Share the Decision to Avoid Regret

Sometimes there is only one appropriate treatment option for a patient. Consider someone with a knife stuck in their carotid artery. Nothing other than surgery is discussed. Many clinical problems aren’t like this and can be addressed by several different options with roughly equivalent outcomes. The choice of available options is “preference sensitive”, meaning that it is influenced by values unique to the decision maker. In the past, the focus has been on the values that physicians believe to be important – preference given to a status quo option with extensive evidence base, that maximizes long-term survival and does not pose a high risk for the given patient. Patients are interested in these issues but have additional priorities that may follow a different rank order – get back to work as soon as possible, avoid disfiguring incisions, maximize quality of life over longevity, etc. Patients that consent to a treatment that results in an outcome inconsistent with their rank order of values could feel that they were not heard when the decision was made. More importantly, they might experience regret about their lack of assertiveness at that time and feeling that things could have been better. The model that best reconciles conflicting values between the provider and patient is for them both to share the final decision about the treatment choice.

An organization called the National Quality Forum was established in 2000 that helped spur a variety of quality improvement initiatives in healthcare such as a reduction in hospital infections, improving maternity care and reducing readmissions. The promotion of shared decision making between physicians and patients was also a central feature of their initial mission statement in 2000. However, the past 16 years have shown painfully slow progress in this area. A recent national survey of US adults showed that the majority perceive benefit from shared decision making but only a minority of health decisions were being made this way. A reasonable interpretation is that physicians are more willing to work on improving the quality of their care than their conversations with patients. Most understand the conceptual basis for shared decision making, but remain like the “beneficent” parent who cannot quite accept that his child is now an adult and reassert that the physician knows best. Starting with Hippocrates, the culture of medicine has been rooted in a paternalistic tradition against full disclosure of all options and risks to patients. For a patient to chart his course understandably, familiarity with therapeutic alternatives and their hazards is essential.

This disconnect in values and motivations of physicians and patients inevitably causes problems. An article in Forbes magazine last week (June 24, 2016) provides fascinating insight into the informed consent process. ( This VA study recorded the information provided during the consent process for surgical prostatectomy and compared 3 sites that had access to a surgical robot with 1 that did not. All clinic visits at the 3 sites with a robot described robotic prostatectomy as the standard of care. In contrast, 78% of clinic visits arranged to talk about prostatectomy at the remaining site excluded any discussion of robotics as an alternative. During the 22% of times that it was mentioned (often prompted by patient questions), physicians downplayed its advantages more than the other 3 sites.

The article starts off with the following quote: “Patients often rely on doctors for info about treatment alternatives. Unfortunately, that info isn’t always objective.” A difference between centers suggests that the information they were given was not objective. It is possible that information given to those that underwent open surgery was slanted to assure that they accepted the alternative favored by their doctor. Few doctors are trained to understand things from patient’s non-medical perspectives. Absent this training, the physicians that obtained informed consent during this study might have found it hard to avoid imposing their values or substituting their level of risk aversion onto their patients. It should raise a red flag that no patients at the hospital without a robot made a decision to be referred out and choose robotic rather than open surgery. Their decision for open surgery seemed foreordained by the type of information given and way it was phrased. This complied with the letter of the informed consent doctrine but not its spirit. The lack of disclosure of the robotic approach as a reasonable alternative to an open surgery in effect limited their right of choice to a veto power over the one treatment that was recommended.

Even if the robot offered no real advantages for prostatectomy, it would not change the fact that the majority of practicing urologists in the US use robotics for this purpose. That makes a description of the robotic alternative as standard of care for appropriate informed consent discussion at the VA or any other hospital. In fact, the actual advantages of robotics are well documented for prostatectomy. The only rationale for these urologists to mention only the open surgery was they assumed patients would not choose to transfer to another VA facility or other hospital that offers robotics. Such an assumption is on shaky ground since the field of less invasive surgery has been built on the willingness of patients to seek out those facilities that have offered less invasive surgery. Assuming that no patient would want to do that goes against the natural human instinct to avoid surgery that is more invasive than otherwise required. The trust that patients put in this (inadequate) consent coupled with the asymmetry of information meant that patients didn’t even realize they lost their autonomy to decide.

The Forbes article itself closes with an alternate and more benign conclusion, giving physicians the benefit of the doubt about their motives: “In shifting their description of robotic surgery based on availability of that technology, they were probably trying to ease patients’ concerns. There are far greater sins than that.” It is reasonable to manage patient’s emotions and expectations. Decision making about serious health issues can be agonizing. Concern for managing emotions is appropriate and even laudable. Open surgery has been around longer than robotics. There are patients that may not want an innovative approach and prefer the one with the longest track record. Prior evidence suggests that many referring doctors are unwilling or unable to share information about underutilized services such as less invasive surgery (Cooper et al., BMJ 2014). While the exact reasons for this are unclear, one possible contributor is that many of these providers may be unable to provide confident advice about a novel procedure for which they had no personal experience. The problem for the VA urologists in this study is that these issues should exist evenly among the 4 hospitals and not be influenced by whether or not the hospital had a robot.

Everett Rogers first described in the 60’s that the enthusiasm for any new idea is highest in early innovators/visionaries, a bit less in the majority, and lowest in late laggards. Every urologist will lie on this spectrum until the life cycle of this innovation is fully complete. Where they lie is defined by their information needs and demands on the robotic innovation. Every hospital has visionaries that race forward to gain the “first mover advantage” balanced against laggards that await large randomized clinical trials before giving their support. This means that some doctors will explain the advantages of robotics very differently than others based on honest differences in opinion and where they lie on the spectrum. The unexpected finding of the VA study was how a hospital’s decision to buy the robot was associated with a systematic shift in how their urologists described robotic surgery. The lack of a robot seemed to create more late laggards at that hospital that were more critical of the evidence for robotics. This phenomenon is not predicted by the Rogers curve and signifies the potential for bias and/or conflict of interest.

Doctors without access to robotics may not have brought this topic up out of concern that patients wanting a referral to another center might be put through a hardship. There is a natural tendency to favor inaction (in this case not mentioning robotics) over action (bringing up this topic and facilitating a referral). We view injury that results from inaction as less causal and less blameworthy than acts of commission. The impact of this “omission bias” is that people tend to view a death from a vaccine as much worse than a death from not getting vaccinated and criminal laws punish those that do harm far differently that those that fail to rescue. Medicine’s most important oath – “first do no harm” – also reflects this bias. Closely related is the tendency to stick with the status quo, which stems from underestimating the risks of the old relative to the new. Open prostatectomy has been around for decades so all physicians know what to expect. Obtaining equally reliable information about the new robotic option (the antidote to omission and status quo bias) comes when physicians refer their patients for the procedure and see how they do. This is far more practical when there is a robot is located where those physicians work.

While the use of robotic technology had no obvious direct financial impact on VA urologists, an indirect financial conflict of interest can never be ruled out when discussing one of the most expensive capital equipment in a hospital operating room. The decision not to purchase the robot at one out of 4 VA hospitals often creates more “skin in the game” than is obvious at first glance. Perhaps the decision not to pursue robotics creates pressure at this one hospital to keep up their number of open prostatectomies. Past contracts of doctors employed by HMOs clearly illustrate the wide range of financial tools and tactics available to hospitals to influence doctor behavior. The vast majority of physicians asked about cost-control arrangements at HMOs found them to cause unacceptable conflicts of interest (ARCHIVES INTERNAL MED. 2000; 649, 651-3). Urologists at the hospital with no robot may not know how to do robotic surgery and are not anxious for their hospital to jump into this arena and be replaced by new surgeons with robotic skills. This also would stimulate the number of open cases. The study provides no evidence that self-serving financial or political incentives caused actual conflict with the welfare of these VA patients. At the same time, the tapes of the consent processes that were studied seemed to yield nothing to rule out a potential for this type of conflict.

Courts have been particularly concerned about physician conflicts of interest. Informed consent can be ruled invalid without disclosure of “any interest that the physician has that conflicts with, or even potentially conflicts with the physician’s fiduciary duty to that patient.” (Moore v. Regents of University of California). Ethical rules of the AMA state “If a conflict develops between the physician’s interest and the physician’s responsibilities to the patient, the conflict must be resolved to the patient’s benefit.” Disclosure of a potential conflict allows the patient to consider the surgeon’s competence to recommend the proposed surgery. At the minimum, the duty to inform required urologists without access to a robot to at least mention a robotic alternative. Optimally, this would be accompanied by helping patients interested in robotics to obtain a second opinion and/or explain why the robotic approach is not preferred. Instead of either of these choices, 78% of urologists chose not to bring it up at all. This was the most effective way to assure that patients stick with the doctor’s choice of open surgery – if these patients came across favorable information about robotics afterwards, it was most likely ignored. However, it only added to concerns about a potential conflict of interest and risked the trust of any patient that would have been interested in robotics. Without trust, there is no physician-patient relationship, only a merchant and a customer, each with competing interests, dealing at arms-length in a commercial transaction.

It is unlikely that the VA patients in this study had knew the expected outcomes of either the open or robotic option, so their decisions weren’t likely influenced by the same biases that may have swayed one group of urologists vs. the others. The hospital where patients ended up receiving care would have no influence on the biases influencing their decisions. Moreover, the concept of “status quo” has a different definition from the perspective of the patient. Patients maintain their own status quo by findings the least possible invasive option. Not consenting to a recommended invasive procedure is essentially choosing the least possible invasive option. It is far more common to see patients refuse a physician’s advice for surgery than to find a single patient demand a treatment more invasive than what was recommended. It has been shown that patients tend to reject recommended invasive surgeries as they are given greater amounts of relevant information. This last fact alone documents a bias of patients against invasive surgery that is generally greater than their physicians.

What this means is that patients as a group tend to be shifted towards the earlier part of the adoption spectrum for any proposed less invasive technique than their physicians. It is a well-recognized marketing fact that demand by patients (not physicians) has been the driving force behind the development of less invasive options in every surgical subspecialty. This puts patients and physicians on opposite sides of the chasm that typically separates the early innovators that kick-start an innovation and the more pragmatic majority that control its long fate. Many patients that I’ve talked with about robotics see their doctor’s demands for more research or better designed randomized trials as a sign of stalling. To them, they feel that to only way to get beyond the current system is to act without complete information. Pragmatic surgeons, on the other hand, see these patients as victims of the hype that often accompanies a new fad. Some patients end up choosing the “time-tested” approach, but they all say that objective and complete information about the robotic option was material to that decision.

Two decades ago, a similar debate raged about less invasive treatment of breast cancer. Women who were actively involved in choosing between the conventional approach – modified radical mastectomy – or a new less invasive treatment had significantly higher overall quality of life at follow-up than women who had passive involvement. After it was shown that both yield equal survival, many surgeons continued to offer only the conventional approach and never mentioned less invasive options to their patients. The issue soon became political, prompting federal regulations and 14 states to enact legislation that mandates surgeons to inform patients of all viable alternatives to treat breast cancer. Some even required patients to verify that they received the specific information mandated by the statutes.

Virtually all the mastectomy surgeons were offering their patients what they believed to be best and their patients most likely did not end up worse off. Nonetheless, legislation was required to protect the right of women, not surgeons, to choose what happens to their body. This is a “dignitary interest” with roots in a decision by a British court in the 1300’s that held a person to be entitled to legal protection of their mental tranquility even with no proof of physical injury by the defendant. Outside of this type of rare legislation, it is negligence statues, not dignitary interests, which provide the legal standard for complaints about informed consent. The plaintiff must prove that physical harm resulted from the failure to disclose all alternatives. It is not enough to prove that the information that was disclosed would have been material to the decision of the plaintiff patient, instead the interests of a theoretical “reasonable patient” are considered. The end result of all these hurdles is that the courts end up being a very poor catalyst to motivate surgeons to change.

The system is built to conserve the existing status quo and mistrust the new. Preferred provider lists are based on the frequently incorrect assumption that all surgeons provide the same list of offerings. When the proves not to be the case, insurance companies don’t allow for patients to go to hospitals outside their preferred provider list without substantial out of pocket costs. Ironically, decision aides developed to promote shared decision making only discuss the conventional options and rarely discuss robotics. As proof that this system is working the way it was designed, Dr. Marty Makary, a surgical oncologist at Johns Hopkins, demonstrated in a survey in 2014 that 25% of hospitals in the US don’t offer less invasive GI surgery even though the complication rates are less than half the open method.

Even guys like Dr. Makary are not without sins. His publications criticize robotics, pointing to the false advertising present on many hospital websites that uses unsubstantiated claims provided by the robot manufacturer. He concludes that “hospitals need to be more conscientious of their role as trusted medical advisers and ensure that information provided on their websites represents the best available evidence.” Yet Dr. Makary’s own website at Johns Hopkins describes a highly complex operation called a laparoscopic whipple to be “generally able to reduce blood loss and risk of infection for the patient.” There is no questioning his unique expertise in laparoscopic surgery and proficiency at this procedure but the evidence to support this statement on the website is by definition weak when this procedure still in its early experimental phase.

If patients interested in novel options can’t expect the straight story from innovators like Dr. Makary, they certainly should not expect help from surgeons that don’t offer robotics, lawyers bound by the rules of medical negligence or a system designed to protect the conventional approach. The VA surgeons in this study were not paid by fee-for-service and knew that their consent process was being recorded. Imagine how the conversation goes when there is a profit motive and no recording?

This leaves the patient with one last line of defense: shared decision making. Where we are today on this issue won’t have any impact. Something is needed to spur a fundamental change in the conversations surgeons have with their patients to bring it in line with the principles of shared decision making. I believe the main reason that this improvement initiative has yet to take off is that there are no good ways to measure its effect. A fundamental rule taught in the first week of most business schools is: “if it can’t be measured, it can’t be improved.” In order to improve progress in advancing the principles of shared decision making, we should start by measuring its impact on patients. Some patients that defer to their physician’s judgment will end up unwittingly choosing an option that they later discover was not an optimal fit for their personal values. Such dissonance causes regret. Regret is a powerful emotion evoked by a sense that things could have been improved by a better choice. Because of its disproportionate influence on patient decisions, regret reflects an effective decision just as closely as mortality and complication rates reflect effective technical performance.

I propose the following litmus test for effective shared decision making. Explain to a patient 6 months after their surgery what are all their surgical options were (presumably a “refresher” course). Find out if this knowledge causes any of them to regret their choice. Patients with outcomes that weren’t as good as they expected are likely to be disappointed but regret suggests someone that chose an option that they didn’t fully understand at the time when currently compared to the alternatives. If this evidence is made transparent, I’d bet $100 that it would cause a sea change. Surgeons subjected to this type of scrutiny would then provide the same level of meticulous concern for informed consent as they do the technical aspects of their surgery.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s