Perspectives on Safety
In Conversation with…Thomas H. Gallagher, MD
Editor's note: Thomas H. Gallagher, MD, is Associate Professor in the Department of Medicine and the Department of Medical History and Ethics at the University of Washington in Seattle. Dr. Gallagher's current research covers the disclosure of medical errors, examining patients' and doctors' attitudes about disclosure, how best to train providers to disclose and apologize for errors, and how to create a system that promotes appropriate disclosure. We asked him to speak with us about new developments in the field of patient disclosure and apologies.
Dr. Robert Wachter, Editor, AHRQ WebM&M: What got you interested in the topic of disclosure?
Dr. Thomas Gallagher: I've always been interested in difficult conversations between doctors and patients. Some of the early work that I did looked at doctor–patient conversations about financial conflicts of interest in managed care. That led to my interest in an even more challenging conversation, which is how doctors and patients talk with one another when there's been a problem in their care. The more I've worked on this topic, the more interesting and complex it becomes. We're only just scratching the surface in this area.
RW: In your practice, is this an area you've struggled with or an area in which you've done particularly well over time?
TG: I think it's an area that all physicians struggle with. I have yet to have a single physician act like disclosing errors to patients is easy or isn't a relevant issue for them. Part of my interest in the topic came from experiences when I was a resident. The literature at that time was trying to highlight disclosure as an ethical and moral imperative and that made it appear to be a relatively straightforward undertaking. You just went and told the patient what happened. As a resident I saw that, even when a clear-cut error had happened, thoughtful people could really struggle with whether it should be disclosed and, if so, how much information should be shared with the patient.
RW: When you talk to practicing clinicians about disclosure, what are the main challenges they raise?
TG: It's clear that health care workers endorse the concept of disclosure. Some people are trying to improve the practice of disclosure through what you might call moral exhortation. They try to remind health care workers that this is a professional and ethical duty, and then the implication is that we need to up our quotient of moral courage and just go do what we know is the right thing to do. But when you talk with health care workers—and this is what we've found both anecdotally as well as in the surveys—they really endorse the concept of disclosure, but struggle with its implementation. Among the key barriers we have encountered, one is clearly worry about the impact of disclosure on litigation. But that was not the only barrier, and for many clinicians it was not the most important barrier. Lots of clinicians really struggle with the concern that sometimes disclosure may not be beneficial to the patient. They worry that disclosure in some circumstances might harm the patient or their family, make the patient unduly anxious, or lower their compliance and their trust. So there are circumstances in which health care workers believe they're being patient centered in their decisions about disclosing limited information or not disclosing anything at all. The other barrier is the challenge of how awkward and uncomfortable these conversations are. Most physicians have told us that they haven't had any formal training in disclosure and are uncertain of just what to say.
RW: What has the research told us over the last 5 years about how patients react to disclosure, and what is its impact on malpractice?
TG: There are areas where the research is very clear and areas where the research is still in development. It's clear across a number of studies that patients want harmful errors disclosed to them, even when the harm is relatively minor. However, our research suggested that patients also worry that health care workers might hide errors. Patients think this is just human nature—if an error's happened, of course, health care workers might want to keep that to themselves. It's also pretty clear what information patients want, at least in broad brush strokes. Patients want someone to explicitly tell them that there's been an error, share a little bit of information about what the error was and what its implications are, and talk about why the error happened and how recurrences will be prevented. Those last two matter a lot to patients because they want to sense that a lesson has been learned from the event, and they want someone to apologize.
The literature is less clear on how those general preferences play out in different situations and patient populations. For example, we know very little about how preferences for disclosure might be different in different cultural groups. We also know very little about how patients want disclosure to happen in the moment. One of the major shortcomings of the research on disclosure has been that most researchers have either talked with patients about disclosure of hypothetical errors or have talked to patients long after they experienced an actual event. We know much less about how a patient who has just experienced an adverse event or error would want disclosure to take place.
Health care workers clearly endorse the concept of disclosure, but really struggle with the implementation so are unsure what words to say. Our research has shown, for example, that physicians are unsure even about basics of disclosure content, like should you say the word "error"? How much information should you give the patient about the error? If you're going to offer an apology, should it be an expression of regret ("I'm sorry about what happened to you") versus a more formal apology that admits the error and accepts responsibility? It appears that internal medicine physicians approach disclosure differently from surgeons. For example, internal medicine physicians are more likely to share information about the error and to offer a formal apology. Surgeons are very enthusiastic about the concept of disclosure but provide much less information about the event to the patient.
We also know that physicians' disclosure attitudes are not as closely connected to the external malpractice environment as people once thought. We did a study that involved physicians both in the United States and in Canada. Canada's malpractice environment is less onerous than in the United States. Yet Canadian and U.S. physicians' disclosure attitudes are much more alike than they are different.
We know a little bit about the general impact of disclosure on litigation, but that literature is pretty early in its development. Some studies suggest that effective disclosure and apology might have a beneficial effect on litigation. Those studies have basically asked patients who have sued, "Why did you sue?" A common reason is, "I felt like the truth was hidden from me" or bad doctor–patient communication skills. So that literature would suggest that better disclosure might reduce litigation. We know from the field that a number of organizations have adopted policies of open disclosure and coupled them with early offers of compensation and have seen positive results on their overall litigation experiences. We describe some of those programs in more detail in our review article. But there is another side to the story. The vast majority of patients who are injured by their care never sue. Some health law scholars worry that one reason the number is so low is that patients were unaware that a medical error caused their injury and that if we start disclosing more, litigation will increase. Just because you say you're sorry does not mean that the patient won't sue you; there will still be cases in which disclosure precipitates litigation. However, I think that disclosure, on the whole, will likely have a beneficial effect on litigation.
RW: Talk about the role of laws that many states are passing regarding this issue, holding people harmless for apologies. What is the role of these kinds of laws?
TG: One approach that 35 states and the District of Columbia have taken is to enact laws that try to promote apology by protecting part of the apology and disclosure statement from being used as evidence in litigation. The important thing to know about these laws is that the protection varies widely. More than half of those states only protect the expression of regret. They protect the words, "I'm sorry," but they don't protect the statement about what happened. If you admit liability elsewhere in the disclosure statement, it's still admissible. Four states protect the entire disclosure and apology statement from being used in litigation. But even in those states, this doesn't mean you cannot be sued because you disclosed an error. That disclosure statement cannot be used in court, but the patient can still use independent means to prove that negligent care was provided. Because of the limited protection that these laws offer, their impact on the practice of disclosure and apology is likely to be relatively modest. Eight states require the disclosure of some adverse events to patients—usually the most serious of these events. Two of those states actually require that the disclosure be done in writing. There are two things that are interesting about these disclosure laws. One is that in a vast majority of the disclosure laws, the disclosure burden is on the institution rather than on the individual health care provider. This is a shift we're going to see over the next few years—more and more disclosure being considered an institutional responsibility. Second, these disclosure laws don't provide any input about what the content of disclosure should be, and it's unclear how these laws are going to be enforced. So, like apology laws, they're a useful public policy endorsement of the importance of transparency, but I'm not sure that they're likely to have a major impact on how things are happening at the bedside.
RW: Another challenge is that medicine is a team sport. Often, multiple physicians and scores of providers have touched the patient. How do we deal with that, and whose ultimate responsibility is disclosure?
TG: It's a very interesting area, which is how do you talk with patients about other health care workers' errors. Like other areas of disclosure, we really are just at the beginning of trying to understand how disclosure works as a team activity. We completed some focus groups with nurses, asking them about their disclosure attitudes. It's clear that nurses want to be involved in disclosures, in part because they worry that if they're not in the room, the doctor will blame them for the error. Now, if disclosure is ultimately an institutional responsibility, institutions will need to have some formal mechanism for helping teams sort through these events and then formulate a disclosure plan. What I think is so interesting is that disclosure in many respects is a conflict between the patients' right to know this information and health care organizations' or health care workers' need to protect themselves from the negative consequences that might flow from disclosure. When there are conflicts of interest in other walks of life, we don't leave it to the people involved in the conflict to sort out whether or not there's a conflict and, if so, what to do. We have external entities or someone who is more neutral who can help sort out if there is a problem and if so, how to handle it. When I've been involved in helping institutions sort through some of these complicated cases, what's so striking is how difficult it is for the institutions to put their self-interests aside and really approach the disclosure process in a patient-centered way. I wonder if over time there might be a role for patient advisory committees that many institutions have in helping the institutions take a more neutral approach to deciding whether and when to disclose these events to the patient. Other institutions are relying more heavily on their ethics committees to help them with these complicated decisions. It will be interesting over the next few years to see how institutions, as they shoulder this burden of disclosure, try to devise mechanisms for bringing some objectivity and patient-centeredness into the process.
RW: What have you learned about providers' needs in terms of training? And what evidence do you have that people can get better at this and that their attitude toward disclosure changes after they've been trained?
TG: There are exciting developments in this area of training. The National Quality Forum issued a Safe Practice about disclosure last year. The practice calls on institutions to provide expert disclosure coaches who can help health care workers around the clock with just-in-time advice about difficult disclosure dilemmas. Many institutions are taking innovative approaches to training those coaches and making that resource available. We have an AHRQ-funded grant right now using simulation to train health care workers in team disclosure and communication skills. Embedded in the simulation is a disclosure coach, who is trained to provide that type of feedback. The health care workers who participate in the simulation really enjoy the training experience. They love the input from the coach. I think they see that this is a skill that you can improve with some practice and some feedback in a safe setting. I'm envisioning that over time health care workers will get background education in disclosure, and that those who are most likely to do disclosure, such as surgeons and other invasive specialists, will get more intense training, probably using simulation. For all health care workers, there will be the availability of just-in-time training provided by medical directors or risk managers who can help them formulate an effective disclosure strategy.
RW: What are common mistakes in disclosure?
TG: Lots of people assume that the most common mistake is hypo-disclosure, by which we mean not disclosing enough information to patients. And that clearly is a problem. But we also see health care workers experiencing the opposite problem, which is hyper-disclosure—telling the patients way too much information, more information than the patient would want to hear. Or the worst, we call mal-disclosure—disclosing information that turns out to be incorrect. That happens because health care workers aren't aware of the institutional resources to help them in their disclosure process. They rush into the patient's room. They tell them, "I feel so terrible. We made this horrible mistake." They offer profuse apologies. And then when the dust settles and a thorough analysis of the event has been conducted, they find that it wasn't an error after all, or it was an error but the error didn't harm the patient. And then you're in the terribly difficult situation of having to go in and un-ring the disclosure bell. You can imagine how mixed up that leaves the patient. The bottom line is, health care workers need to balance this urge to tell the patients what happened right away with availing themselves of the institutional resources to really help them understand what should be said and to conduct a thorough analysis of the event. Of all of the things we owe the patient, we certainly owe them information that's accurate and truthful, and many health care workers are unaware how difficult it can be to get that information.
RW: Let's say a patient on your watch experienced a terrible medication error and was now in the ICU, and you were going to speak to the family. How would you start that conversation?
TG: Health care workers ought to see disclosure as involving at least two phases. The initial conversation often is relatively limited. It involves going in, quickly ascertaining what the patient knows about what's happened, and then, in a very straightforward way, trying to describe what is known at that time about the events. If it's clear that it was an error, I think it's fine to say that an error has happened or let the patient know it appears that this was a preventable adverse event—only when it's absolutely clear that's been the case. Here's what we know about what happened, and what we've done to mitigate its impact on you. I think an expression of regret about what happened at that point is also appropriate. Then let the patient know we're going to look into this event carefully and we'll be back in touch to let you know what we've found. Then ask the patient what questions they have. The key thing is, after that initial conversation, make sure that there's frequent follow-up with the patient to let them know what we've found about what took place. Here, the balance is sharing the results of error analyses with the patient versus the institution wanting to keep some of that information shielded under their peer review protection. It's a delicate conversation that takes place between the health care workers and the risk managers and the institution itself. But that follow-up conversation usually is an opportunity to again explain what happened. Share the results of the analysis to the extent that you're able. And when it was an event that was due to a clear-cut error, a formal apology that accepts responsibility is appropriate at that time. Then again providing the patient an opportunity for questions. It's important for health care workers to realize that this is a multi-step process.
RW: I imagine that some patients are comforted by the honesty and by what you and the organization are doing to mitigate the harm to that particular patient and fix the problem so it doesn't happen again. I'd also imagine that some patients will lose trust in the provider or in the organization. How do you deal with that?
TG: We've watched a lot of physicians take different approaches to disclosure in some of the research that we've done. I think that health care workers are most effective when they're very direct on this trust issue. For example, when a harmful error has taken place in the operating room and a second procedure is required, I've seen some surgeons in a very skillful way come right out and say to the patient, "I know how important it is that you be able to trust your surgeon. I am hopeful that you'll allow me to do the second procedure and I feel qualified to do so, but if you don't feel comfortable with that I'm happy to arrange for one of my colleagues to come by and talk with you about doing the operation themselves." It's much better for health care workers to be up front and address that trust issue because, you're right, patients will appreciate the honesty and forthrightness, but they may have reduced trust in your competence. Dealing with that issue directly, the majority of the time, patients will feel comfortable proceeding with the original health care team. However, it won't be across the board, and making arrangements for patients to receive care in a different setting keeps the patient's needs front and center in the disclosure process.