RW: If you were the CEO or chief medical officer of a hospital and building a program to support second victims was going to cost significant resources, would this rise up high enough on the agenda to spend some scarce resources?

AW: Absolutely. We're doing this now at Johns Hopkins. Some internal survey data suggests that, if nothing else from a strictly pecuniary point of view, there are a lot of burnt-out health care workers, a lot of absenteeism, and a lot of staff turnover because people are not getting the support they need. That by itself might be a convincing argument. And besides that, it's the right thing to do. So I think this is one of those rare instances in which it's quite easy to see how an institution might benefit from investing in improving recognition and systems to deal with the problem of second victims. Our own plan involves a three-pronged approach of awareness, intervention, and prevention. We need to increase awareness of the problem because all health care workers have the opportunity to be the first responder to colleagues in these incidents. We are trying to develop some interventions that may include a crisis response team, peer response, or a reinforced peer support system. We'd also like to do some prevention to improve resiliency. Of course we want to reduce adverse events, but we also want to look at the current system and see if it's promoting or allowing more damage from the second victim problem, such as more staff burnout.

RW: Do you foresee any major changes in this landscape in the future?

AW: I think 5 or 10 years out, we should have a more comprehensive strategy for handling adverse events. We now have policies and standard operating procedures for handling the patients. We need to have policies and procedures for handling health care workers as well. And we need training and support systems including just-in-time support for both patients and their families, and health care workers. If we have a more comprehensive strategy, we'll have a better chance of learning from errors, taking care of our patients, and taking care of each other.