Right and Wrong

Most physicians take being right very seriously. We take pride in our work, relish using our hard-won knowledge to help patients, and do not like to be told that we are wrong. Likewise, it can be difficult for us to cope with uncertainty. We want to know whether we are right or wrong. When we transmit this passion for clarity to medical students and residents, we make at least two assumptions. First, we assume that they should dislike uncertainty as much as we do. Second, we assume that we know the correct answer. There is no question that being right is a good thing, and much to be preferred to being wrong. Yet being right is not the only criterion, and often not even the most important criterion, by which to assess medical excellence.

Arriving at the correct diagnosis or prescribing the right treatment does not completely discharge the physician's responsibility. We do learners a profound disservice if we lead them to suppose that their primary mission is never to make a mistake, never to get caught not knowing something. Too much emphasis on getting the right answer may in some cases actually undermine the full development of a physician. To see why this is so, we need to examine the role of correctness in medical training,identify some of its deficiencies as medicine's holy grail, and develop an expanded vision of medical excellence that extends beyond merely getting the right answer.

The desire to get the right answer has many roots. It stems in part from our generic preference as physicians for situations where our roles are clearly defined, we have direct personal influence over outcomes, and where we receive prompt and unequivocal feedback on our performance. We cut our teeth in classrooms, where expectations were clearly specified at the beginning of each term, performance was regularly assessed by clearly scored examinations, and we knew exactly where we stood in the course. The best students were the ones who answered the most questions correctly, and we had but to compare our responses to an answer key to know which ones we missed. The higher our examination scores, the better we were doing.

This attitude persists and gets intensified in subsequent medical training. When we evaluate our learners, we tend to focus on those aspects of performance that can be readily quantified, especially in the format of multiple-choice examinations. On such examinations, one answer is always right, and the remainder are always wrong. When medical students and residents discuss cases, we tell them when they get it wrong. If they are not "on the right track," we let them know. The ideal case from the learners' point of view seems to be one with a clear-cut correct answer, where the history, physical examination, and laboratory findings all point to a single diagnosis.

Right and wrong have great methodological appeal. If we ground our vision of medical performance in such a paradigm, it becomes much easier to measure how well we are doing.We can show physicians-in-training 10 cases or 100 cases, and see how many they get right. We can plot their sensitivity and accuracy and compare them to those of others at the same level of training. Yet what gets omitted is something like practical wisdom, the ability to relate the material tested on examinations to the much more complex clinical context of patient care. If learners focus their energy on performing well on the examinations, they may become better and better at taking tests, but not necessarily better physicians.

In the real-world practice of medicine, the correct answer is often unknown. In some cases, the radiologist may be the arbiter of truth, by saying whether a bone is fractured. In other cases, it may be the pathologist, whose tissue analysis establishes the diagnosis. In many cases, the natural history of disease and the response to therapy provide the best feedback on the accuracy of our diagnostic hypotheses. In most cases, no independent and irrefutable assessment of the correctness of our judgments is ever made available to us. Because most injuries and illnesses tend to improve on their own, this means that we often never know whether we were right.

In many cases, learners have little more in the way of correct answers to rely on than what their teachers assert to be the case. The medical student may hear a heart murmur, which the attending physician denies to be present. Who is correct? We typically assume that the more senior physician is the more reliable judge, but we have no independent answer key by which to grade their responses. No clinical follow-up or pathological verification is ever obtained. To some degree, learners and educators function as co-conspirators in a plot to preserve our mutual faith in the paradigm of correctness. Learners need an answer key to feel that medical education rests on an objective foundation, and educators need to believe that our judgments are reliable. Being wrong is bad, but supposing that no one knows for sure is even worse.

The tyranny of correctness can narrow the focus of medical education to a dangerous degree. It can distract us from the vital role in medical reasoning of the larger clinical context. More than knowing whether we were right or wrong, we need to become skilled investigators, who know how to ask good questions. What should we be looking for, and why? In many cases, key pieces to the diagnostic puzzle are found in multiple domains that become apparent only if we effectively investigate them. If the correct answer on an examination is the figure, the larger clinical context of the patient is the ground. How we perceive, describe, and interpret any finding depends on the background against which it is projected.

The paradigm of correctness offers a stripped-down version of medical care, in which physicians are likened to computers that receive input and spit out differential diagnoses. But what questions have produced the input? Were the appropriate questions asked? Were the appropriate tests performed? What decision are we trying to make, and what are the implications of different diagnostic results for the patient's management? The practice of medicine is less like computation and more like a social investigation that involves multiple perspectives and multiple actors. Our performance is shaped not only cognitively, but professionally and institutionally. The goal is not to avoid making mistakes but to contribute as much as we can to the care of our patients.

Patients want accurate diagnoses, but they also want a whole lot more. They want to regain or preserve their health, and to lead full and long lives. Likewise, our medical colleagues want accurate diagnoses, but we esteem correctness less highly than effective patient management. The correct diagnosis is merely a tool that we can use to do well our larger job of caring for patients. We want to be accurate, yes, but it is at least equally important that we be relevant. We can tell patients the right answer without ever really getting across to them what they need to hear or making a real difference in their lives. If we do not keep our eyes on this larger prize, we can produce medical charts that are totally accurate and completely useless, because they do not get at the real problem.

We need to help learners acquire an appropriate sense of proportion about correctness and accuracy. If we fail to appreciate the larger clinical context, we may err in defining the degree of accuracy we need to pursue. When findings are almost certainly benign or there is little we could do about them, it may be less important to nail down a precise diagnosis. A brain biopsy is probably not warranted in every case of suspected Alzheimer's disease, even though it would go a long way toward eliminating any uncertainty about patient management. In other cases, such as suspected child abuse, nothing less than the most rigorous diagnostic work-up is appropriate. Mere precision for precision's sake is not our goal. Instead we need to pursue the degree of certainty that the clinical context warrants.

The correctness paradigm can also distract our attention from providing good service to colleagues and patients. Most of us could take steps to improve the efficiency, cordiality, and usefulness of the services we provide. Getting the right diagnosis is an important link in the medical value chain, but a chain is only as strong as its weakest link, and people may shun our services for reasons other than mere inaccuracy. What can we do to build better collaborative relationships between the members of our healthcare teams, such as improving the two-way sharing of perspectives between different specialists involved in a patient's care?

The single-minded pursuit of correctness may also undermine the cultivation of important academic perspectives. In some cases, there is more to know than the existing textbooks and journal articles, our de facto answer key, can assess. If we look beyond merely getting every question right, we can address an even more important question: what opportunities are before us to advance medical knowledge? The information in the textbooks of today needs to be improved upon, and that will require a willingness to engage with the unknown, to venture where existing answer keys can no longer guide us. We need to approach our clinical work with more than a determination not to be wrong. We need skepticism, curiosity, and creativity.

If our medical education programs are going to carry us beyond mere correctness, we need to cultivate a more complete model of medical excellence. We should encourage learners to devote as much or more time and energy to asking good questions as getting the right answers. If they are really thinking for themselves, they will not always be content merely to accept educators' opinions as irrefutable truth. Instead they will place less reliance on conformity and more on intellectual rigor. We should spoonfeed them less and send them out foraging more. When is the available information insufficient, and how can they go about pursuing it? When can uncertainty or groundless certainty be exploited for educational and investigative purposes?

Errors are not medicine's cardinal sins. In many cases, we should treat errors not as failures but as opportunities for discovery. In the real world, the best physicians among us learn more from our mistakes than from our successes. We must scrupulously guard against a culture that treats error as intolerable and embarrasses or even punishes every mistake. In these settings, no one learns from their own mistakes, let alone the mistakes of others, and the failure to learn is a sign of approaching obsolescence. Such an attitude is inimical to the spirit of inquiry and the quest to continually improve the quality of our practice.

In many cases, we would do medical students and residents a favor by presenting them problems to which the answers are already available. Too often, learners otherwise devote so much energy to getting the right answer that other important aspects of a case get neglected. Correctly diagnosing a patient's congestive heart failure may be less important than elucidating the psychosocial features of the patient's home life that must be addressed by any successful treatment regimen. Another equally valuable approach is to withhold the "correct" answer indefinitely, so that learners never find out whether they got it right. This enables learners to become more effective monitors of their own performance in ways that are more reflective of the real-world practice of medicine. We need to learn how to live with, and to optimize our management of, uncertainty.

To be sure, we want to educate physicians who actively audit the accuracy of their performance, and we should do our best to equip them to do so effectively. By immediately telling them whether they were right or wrong, however, we may stunt their own process improvement approaches.

We also need to evaluate learners in ways that transcend mere correctness. Scores on most standardized tests, our favorite evaluation technique, neglect vital factors of medical excellence. For example, how effective are learners as consultants, at eliciting key information from patients, and as investigators and educators? Systems of evaluation and reward should be sufficiently balanced and comprehensive that they reflect a complete view of medical excellence. To do otherwise is to distort both the educational process and its product.

0 0

Post a comment