Portal de Periódicos da CAPES

A Conversation with … Charles L. Bosk PhD, Expert on Surgical Education and Medical Error, and Author of Forgive and Remember: Managing Medical Failure

2020; Lippincott Williams & Wilkins; Volume: 478; Issue: 6 Linguagem: Inglês

10.1097/corr.0000000000001184

1528-1132

Seth S. Leopold,

Innovations in Medical Education

Perhaps you remember the book (or the movie) Gorillas in the Mist [6]? In it, primatologist Dian Fossey PhD, spent years in the rainforests of Africa living alongside the great apes, observing their social behaviors. In this month’s “A Conversation with…”, I visit with Charles L. Bosk PhD, a sociologist whose first book might have been called Surgeons in the Mist. It isn’t (it’s called Forgive and Remember: Managing Medical Failure [1]), but it has many things in common with Dian Fossey’s classic work: Dr. Bosk spent 18 months eating, rounding, and commiserating with surgical residents and their attendings in the 1970s with the goal of learning something about their (our) social structure.Charles L. Bosk PhDIn the mists of a Chicago city hospital, surgeons were the gorillas, and Dr. Bosk was the “primatologist”. His inquiry then—and a large part of what he’s investigated in the four decades since—involved the qualitative study of medical error. Among other things, in Forgive and Remember [1], Dr. Bosk wanted to see what kinds of mistakes trainees could make that would result in a resident being fired. The work was done in the era of pyramidal programs; for the younger reader, pyramidal programs hired more residents than they could graduate, such that firing some residents each year was part of normal business. I read Forgive and Remember [1] as a junior resident; it changed the way I thought about learning. As an attending, it’s helped me to understand why some seemingly small things create deep tensions between residents and their faculty. In general, Dr. Bosk found that technical errors and judgment errors often were remediable, but normative or “moral” errors—like lying and laziness—typically were not. Surgeons probably are less surprised by this than Dr. Bosk was; I think surgeons know that it’s our job to teach psychomotor skills and thoughtful decision-making, and we generally believe that a 28-year-old person who practices incomplete honesty is not someone we want in a surgical training program. But I still found some of his findings surprising, even eye-opening. For example, in his taxonomy of error, Dr. Bosk described a phenomenon he called quasi-normative error. A quasi-normative error occurs when there may be several equally valid treatment approaches according to the best-available evidence, but an attending has previously articulated a preference to a trainee, and despite this, the trainee chooses a different approach while caring for that attending’s patient. Dr. Bosk found that quasi-normative errors—in essence, failures to honor idiosyncrasies and eccentricities of attendings—were as bad normative defects like dishonesty. Residents got fired for committing quasi-normative errors. The given reason for these dismissals was that the commission of quasi-normative error conveys that the trainee believes he has nothing more to learn from the program, thus violating the key shared value that residents are in a training program to learn. The more-astute reader will notice that in the last sentence, just above, I used the male pronoun. So did Dr. Bosk; in that time and place, surgeons and residents were all men. They often referred to nurses as “girls”. Racial stereotypes and heavy biases run through some storylines (these were the surgeons’ biases, not Dr. Bosk’s). The dialogue he recorded is of its time and place. Reading it again in advance of this interview reminded me how much has changed for the better in those areas, but also how much still needs to change. If you haven’t already, consider reading Forgive and Remember [1], along with the rest of Dr. Bosk’s oeuvre about surgical and medical errors. But regardless, don’t miss the fascinating conversation with him that follows. Seth S. Leopold MD: You described a kind of error that has stuck in my mind since I read about it: Quasi-normative error. Your study found that quasi-normative errors were as bad as what you called normative and moral errors, like lying and laziness; you saw residents get fired for quasi-normative errors. This raises an important tension: On one hand, we want our trainees to develop independent judgment as they advance in training; we don’t want to educate people to become dogmatic. And yet, a resident who disregards his/her attending’s preferences makes “a claim that his judgment is as adequate as his superior’s … [and] risks his reputation as a trustworthy recruit”[1]. Where is the balance point on this? Charles L. Bosk PhD: Quasi-normative errors were something that I discovered while trying to make sense of my field diaries. They are a type of resident misstep that I would define more broadly today than I did in 1979 when the term first appeared in print. In 1979, after all, I had a much thinner observational base. I now think of quasi-normative errors as the noise that has the potential to neutralize the signals that are sent when residents are given negative feedback about their performance. Resident evaluation is now more systematic and formal than when I did the field research for Forgive and Remember [1]. Then, I was given access to the personnel files of the junior residents being considered for advancement to senior residency. I was most impressed by how thin the files were. Typically, if a resident evaluation was longer than a sentence, the resident was in deep trouble. Today, I know not to bother to ask for such data, as I would not be given such access. But if I had access to such files today, the very opposite would be true—I would be drowning in data. But more data points about performance, does not mean that we have more information about performance. In fact, the wealth of data points is just as likely to create a mixed record as show uniform excellence. The assumptions we use to transform data about performance into meaningful interpretations go unexamined. Are some dimensions of resident performance more reliably measured than others? Do some assessments get weighed more heavily than others? What I see today is that quasi-normative errors sometimes signal a resident who fails to play well with others in a way that negatively affects patient care. At other times, quasi-normative errors are a way for biases and prejudices that standardized and objective evaluations are said to eliminate to creep back into resident evaluation. Certain heuristic misfirings like the halo bias—the more one looks the social role they are being asked to play, the more their performance will be judged favorably—have a differential impact on who is vulnerable to having their performance evaluated negatively on the basis of quasi-normative errors. Quasi-normative errors are confusing: Sometimes they signal a real problem; and at other times, they are markers of bias, conscious or unconscious, on the part of those providing the negative assessment. Paradoxically, the formalization of evaluation increases the noise quasi-normative error sends about professional behavior and mutes the signal that normative errors are intended to send. Let me describe the process that I see operating to confuse quasi-normative error and normative error. The Accreditation Council for Graduate Medical Education requires that programs train and measure residents in six core competencies. One of these competencies—“professionalism”—lacks any objective correlate. When my research group was studying the implementation of the 80-hour work week between 2007-2011, we frequently heard from residents who said that criticisms of their professionalism were unfair on three counts: (1) The sample or incident was so unrepresentative that the criticism was overly broad and harsh; (2) the resident was being held to a standard that the faculty member was seen as not meeting themselves; or (3) the criticism being made had no connection to any plausible definition of professional behavior. Formalized feedback is as likely to aggravate tension as relieve it in those situations when residents disagree with criticism of their professionalism. So, to return to your question, formalization and standardization have done little to reduce the diversity of operational definitions of professionalism, which are used to assess performance. A lack of formal standards and the routine breaches of those making assessments has made definitions of professionalism all the more contested in performance assessment. Dr. Leopold: If I’m following you here, the subjective construct you described as quasi-normative error in resident education has been replaced by a broader—but vaguer and even-more subjective—assessment construct called “professionalism”. I’m sure we agree that we don’t want to graduate residents who habitually violate professional norms (quasi- or otherwise), and that we want our surgeons to display professionalism. From what you’ve learned, how can we make this process more objective? Dr. Bosk: I don’t think the way to improve the process is to make it more objective. Often, what constitutes “professionalism” is in the eye of the beholder. There simply is no objective standard nor can there be. Whether behavior is professional or not is a normative judgment made within a situational context. instead of objectivity, what is needed is reflective discussion about the judgments being made. I think in graduate medical education, people have gotten so caught up in developing performance measures that they forget to interpret them in ways that those whose behavior is being praised or criticized understand why. Dr. Leopold: One of my favorite parts of Forgive and Remember [1] is the chapter about Morbidity and Mortality (M&M) conference. After reading your book, I’ve asked countless residents (and partners) what they think the role of M&M is; nearly all have responded “to try to learn from mistakes so we don’t repeat them.” If that really were the case, we’d run a quality-improvement conference similar to when the National Transportation Safety Board runs inquiries into airplane or train crashes—in a no-shame, no-blame, just-the-facts manner. As you point out, M&M is anything but objective, and it’s really a vehicle that surgeons use to model important professional norms. Can you explain this in a bit more depth, and perhaps nuance your comments with what you’ve learned about this in the nearly 40 years since you wrote Forgive and Remember? Dr. Bosk: The M&Ms that I have attended more recently try, and sometimes fail, to be run as more cordial affairs, more along the lines of a quality-improvement conference. The shift to a language of system error aligns poorly with an ethic of personal, individual, and professional accountability that M&Ms seek to instill. I think it important that M&M’s have a personal, relational, and professional dimension. The lay person in me—the eventual patient we all become—wants physicians to feel badly after deaths and complications and to think about whether any changes in treatment might have made a difference. In my experience, it is impossible for anyone in any field anywhere to learn anything without reflecting on how their behavior contributed to some unwanted outcome. M&Ms remain worthwhile so long as they provide useful narratives that allow the audience to understand how a death or complication occurred as well as what, if anything, might have been done to prevent it. There is an undeniably performative element to becoming a surgeon. M&Ms allow surgeons to display humility while explaining adverse outcomes. It matter less how residents and senior surgeons learn from mistakes than does the fact that there is a weekly forum that provides an opportunity to do so. In its way, being forced to take one’s lumps in an M&M conference addresses the fact that even when errors occur at the large-systems level, people are involved. I believe that the performative element of the M&M conference acts to inoculate surgeons from becoming second victims of the deaths and complications. In my view, the current epidemic of burnout among physicians and nurses is a consequence of there being too few safe spaces for discussing the affective dimensions of working in environment that requires time-pressured decision-making under conditions of unresolvable uncertainty. Providing care is demanding work at every level—work that is complex, cognitively, technically and emotionally. M&Ms individualize deaths and complications; this is something important that needs to be done before the patient becomes a deidentified element in a more complex algorithm that yields an evidence-based best practice. Before moving on to more impersonal ways of viewing untoward outcomes, thinking of them individually is a way to honor the human element in the doctor-patient relationship. Dr. Leopold: That was not the answer I was expecting! Fantastic. So from where you sit, M&M continues to serve an invaluable psychosocial role, and if there should be formal, systems-level quality improvement approaches, these should occur elsewhere. If the meta-goals of M&M are to provide a safe space, to mitigate burnout, and a means to keep surgeons from becoming second victims of their complications, are there approaches we might adopt that can help us to do this more effectively? Dr. Bosk: I think it is the responsibility of senior surgeons to create that safe space. However, creating it is easier said than done. I think that the M&M conference, by its very nature, can never be a completely safe space. Yet I believe confession is good for the soul. I think Balint- type groups are not just for medical students. Providers need to find ways to deal with the affective dimensions of their work. We are currently narrowing rather than widening the ways to make this possible. Dr. Leopold: Bringing us more into the present, you’ve written thoughtfully about checklists. Surgeons are asked to use checklists as a part of normal care; to make them as much a part of surgery as scalpels and sutures. We often hear the aviation metaphor for checklists (“pilots use them in a similarly complex system”). Yet you seem skeptical[3], and you suggest this approach exposes a number of “blind spots” in the science of safety[4]. Why is that, and, more importantly, how can surgeons use your observations to take better care of their patients? Dr. Bosk: First, I am baffled that anyone thinks checklists are an innovation. I cannot remember observing a morning rounds where a service did not construct a daily checklists; and I first started observing rounds in 1973. Timeouts, huddles, briefings and debriefings, and Plan-Do-Study-Act cycles—all have been added to the quality improvement and patient safety toolbox. All are likely to serve some useful purpose. But as soon as they become mindless routines, they become a source of heedless error. Improving quality and reducing preventable complications requires constant innovation. When quality and safety metrics are good to begin with—and I recognize how evaluative and subjective a term like good is—campaigns to improve care can become counter-productive. I think this has already happened with regard to many measures intended to improve patient safety. To make checklists work to reduce Central Line-associated Bloodstream Infection first in Michigan, and, then nationally is a much more complex story than “here’s the checklist, use it and all your troubles will go away.” My work on checklists is dedicated to showing that more is needed to reduce hospital-acquired infections than better checklists. My problem with widespread advocacy of checklists as a tool for improvement is that they represent an overly simple solution to a fantastically complex operational question—how do we improve the quality of care and reduce harm to patients? How do we do this as advances create increasingly brittle patients with multiple co-morbidities? Dr. Leopold: Recently, you wrote “simple narratives of policy change fail to address the complexities of the problems that they are designed to remedy” and “many of those who possess[ed] certainty about how to fix the broken parts of the health care delivery system would be better served if they possessed more humility, [and] less swaggering confidence in the rightness of their prescriptions”[2]. In that same article, you also point out—somewhat discouragingly—that the things that bother physicians now have not changed much over the four-decade span of your career: Our inability to provide access to good care for all patients, the need for tort reform, and too much regulation (except when there is too little). As a keen and lifelong observer, what do you see as the most-promising avenues through which fundamental improvements to the profession can come? Dr. Bosk: I am far from pessimistic about positive change. I think before we excoriate health care for its failures, we need to appreciate that many of those failures are the unintended consequences of spectacular success. I am going to try to capture what I think are the most-promising avenues in three overly glib formulations. First, identify and eliminate perverse and contradictory incentives that hide in plain sight. For example, why do we reward providers for short lengths of stay while punishing them for costly 30-day readmissions? How can we cut costs at the same time that we provide more surgery to patients with higher levels of acuity? How do we untangle the conflicting interests of individual patients, providers and society at large? Second, I cannot help be skeptical about the use of technology to solve the problems created by technology. Suppose the claim that the root cause of 80% of preventable adverse events is faulty communication is correct. How likely is a new app to correct faulty communication? We need to correct the erosion of face-to-face interactions among providers and between providers and patients if we are trying to reduce communications errors. We are moving rapidly in the other direction. Just as a map is not the territory it depicts; an electronic medical record is not the patient. Third, we need to recognize that the population most likely to experience a preventable adverse event in the hospital includes people who are most vulnerable to harm outside it. The poor, the elderly, the admission from the emergency room, and the most acutely-ill are the groups most likely to experience a preventable adverse event [5, 7]. If we want a healthcare system that is more efficient and reliable at a population-level, we need to make investments in care that pay more attention to the cradle to and less to the grave. We need to insure access to adequate nutrition, housing, and education to poor and vulnerable children and families, if only to avoid the lifetime of health disparities, reduced opportunities, and psychic distress created by deprivations that exist prenatally and continue for a lifetime. At the other end of the lifespan, we need to think more clearly about how balance patient autonomy with collective equity. We need to recognize that as we expand access—and that seems inevitable—each of us will need to learn to share the healthcare pie. Dr. Leopold: How has your career of observing physicians changed what you look for in a doctor when you or a family member needs one? Dr. Bosk: Not much. I think that I have always wanted the same thing in a physician. I look for someone who is able to respond as a human being, who makes eye contact, who responds to unvoiced anxiety, as well as anticipates and answers questions that I am not yet to formulate. I want a physician who doesn’t look irritated when my concerns force him or her to spend more than my allotted time according to some performance metric. When any member of my family or I have needed surgery, surgeons often ask me if I want them to do anything special. I am never sure what they mean by asking this. As quickly as I can, I tell them to do whatever they normally do, the way they always do it.