Handbook of Consultation-Liaison Psychiatry

26. The Emergency Department Setting

Seth Powsner


26.1 Overview

26.1.1 Expulsion from the Social Matrix

26.1.2 Vignette

26.2 Evaluation

26.2.1 Medical Clearance

26.2.2 EMTALA Emergency Medical Transfer and Active Labor Act

26.2.3 Psychiatric Evaluation

26.2.4 Diagnostic Considerations

26.3 Treatment

26.3.1 So-Called Agitated Patients

26.3.2 Alcohol Withdrawal and Sedative/Hypnotic Withdrawal

26.3.3 Anxiety

26.3.4 Catatonia

26.3.5 Conversion Disorder

26.3.6 Intoxication

26.3.7 Major Depression

26.3.8 Mania

26.3.9 Psychosis

26.3.10 Violence and Suicide

26.4 Confidentiality

26.4.1 Realities

26.4.2 Tarasoff Considerations

26.4.3 Child Abuse, Gunshot Wounds, and Local Rules

26.4.4 Health Insurance Portability and Accountability Act of 1996 (HIPAA)

26.5 Special Problems in Emergency Departments

26.5.1 Admission Screening

26.5.2 Patients Brought from Jail

26.5.3 Untruths

26.1 Overview

Consultants may be called to an emergency department for a variety of reasons. Most requests are like those from a general hospital ward. However, two issues are notable: a broader definition of psychiatric emergencies and greater concern about patient rights. These issues stem from an emergency department's lack of a buffer from its surrounding community: patients come as they are, whether pushed, or just so inclined. There is little or no time for patients and emergency department staff to come to any understanding. In this absence of a traditional physician patient relationship, consultants may be forced to change their usual approach.

Psychiatric emergencies now include patients who are depressed, disorganized, odd, or acting badly for no obvious gain. Psychiatric emergencies traditionally meant patients going berserk: yelling, screaming, likely to hurt themselves or others. The newer, broader definition follows in part from a better appreciation of the morbidity of untreated psychiatric illness. And it follows in part from a fear of liability for homicidal public violence, perhaps as part of a suicide attempt. Ever since the Columbine High School massacre, Americans have become leery of any adolescent talk or behavior suggesting depression or self-destructive urges. Widely publicized workplace shootings and maternal infanticides have further increased public fears of mental illness.

Concern for the patient's right to accept or refuse medical treatment is a frequent trigger for emergent consultations. Old attitudes were simpler: "If you want treatment, walk in: if you don't want treatment, walk out." Such attitudes are very fitting for a country of frontiersmen. However, there are few frontiersman left, and more urbane citizens worry that the complexities of medical treatment will elude anyone whose cognitive capabilities are impaired, by mental or by medical illness. Psychiatric consultants now find themselves cast as arbiters of medical choice.

An emergency department setting does entail other specific issues: time constraints, incomplete histories, overt patient intoxication, admission screening, ambiguous patient status, shifting treatment personnel. Moreover, some emergency departments expect their psychiatric consultants to assume responsibility for cases, though these consultants may have neither dedicated nursing staff nor dedicated psychiatric beds. General hospital consultation skills are helpful in an emergency department, but when assuming responsibility for cases, good inpatient treatment skills can become critical.

Because psychiatric consultation to an emergency department has so much in common with consultation to general medical and surgical wards, this chapter focuses on the areas of divergence. The nature of patients coming to emergent psychiatric attention is the first topic. Then, a clinical vignette is presented for purposes of discussion throughout this chapter. Discussions of medical clearance, evaluation, and treatment follow.

26.1.1 Expulsion from the Social Matrix

People end up in an emergency department because they have been expelled from their social matrix. Almost everyone lives in some sort of community, some social matrix. Neighbors, if not immediate family members, surround most people as they begin their day. Working people travel to another segment of their matrix for part of the day. There are usually other places to eat, shop, or find entertainment. There are different people in different segments of each matrix, different expectations in different segments. However, there are expectations, there are limits on behavior; failure to meet expectations, failure to abide by limits, leads to expulsion.

Suicidal comments commonly lead to expulsion. Family and friends were probably more encouraging and tolerant in the past. These days they may worry about murder-suicides or another Columbine massacre. They often push patients to psychiatric attention.

Some patients do come on their own. They complain they are anxious, overwhelmed, or depressed, even suicidal. However, only a fraction come on their own, truly believing that no one around them is concerned. And many of them first seek care in a clinic or office setting.

Criminal behavior usually results in transport to jail -a hardened, closed segment of the social matrix. Only a small fraction of these people come to psychiatric attention. On the other hand, expulsion for unacceptable behavior that serves no obvious criminal gain, or is just odd, usually results in transport to emergency psychiatric attention. One patient was brought to our emergency department for waving a snow shovel at passing cars. He had not actually hit a car or driver. But there was very little snow on the ground and no conceivable purpose to his actions. Eventually, we located his mother who told us that he suffered from schizophrenia, had stopped taking his medication, and wandered away from home. (We returned him to his caregivers in a nearby state.)

Communities could make other arrangements for their strangely behaved: a local mental health center could operate an around-the-clock intake service. However, these days, managed care abhors the labor costs associated with around-the-clock service. Hospitals, police departments, fire departments, and fast food restaurants are the only around-the-clock operations found in most communities. Of these, only hospitals and police feel obliged to deal with disturbed patients.

26.1.2 Vignette

Psychiatry was called to consult on a middle-aged, married, Caucasian woman with no formal psychiatric history. Her husband brought her to our emergency department Monday evening because she could not walk, again. Generally healthy, even athletic, she first had trouble walking a few weeks ago, after a pet was lost during a family outing. The pet was eventually found. However, there have been some continuing financial stresses, and she was very ambivalent about a recent, milestone birthday.

She first complained of leg weakness on the day after their outing. Her husband took her to a small hospital near home. The workup was unrevealing. She was released. She suffered a recurrence a few days later, which led to an admission and an extensive workup (including magnetic resonance imaging and lumbar puncture). All tests were normal.

About a week has passed since discharge. Neither patient nor husband could describe any other unusual events. Staff at the first hospital faxed all available test results to us. Her blood count, sedimentation rate, glucose, and a few other easily obtained lab tests were rechecked. There seemed to be very little possibility of any rapidly progressive or new disease; no abnormalities were found.

Patient and husband were cooperative. Both were concerned about her inability to walk, both were a bit exasperated that there was no diagnosis. Her husband eventually became tired of all the time spent waiting around our hospital; he went home to relieve their baby sitter. Our patient seemed a bit subdued, affect a bit flat, but otherwise entirely normal. No other stresses or conflicts were elucidated. Family history revealed no mental illness and no neurologic illness.

The emergency medicine staff initially requested psychiatric consultation, but later suggested that this patient be moved from their area to our locked unit (within the emergency department). No one believed she was a safety risk. Since space was not tight in our medical area, transfer was delayed for a few hours for a trial of oral lorazepam. She was given 1 mg along with the suggestion that her weakness was likely due to stress, for which lorazepam might be helpful. She was also told that if she could not walk, she would have to be admitted to the psychiatry department, since we could find no reason to admit her to a medical service. (All of this was communicated in a very matter-of-fact manner.)

After about 90 minutes, the patient was reexamined and was now able to move her feet a bit. After about 2 hours, she could stand on her own. When called, her husband sounded tired, but willing to fetch her, provided she could walk to use the bathroom on her own. This was relayed to her. About 20 minutes later she felt ready; she walked 20 feet to the nearest bathroom. She was discharged with a referral to a psychiatric clinic near her home.

26.2 Evaluation

26.2.1 Medical Clearance

The above vignette raises a number of issues in emergency department consultation. First and foremost is the issue of medical clearance. Colleagues talk of patients being medically cleared as if it were a routine process like being disinfected or immunized. Unfortunately, it has more in common with security clearance, a process of looking into someone's history for clues of disloyalty or past criminal behavior that might lead to future security breaches or outright spying. There is no reliable lie detector, no reliable medical illness detector. There is no simple collection of medical tests to ensure the absence of medical illness affecting mental function or behavior (Allen et al., 2005; Lukens et al., 2006).

The odds of a mental illness being due to a general medical condition are reduced if a patient's urinalysis is benign and blood tests are all within normal limits (white blood cell count, hematocrit, glucose, blood urea nitrogen, creatinine, and electrolytes). That leaves thyroid disease and liver disease untested, and even testing for those leaves very pertinent conditions like multiple sclerosis and B 12 deficiency untested. The list goes on and on. Multiple sclerosis, vasculitis, even Wilson's disease are all possible, although uncommon, in a previously healthy adult. As our vignette demonstrates, a thoughtful workup for acute neuromuscular disease ranges far beyond anything reasonable in an emergency department; it used to take a 2-day medical admission.

Viewed from another perspective, our patient's extensive workup demonstrates that most adults who report they have been healthy are in fact healthy, at least as far as simple blood tests reveal. Medical history, vital signs, and physical examination are the most useful steps to detect medical illness causing psychiatric symptoms.

What about uncooperative patients, or those very disturbed patients who cannot give a coherent medical history? They represent a true challenge for emergency medicine physicians and psychiatrists alike. The history still provides the most helpful information, though it may have to be obtained from ambulance staff, police, friends, family, and old records. The physical exam gives a clue about trauma, systemic illness, and recent living conditions. For elderly patients, urinalysis may reveal unsuspected urinary tract infection. For younger patients, alcohol breath testing, finger-stick glucose measurements, and urine toxicology screening are the laboratory tests most likely to provide a clue about their acute behavioral disturbances.

26.2.2 EMTALA-Emergency Medical Transfer and Active Labor Act

EMTALA (part of COBRA, the Consolidated Omnibus Budget Reconciliation Act of 1986) has made medical screening exams mandatory for hospital emergency departments, regardless of chief complaint (American Academy of Emergency Medicine, 2006). Free-standing psychiatric walk-in clinics, even some attached to medical clinics, would not normally perform physical exams, check vital signs, or draw blood for laboratory tests. Nonetheless, EMTALA mandates medical screening in an emergency department. Patients with only psychiatric complaints are not exempt. Luckily, EMTALA's requirement is not detailed, so a short examination, expanded only based on a patient's physical complaints, seems appropriate.

26.2.3 Psychiatric Evaluation Overview

Once a patient can be considered cleared from a medical standpoint, psychiatric evaluation becomes the main task. Psychiatric evaluation in an emergency department is very similar to psychiatric evaluation on a medical or surgical service. The primary difference is in the nature of a patient's story. Consultation for a medical patient requires an understanding of the patient's medical history and treatment with particular attention to aspects that affect brain function and to aspects that resonate with prior psychological experiences. On the other hand, psychiatric patients arriving in an emergency department usually have no medical illnesses and no active treatments. The Story

The story, critical to a psychiatric patient's arrival in an emergency department, is the story of their expulsion from their social matrix. Did they do something dangerous? Did they say something? Who became concerned? Is this a change? How long has this been happening? If this information is not available, consultation must proceed cautiously, if at all. Occasionally, the available information is misleading. Ex-boyfriends and ex-girlfriends have been known to falsify reports to the police of odd or dangerous behavior. Their hapless victims arrive in our emergency department, quite surprised. Once calm, they can usually provide some collateral source of information to support their request to be released.

The story is also critical because symptoms and diagnosis do not always determine treatment. A patient may hear voices, telling him he is no good, telling him that he ought to die, but he hears them chronically and ignores them. If such a patient walks into an emergency department, on his own, for an intractable cough, he may need testing for tuberculosis, but he does not need psychiatric admission. However, another patient with exactly the same complaints, sent from jail because he has been banging his head bloody on cell bars, likely needs psychiatric admission.

A patient's history should make sense as a story. The patient in our vignette might have recounted something like this: "Things have been tough around our house since my company relocated out of state. I had a good job, but my husband's work pays more, and he's got a lot of seniority, we couldn't afford to move. Problem is, I haven't been able to find work. I haven't got any special skills. Money's been tight. I was fretting about the cost of our camping trip. Then our dog ran away. I couldn't stand it. As soon as we found the dog, I insisted we go home. I couldn't sleep. I was so weak the next day."

Such a story would make perfect sense as part of our vignette. It offers an entirely plausible sequence of events. Its wording provides a nice linguistic connection to our patient's symptom. Alas, no such recitation was part of either our patient account or her husband's.

Patients occasionally offer stories that make no sense from a psychological or social perspective: "Everything was fine till today. This evening I couldn't walk. Why do you keep asking me how I'm getting along with my family?"

A patient might suddenly become lame, without any change in his or her life, but not based on psychological issues. Such a story only makes sense if the untold prologue goes something like this: "Ms. Jones had a small ventricular septal defect that was never documented. She failed to keep an appointment for an echocardiogram years ago, ordered by a physician who heard a murmur. Ms. Jones believed her health was fine. She complained that their local cardiologist wanted a new car, that's why he recommended a fancy test."

Some stories require detective work. Patients occasionally say: "I've always been depressed. Been that way my whole life. Today's no different. Couldn't take it any more. I came to the emergency department." This is not a story. As children, even these patients would never have accepted the equivalent bedtime story: "The prince was riding around his kingdom. He turned into a frog. The end." It falls to a consultant or emergency department staff to call the patent's friends or family in search of an undisclosed offense toward an undisclosed witch.

Some stories never achieve coherence. Our lame patient's vignette makes unsatisfying human drama. No doubt, both husband and wife omitted some critical details. Was there a dispute, an affair? It was never revealed to us. Treatment had to proceed with a nonspecific intervention. Social Review of Systems

A social history and a developmental history can be very helpful for emergent psychiatric evaluation. The challenge is to stay on task, to understand why a patient has been extruded from their social matrix, why at this juncture. Reducing social history to an inventory of substance use, or reducing developmental history to a timeline of infant milestones, is not informative. It is more helpful to think in terms of patients' progress and their place in society: a review of their social systems. Did the patient start poor and later climb the corporate ladder? Was the patient born into riches, but slipped into a life on the streets and in homeless shelters? These contrasting trajectories suggest different diagnostic possibilities, though either patient might initially come to our attention as a man found wandering by the police.

For the social review of systems, first come questions about our patients' start in life. What was their family of origin like? Where did the patient grow up: in a ghetto, a rough-scrabble rural area, a wealthy suburb? Modem American demographers would opine that the mother's zip code at the time of the patient's birth provides a good clue. However, it is friendlier to ask: "Where were you born? Where did you grow up? What did your parents do for a living?" Indeed, many patients will talk at length about childhood, their family, friends, accomplishments, and disappointments, yielding everything needed to understand their current crisis.

Knowing the patient's starting point in life, we can then ask what kind of school they attended, how far their education progressed. These answers round out our picture of the patient as a child; they give us a sense of how he or she performed at society's first task: being a student. Traditional developmental milestones are not very helpful in evaluating adults in an emergency department; if they were not successful at walking, talking, and toilet training, they would probably arrive from a supervised living arrangement, complete with a report on their disabilities.

Next come questions about young adult challenges: work, military service, finding a spouse. Now, it may turn out that the patient succumbed to schizophrenia or substance abuse at this critical stage, which changes the rest of our expectations. Or the patient may have had a more benign life, in which case we can ask how she or he did relative to their parents.

Finally come questions about current social function: home, work, family. These are traditional elements of a social history, but more meaningfully seen as part of a lifelong social trajectory.

Questions about alcohol, tobacco, and substance abuse may fit better with questions about past psychiatric history than about social history. Americans increasingly view addictions as a type of psychiatric illness. It is only physicians who have been trained to think of a social history as "no tobacco," or "three to four beers per week." Traditional Review of Systems

A review of systems (ROS), in the traditional sense, is a useful addition to the psychiatric evaluation. It can serve as a wrap-up, or a short review of the patient's medical history, head to toe, system by system. It helps ensure that a consultant understands the patient's condition.

An ROS can also be very useful to psychiatrists reimbursed by Medicare and other payers following the Centers for Medicare and Medicaid Services (CMS) 1995 and 1997 Documentation Guidelines for Evaluation and Management Services (Centers for Medicare and Medicaid Services, 2006). The CMS guidelines divide a physician's chart note into components, which are then tallied according to a complicated scoring system. This process has been termed "bullet counting" in honor of the bullet points adorning the many PowerPoint presentations used to explain this system. For a history of present illness (HPI) to rate as an extended HPI, as obtained by many consultants, an ROS must be included (or there must be accurate documentation of time spent on counseling and coordination of care).

Whatever the motivation, it does not hurt to recast the usual psychiatric concern for neurovegetative signs into a broader view of a patient's physiology: anorexia and constipation into questions about general gastrointestinal function, anergy into questions about endocrine function, etc. Recording a patient's answers in the form of an ROS can improve both the consultant's understanding of a case as well as the billing office's rating of a case.

An ROS can also serve the mental status examination (MSE). After questioning patients about eye trouble, ask them to identify three objects of decreasing size. It is not as accurate as using a reading chart, but does provide a quick test of language as well as vision. This maneuver also leads to a test of immediate memory: ask patients to name the objects again, without prompting. Finally, toward the end of the MSE itself, ask once again for the names of the objects, yielding a measure of delayed recall. Mental Status Examination

A formal MSE is a critical task for psychiatric consultants in an emergency department. It is second in importance only to a patient's history for diagnosing psychiatric illness and for distinguishing psychiatric illness from general medical conditions. Even when psychiatric diagnosis seems obvious, a formal MSE provides assurance that other serious conditions do not go unnoticed. For example, a schizophrenic patient brought to an emergency department for bizarre behavior should not be lethargic and disoriented; ingestion or head trauma is a more likely cause of such symptoms in this setting.

The cognitive portion of the MSE is the critical component in an emergency setting. Unfortunately, cognitive testing often receives short shrift: physicians write "A&O x3" (alert and oriented in three spheres) when they have ascertained only that their patient is alert and responds to very simple questions. Consultants can add much by clearly documenting a patient's level of alertness, and then asking, and documenting explicitly, a patient's response to questions about their name, their location, and today's date. Along with these responses, consultants should document whether or not a patient remembers three objects after a few minutes, and the patient's ability to spell a five-letter word backward. These six components are, arguably, the irreducible minimum of an MSE.

Risk for violence is another major concern in emergency department consultations: is this patient a danger to himself or others? Some would argue that this is the most critical aspect of an MSE. However, it is rare for a patient's violent inclinations to become evident solely during an MSE. Comments about suicide or homicide are usually included in the chief complaint or reason for consult. Common practice requires some documentation of a patient's suicidal or homicidal thoughts with an MSE; however, if these are serious considerations, they warrant explication within the HPI.

Deafness may not normally be considered a topic for the MSE, but it occasionally confounds psychiatric consultation. Fast-paced, noisy emergency environments increase the odds that a patient who is hard of hearing will be mistaken for a patient who is disorganized or bizarre. Psychiatrists should have a high index of suspicion when consulted about older patients who are reasonably groomed, without history of psychiatric treatment, and who are alert but give nonsensical answers. Such patients can seem quite fine, and begin their interactions normally enough. They recognize the attempt to converse; they just cannot hear the words. Once this problem is recognized, there are often obvious solutions (quiet examination room, amplifier, writing, etc.).

A complete MSE has a number of other components; psychiatrists rarely fail to comment on mood, affect, hallucinations, and such. These are all useful, and certainly should be recorded. However, it is the cognitive portion and inclinations to violence that merit special attention during consultations in an emergency department. Physical Examination

Under certain circumstances, the physical examination becomes important to psychiatric evaluation in an emergency department. If a psychiatric consultant is expected to be responsible for the physical examination and medical care of patients in an emergency department, that needs to be very clearly understood, along with guidelines for transferring complicated medical care back to an emergency medicine physician.

In addition to a critical role in detecting major medical illness, physical examination may yield objective evidence about a patient's mental state. Dirt, untreated sores, ragged fingernails, and odor are physical findings that augment traditional psychiatric evidence that a patient is confused, distracted, or disorganized. Old neck and wrist scars can support a history of prior suicide attempts. Fresh, deep wounds suggest a greater risk than superficial scratches.

Physical findings may also be of legal importance. Failure to document physical injuries from a suicide attempt may cast doubt on the overall psychiatric evaluation, for example, when a patient challenges an involuntary commitment or a judge considers appointing a conservator.

26.2.4 Diagnostic Considerations

Diagnostic considerations are quite similar in both emergency departments and general hospital wards. Both settings can host a full spectrum of psychiatric problems, ranging from chronic mental illness to side effects of research protocols. However, some problems are more common in one setting than the other. Acute schizophrenic exacerbations are more common in emergency departments. Delirium is more common in medical and surgical units. Still, the diagnostic efforts remain more alike than different.

Separating traditional mental illness from mental illness due to a general medical condition is a diagnostic challenge that warrants extra attention. From an emergency department, patients are often transferred to very different facilities based on diagnosis. Transfer may be to a psychiatric hospital with little capacity to treat conventional medical illness. Or admission may be to a general hospital with no psychiatric wards. Sending patients to the wrong institution can cause significant morbidity and cost.

This challenge to differentiate psychiatric from medical illness is often posed to consultants as a short clinical vignette: "A man was brought by his family because he'd been acting oddly; they became worried and brought him to our emergence department because they did not know what else to do." A subtle tone may hint that the consultee is really quite uncertain what to do. Guidelines follow to help direct diagnostic attention. They are based on my experience in urban settings; thus a priori diagnostic probabilities could be different in other locales. These guidelines assume basic triage findings have not been contributory, such as vital signs, finger-stick glucose, alcohol breathalyzer, inspection for head trauma. These guidelines serve as a starting point for further investigation.

Two critical pieces of data must be available:


Onset/time course

Past history, medical and psychiatric, is the next most important information. Friends and family are often the best source of this information when there is suspected mental illness.

So we must elaborate on our vignette: A 66-year-old retired businessman was brought by his wife because Saturday morning he suddenly started talking to someone who was not there. He has never seen a psychiatrist. His internist is following his high blood pressure.

An alternate elaboration might go as follows: A 22-year-old college student was brought by his parents because they noticed that Saturday morning he was talking to someone who was not there. He had been out very late Friday night with friends. He has never seen a psychiatrist and has generally been in good health.

Neither of these patients is likely to be suffering from new-onset schizophrenia, although, theoretically, both could be. More likely, the older patient has had a stroke; the younger patient has ingested some illicit substance. Careful neurologic exam, and perhaps neuroimaging, are part of the initial workup for the older patient. Urine toxicology screening is an immediate part of the workup for the younger patient. These sorts of considerations above are expanded in the guidelines below. Convenient Categories Onset: Age: Likely Etiologies:

Acute onset (unlikely to be a traditional psychiatric illness)

Children: infection, unrecognized ingestion, trauma

Teens: ingestion (intentional or not), infection, trauma

Young adults: same as for teens, but ingestions are almost always substance abuse; be alert for meningitis/encephalitis depending on living situation

Adults: same as for young adults, but add central nervous system (CNS) vascular events, and ingestions include prescription drug side effects, other iatrogenic effects, including hypoglycemia, and prescription drug misuse

Elderly: infection (urinary tract) and CNS vascular events most likely, otherwise same as for adults

Subacute (anything is possible)

Children: endocrine, metabolic, infection, seizures, subdural (trauma), tumor

Teens: drugs, otherwise same as younger children

Young adults: differential is very broad, anything from lupus to schizophrenia may first manifest itself

Adults:most psychiatric illness would have already declared itself, but infections and inflammatory disease loom larger over this time span, and tumor becomes a possibility

Elderly: same as for adults, but effects of drug changes, drug buildup, congestive heart failure all add to the picture

Gradual (psychiatric illness becomes more likely)

Children: family conflicts, developmental abnormalities, environmental, other

Teens: drugs, family conflicts, pregnancy, psychiatric

Young adults: psychiatric, drugs, autoimmune

Adults: psychiatric, drugs, HIV, tumor

Elderly: dementia, cerebrovascular accident (CVA), B1, deficiency, normal pressure hydrocephalus

Insidious (psychiatric illness remains quite possible)

Children: family conflicts, developmental abnormalities, environmental, other

Teens: drugs, family conflicts, psychiatric, pregnancy

Young adults: psychiatric, drugs, autoimmune

Adults: HIV, tumor, psychiatric, drugs,

Elderly: dementia, CVA, B12 deficiency, normal pressure hydrocephalus

In general, acute changes in behavior or mental status suggest medical illness or ingestion. Gradual or insidious onset reduces the odds of acute medical illness (infection, infarction), but does rule out inflammatory process (lupus), endocrine disease (thyroid), or tumor. Workup for slower onset changes is likely to include brain imaging, thyroid testing, HIV testing, and the center's preferred tests to rule out autoimmune disease; however, this may not really be appropriate in an emergency department itself.

Based on the above lists of rough diagnostic probability, the following tests can be considered (keeping in mind emergency department time limitations).

GLU: Glucose, since not all diabetics give a clear history, yet may be taking hypoglycemic agents

U/A: Urinalysis for the elderly, since delirium or cognitive impairment can be seen with otherwise asymptomatic urinary tract infection; a chest x-ray for relatively asymptomatic pneumonia is an option

WBC: White blood cell count, as a second test for otherwise asymptomatic infection in the elderly and in young patients

EtOH & U-Tox: Alcohol breath testing and urine toxicology screening for ages 13 to 64 (can be selective); substance abuse is often omitted (or denied) when patients give a history

LFT: Liver function tests, for clues about covert alcohol use, poisoning, and drug side effects

BUN/Cre: Blood urea nitrogen and creatinine may reveal early kidney disease

ESR: Erythrocyte sedimentation rate along with C-reactive protein can be helpful if negative; can essentially rule out infectious or inflammatory illness

CT/MRI: Brain imaging with computed tomography or magnetic resonance imaging may prove helpful in patients of ages 26 to 64; old strokes and atrophy are often seen in older patients, but not helpful in making decisions about a particular emergency department visit

LP: Lumbar puncture should be considered if there is any question of central nervous system infection

A number of commonly ordered tests are rarely helpful. Electrolytes are almost never a cause of behavioral disturbance unless there is a history of eating disorder, drinking, or polydipsia. Normal BUN and Cre and the absence of any such history should be sufficient. Likewise for serum calcium, magnesium, and phosphorus. Thyroid testing is very rarely helpful, unless a patient has a history of thyroid dysfunction, and results are rarely available in a reasonable time frame. (In my experience, uremia is a better mimic of major depression than is thyroid disease, and internists usually diagnose thyroid disease long before patients come to psychiatric attention.) Venereal Disease Research Laboratory (VDRL) or fluorescein treponema antibody (FTA) testing for syphilis, along with B12 and folate testing for nutritional deficits, should be considered in puzzling cases. However, these are also unlikely to be available in a timely fashion.

26.3 Treatment

Distractions and time pressure can make emergency department treatment difficult: it is a noisy, busy place. Conventional wisdom holds that the only real options are "treat or street," that is, admit for inpatient treatment or discharge to the street. In truth, simple-minded approaches are a bigger distraction than the noise and activity. Accurate evaluation facilitates efficient treatment, whether or not a patient requires admission. Less-than-thoughtful evaluation risks complications and morbidity.

26.3.1 So-Called Agitated Patients

So-called agitated patients are repeated tests of each consultant's ability to make careful psychiatric evaluations in an emergency department. Faced with an agitated-combative-wild patient, many clinicians reflexively order a mixture of tranquilizers. "Five-two-and-one" is a favorite combination: haloperidol 5 mg, lorazepam 2 mg, and benztropine 1 mg. Few clinicians even wait for registration to confirm patient identity and computerized records to report known allergies; luckily, true allergies to haloperidol or lorazepam or benztropine are rare. Fewer clinicians yet, even in quiet moments, seriously consider the need for benztropine when haloperidol is given with lorazepam; it is no surprise that benzodiazepines are a second- or third-line treatment for parkinsonian side effects. I have never seen dystonia after one injection of haloperidol with lorazepam. Medication is not the first line of defense against violent or dangerous patients. Table 26.1 lists some times to consider (Drugdex, 2006; Eli Lilly, 2006; Pfizer, 2005).

Only anesthetic agents begin to work quickly enough to stop a raging patient. Realistic wild-animal shows (e.g., National Geographic) show chemical dart guns being used from a distance, preferably from a truck. Raging rhinos can cover a lot of ground in the minutes required for modern opiates to take effect. (Difficulties ventilating rhinos in the wild make succinylcholine an unattractive option.)

Physically unmanageable patients may require physical restraints (Rund and Hutzler, 2004). Trained staff, careful planning, and thoughtful facility design can all reduce the number of patients requiring restraints. However, unless local authorities and ambulance crews all agree to send combative patients elsewhere, restraints will be necessary in some cases.

Once a patient is physically safe, it is time to carefully review what is known and what can be determined by exam, perhaps even by interview. (A small number of patients do settle down once restrained.) Thoughtful clinicians consider a number of possibilities: Is this patient intoxicated? Is there evidence of head trauma? Is this patient already taking a sedative? Is this patient known to respond to some specific treatment? Noncompliance/nonadherence is a common cause of relapse; if patients will accept an oral dose of their routine medication, recovery will be underway.

Keep in mind that patients, their families, and our colleagues are all human; in a crisis they may fail to report critical information. One very large, very paranoid, and then very combative patient required the efforts of eight staff to subdue him. Only afterward did his mother reveal that he had jumped out a second-story window that was the real reason she had finally brought him to the emergency department. Initially, she had only mentioned he was acting "differently" for a few days. Given his obvious, initial anxiety, the triage staff slotted him directly for psychiatric evaluation without any check for trauma. You can imagine the staff's anxiety on discovering that the patient they had just wrestled into restraints was at risk for a broken neck. Luckily, there were no fractures and the patient's phencyclidine eventually lost its hold on his thinking.

At the time of this writing, there is no diagnosis of "agitation disorder" or "agitation disorder not otherwise specified" in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV), and the American Psychiatric Association has yet to garner Food and Drug Administration (FDA) approval for drug treatment of signs or symptoms separate from a specific DSM diagnosis. This could all change. Until then, there is no official approval for the concept of drug treatment of an "agitated" patient. We can recognize and keep clear in our minds that we do treat some patients before we are certain of their diagnosis. Careful consideration of a given patient's diagnostic possibilities allows careful choice of treatment, even if it is a combination of drugs (Alexander et al., 2004; Allen et al., 2005; Andrezina et al., 2006; Battaglia et al., 1997; Breier et al., 2002; Breitbart et al., 1996; Broderick et al., 2002; Brook et al., 2000; Eli Lilly, 2006; Food and Drug Administration 2001, 2006; Martel et al., 2005; Pfizer, 2005; Preval et al., 2005; Scahill et al., 2005; Tesar, 1996; TREC Collaborative, 2003).

26.3.2 Alcohol Withdrawal and Sedative/Hypnotic Withdrawal

Most alcohol withdrawal is either directly reported by patients themselves or strongly suggested by histories of alcohol abuse. There are no special considerations in an emergency department; it is best to treat before overt delirium tremens or seizures are manifest.

It is a little more common in an emergency department for patients to claim, "I'm just always anxious." They may hope to keep their addiction a secret. They may hope for a benzodiazepine prescription and then to be on their way. One trainee was thus misled by an entirely pleasant, middle-aged woman who promptly seized when the attending arrived to examine her.

Sedative/hypnotic withdrawal is essentially identical to alcohol withdrawal. Unfortunately, finding a suitable dose of replacement can be a challenge; patients frequently minimize or exaggerate their daily use. There may be some advantage in sticking with whichever agent the patient normally takes.

It is important to differentiate among the sedating agents. Withdrawal from agents that affect y-aminobutyric acid (GABA) receptor complexes leads to symptoms of alcohol withdrawal (benzodiazepines, barbiturates, alcohols). Agents that work elsewhere (antihistamines, antipsychotics) do not treat alcohol withdrawal; indeed, antihistamines can aggravate matters by lowering seizure thresholds. Agents not traditionally considered sedatives (opiates, antidepressants) can independently cause sedation in emergency department patients, further complicating evaluation and treatment.

In my experience, the most common errors in the treatment of alcohol withdrawal are failure to diagnose until quite advanced, and failure to give adequate doses of benzodiazepines. It is easy to dismiss anxiety and mildly elevated vital signs in a middle-aged accident victim, but remain alert for symptom progression and a history of alcohol use. Though standard doses of benzodiazepine (e.g., diazepam 5 to 10 mg) are usually effective, some patients need a lot more (Mayo-Smith et al., 2004).

26.3.3 Anxiety

Patients in an emergency department may be anxious for a wide variety of reasons. Making treatment even more difficult, their physical problems may not be fully known at the time a psychiatric evaluation is requested. A patient going into shock might report "anxiety," a "nervous, queasy feeling," especially if he is already in psychiatric treatment. This requires the consultants to an emergency department to carefully review vital signs, physical findings, and test results, some of which may not yet be integrated into a complete diagnosis.

When anxiety is a manifestation of physical illness or preexisting psychiatric illness, the first course of action should be to treat the underlying problem. Then, check for improvement or worsening. Reflex administration of a benzodiazepine may cloud the picture.

Fear is more common than one might guess, at least based on emergency department conversations. People often prefer Freud's use of anxiety, a psychological signal of inner conflict, to fear, a central nervous system signal of potential mortality (or morbidity). Nevertheless, patients in an emergency department may have very real reason to be afraid, and it may fall to a consultant to point this out.

Fear may respond to facts, family presence, and general reassurance. If these and other simple maneuvers fail, be alert to pain as a critical underlying factor. Pain and fear are supra-additive. Likewise, a little attention to analgesia may yield large improvements. Lastly, fear responds to benzodiazepines, but at a cost in alertness, cognition, and memory.

Anxiety disorders themselves, particularly panic disorder, may present and be first diagnosed in an emergency department. A panic attack is likely to respond to a benzodiazepine, which is a reasonable, immediate intervention. That then leaves the rest of the medical workup (e.g., thyroid tests, which are not immediately available), and psychiatric follow-up (i.e., overall condition, side effects). Unlike medical ward consultation, there may be no option to see a patient again the next day. Practical limitations in follow-up care are an important constraint on emergency department treatment recommendations.

Recommending an antidepressant to treat an anxiety disorder may not be simple in an emergency department. The FDA (2005) has issued warnings on suicidality in patients treated with antidepressants. These warnings might make it seem negligent to prescribe an antidepressant without first establishing follow-up care.

A short, trial course of benzodiazepines is a common intervention, but one that can lead to problems in an emergency department. If quick psychiatric follow-up is not available, it is no longer a clinical "trial"; there is no trained professional to evaluate results. A more pernicious problem is well known to emergency medicine clinicians developing a reputation as a facility that dispenses benzodiazepines. If local addicts discover that panic is treated with a week's worth of Xanax, then there will be a lot of panic attacks to be treated. This does not mean that prescriptions for benzodiazepines should never be dispensed, just that more care is required than in a controlled environment like a medical ward.

Addicts raise other anxiety-related treatments issue in an emergency department. Crack/cocaine and stimulant users may arrive very anxious, due to intoxication. They can be overtly paranoid. A benzodiazepine will usually help. However, be aware that some stimulant abusers do not actually have any sedative tolerance, and may become very sedated. Low-dose antipsychotics may also be helpful. Avoid giving a-adrenergic antagonists in the face of stimulant intoxication as they can increase demands on cardiac output.

26.3.4 Catatonia

The underlying mechanisms of catatonia remain unknown. For emergency department purposes it is reasonable to assume it represents overwhelming anxiety or fear, causing a patient to freeze like a deer in the headlights. This matches the clinical impression that a catatonic patient is awake and alert, not comatose or lethargic. It also leads to use of a benzodiazepine as an immediate intervention. Lorazepam 1 mg IM or IV is usually effective within an hour. Other benzodiazepines should work just as well. Oral doses can be effective but take longer (2 hours or longer).

Keep in mind that catatonia is a sign of some other process, likely an affective disorder with psychotic features. Treatment of the underlying process is necessary to prevent recurrence. Repeated attempts to temporize with a benzodiazepine are likely to fail.

26.3.5 Conversion Disorder

Conversion disorder is an irritation to emergency departments. Other patients who can be shown to be free from physical ailments are not a problem; emergency department staff members are quite happy to rule out myocardial infarction (so long their patients are reasonably cooperative). The underlying problem with conversion disorder is that these patients are not relieved, rarely grateful, and often cannot leave because their symptom is paralysis. A consultant's real challenge may be to get the patient out.

Conversion disorders were apparently quite common among Charcot's patients; they helped spur Freud's development of psychoanalysis. Unfortunately, in urban emergency departments, patients with conversion disorder rarely show much insight or response to interpretation. Some are quite willing to talk, but usually make no connection among affects, anxiety, and physical (dys)function. Many reveal no clear conflict or recent stressor.

Emergency department treatment often devolves to very general interventions: reassurance that there's no evidence of serious medical illness, suggestions that their physical symptoms are likely to remit on their own, encouragement to continue regular activities as much as possible. Patients who show any interest in counseling or any acceptance of the idea that stress might be a significant issue can be referred to a mental health professional. Be alert that some patients may react quite negatively to any implication that their symptoms are all in their head.

When reassurance and referral fail, consultants can recommend a benzodiazepine, for example, lorazepam 0.5 to 1.0 mg orally (or parenterally). Before the discovery of benzodiazepines, earlier generations of psychiatrists would use a barbiturate. It is hard to know whether sedation, anxiety reduction, or cognitive dulling is key. Placebo injections are rarely helpful. In any case, after 1 to 2 hours, there may be sufficient improvement to allow discharge.

Refractory cases may have to be admitted to a psychiatric facility. Few medical or surgical services will accept a patient whose diagnosis is conversion disorder. Inpatient psychiatric services are a bit more tolerant, and refractory cases may prove to have an underlying psychotic illness.

Pseudo-seizures may not be good cases to treat with a benzodiazepine. An apparent response to benzodiazepine administration may confuse both staff and patients. Luckily, these patients usually stop seizing, and then accept output follow-up. Those who do not may have to be admitted to a psychiatric service.

26.3.6 Intoxication

Intoxication per se has not traditionally been considered a psychiatric problem. However, consultees may request psychiatric help for particularly bizarre intoxicated patients. And in some facilities psychiatry does take primary responsibility for substance abuse disorders. So some guidelines may prove useful.

Opiate intoxication and benzodiazepine intoxication are the only two types that can be reversed. Naloxone (Narcan) is an injectable opiate antagonist. Flumazenil (Romazicon) is an injectable benzodiazepine antagonist. Both entail significant risks. They are effective, though their half-lives are short compared to most drugs of abuse. Flumazenil can induce seizures in patients who are dependent on benzodiazepines (a significant population in an emergency department). Naloxone treatment can lead to immediate withdrawal, including severe agitation, vomiting, diarrhea, and cramps.

Cocaine and stimulant intoxication cannot be reversed. It may be moderated with a sedative or an antipsychotic. Be wary of medications causing peripheral vasodilation and thus increased cardiac demand (e.g., (x-adrenergic blockers).

Hallucinogens cannot be reversed. As with stimulants, a sedative or antipsychotic may be a useful temporizing agent.

Alcohol intoxication cannot be reversed. Anecdotal remedies for reducing alcohol levels upon exiting a pub have yielded no reliable approach; likewise for efforts to reduce after-effects the next morning. Medical professionals have not faired any better. However, medical and law enforcement professionals have discovered some approaches that increase morbidity and mortality. Benzodiazepines can augment alcohol's suppression of respiratory drive, even though they themselves have very little effect. Droperidol can lead to sudden death (FDA, 2001). Rapid osmolar changes from IV fluids can lead to central pontine myelinosis. These therapeutic pitfalls suggest the primary goal should be simply to keep staff and patient safe until the patient is sober.

Haloperidol can be tried to render alcoholics less agitated until they have metabolized their alcohol. Be aware that haloperidol is a butyrophenone, like droperidol, and has been reported to cause arrhythmia when given IV (torsades de pointes). Alternatively, a high potency phenothiazine (e.g., fluphenazine) would be just as effective or ineffective. Low-potency phenothiazines, and certainly antihistamine sedatives, risk lowering seizure thresholds.

26.3.7 Major Depression

It is quite unfortunate that antidepressant treatment entails risk of suicide during the 3 or 4 weeks required for full effect. Major depression, episodes of depression in bipolar disorder, and postpartum depression may present first to an emergency department. Once the diagnosis is made, it is very tempting to prescribe an antidepressant and send the patient home, just as our medical colleagues do with antibiotics for infections. However, the FDA (2005) clearly publicized the risk of suicidality, if not suicide itself.

Arranging reliable follow-up is now the biggest challenge when starting patients on antidepressants. Follow-up must be available, and there must be good reason to believe the patient will actually go for follow-up. Odds are less than 50% that patients will keep a psychiatric clinic appointment that is simply scheduled for them (some clinics and practitioners do not accept appointments unless patients call on their own). This suggests that a family member or friend should be involved to ensure appointments are kept, and perhaps to bring the patient back if symptoms worsen.

All of the above can seem quite frustrating to consultants. However, the current situation is much improved from the days when only tricyclic and monamine oxidase inhibitor antidepressants were available. Back then, patients could return to an emergency department dead, or near death, accompanied by an empty pill bottle bearing the prescribing psychiatrist's name.

If reliable follow-up treatment can be arranged, the choice of an antidepressant should be given to the follow-up clinician, who must deal with the complications. Most clinicians start with a selective serotonin uptake inhibitor.

26.3.8 Mania

Mania's treatment may be considered in three phases: immediate, episodic, and long-term prophylaxis. Floridly manic patients usually trigger a request for psychiatric intervention. Their diagnostic possibilities are similar to any inpatient consultation request, although there may be more pure mania in an emergency department population and more stimulant intoxication. Immediate treatment of uncomplicated mania can begin with almost any sedative; benzodiazepines continue to be a favorite.

Manic patients may require a lot of sedation immediately, and some will require an antipsychotic. In an otherwise healthy manic, any sedating antipsychotic would likely work by itself. Chlorpromazine (Thorazine) has a long track record. Olanzapine (Zyprexa) is a new favorite, along with Ziprasidone (Geodon). All three are available in parenteral forms; all three have their drawbacks. Chlorpromazine injections can be locally irritating and require a relatively large injection volume (only available in 25 mg/mL solutions). Olanzapine must be reconstituted before injection and has a long half-life. Ziprasidone carries a precaution about QTc prolongation, which may be difficult to check in a floridly manic patient; that is, avoid use in the presence of other drugs known to prolong QTc, in the presence of arrhythmia, or in the presence of electrolyte abnormalities. All of this contributes to the continued popularity for combination therapy with haloperidol plus lorazepam (Allen et al., 2005).

While most patients respond to any of the above approaches, it is safest to use a medication that has helped the patient in the past. Using a previously successful medication reduces the odds of a new, adverse drug reaction, and makes a start toward treatment of the episode itself.

Clearing an episode of mania is not easy. It usually takes days or weeks. Sedatives alone are unlikely to succeed. An antipsychotic or a drug such as lithium, also effective for prophylaxis, is usually needed. Unfortunately, none of the proven prophylactic agents are likely to be effective quickly enough for emergency department purposes; manic patients must usually be admitted to a psychiatric service.

Unfortunately, restarting patients on lithium or valproic acid (Depakote and others) requires attention to potentially severe side effects. Lithium can destroy renal function: check (baseline) blood urea nitrogen (BUN) and creatine (Cre). Valproic acid has direct liver toxicity and affects hematopoiesis; do (baseline) liver function tests (LFTs) and complete blood count (CBC). Serum levels of both should be monitored. It is usually safe to give one or maybe two doses without test results. That allows restarting treatment but delaying blood drawing until after the patient is a bit more cooperative.

26.3.9 Psychosis

Treatment of psychotic patients in an emergency department is essentially the same as their treatment on a medical or surgical service. (Here we mean psychotic in the restricted sense of a patient suffering from a psychotic spectrum disorder, e.g., schizophrenia.) Treatment is simplified by the fact that most psychotic patients presenting to an emergency department are physically healthy, and many are suffering recurrence of an established mental illness, with an established treatment regime. They will respond to a resumption of their usual medications.

Two caveats are in order: some psychotic patients have serious medical illness, and some do not actually need treatment at all (at least not any new or additional treatment).

It seems emergency department staff members are inclined to assume that any apparently psychotic person who appears healthy is physically healthy. Statistics would support their contention: most such patients do not need any emergent medical attention. However, an occasional psychotic patient will prove to have diabetic ketoacidosis or even internal trauma (e.g., splenic tear). These patients may omit critical pieces of medical history, such as a diagnosis of diabetes or a recent auto accident. When consulting to an emergency department, keep in mind that the patient has not been under continuous medical scrutiny as would be implicit on a hospital ward.

Occasionally, some psychotic patients do arrive not needing any treatment. This usually happens when a patient stumbles into a new neighborhood, or a new clinician's office, or is visited by a new, temporary nurse. Such patients are at their chronically disturbed baseline. The challenge to consultants is to curb their therapeutic enthusiasm until they have contacted someone who really knows the patient.

26.3.10 Violence and Suicide

Violence is not treated per se (except in the movie A Clockwork Orange). However, consultants are frequently asked to evaluate patients for their potential to do violence: their risk to hurt themselves, or their risk to hurt others. If there is a significant, imminent risk, dangerous patients are usually admitted to a psychiatric unit.

This chapter does not explore the myriad aspects of suicide and homicide risk assessment. I favor looking at each patient's story; however, there has been much written on this subject, and this chapter aims only to review differences and similarities between hospital ward and emergency department consultation (American Psychiatric Association, 2003; Paris, 2006).

Emergency department consultations around potential violence are complicated by time pressures; patients have been observed for only a short period of time in an emergency department, and emergency staff would like patients discharged in a short period of time. Thus, a patient's angry, off-the-cuff, statement that "I wish I were dead" comes to be seen as the chief complaint, rather than an expression of irritation, uncomfortable examinations, and delayed treatment. A consultant's task is simplified by finding some collateral source of information about the patient's state of mind.

Many patients settle down if the consultant addresses the current problem in a matter-of-fact manner. For example: "Ever since Columbine, people get very worried whenever someone makes any kind of comment that smacks of suicide. Are you really out to kill yourself?" Assuming the answer is no, the consultant can follow-up: "I'm inclined to believe you. Is there anyone I can call to confirm your story?" Most patients do not want to be held in a psychiatric facility, so they will provide the number of a friend or family member to contact. Occasionally, there is a personal attorney available to provide information.

A similar approach to patients who make threats against others is appropriate to an emergency department setting. However, it is not uncommon to hear nonpsychiatric threats of violence from emergency department patients. It is not traditionally considered a psychiatric problem if a patient talks of killing his mother because she has called the police or changed the locks on door because he pawned her television set to buy drugs. However, it is assumed to be a psychiatric problem when a patient talks of killing his therapist because she is colluding with his mother to control his thoughts through TV broadcasts. Many psychiatric services will not admit the first patient, whose diagnoses are cocaine dependence and antisocial personality disorder. They will also balk at admitting someone for violent threats in the course of a domestic dispute or divorce.

Consultants should contact their lawyer or their facility's lawyer for directions in handling a nonpsychiatric risk of violence. These and other violence-related issues are better reviewed ahead of time, rather than at 2 a.m. in a busy emergency department. What are the local standards for risk or imminent risk? What are the local standards about confidentiality when the risk of life is involved? What are the local options for psychiatric and nonpsychiatric control of a patient?

26.4 Confidentiality

Traditional emphasis on protecting patient confidences must often be tempered in an emergency department. Consider a person who was dancing nude in the park at midday. Local police send him to your emergency department, and since it was a public act, there is no confidence to keep. Then there are acts that must be reported in many locales, for example, suspected child abuse or gunshot wounds. And certain threats are no longer to be kept secret, because of legal policy stemming from the Tarasoff case. Herbert and Young 2002 reference provides a very good review of the case. These limits on confidentiality may arise during consultation to general hospital wards, but arise more frequently in an emergency setting.

26.4.1 Realities

Public acts cannot be kept confidential. If the patient's parents called for an ambulance because the patient had cut his wrists, the consultant can call the family for further details about the patient's psychiatric problem without violating confidentiality because the family knows about it. On the other hand, a consultant would have little reason to make this event more public, say, by calling this patient's employer.

The spirit of keeping patient information confidential can be upheld. Suppose our nude dancer had simply overindulged. It would be appropriate to ask him whom to call for assistance and transportation. It is not necessary to immediately call his family. Not all local newspapers publish a complete police blotter; detailing every citizen's petty encounters with authorities. The patient may have a friend who is more understanding and discreet than his spouse or parents. Some events can be kept less public than others.

Private, voluntary patient contacts should be kept confidential. If patients come to the emergency department on their own for a problem that is not overtly dangerous (such as worsening anxiety) or for an understandable reason (for example, it is August and their psychiatrist is on vacation), and if they appear able to fend for themselves, then there is every reason to keep their visit confidential. Patients do not forfeit confidentiality just because they come to an emergency department rather than a private office or clinic.

It is the nature of a patient's acts and risks that determine the reality of confidential treatment. Information about private acts entailing no risk should be kept in confidence. This is true even if the patient has been openly sent to your emergency department for some other reason. However, if there is reason to believe a patient is at risk for harming himself or others, then the value of maintaining confidences has to be weighed against the apparent risks. The patient may be able to help minimize the loss of privacy by choosing which person is to be contacted, as discussed in the earlier subsection on violence.

26.4.2 Tarasoff Considerations

Most psychiatrists believe they have an obligation to warn their patient's potential victims. Tarasoff a Reagents of the University of California is the case that most cite, though any such obligation stems from a number of related decisions. And, in truth, it has not been easy for the courts or society to balance the need for privileged communication against the need for personal safety. Practitioners should recognize there are many legal fine points and local variations in this matter (Herbert and Young, 2002).

There is often a simpler approach to so-called Tarasoff cases, namely, to ask is someone likely to be hurt. If the answer is yes, then ask what can we do to prevent it. In other words, focus attention on preventing harm rather than on legal complexities concerning specific warnings. Admitting patients on the grounds that they are dangerous to others is one way to reduce risk. This approach may also simplify discussions with your own legal counsel. Are there grounds to force admission? If not, how compelling is the evidence that a warning is needed? (see earlier subsection on violence).

26.4.3 Child Abuse, Gunshot Wounds, and Local Rules

There is a long tradition of state statutes requiring doctors to report certain medical problems, such as gunshot wounds, active tuberculosis, and other infectious disease. In such matters, public safety is held to be more important than individual privacy. Unfortunately, there is much variation state to state. Physicians can expect to have to report suspected child abuse in all 50 states. Elder abuse reporting requirements are increasingly common. Whatever the locale, its emergency departments are a uniformly likely place for such problems to become evident. Consultants should be aware of their local requirements and their institution's mechanism for handling mandatory reports.

26.4.4 Health Insurance Portability and Accountability Act of 1996 (HIPAA)

The HIPAA privacy rules apply equally in emergency departments as on hospital wards. They have caused many procedural changes in an effort to ensure patient privacy. However, they were not meant to impede clinical care, to usurp state laws, or to shield physicians from mandatory reporting (as noted above). The HIPAA rules have little impact on the confidentiality issues that loom largest for clinicians, for example, who to warn, when to contact family or friends for critical clinical information.

26.5 Special Problems in Emergency Departments

There are a few clinical problems that arise frequently in emergency departments, largely because emergency departments serve as a screening area for their hospitals and sometimes other local services. These issues must be resolved before a patient goes to a general ward, and thus are less commonly a subject of general hospital consultations. A request to evaluate a patient, for psychiatric admission, is the most obvious such consultation. Another frequent problem is to evaluate a suicidal patient brought from police lock-up, a situation in which secondary gain raises serious questions about veracity. Other problems may arise depending on local circumstances.

26.5.1 Admission Screening

Screening admissions is a multifaceted task: to ensure patients are directed toward appropriate treatment, to ensure there are beds available, and to ensure implicit and explicit criteria admission are met. Some patients have an obvious, easily verified need for inpatient psychiatric treatment. Others may need admission to a medical ward or a med-psych ward first; comorbidities like diabetes or emphysema may require treatment unavailable on some psychiatric wards. Psychiatric consultants to an emergency department are usually expected to make these clinical decisions. Checking bed availability would seem to be a clerical task, but it can devolve directly to consultants. Likewise, consultants may have to complete admission checklists (diagnosis, illness severity, risk of violence, medical needs, insurance status). Unfortunately, institutional politics may lead to a complicated collection of implicit admission criteria, requiring consultants to become negotiators or facilitators.

It is tempting to view admission screening simply as a nuisance, but this is to ignore fundamental organizational needs. Emergency departments buffer hospital inpatient units against the hour-by-hour variation in patient arrivals and case-by-case variation in clinical needs. If an emergency department does not serve this function, then the hospital has to commit other staff and resources to provide it. Unfortunately, from the patient's point of view, an emergency department can become a barrier to psychiatric inpatient treatment. Consultation requests for evaluation under these circumstances may be better understood as requests to help patients over this barrier.

26.5.2 Patients Brought from Jail

Patients brought from jail to an emergency department are problematic. There is the issue of secondary gain (a temporary reprieve from their jail cell). There may be a limited commitment to telling the truth. There is the sad fact that a number of our chronically mentally ill land in jail for lack of more suitable treatment settings. Consultants are nevertheless expected to evaluate and recommend treatment.

Examining a patient may or may not provide useful information under these very constrained circumstances. Some patients are grossly unkempt or obviously psychotic, and even the prison authorities do not want them back in jail. A small number lack guile, or experience, and let it slip that they are fine, their real complaint is the wait over a long weekend until a judge will be available to set bond. That leaves a large fraction whose moods fall somewhere between unhappy and miserable, who may be disturbed, or who may simply impulsive enough to harm themselves. For this large fraction, an interview leads to no direct conclusion.

One helpful tack is to obtain as much information as possible from authorities, lawyers, family, and old records. Local authorities should know what sentences have been passed, what charges are pending, and when court proceedings are scheduled. This provides a framework in which to consider a patient's behavior. Police may also have specifics about a patient's behavior when apprehended. A patient's lawyer, or prosecutor, can occasionally offer specific arrangements for a patient's care. Family may be able to describe a patient's behavior before his or her legal entanglement. Old medical/psychiatric records can provide any number of clues, including a history of similar behavior under similar circumstances. All of this information may be even more useful if it is available before interviewing the patient.

Depending on the circumstances of a case, it is worth asking a patient whether he or she wants to be admitted to a psychiatric ward, with an understanding that this would only delay, not reduce, jail time. If the answer is no, then there is an opening to discuss the patient's ability to keep from hurting him- or herself when returned to jail. If the answer is yes, then the implication is that this patient is unhappy enough to prefer more time in confinement. Unfortunately, the answer may be yes because some patients prefer any alternative to jail, and because some have discovered that any delay works to their advantage in the courts.

In some cases, all available information leads one to be concerned that a patient may try to hurt him- or herself on return to jail, but clinically does not lead one to believe that the patient requires admission to a hospital psychiatric service. It often helps to review such cases, keeping in mind the social matrix: Why has it expelled this patient? What rearrangements would allow it to accept this patient back? Perhaps extra precautions can be put in place in jail. Perhaps there is some way to expedite court action. Perhaps an inpatient psychiatric admission is not such an objectionable alternative. In very rare circumstances, perhaps the patient should remain in the emergency department under guard until a judge is available. Social engineering may yield a better solution than traditional psychiatric approaches.

26.5.3 Untruths

Trainees often ask how do you know what the patient says is true. The short answer is you don't. The longer answer is we don't need the truth, just a good story. Clinical practice yields unbelievable coincidental events that prove to have happened, and entirely plausible reports made of whole cloth. There is no easy way to be certain.

Questions about truth may arise more often in an emergency department for lack of an established doctor-patient relationship. Patients are probably no less reliable in this setting, just less well known. A known, compulsive liar is not a problem; his statements can be ignored in favor of objective signs and collateral reports. Patients from jail accentuate this issue: they are a mix of unsuccessful sociopaths who will say whatever is convenient, and otherwise upright citizens (only accused of misdemeanors) who would never mislead a physician.

Clinicians hell-bent on finding the truth should remind themselves that there is no reliable way to get the truth out of anyone. If there were a reliable technique, or drug, the Central Intelligence Agency and its less savory counterparts would use it. And any such approach would be used so frequently that details would inevitably become public. Public lore only maintains that anyone can be made to talk, but their statements may not reflect the actual state of our world.

The cultural context is critical for any discussion of truth and the value of truth-telling. Dominant social norms dictate that patients skip over most toilet activities when answering the question, What happened today? Such omissions are considered good manners, not deliberate lies. However, social norms around exaggeration and discreet social lies may be in flux because of many widely reported incidents of politicians and other public figures being caught in telling lies. In this cultural context, how much truth can we expect from our patients?


Alexander J, Tharyan P, Adams C, John T, Mol C, Philip J. Rapid tranquillisation of violent or agitated patients in a psychiatric emergency setting. Pragmatic randomised trial of intramuscular lorazepam v. haloperidol plus promethazine. Br J Psychiatry 2004;185:63-69.

Allen MH, Currier GW, Carpenter D, Ross R, Docherty JP Treatment of behavioral emergencies. J Psychiatr Pract 2005;11(suppl 1):5-108.

Andrezina R, Josiassen RC, Marcus RN, et al. Intramuscular aripiprazole for the treatment of acute agitation in patients with schizophrenia or schizoaffective disorder: a double-blind, placebo-controlled comparison with intramuscular haloperidol. Psychopharmacology (Berl) 2006;188:281-292.

American Academy of Emergency Medicine. http://www.aaem.org/emtala/, 2006.

American Psychiatric Association. Practice guideline for the assessment and treatment of patients with suicidal behaviors. Am J Psychiatry 2003;160(11 suppl):1-60.

Battaglia J, Moss S, Rush J, et al. Haloperidol, lorazepam, or both for psychotic agitation? A multicenter, prospective, double-blind, emergency department study. Am J Emerg Med 1997;15:335-340.

Breier A, Meehan K, Birkett M, et al. A double-blind, placebo-controlled dose-response comparison of intramuscular olanzapine and haloperidol in the treatment of acute agitation in schizophrenia. Arch Gen Psychiatry 2002;59:441-448.

Breitbart W, Marotta R, Platt MM, et al. A double-blind trial of haloperidol, chlorpromazine, and lorazepam in the treatment of delirium in hospitalized AIDS patients. Am J Psychiatry 1996:153:231-237.

Broderick KB, Lerner EB, McCourt JD, Fraser E, Salerno K. Emergency physician practices and requirements regarding the medical screening examination of psychiatric patients. Acad Emerg Med 2002:9:88-92.

Brook S, Lucey JV, Gunn KP, Ziprasidone IM Study Group. Intramuscular ziprasidone compared with intramuscular haloperidol in the treatment of acute psychosis. J Clin Psychiatry 2000:61:933-941.

Centers for Medicare and Medicaid Services. http://www.cms.hhs.gov/MLNEd WebGuide/25_EMDOC.asp, 2006.

Drugdex® System. http://www.thomsonhc.com. Greenwood Village, CO: Thomson Micromedex, 2006.

Eli Lilly http://www.fda.gov/medwatch/safety/2006/Aug_Pls/Zyprexa_Pl.pdf, 2006.

Food and Drug Administration. Rockville, MD: http://www.fda.gov/bbs/topics/ ANSWERS/2001/ANS01123.html, 2001.

Food and Drug Administration. Rockville, MD: http://www.fda.gov/cder/drug/ antidepressants/default.htm, 2005.

Food and Drug Administration. Rockville, MD: http://www.fda.gov/cder/drug/ InfoSheets/HCP/aripiprazoleHCP.htm, 2006.

Herbert PB, Young KA. Tarasoff at twenty-five. J Am Acad Psychiatry Law 2002: 30:275-281.

Lukens TW. Wolf SJ, Edlow JA, et al. Clinical policy: critical issues in the diagnosis and management of the adult psychiatric patient in the emergency department. Ann Emerg Med 2006:47:79-99.

Martel M, Sterzinger A, Miner J, Clinton J, Biros M. Management of acute undifferentiated agitation in the emergency department: a randomized double-blind trial of droperidol, ziprasidone, and midazolam. Acad Emerg Med 2005;12:1167-1172.

Mayo-Smith MF, Beecher LH, Fischer TL, et al. Working Group on the Management of Alcohol Withdrawal Delirium, Practice Guidelines Committee, American Society of Addiction Medicine. Management of alcohol withdrawal delirium. An evidence-based practice guideline. Arch Intern Med 2004:12(164):1405-1412.

Paris J. Predicting and Preventing Suicide: Do We Know Enough to Do Ether? Harv Rev Psychiatry 2006;14:233-240.

Pfizer Roerig. New York: http://www.fda.gov/ medwatch/safety/2005/aug_PI/ Geodon_ PLpdf, 2005.

Preval H, Klotz SG, Southard R, Francis A. Rapid-acting IM ziprasidone in a psychiatric emergency service: a naturalistic study. Gen Hosp Psychiatry 2005;27:140-144.

Rund DA, Hutzler JC. Behavioral disorders: emergency assessment. In: Tintinalli JE, Kelen GD, Stapczynski JS, eds. Emergency Medicine: A Comprehensive Study Guide, 6th ed. New York: McGraw-Hill, 2004.

Scahill L, Blair J, Leckman JF, Martin A. Sudden death in a patient with Tourette syndrome during a clinical trial of ziprasidone. J Psychopharmacol 2005:19:205-206.

Tesar GE. The emergency department. In: Rundell JR, Wise MG, eds. Textbook of Consultation-Liaison Psychiatry. Washington, DC: American Psychiatric Press, 1996:914-945.

TREC Collaborative Group. Rapid tranquillisation for agitated patients in emergency psychiatric rooms: a randomised trial of midazolam versus haloperidol plus promethazine. Br Med J 2003:327:708-713.

If you find an error or have any questions, please email us at admin@doctorlib.info. Thank you!