Editor's note: Michelle Mello is Professor of Law at Stanford Law School and Professor of Health Research and Policy at Stanford University School of Medicine. She conducts empirical research into issues at the intersection of law, ethics, and health policy. We spoke with her about legal issues in patient safety.

Dr. Robert M. Wachter: What got you interested in the confluence between law and health care?

Ms. Michelle Mello: I got interested in this area from personal acquaintance with physicians and understanding their narratives of how legal pressures affected their clinical practice. It became clear very quickly from hearing these narratives that the law is working in ways that, at least to physicians, are counterproductive and that beg the question of how we can make law work better for medicine—not just for doctors but for the quality of medicine.

RW: Tell us a little bit about what health care people don't understand about law and lawyers, and vice versa.

MM: The number one misperception that physicians have about malpractice law and lawyers is how much risk they are at and specifically the role of the plaintiff's lawyer in creating that risk. Most physicians tend to radically overestimate the probability that they will be sued out of any given instance of negligence. One study found that it was off by about a factor of almost 30. In reality, we know from pretty well established evidence that only about 2% or 3% of instances of negligent injury ever result in a malpractice claim. But physicians perceive that they're under siege by the malpractice system, and they blame plaintiff attorneys for this.

Another common misperception is that plaintiffs' attorneys are out there trolling around searching and searching to find anyone who could be scraped up as a plaintiff. Plaintiffs' attorneys turn away about 90% of the potential claimants who contact them for a variety of reasons that have to do with the strength of the case and how easy it will be to make that case in court and also how large the damages are. It's actually quite hard as a claimant to find representation. It's impossible to bring a malpractice claim without representation. That leads to the reality that most instances of negligent injury never are addressed in the legal system.

RW: What do lawyers misperceive about doctors?

MM: It's hard for me to say for sure. I don't tend to consort too much with plaintiffs' attorneys, but I suspect that they tend to underrate the extent to which physicians are working very, very hard not to hurt their patients—not only out of fear of malpractice liability but because underneath every malpractice suit is a patient that they cared about who got hurt. Physicians have very strong intrinsic motivation to avoid hurting their patients. They don't really need the threat of liability to make them practice safely. The perception of some malpractice attorneys may be that physicians' default will be to be careless, and it is only through the intervention of tort liability and the plaintiffs' bar that we keep physicians practicing honestly and safely.

RW: Tell us a little bit about the underpinnings of malpractice and tort legislation. What's it supposed to be accomplishing?

MM: The common law has evolved to try to serve three functions. One is to get compensation to people who have a meritorious claim. The second is to provide some sense of corrective justice, which is a squishy concept, but refers to the notion that having your day in court, telling your story, and having somebody held accountable provides a sense of restoration to somebody who has been harmed—in a sense, makes them whole. The third function, and arguably the most important, is deterrence—that is, to encourage medical care providers to engage in the socially optimal level of precaution taking. Not to prevent every accident, because that would be costly in ways that we don't want to encourage, but to practice reasonable care and avoid foreseeable accidents. Malpractice law is intended to make health care safer.

RW: Malpractice law has been around for a long time, and physicians have complained about it for a long time. How did the dynamic of that relationship change when the patient safety field became a thing 15 years ago or so?

MM: There have been a couple of things. One is that people became much more sensitized to the prevalence of medical errors. Many malpractice attorneys even started using the Institute of Medicine's report on medical error in their marketing materials: "Did you know that between 44,000 and 98,000 people a year are killed due to medical errors?" This notion that medical error was a commonplace occurrence and something that patients should routinely think of when trying to explain a bad outcome in the hospital became much more common after the publicity over the IOM report and the plaintiffs' bar making use of that. The other thing that has challenged malpractice plaintiffs in recovering compensation is that, with the advent of the patient safety movement and additional research on the causality of medical error, we become more aware that errors are due to many causal factors, and that they cannot always be immediately pinned on an individual health care provider, much less on fault by an individual health care provider. Trying to recover for errors caused by systems breakdowns and systems of care can be challenging. It's not easy to hold hospitals legally accountable for care, and when the error occurs because of an interaction between humans and their environment, that too can be difficult to explain to a jury in a way that results in a judgment for the plaintiff.

RW: Does that actually come up in court where in the old days, it would have been logical for everyone to point at a physician and say this was an error and therefore that's the appropriate focus, but today people are thinking more in terms of systems and understanding its complexity?

MM: I think the sophisticated attorneys are. It's important to remember that only a very small percentage, probably less than 5%, of malpractice cases ever get to trial. So these things play out in earlier stages of litigation. First, when a plaintiff's attorney is deciding whether to take a case. Second, when he's talking to the hospital or the insurance company about settlement. These parties should be sophisticated enough to know that reality has become much more complicated, the more we know about medical error, the less black and white error typically tends to be. On the one hand, that may sometimes lead plaintiffs' attorneys to settle more regularly for smaller sums of money if they think it's going to be harder to make a case should it go to trial if systems errors are involved. On the other hand, we know that when parties have different perceptions of what occurred, when there's greater uncertainty in a case, that actually tends to push the parties away from settlement. So there are a variety of effects that this complicating story may have on what malpractice cases get brought and how they get resolved.

RW: One of the rationales you raised for the malpractice system was deterrence. It strikes me that prior to 2000, the malpractice system may well have been the only tangible vehicle to promote deterrence. Today, many more parties are involved in trying to affect policies that lead to deterrence—whether it's public reporting of errors or pay-for-performance programs or no-pay-for-errors programs. How does that change the dynamics of the malpractice system?

MM: It makes it a little harder to justify the system that we have. The evidence for tort deterrence in health care has always been thin. A few studies published in the last few years suggest that, in a few narrow areas, there may be a salutatory effect. But other studies have found no effect. Plaintiffs' attorneys and others who strongly advocate for the preservation of full tort remedies tend to point to deterrence as the rationale, in part because the other two are so weak. The idea of having your day in court for corrective justice is a fiction: most plaintiffs never reach court. While real, the compensation function could be achieved much more efficiently through other systems like administrative compensation. So if you're going to advocate for this adversarial system, it must be on the basis of deterrence. That has become harder to do both because of the thin evidence base that tort works for that and because there are alternative ways to reach that goal, as you've indicated.

RW: What are some alternative ways, and which ones do you find the most attractive?

MM: To me, the most attractive is the administrative compensation idea. That's the notion that we give up on the judicial courts as the primary arbiters for malpractice claims. There are too many problems with it. It's too expensive. Move it to an administrative process similar to what we might do for worker's comp claims, social security disability insurance, vaccine injury claims—a process by which experienced and skilled adjudicators are assisted by precedent, written opinions, guidelines, and other decision aids to try to improve the consistency of the process and the predictability. That has worked very well in a relatively small number of relatively small countries. It's a big idea to take to a big country like the United States, and understandably there has been a lot of reluctance to do so. But I do find that to be the most promising on the merits.

RW: How would that work in the United States?

MM: There are a number of ways you could set this up, but probably the most feasible way for the US would be to keep things basically as they are at the earliest stage, which is to say we keep private insurance companies and we allow them to make decisions on the initial claim that has been filed. But we ask them to do so in the context of a system that has shifted the standard of care from just imposing liability when care was negligent to a broader standard that compensates patients whenever an outcome could have been prevented in the hands of an experienced specialist or the best specialist in the field. So what it offers to patients is a wider door for obtaining compensation without having to demonstrate that the provider is at fault. And what it offers to providers is the ability to ensure their patient receives compensation without having to endure an adversarial battle in which the provider is stigmatized as having been negligent. The hope is that in the aggregate, the cost nets out because you can save so much on the administrative costs of the system. Instead of filing a claim in court if the case wasn't resolved at the earlier stage where the insurer reviews and makes an initial decision, then it would go before an administrative panel probably convened by the state, which would hear this case using these decision aids.

RW: Does the administrative resolution system say something different than the malpractice system about this tension between individual liability and negligence and system liability and negligence? Because, at least theoretically, if there's compensation when a patient is harmed, that harm may be from a system flaw rather than an individual flaw.

MM: Yes, it certainly could. So if we're adjusting the standard of care, it could also be widened to embrace situations in which the deviation from care was at the system level. Many times when we talk about this possibility, it becomes useful to think about a second reform that could go alongside administrative compensation: enterprise liability. As I alluded to earlier, right now it's very difficult to hold health care institutions accountable for deviations in care when those institutions don't directly employ the person who made the error, as is the case in most hospitals in the US, particularly community hospitals. And hospitals don't practice medicine, so unless the failure on the part of the hospital itself can be identified, it's very difficult to hold those institutions accountable. An enterprise liability system would replace or supplement individual liability with expanded liability for health care institutions and the recognition that they are the ones in control of the system. So if we want to send incentives to the party that's in the best place to make the kind of improvements that will really make care safer, we have to redirect legal accountability for error.

RW: In an administrative compensation system, is there a protocol that says a certain kind of injury pays a certain amount? It's not negotiated, but we have decided that a broken tooth during an intubation pays out $10,000? Is that the end game here?

MM: It could be. The advantage of this system is that we can decide collectively how we want damages to be awarded. Right now when a case goes to a jury, we say to a jury, the plaintiff has shown what their economic losses are. They've brought in economic experts to show you how much lost income and medical expenses they have. But when it comes to pain and suffering, award what you think is reasonable. And that's all the instructions they get. When we centralize these decisions in an administrative agency, a number of other possibilities develop. We can use decisions in past cases, which would be recorded in a database as a guidepost, or we can get together and have some commission develop guidelines for noneconomic damages. That idea appeals to me greatly. While I think economic losses should be fully compensated, it would be useful for us to deliberate about what we think the value of different common types of injuries would be and maybe a dollar range based on the severity of the injury provided to these decision makers makes sense.

RW: When people look at the costs of the current malpractice system, two big targets are contingency fees for the lawyers and expert witness fees for a lot of people, including physicians. How would both of those be handled in an administrative system?

MM: The administrative system could be designed so that for simple claims you didn't need a plaintiff's attorney. That's the idea of most administrative systems. Right off the top, there's the potential for a large savings in attorneys' fees. The need for experts is also greatly reduced because this is not an adversarial system where we pit two hired guns against one another. Rather, the thinking is we could have some experts employed by the administrative compensation system. That's how it's done in other foreign countries. There are people who work for the system and who are meant to be neutral and who have substantial experience analyzing claims that can provide those opinions.

RW: What are the chances that we move in this direction in the current political environment?

MM: Several years of discussing this idea with policymakers makes me not very optimistic that we will move in this direction anytime soon. There is a very strong attachment from legislators of both parties to the idea of a jury-based system. That's unsurprising given the number of legislators who are lawyers themselves and who are steeped in the culture of the notion that adversarial clashes tend to produce the truth. They worry, and their constituents worry, that the system would be captured by industrial concerns and couldn't be fair to claimants, and they truly believe that plaintiffs' attorneys and juries contribute to fair outcomes and ensure fair representation. Of course many of these legislators also benefit from campaign contributions by the plaintiffs' bar. Substantial resistance has kept this proposal from moving forward. It's a shame because a lot of the suppositions made about the role of juries and attorneys in the system are not empirically accurate. For example, far from protecting the interest of plaintiffs, juries tend to vote against them in four out of five malpractice trials.

RW: As you said, such a small percentage of events that involve negligence actually lead to financial benefit to the plaintiff. One could argue that the malpractice system actually doesn't pay out enough.

MM: That's right, and that's another reason why insurers have been hesitant to embrace this proposal. While some buy the arguments that by greatly reducing the exorbitant overhead costs of the system, which run about 55 cents on the dollar, we could actually compensate a lot more people for the same price, others really worry about making it easier for people to bring claims. Because actually, the result that we'd like to see is that more who are entitled to compensation under the rules that we've set up do come forward and do have an ability to obtain redress in the legal system.

RW: I want to shift to one of the other trends in the field, which is apology and disclosure. What's your read of the last 10 or 15 years of history in that part of the malpractice world?

MM: In terms of institutional programs, the current parlance for those is communication-and-resolution programs, in which health care institutions and their liability insurers commit to disclose adverse events to patients, provide a full explanation of what happened, and provide an apology that is appropriate to why the event happened—either "I'm sorry this happened" or "I'm sorry that we made an error." Then, if an error did occur, they proactively offer compensation to patients without requiring them to file a claim. I was an initial skeptic of those programs, but over time I've become a real believer that, when they are done well, they are extremely powerful and effective. The challenge is doing them well.

It is a massive culture shift for institutions to move from a traditional "deny-and-defend" posture to a culture in which not only do health care providers feel supported in disclosing error, but the insurance company is also ready to write checks in cases where patients are not asking for money. It's one thing to do this where the insurer knows that they're going to pay on this claim eventually, and they can save money by settling in a more expeditious manner. It's quite another to say to an insurer, "Here's a patient who got hurt through error. The patient seems pretty happy with the disclosure and the apology, is not presently asking for money, but we'd like you to make an offer anyway." Some are ready to do that and some are not. This program absolutely is worth continuing to pursue, but it really requires a deep commitment from the highest levels of the organization and its insurer if you're going to do it right and make it work.

RW: Interesting that you raise the insurer. Those of us on the front lines tend not to think about the insurer being such a major player in this. It's the health care system dealing with the patient and the patient's attorney. But as you correctly say, the intermediary of the insurer is a very, very important player here. What about that do we generally not understand?

MM: A lot of the discussion about these programs has gone on by those of us who sit in academic medical centers that are self-insured and have a lot of control over what happens with their clinicians. And that, again, does not represent the experience of most hospitals in the US. Most of them have a separate insurer and what happens is they do the initial investigation, and they have the primary interface with the patient early on, but then hand off that case to the insurer to make the call about what to do about compensation. If the insurer is not on board, it's very easy for things to fall off the rails. Frankly, it's easy for that to happen because the insurer has a different mission from that of the hospital. It's relatively easy for hospitals to say, "Our mission is to take care of patients. That caring relationship doesn't end when an injury occurs, particularly if the injury is our fault. We're going to continue to take care of patients." They have to see and deal with these patients at the bedside after something goes wrong. Risk managers and claims folks within hospitals often see the advantage of this approach, not just financially but because it feels like the right thing to do when you're confronted with the patient. Now think about an insurer who is distant from the point of care, doesn't have boots on the ground, doesn't see or interact with the patient in a meaningful way, and whose mission is to minimize financial loss for the people who have entrusted their liability to them. It's very difficult to convince an insurer like that to get on board with this program.

RW: It sounds like you were somewhat skeptical and you have changed your thinking. Do you have any sense that is happening in the insurer world?

MM: I do. As more and more successful experiments are conducted, there is a lot of optimism about this approach and a great deal of interest. And there are inquiries now about, how do we do this right? What do we need to know to avoid outcomes that have been seen in institutions where it hasn't worked and emulate the outcomes in institutions where it has worked?

RW: You've written about electronic health records as one of the sea changes in medicine over the last 10 years. How does that change the nature of the issues that you think about and study?

MM: EHRs are an inevitability. Everybody is going to get there eventually, and the question is how do we prepare for that transition in a way that makes health care safer and that makes liability costs lower rather than higher? Electronic records reduce malpractice liability risk in many ways. They reduce the underlying safety problems that can lead to injuries that lead to claims. And that is a wonderful thing. In other ways they can elevate the risk of having to pay out once an incident does occur. Documentation tends to be much better. There is more of an evidentiary trail for the plaintiff to obtain and follow. Then, we have a few cases in which EHR systems themselves have had problems that have led to injuries and have led to claims. It also tempts users to engage in behaviors that are not great from a malpractice perspective, like cutting and pasting medical histories rather than taking the time to rewrite them. Then there's the possibility that plaintiffs' attorneys and their clients will come to expect that when a provider enters a room to see a patient he or she has thoroughly reviewed the medical record. In the age of EHRs, that record is extensive, and it may not be realistic to expect that that is going to be what constitutes reasonable care going forward. On balance, at this point few think that EHRs are a bad thing or that we've gone in the wrong direction. But some of these risks have to be considered when institutions are selecting their systems, maintaining their systems, and instructing their care providers about how to use those systems responsibly.

Disclosure: Dr. Wachter serves on the board of The Doctors Company, which provides medical malpractice coverage for physicians.