Harwood-Nuss' Clinical Practice of Emergency Medicine, 6 ed.

INTRODUCTION

The Approach to the Patient in the Emergency Department

Robert L. Wears

… in the physician or surgeon, no quality takes rank with imperturbability … Imperturbability means coolness and presence of mind under all circumstances, calmness amid storm, clearness of judgment in moments of grave peril, immobility, impassiveness, or, to use an old and expressive word, phlegm. It is the quality which is most appreciated by the laity though often misunderstood by them; and the physician who has the misfortune to be without it, who betrays indecision and worry, and who shows that he is flustered and hurried in ordinary emergencies, loses rapidly the confidence of his patients.

—Sir William Osler, Æquanimitas

It is one thing to practice medicine in an emergency department (ED); it is quite another to practice emergency medicine. The effective practice of emergency medicine requires an approach, a way of thinking that differs from other medical specialties.

Its physical limitation to the ED is perhaps the least important of emergency medicine’s characteristics. Five factors have led emergency physicians to develop a unique approach to the patient: the pressure of time and volume, the variety of conditions faced, the paucity of information, the limitation of therapeutic options, and the constraint of disposition.

TIME AND VOLUME PRESSURE

More than any other specialists, emergency physicians experience clinical work as turbulent flow in a constricted channel. In a true emergency, seconds to minutes may make the difference between life and death or serious disability. In these situations, emergency physicians, contrary to much of their previous training, must be prepared to “treat first and ask questions later.”

Also, the time available for an emergency physician to evaluate and think about any given patient is severely limited by the demands of other patients being managed concurrently. In most practices, during the busiest part of the day an emergency physician has on average only 10 to 15 minutes per patient for evaluation, testing, treatment, disposition, and documentation. Whether this is adequate or not is immaterial; the reality is that no more time is available.

The combination of time and volume pressures forces emergency physicians to be much more aware of priorities among patients. Although trauma surgeons occasionally face triage conditions, emergency physicians are the only practitioners who routinely make (at least mental) triage decisions every day.

VARIETY OF CONDITIONS

As a specialty in breadth rather than depth, emergency physicians must manage a wider variety of conditions than any other specialists, save perhaps family practitioners in remote settings. In addition, unlike other specialists, emergency physicians must shift cognitive frames rapidly. Practicing emergency medicine is like carefully lining up a putt, then dropping the putter, picking up a tennis racket to return a volley or two, quickly sidestepping an onrushing tackler, and then returning to sink the putt.

PAUCITY OF INFORMATION

Emergency physicians frequently deal with episodes of real or perceived crises in patients of whose past course they are unaware. Old records are often unavailable, electronic records may be available but of doubtful provenance (1), and patients’ memories are limited in scope and reliability. The information-gathering options available in the ED are similarly limited. Only a small subset of the vast diagnostic armamentarium is available to the emergency physician within a reasonable time.

LIMITED THERAPEUTIC OPTIONS

Options for treatment are limited as well. Often, emergency physicians can provide only temporizing or symptomatic treatment, whereas definitive management must be deferred to another specialist. In addition, in emergency situations, the tolerance for therapeutic failures or misadventures is more limited than in nonemergencies.

Although one may not realize it when viewing the chaotic, turbulent activity of a busy ED, emergency medicine possesses a certain elegance enforced by these constraints. There is an attractive intellectual simplicity in meeting the challenge of providing timely, accurate care using primarily one’s own hands and brain, supplemented by a few, limited laboratory and imaging tests.

CONSTRAINT OF DISPOSITION

Every patient interaction an emergency physician has must soon be ended. This forces emergency physicians to focus on the “bottom line.” No matter how uncertain the diagnosis or how much extended observation or testing might help, every patient encounter in the ED ultimately reduces to three binary decisions: Is the patient sick or not sick? If sick, should I treat or not treat? Should I admit or discharge? These questions must be answered despite the lack of definitive information, the lack of time to collect or consider additional information, or the lack of availability of advice from colleagues and consultants.

A FRAMEWORK FOR APPROACHING THE ED PATIENT

There are many sources that outline the mechanics of history taking and physical examination, and this section does not discuss these areas in detail. Instead, it concentrates on how they are different in emergency medicine.

The first question to be asked on initial approach to a patient in the ED is, “Do I need to resuscitate this patient?” Corollaries to this question are, “How great is the threat?” and “How soon must I act?” The first question is usually answered in the first few seconds at the bedside and is as often based on an overall gestalt of the chief complaint and the patient’s general appearance as on specific complaints or vital sign abnormalities.

In true emergencies, the usual sequence of history, physical examination, laboratory testing, and treatment is altered by the need to take rapid action. In the most extreme cases, the sequence becomes treatment, physical examination, laboratory testing, and history.

When action is urgently required, the emergency physician must commit to it unhesitatingly. This requires mental preparedness; there is no time to ponder the benefits and risks of the various options for managing acute upper airway obstruction. The thinking required must be invested ahead of time, and a plan of action internalized by the physician before the situation requiring it arises.

If action is to be effective, it often must be initiated before all the information bearing on the decision is available. This requires a “bias toward action” in the emergency physician’s mind. While this is occasionally a source of criticism from other specialists who fail to understand emergency medicine, in the proper circumstances it can save lives. A simplistic example of bias toward action is putting a patient with chest pain on a cardiac monitor on arrival to the ED; clearly the diagnosis of heart disease has not yet been made, so it is not certain that the monitor is truly “needed,” but we anticipate a potential need and act without waiting for confirmation. A more dramatic example is sudden upper airway obstruction, where the decision to perform a cricothyroidotomy must be made promptly; to wait until the need is clear may be to wait too long.

Part of the art of emergency medicine is the ability to reliably discriminate between cases requiring urgent action and those allowing a more measured approach. Maintaining a bias toward action should not be used as an excuse for an indiscriminate, “shoot-from-the-hip” approach to the patient; it is a tool that the emergency physician must learn to use effectively.

BEDSIDE EVALUATION

If resuscitation or other urgent interventions are not required, the usual framework for patient evaluation (history, physical, laboratory, treatment, disposition) can be used. However, in the ED, certain factors require more or less emphasis or alteration (Table INT.1).

Patients and physicians in EDs are generally complete strangers, and thus patients have legitimate questions about the emergency physicians they have just met: Are they capable? Can they be trusted? Consequently, the first task faced by the emergency physician, even before information gathering begins, is to build a working relationship with the patient right away (2).

The initial contact the physician makes should be social—an introduction. This should include everyone in the room, and should be accompanied by physical contact—a handshake or touch. This establishes a tone of personal respect for the patient and helps enlist the family, if present, in an alliance with the physician. As early as possible the physician should show empathy with the patient by acknowledging recognition of an apparent need. This can be done by an offer to relieve apparent pain, if appropriate.

The body language of the physician–patient encounter is important. Often in the ED, the standing physician presides over the supine, frequently disrobed patient, presenting an authoritarian image that can be threatening to patients. As often as possible, therefore, the physician should sit while conducting the initial interview, ideally bringing his or her head to the same level as the patient’s (3). This is especially important in dealing with children, but should not be neglected in adult patients.

The information-gathering phase of the encounter should begin with a general, open-ended approach, such as, “How can I help you?” or “Tell me your story.” More direct approaches, such as “What is your problem?” risk being misinterpreted. It is important to remember that both parties, not just the physician, are interested in gathering information at this time. Patients’ desires for information, explanation, and reassurance are great, but are seldom met by physicians (4,5). One way emergency physicians can meet this almost insatiable need for information early in the encounter is by commenting about findings during the physical examination: “Your heart sounds normal,” or “Your throat looks red” (2).

Given the time pressures of emergency medicine, physicians typically fear losing control of the interview and attempt to keep the patient focused and on target. Observations of physician–patient interactions show that patients are interrupted by the physician on average about 18 seconds after they begin to relate their history, but that if uninterrupted, 80% can get their story out within 1 minute (6). Physicians’ time concerns are real, but if the physician dominates the interview, the patient will become passive and volunteer little (6); the physician will then have to spend additional time extracting information later. In addition, the chief complaint may not always be mentioned first; failure to wait for the true chief complaint to emerge can lead to false trails and lost time (2). Considering the gain in rapport with the patient and the potential for mischief with early interruption, it seems wise for emergency physicians to discipline themselves to be quiet for the first 60 seconds of the interview.

The introduction of electronic records is a potentially intrusive factor in patient–physician communication. Initially there were concerns that physicians might be perceived as paying more attention to the computer than to the patient, although as this technology becomes more ubiquitous, patients will likely come to treat the physician–computer interaction as routine.

However, it is important that the bullet point, checkbox nature of electronic charting not be allowed to prevent the physician and patient from codeveloping the narrative of illness (or injury). Understanding about that nature of a patient’s problem is acquired iteratively; early information leads to the development of a tentative structuring and categorization which guides further investigation, but which is itself modified according to newer information; the structure and categories change as understanding improves. But if the full narrative never comes into existence, this cycle of increasing understanding is broken, and the physician is left with a model of the patient’s problem that has been prematurely frozen in an unfinished and perhaps even erroneous state (7,8). Thus, it is good practice to take a history and perform a physical examination before interacting with a charting template.

DISPOSITION

Once the initial assessment is complete, the physician should outline the plan to the patient, and should give an honest estimate of about how long it will take. It is also useful to provide some anticipatory guidance about the probable outcome, particularly in complex, chronic problems that are unlikely to be resolved in an ED evaluation. During the wait for laboratory results, consultations, or observation, the patient and family should be updated periodically on progress and asked if there is anything they need.

Finally, once a disposition is reached, the plan should be carefully and clearly communicated to the patient and family. Particularly if the patient is to be discharged, specific instructions about follow-up and reasons to return to the ED should be covered. Because patients are likely to have difficulty remembering a series of detailed explanations or instructions (9), it is most helpful if they are given instructions in writing. Having a printed instruction sheet does not relieve the physician of the obligation to instruct the patient, but it does help to reinforce those instructions once the patient has left the ED.

A FRAMEWORK FOR DECISION MAKING IN EMERGENCY MEDICINE

Medicine is essentially about resolving ambiguity through action; in emergency medicine, this process is intensified. However, a great deal of medical decision making occurs at a subconscious level where it is inaccessible to the physician. Although formal decision-making models do not describe physicians’ mental processes well, they can provide some useful insights into the nature of decision making in the ED.

MANAGEMENT, NOT DIAGNOSIS

Traditionally, medicine has focused on diagnosis as the clinician’s central important task. Emergency medicine makes it clear that this is erroneous: In medicine in general, and especially in emergency medicine, the central task is not diagnosis, but management; not so much “what is going on?” as opposed to “what should we do now?” Often a diagnosis cannot be made under the constraints of an ED evaluation; the great insight that emergency physicians have contributed to their colleagues in other specialties is the notion that there need not always be a diagnosis. If one can be made, it is extremely helpful, but if not, decisions must still be made and actions must still be taken.

In addition, patients want more than simply a diagnosis. They want explanation and reassurance. For example, parents will not be satisfied to know that their child’s abdominal pain is caused by gastroenteritis; they want to know that it is not appendicitis. Simply providing the diagnosis, however correct, dismisses the parents’ concerns and leaves them unsatisfied. By empathizing with their fears, the physician can strengthen rapport with the family and take a position as their friend and ally, rather than as a remote authority. Thus the intellectual task is to devise a reasonable plan of management, which may include, but not be limited to, making a diagnosis.

DECISION THRESHOLDS

The concept of “decision threshold” (10) clarifies some of the processes at work when emergency physicians make decisions and adds specificity to the often used but somewhat vague concept of “index of suspicion.” Consider the following hypothetical scenario of the simplest possible situation in emergency medicine: There is only one disease under consideration and there are only two possible actions—to treat or not treat—based on the emergency physician’s assessment of the patient. In this scenario, there are four possible outcomes: The patient either has the disease and may be treated or not treated, or does not have the disease and is treated or not treated. The optimal management decision depends on the values of these outcomes and on the probability that a particular patient (or group of patients) actually has the disease in question.

If the patient is almost certainly nondiseased (i.e., the probability of disease is near zero), then the correct decision is to not treat, because treatment entails costs and risks of its own. Conversely, if the patient surely has the disease (probability near 1), then the correct decision is to treat, because the costs and risks of treatment are outweighed by the negative consequences of failing to treat. Therefore, there must be some threshold probability between 0 and 1 at which the decision is a toss-up; that is, either choice will produce about the same outcome. This treatment threshold need not (and often does not) coincide with the diagnostic thresholds that clinicians traditionally have been taught. Consequently, the clinician’s task is not necessarily to make a diagnosis, but to judge whether the probability of disease is clearly over or clearly under the treatment threshold, and to act accordingly.

The threshold approach can be extended to more complex problems. For example, a more realistic scenario allows use of a diagnostic test to improve the physician’s initial assessment. Now there are two decision thresholds. At very low probabilities of disease, it is better not to test and not to treat, because the outcome of large numbers of false positives erroneously treated outweighs that of the occasional diseased patient detected. Similarly, at very high probabilities, it is better not to test, but to treat everyone, because the consequences of large numbers of false negatives erroneously denied treatment outweigh the occasional nondiseased patient spared unnecessary treatment. At intermediate probabilities, testing and treating only those patients with a positive test will produce the best overall outcome. Thus, the clinician’s task is to decide whether a given patient’s probability of disease lies either below the no treat–test threshold, or above the test–treat threshold. Testing is useful only in changing the management decision in the area between the two thresholds. While there may be other benefits to testing besides producing a change in management (e.g., reducing uncertainty or clarifying prognosis) (11), the threshold approach provides a rationale for understanding physicians’ differing choices when faced with seemingly similar clinical scenarios.

In framing decision making in this way, it is important to note that a very broad idea of “test” is used. Anything that helps physicians revise their assessments is considered a test in this sense, so therapeutic trials, periods of observation, clinical scoring systems, or practice guidelines could all be considered “tests” because they help physicians sort out the probabilities of diseases. There is interesting evidence that physicians’ diagnostic and management performance improves when they can take advantage of the nonverbal cues that accompany face-to-face communication, even though this “test” occurs largely at a subconscious level (12).

One drawback of the threshold concept is that it can taken to the extreme and result in a very narrow, overrational approach to problem solving (13). However, the way in which emergency physicians resolve the question of whether or not to act (or what actions to take) seems to be by the use of naturalistic or “event-driven” decision making (14). This contrasts strongly with the algorithmic or analytical decision-making process taught in medical school. Naturalistic decision making is effective in settings characterized by ill-defined problems; dynamic, uncertain environments; shifting, unclear, or competing goals; tight but nonlinear or hidden coupling between actions and their effects; time pressure; high stakes; and scant opportunity for learning by trial and error.

The decision-making process can be divided into several phases. First, the physician classifies the current situation as typical or atypical, based on the matching of critical cues to stored patterns or schemata. If the situation is typical, the stored patterns will evoke a customary set of responses. If it is not typical, variations of customary responses are considered first, and then novel, “one-of-a-kind” responses are considered. Clearly, experience is critical in developing a sufficiently rich set of stored schemata such that the vast majority of clinical problems can be recognized and dealt with by such a preplanned response.

Next, the physician engages in serial evaluation of the available courses of action, beginning with the most typical response, and evaluating each by mentally simulating the expected outcomes. Once a sufficiently satisfying action is discovered, it is implemented. In other words, there is generally not a search for the best of all possible responses, but for a sufficiently good response. The payoff for accepting a good, but not necessarily best, response is that decisions can be quick and almost effortless. In fact, a great deal of physicians’ expertise seems to lie in their ability to constructively perceive the problem, which seems to lead automatically to a solution without much conscious effort. This contrasts sharply with the slow and laborious analytic approach of exhaustively considering all the possibilities and eliminating them one by one until the best option is identified.

Learning this method of decision making is not easy. It can be gained with experience supplemented by feedback and reflection, and is generally communicated by narratives of meaningful cases (15). It does not appear to be attainable through application of formal analytical methods. Such methods have their place, but they have not proven useful in helping medical students or residents to become physicians.

Common Pitfalls

There are three common pitfalls in the approach to the patient in the ED.

• The first is tunnel vision. Here the physician, in single-minded pursuit of the disposition, fastens onto a single complaint (typically the first offered) without waiting to be sure that the patient’s chiefcomplaint has been elicited. A related error is the premature closure of hypothesis generation (e.g., assuming that chest pain is caused by myocardial ischemia without first mentally ruling out the possibilities of aortic dissection, pulmonary embolism, esophageal rupture, or pericarditis). This problem often afflicts the less experienced, who, once having reached a characterization of the problem, are reluctant to reopen consideration of it (16).

• The second pitfall is just the opposite: an inability to see the forest for the trees. Here the physician is unable to rank findings in any order of importance, is reluctant to close off hypothesis generation, or is unable to integrate the findings into a small number of possible explanations. This is frequently manifested by procrastination in decision making, perhaps hidden behind a cascade of laboratory testing. The ability to simultaneously entertain a modest number of possible explanations for the patient’s problem without prematurely settling on one, or letting the number in contention grow too large, is the “golden mean” between these two problems.

• The final pitfall is failure to attend to the patient. Time “gained” by failing to attend to the social interaction with patient and family is likely to be lost later, as missing information may not be extracted except by persistent questioning, leaving the physician at risk of pursuing many false leads before happening across the true trail. More important, perhaps, is to remember that the doctor–patient interaction remains at its most fundamental level a meaningful social encounter, on which all healing and care, in the broadest sense, are based.

REFERENCES

1. Hirschtick RE. A piece of my mind. Copy-and-paste. JAMA. 2006;295(20):2335–2336.

2. Rosenzweig S. Emergency rapport. J Emerg Med. 1993;11(6):775–778.

3. Murphy A. To sit or not to sit: A question of cultural performance. Ann Emerg Med. 2008;51(2):194–196.

4. Adamson TE, Tschann JM, Gullion DS, et al. Physician communication skills and malpractice claims. A complex relationship. West J Med. 1989;150(3):356–360.

5. Waitzkin H. Doctor–patient communication. Clinical implications of social scientific research. JAMA. 1984;252(17):2441–2446.

6. Beckman HB, Frankel RM. The effect of physician behavior on the collection of data. Ann Intern Med. 1984;101(5):692–696.

7. Resnik P. Language technology, electronic health records, and the clinical narrative. http://www.umiacs.umd.edu/∼resnik/resnik_sxsw2012.pdf. Accessed April 10, 2012.

8. Resnik P, Niv M, Nossal M, et al. Communication of clinically relevant information in electronic health records: A comparison between structured data and unrestricted physician language. In: Computer Assisted Coding 2008; 2008.

9. Vashi A, Rhodes KV. “Sign right here and you’re good to go”: A content analysis of audiotaped emergency department discharge instructions. Ann Emerg Med. 2011;57(4):315–322.e311.

10. Pauker SG, Kassirer JP. The threshold approach to clinical decision making. N Engl J Med. 1980;302(20):1109–1117.

11. Fryback DG, Thornbury JR. The efficacy of diagnostic imaging. Medl Decis Making. 1991;11(2):88–94.

12. Friedman CP, McNutt RA, De Bliek R, et al. Visual information and the diagnosis of chest pain. Acad Med. 1994;69(10 suppl):S28–S30.

13. Berg M. Rationalizing Medical Work: Decision Support Techniques and Medical Practices. Cambridge, MA: MIT Press; 1997:238.

14. Klein G, Orasanu J, Calderwood R, et al, eds. Decision Making in Action: Models and Methods. Norwood, NJ: Ablex Publishing Company; 1993:448.

15. Klein G. Sources of Power: How People Make Decisions. Cambridge, MA: MIT Press; 1998.

16. Schubert CC, Denmark TK, Crandall B, et al. Characterizing novice-expert differences in macrocognition: An exploratory study of cognitive work in the emergency department. Ann Emerg Med.2012;61(1):96–109.