Expert Consensus Guidelines on the Treatment of Behavioral

Journal of Psychiatric Practice Vol. 11, Suppl. 1 November 2005 5

Introduction: Methods,

Commentary, and Summary

Objectives. Due to inherent dangers and barriers to research in emergency settings, few data are available to guide clinicians about how bestto manage behavioral emergencies. Key constructs such as agitation are poorly defined. This lack of empirical data led us to undertake asurvey of expert opinion, results of which were published in the 2001 Expert Consensus Guidelines on the Treatment of BehavioralEmergencies. Several second-generation (atypical) antipsychotics (SGAs) are now available in new formulations for treating behavioralemergencies (e.g., intramuscular [I.M.] olanzapine and ziprasidone; rapidly dissolving tablets of olanzapine and risperidone). Critical ques-tions face the field. The SGAs are significantly different from the FGAs and from each other and have not been studied in unselectedpatients as were the FGAs. Can the SGAs can be thought of as a class, do all antipsychotics have similar anti-agitation effects in differentconditions, and, if equally effective, what limits might their safety profiles impose? Should antipsychotics be used more specifically to treatpsychotic conditions, while benzodiazepines (BNZs) alone are used nonspecifically? Few data are available concerning combinations ofSGAs and BNZs, and findings concerning the traditional combination of haloperidol plus a BNZ may not be relevant to combinationswith SGAs. The culture is also evolving with more emphasis on patient involvement in treatment decisions. An international consensushas been developing that calming rather than sedation is the appropriate endpoint of behavioral emergency interventions. We undertooka new survey of expert opinion to update recommendations from the earlier survey.

Method. A written survey of 61 questions (1,020 options) was mailed to 50 experts in the field, 48 (96%) of whom completed it. The sur-vey sought to define level of agitation at which emergency interventions are appropriate, scope of assessment depending on urgency andpatients’ ability to cooperate, guiding principles for selecting interventions, and appropriate physical and medication strategies at differentlevels of diagnostic confidence for a variety of provisional diagnoses and complicating conditions. A modified version of the RANDCorporation’s 9-point scale for rating appropriateness of medical decisions was used to score most options. Consensus was defined as anon-random distribution of scores by chi-square “goodness-of-fit” test. We assigned a categorical rank (first line/preferred, secondline/alternate, third line/usually inappropriate) to each option based on the 95% confidence interval around the mean. Ratings were usedto develop guidelines for preferred strategies in key clinical situations. This study received financial support from multiple sponsors, withthe panel kept blind to sponsorship to reduce possible bias. Medication ratings were based on responses of only those respondents withdirect experience with each drug. In reporting practice patterns, the panel was asked to respond based on actual data rather than estimates.

Results. The expert panel reached consensus on 78% of the options rated on the 9-point scale. The responses suggest that physicians canmake provisional diagnoses with some confidence and that pharmacological and nonpharmacological interventions are selected differen-tially based on diagnosis and other salient demographic and medical features. BNZs are recommended when no data are available, whenthere is no specific treatment (e.g., personality disorder), or when they may have specific benefits (e.g., intoxication). No single SGAemerges as a nonspecific replacement for haloperidol; instead, different SGAs are preferred in various circumstances consistent with cur-rent evidence.To the degree that haloperidol is recommended, it is almost always in combination with a BNZ; haloperidol alone is pre-ferred only in the medically compromised. In contrast, the SGAs are more often recommended for use alone, and the panel would avoidcombining BNZs with some SGAs. Oral risperidone alone or combined with a BNZ receives strong support in a variety of situations. Oralolanzapine was rated very similarly to risperidone, with slightly higher ratings than risperidone in situations where it has been studied (e.g.,schizophrenia, mania) and slightly lower ratings where it has not been studied or safety may be a concern; there was less support for com-bining oral olanzapine with a BNZ. For oral treatment of agitation related to schizophrenia or mania, olanzapine alone, risperidone aloneor combined with a BNZ, and haloperidol plus a BNZ are first line, with strong support also for combining divalproex with the antipsy-chotic for presumed mania. Oral ziprasidone and quetiapine generally received similar second-line ratings in most situations. If a parenteralagent is needed, I.M. olanzapine alone received somewhat more support than I.M. ziprasidone alone; however, there was more support forI.M. ziprasidone alone or combined with a BNZ than for I.M. olanzapine plus a BNZ, probably reflecting safety concerns. For example,for a provisional diagnosis of schizophrenia, first-line parenteral options are I.M. olanzapine or ziprasidone alone or I.M. haloperidol orziprasidone combined with a BNZ. Neither of the new parenteral formulations received as much support as traditional agents (I.M. BNZs,I.M. haloperidol) when no data are available or the diagnosis involves medical comorbidity or intoxication. When initial intervention withrisperidone, ziprasidone, or haloperidol is unsuccessful, the panel recommended adding a BZD to the antipsychotic. However, when ini-tial treatment with olanzapine or quetiapine is unsuccessful, increasing the dosage is recommended. Perphenazine was consistently ratedsecond line and droperidol and chlorpromazine received third-line ratings throughout.

Conclusions. Within the limits of expert opinion and with the expectation that future research data will take precedence, these guidelinessuggest that the SGAs are now preferred for agitation in the setting of primary psychiatric illnesses but that BNZs are preferred in othersituations. (Journal of Psychiatric Practice 2005;11 [Suppl 1]:1–108)

KEY WORDS: behavioral emergency, agitation, psychosis, mania, expert consensus guidelines, antipsychotics, benzodiazepines, sedatives,restraints

MICHAEL H. ALLEN, MDGLENN W. CURRIER, MD, MPH

DANIEL CARPENTER, PhDRUTH W. ROSS, MA

JOHN P. DOCHERTY, MD

WHY A REVISION?

Behavioral emergencies are complex, dynamic situations. The diag-nosis is often unknown or provisional at best. There is a sense ofurgency, limited time for decision-making, and a need to interveneimmediately despite limited data and change course rapidly as newinformation becomes available, including responses to prior inter-ventions. Whether in the emergency room or inpatient psychiatricsetting, immediate assessment and effective intervention can reducedanger to patients and staff and promote faster recovery. At the sametime, any action or inaction may have serious adverse effects. Evenan objectively good response may leave patients feeling traumatizedand angered by the process. It is important for clinicians to remem-ber they must first do no harm.

Unfortunately, high-quality, empirical data on the most effectiveand appropriate methods of managing behavioral emergencies arelimited. This is largely due to practical and ethical difficulties in per-forming controlled clinical trials in agitated patients in emergencysettings. For example, it is often impossible or impractical to obtainconsent for participation from severely agitated patients, and yetthese are the very patients for whom information on appropriatetreatments is most needed. It was for this reason that we undertookour original survey of expert opinion on behavioral emergencies, theresults of which were published in 2001 in The Expert ConsensusGuidelines on the Treatment of Behavioral Emergencies.1

Several new options for treating behavioral emergencies havebecome available since those guidelines were published. At the timeof the first survey, only conventional antipsychotics (e.g., haloperi-dol, perphenazine) were available in intramuscular (I.M.) formula-tions, whereas two second-generation (atypical) antipsychotics(SGAs), olanzapine and ziprasidone, are now available in fast-actingI.M. formulations as well as fast dissolving tablets of risperidone(Risperdal M-Tab) and olanzapine (Zyprexa Zydis).2

In addition, agitation is still poorly understood. It has beendescribed by Lindenmayer3 as a transnosological syndrome andagents that have historically been available, including first generation(conventional) antipsychotics (FGAs) and sedatives, have been usedacross disease states. In our previous survey, benzodiazepines (BNZs)were recommended when no diagnostic information was available.FGAs have been used somewhat more narrowly, not due to lack ofefficacy in different states, but due to poor tolerability comparedwith BNZs. The FGAs have also been the treatment of choice foragitation associated with delirium. Not only is there evidence thatboth BNZs and FGAs are effective in multiple diseases but there islittle evidence that one is superior to the other.

However, the U. S. Food and Drug Administration (FDA) hasbeen unwilling to accept a nonspecific “pain and fever” model of agi-tation. To do so, it would have to be demonstrated that symptoms ofagitation are universally recognized, readily measured, have a wellunderstood pathophysiologic basis, and respond similarly to symp-tomatic drug treatment, apart from the diverse disease states that

may lead to the symptom.4 Consequently, new drugs are studied inthe context of particular diseases and are then said to be indicated foragitation associated with a particular disease entity such as schizo-phrenia or mania. Thus, new agents have been demonstrated effec-tive only in well characterized and homogeneous populations. Ofcourse, patients in emergency settings rarely present with a single,clearly defined diagnosis; instead, comorbid medical and substanceuse disorders are highly prevalent. Agitation may present in manydifferent conditions, and treatments must be chosen based on limit-ed data. Treatments must be safe enough to use in agitated patientswho are assumed to have serious undiagnosed complications that canonly be assessed after the agitation is successfully managed. In somefraction of cases, which varies by setting, agitation will later bedetermined to be associated with a primary psychiatric disorder forwhich antipsychotics might be indicated as a routine part of care. Inthose situations, using antipsychotic medication for agitation mayspeed recovery by a brief period. In other circumstances, antipsy-chotics may prove effective but unnecessary and will be discontin-ued. This situation raises questions as to whether the SGAs will befound to be effective in most forms of agitation and sufficientlybenign that unnecessary exposure is of little consequence. If so, non-specific use of antipsychotics may turn out to be the most efficientoption, since treatment of patients who ultimately require antipsy-chotics will be hastened and others will not be harmed.

However, the SGAs appear quite different from the FGAs inimportant ways, and their safety and efficacy have not been demon-strated in many relevant circumstances. While data are emerging thatsuggest there may be a cumulative risk of adverse health conse-quences with chronic exposure to some SGAs, the health conse-quences of a single dose in the context of p.r.n. use are completelyuncharacterized. If the SGAs cannot be used nonspecifically, it isimportant to know what physical and psychiatric findings might bequickly identified on the basis of a limited examination to facilitatechoice of a specific agent.

Furthermore, if the newer agents are more tolerable, the rationalefor combining an antipsychotic with a BNZ may be obviated. FGAswere often combined with BNZs to achieve sedation without use ofexcessive antipsychotic doses and to ameliorate side effects of theFGAs, principally haloperidol. This may be unnecessary and unwisewith the SGAs. If the newer agents are in fact more tolerable, a betterstrategy may be to increase the dose of the SGA rather than risk com-bining these more pharmacologically complex agents with BNZs.

Finally, there is the issue of the role of sedation in the managementof agitation. An international consensus has been developing thatcalming rather than sedation is the appropriate endpoint of inter-ventions for behavioral emergencies.5–7 Recent years have also seen agrowing emphasis in both the clinical and lay literature on increas-ing patient involvement in treatment decisions for behavioral emer-gencies to the greatest extent possible, both through use of advancedirectives and input during the emergency.1,6–8 This would seem tofurther erode support for combinations with sedative medications.

Journal of Psychiatric Practice Vol. 11, Suppl. 16 November 2005

Expert Consensus Guideline Series

In order to update recommendations to reflect the rapid evolutionand new concerns of the field, we undertook a new survey of expertopinion concerning the most appropriate goals of emergency inter-ventions and best choice of specific agents.

EXPERT CONSENSUS AND PRACTICE GUIDELINES

The sheer number of possible combinations and sequences of avail-able treatments for many diseases makes it difficult to provide com-parative recommendations based entirely on clinical trial data.9,10 Amethod for describing expert opinion in a quantitative, reliable man-ner to help fill gaps in evidence-based guidelines has been developedand applied to a variety of psychiatric disorders.11–25

Creating the Surveys

Based on a literature review and the 2001 survey results, we identi-fied key decision points in the management of behavioral emergen-cies and feasible options for intervention, highlighting key clinicalquestions not adequately addressed or definitively answered in theliterature.26 A written survey was developed with 61 questions(1,020 options) asking about types of assessments, most appropriateinterventions, and tailoring selection of interventions to the mostlikely etiology of the behavioral dyscontrol. We also asked about nextsteps if there is an inadequate response to initial intervention as wellas safety and tolerability issues, such as management of behavioralemergencies in children, the elderly, and patients with medical com-plications. The survey took 2 or more hours to complete, and weprovided a $500 honorarium.

The Rating Scale

For 854 options in the survey, we asked raters to evaluate appropri-ateness by means of a 9-point scale slightly modified from a formatdeveloped by the RAND Corporation for ascertaining expert con-sensus.27 For the other questions, we asked respondents to fill in ablank (e.g., write in a dosage) or check a box endorsing one of sev-eral alternatives. We asked the raters to consider the best possibleapproach for the first few hours of intervention in order not to havea negative impact on the clinician’s ability to diagnose and treat thedisorder in continuing care. We asked the panel to draw on boththeir knowledge of the research literature (we did not provide a lit-erature review) and their best clinical judgment in making their rat-ings, but not to consider financial cost. We instructed respondents torate only those medications with which they had direct experienceand to base their responses concerning practice patterns on actualdata rather than estimates. In order to take as conservative anapproach as possible, ratings of each medication were based on theresponses of only those experts who had direct experience with thatdrug. We presented the rating scale to the experts with the anchorsshown in Figure 1.

Figure 2 shows Survey Question 6 as an example of our questionformat.

Composition of the Expert Panel

We identified 50 leading American experts in psychiatric emergencymedicine from several sources: board certified members of theAmerican Association of Emergency Psychiatry who have adminis-trative responsibilities for a psychiatric emergency service (PES) aswell as academic affiliations, individuals who have published researchon emergency psychiatry or psychopharmacology, and emergencyphysicians recognized for their contributions to the treatment ofpatients with mental illness in general emergency settings. More than90% of the panel were board certified and affiliated with academicinstitutions. We received responses from 48 (96%) of the 50 expertsto whom the survey was sent, who practice in 17 different states. Ofthe respondents, 47 hold an MD degree and 1 a DO degree. 77%are male. The experts’ mean age was 48 years (SD 8.6, range 31–69),with a mean of 17 years in practice (SD 8.6, range 3–45) and a meanof 16 years in emergency psychiatric care (SD 7.2, range 4–35).100% reported that their practice involved clinical or research workwith agitated patients. 65% reported spending at least half theirwork time seeing patients. 45% practice in a psychiatric (15%) orgeneral hospital (30%) emergency service; 28% practice on an inpa-


TREATMENT OF BEHAVIORAL EMERGENCIES 2005

Figure 2. Sample survey question

6. Goals of emergency intervention. Assume you have decided toundertake an emergency intervention for a patient with clini-cally significant agitation. Please rate the following as appro-priate goals for such an intervention.

Calming the patient without sedation 1 2 3 4 5 6 7 8 9

Mild sedation to the point of drowsiness but not sleep 1 2 3 4 5 6 7 8 9

Sleep or heavy sedation 1 2 3 4 5 6 7 8 9

Figure 1. The rating scale

Extremely 1 2 3 4 5 6 7 8 9 ExtremelyInappropriate Appropriate

9 = Extremely appropriate: this is your treatment of choice

7–8 = Usually appropriate: a first-line treatment you would oftenuse

4–6 = Equivocal: a second-line treatment you would sometimesuse (e.g., patient/family preference or if first-line treatmentis ineffective, unavailable, or unsuitable)

2–3 = Usually inappropriate: a treatment you would rarely use

1 = Extremely inappropriate: a treatment you would never use

tient service, either in a general or psychiatric hospital; and 28%practice in both emergency and inpatient services. 71% of theexperts had participated in clinical research studies during the pre-ceding 5 years. The respondents reported the following percentagesof patients by diagnostic group:

Mean SD Range% % %

Psychotic disorder 31 18 1–80Major depression 22 12 1–60Bipolar disorder 16 9 1–40Primary substance abuse 16 12 2–50Dementia 8 14 1–75Other Axis I disorder 8 7 1–32Axis II disorder only 8 11 1–57No psychiatric disorder,

required social services 4 3 1–10

42% of the patients served by the experts’ services over the previousyear had a secondary substance use disorder. 20% of their psychiatricemergency services evaluate fewer than 250 patients each month,29% evaluate 250–500 patients each month, and 51% more than500 patients. The respondents reported that a mean of 36% ofpatients in their services were treated involuntarily. The authorsacknowledge that many panel members were drawn from urban aca-demic medical centers, which may affect the applicability of theirrecommendations for rural settings.

Financial Support

This survey study was supported by grants from multiple sponsors:AstraZeneca Pharmaceuticals LP, Janssen Pharmaceutica ProductsLP, and Pfizer, Inc. The experts were kept blind to the sponsorshipof the survey in order to reduce any possible bias.

Data Analysis for Options Scored on the Rating Scale

For each option, we first defined the presence or absence of consen-sus as a distribution unlikely to occur by chance by performing a chi-square test (p < 0.05) of the distribution of scores across the 3 rangesof appropriateness (1–3, 4–6, 7–9). Next we calculated the mean and95% confidence interval (CI). A categorical rating of first, second, orthird line was designated based on the lowest category in which theCI fell, with boundaries of 6.5 or greater for first line, and 3.5 orgreater for second line. Within first line, we designated an item as“treatment of choice” if at least 50% of the experts rated it as 9.

Displaying the Survey Results

The results of Question 6 (Figure 2) are presented graphically inFigure 3. The CIs for each option are shown as horizontal bars and

the numerical values are given in the table on the right, including thenumber of experts who rated each option.

The Ratings

First-line treatments are strategies that received highest rat-ings when the experts’ responses to the survey were statistically ana-lyzed. These are options the panel feels are usually appropriate asinitial treatment for a given situation.

Treatment of choice, when it appears, is an especially strongfirst-line recommendation (rated “9” by at least half the experts). Inchoosing between several first-line recommendations, or deciding touse a first-line treatment at all, clinicians should consider the overallclinical situation, including the patient’s prior response to treatment,side effects, general medical problems, and patient preferences.

Second-line treatments are reasonable choices for patientswho cannot tolerate or do not respond to the first-line choices. A sec-ond-line choice might also be used for initial treatment if the first-line options are deemed unsuitable for a particular patient (e.g.,because of poor previous response, history of side effects, generalmedical contraindication, potential drug interaction, or if the expertsdo not agree on a first-line treatment). To differentiate among sec-ond-line options, we label items whose CIs overlap with the first-linecategory “high second line.”

Third-line treatments are usually inappropriate or used onlywhen preferred alternatives have not been effective.

No consensus. For each item in the survey, we used a chi-square test to determine whether the experts’ responses were ran-domly distributed across the 3 categories, which suggests a lack ofconsensus, indicated by an unshaded bar in the survey results.

Statistical differences between treatments. While we did not per-form tests of significance for most treatments, the reader can readilysee whether CIs overlap (roughly indicating no significant differencebetween options by t-test). The wider the gap between CIs, thesmaller the p value would be (i.e., the more significant the differ-ence). In some questions there are important differences within lev-els that we point out. However. differences within levels are often notsignificant from a statistical perspective. Also, there are sometimes nostatistical differences between choices at the bottom of first line andthose at the top of second line.

From Survey Results to Guidelines

After the results were analyzed and ratings assigned, the experts’ rec-ommendations were used to develop user-friendly guidelines. Twolevels of recommendations are presented: first-line/preferred and high

*



second line/alternate. When the guideline gives more than one treat-ment in a rating level, they are listed in order of mean scores. Forexample, results of question 6 above appear on page 63 and are usedin Guideline 2D: Goals of Emergency Intervention. (p. 30). As indi-cated by the black box with the asterisk in the results and bold italicsin the guideline, calming without sedation was the goal of choice,being rated 9 by more than 50% of the panel. Mild sedation to thepoint of drowsiness but not sleep was high second line, while sleep orheavy sedation was not endorsed as an appropriate goal (third line).

Degree of Consensus

Of the 854 options rated on the 9-point scale, consensus was reachedon 663 options (78%) as defined by the chi-square test. When thereis no first-line recommendation, the highest rated second-line optionis presented as the “preferred” treatment in the guideline.

RESULTS AND COMMENTARY

The clinical problem addressed in these guidelines differs from othersfor which these methods have been used.11–25 Most treatment algo-rithms on which previous expert consensus guidelines were basedbegan with a diagnosis and assumed a typical physician-patient rela-tionship. However, behavioral emergencies are defined by disruptionin the physician-patient relationship with resulting deficiencies inassessment and patient participation. Many questions involve forceddecisions based on assumptions about urgency, cooperation, amountof available information, diagnostic confidence, and individual riskfactors. In a sense, the goal of these interventions, a confident diag-nosis and informed consent, is the starting point of other guidelines.In the following sections, we summarize key recommendations fromthe guidelines on management of behavioral emergencies and discusshow the recommendations relate to available research and how theyhave changed since the first survey. Complete survey data appear onpages 60–108 and the guidelines based on those data are presented onpages 26–59.

Defining Clinically Significant Agitation

We presented the panel with a number of putative definitions of agi-tation based on the literature, from which they identified the fol-

lowing as central to the concept of clinically significant agitationrequiring intervention in the emergency setting: abnormal and exces-sive verbal, physically aggressive, or purposeless motor behaviors,heightened arousal, and clinically significant disruption of thepatient’s functioning. In the first survey, the panel gave increasingsupport to the use of an emergency intervention (involuntary med-ication or physical restraint) as patient behavior suggests an increasedpotential for violence, with 100% indicating that it would almostalways be appropriate for a patient who is directly threatening orassaultive, while the majority did not consider such interventionappropriate for a patient who displays only a refusal to cooperatewith unit routine or intense staring.1

In the current survey, to help define the syndrome of agitation andfurther clarify indications for emergency intervention, we provided alist of target behaviors (taken from the literature and various ratinginstruments for agitation) and asked the experts to select 10 featuresthey considered most typical of clinically significant agitation requir-ing emergency intervention. The following 7 behaviors wereendorsed by 60% or more of the panel: explosive and/or unpre-dictable anger; intimidating behavior; restlessness, pacing, or exces-sive movement; physical and/or verbal self-abusiveness; demeaningor hostile verbal behavior; uncooperative or demanding behavior orresistance to care; and impulsive or impatient behavior or low toler-ance for pain or frustration. Of note, 5 of these 7 are among the 14items on Corrigan’s Agitated Behavior Scale.28

Goals of Emergency Intervention

The panel strongly endorsed calming the patient without sedation asthe most appropriate goal of emergency intervention (rated first lineby 92%) and gave high second-line ratings to mild sedation to thepoint of drowsiness but not sleep. They did not endorse sleep orheavy sedation as an appropriate goal of intervention (rated third lineby a majority) These recommendations are consistent with theClinical Guideline on Schizophrenia from the National Institute forClinical Excellence (NICE) in the United Kingdom, which statesthat “the aim of drug treatment in such circumstances is to calm theperson and reduce the risk of violence and harm” and that “an opti-mal response would be a reduction in agitation or aggression with-out sedation, allowing the service user to participate in furtherassessment and treatment.”5



Figure 3. Results of survey question 6

9 5 % C O N F I D E N C E I N T E R V A L S Tr of 1st 2nd 3rdThird Line Second Line First Line N Avg(SD) Chc Line Line Line

Calming the patient without sedation 48 8.4(1.3) 75 92 8 0

Mild sedation to the point of drowsiness but not sleep 48 6.8(1.4) 17 54 46 0

Sleep or heavy sedation 48 3.7(1.8) 0 10 31 581 2 3 4 5 6 7 8 9 % % % %

*

Patient Involvement in Care

In the first survey,1 the panel stressed the importance of collaborationbetween patient and clinician whenever possible in achieving the bestshort- and long-term outcomes. They also emphasized the impor-tance of safety concerns in the short-term (e.g., control of aggressivebehavior and protecting the patient and community). These ratingsagree with findings in a recent survey/focus group study of con-sumers, which reported that consumers place great emphasis on beingtreated with respect and being allowed to participate in decisionsabout their treatment to the greatest extent possible.8 This can beaccomplished by suggesting that the patient might feel better if he orshe took some medication and asking if there is a medication he orshe would prefer. Failing that, the patient can be offered a choicebetween two medications or, as a last resort, can simply be asked toassent to a medication (e.g., “Would you be willing to take…?”).

Initial Assessment

The goal in a behavioral emergency is to facilitate resumption of amore typical physician-patient relationship, obtain informed consentif possible, and promote the best possible long-term treatment out-come. Acute agitation interferes with assessment and treatment at atime when immediate intervention appears needed because of dan-gerous behavior or warning signs of such behavior, and yet assess-ment plays a critical role in selecting the most appropriateintervention. Recommendations for assessment based on the expertsurveys are summarized below.

Medical assessment. In the first survey, the panel indicated that a keystep in initial evaluation is identifying any medical conditions thatmight be causing or contributing to agitation. This is especiallyimportant because available data suggest that delirium, in particular,should be managed based on its underlying etiology if it can be iden-tified. The earlier panel also indicated that it is most important in theinitial medical evaluation to obtain vital signs and a medical history,perform a visual examination of the patient, a urine toxicology screen,a cognitive examination, and a pregnancy test if the patient is awoman of childbearing age.1 There was somewhat less support formore complete physical examinations, possibly reflecting constraintson time and availability of personnel. The level of examination willdepend on the patient’s specific signs and symptoms, with more com-plete evaluations indicated in some circumstances or later in treat-ment. We asked some additional questions in the current survey torefine these recommendations. The panel in the current surveystressed the importance of assessing the following in the initial assess-ment: head trauma, respiration, heart rhythm, color, smell of alcohol,diameter/reactivity of pupils, lacerations, nuchal rigidity, and frac-tures; they would also obtain glucometry and urine for toxicology.Despite the interest in toxicology, they did not endorse nystagmus,reflexes, and tongue fasciculations although these may provide evi-

dence of intoxication or withdrawal. The panel endorsed the Mini-Mental State Examination29 as the preferred tool for assessing cogni-tive function in the emergency setting, with the Clock Drawing Test30

and the Confusion Assessment Method (CAM)31 second-line options.Methods of cognitive screening have been reviewed elsewhere.32,33

These recommendations generally agree with evidence-based clini-cal policies being developed by the American College of EmergencyPhysicians (ACEP) (with the participation of two of the authors ofthese guidelines) for assessment and management of psychiatricpatients in emergency departments.7 The ACEP policy endorses theCAM for cognitive assessment and indicates that diagnostic evaluationshould be directed by the history and physical examination, whichshould be used to identify patients at risk. Routine testing is not rec-ommended; in fact, no procedures are recommended in the absence ofindications from the history and physical examination. The ACEPclinical policy committee specifically addressed breath alcohol con-centration and urine toxicology; they recognized that urine toxicologymay be important for psychiatric diagnosis and disposition but foundno evidence that it would alter emergency medical care and felt itshould not delay transfer to a more appropriate setting. The currentexpert panel, of whom 10% were emergency physicians, did indicatethat substance use would alter the management of a behavioral emer-gency, though there are no studies that compare different strategies foragitation in the setting of substance use.

Psychiatric assessment. In the first survey, we ascertained that a briefassessment leading to a general category of diagnosis is adequate formanaging agitation. We tested this by asking the panels in both sur-veys how they would manage an agitated patient in need of immedi-ate treatment but with no data available to guide the decision. In thissurvey, we also presented vignettes typical of actual situations encoun-tered in emergency settings. We then asked the panel to rate the mostappropriate interventions based on available information to determineif they could form a provisional diagnosis and if that diagnosis wouldinfluence or determine their approach to treatment or if they wouldpursue the same course regardless of available information. If the sametreatments emerged consistently regardless of differing presentations,this might suggest the existence of a nonspecific approach.Alternatively, differences between strategies used in different situationsmight suggest more specificity in treatment selection.

Interventions When No Information Is Available

There are situations in which immediate intervention is needed toassess the patient or to ensure safety of patient and others but insuf-ficient data are available on which to base even a provisional diagno-sis (i.e., the only sign or symptom entering the decision-makingprocess is agitation). This seems to fit the definition of nonspecifictreatment rather than chemical restraint, as indicated in the first sur-vey.1 To proceed safely with assessment when no additional informa-tion is available but the patient is willing to take oral medication, the



panel recommended lorazepam as agent of choice, with risperidoneanother first-line option, and olanzapine, haloperidol, and quetiap-ine high second line. If I.M. medication is required before an assess-ment can be done, I.M. lorazepam is again agent of choice, with I.M.ziprasidone, olanzapine, and haloperidol other first-line options.This preference is unchanged from 2001 when an oral or I.M. BNZalone was preferred if one must intervene with medication to man-age agitation when insufficient information is available to make evena provisional diagnosis. BNZs are also preferred when there is nospecific pharmacologic treatment (e.g., conduct disorder, stimulantintoxication), presumably because the goal is to achieve time-limitedsedation while avoiding undue risk of extrapyramidal symptoms(EPS) and other side effects associated with antipsychotic drugs.(BNZs are also recommended for alcohol-related agitation.)

Interventions Selected Based on Limited Information

We presented several case vignettes and asked what provisional diag-nosis the panel would make based on the information given, howmuch confidence they would have in this diagnosis, how much thediagnosis would guide treatment selection, and what initial nonphar-macologic interventions and medications they would recommend.

Patient 1: The patient, a young white male who appears to be in hisearly 20s, presents to the PES in a highly agitated state. He is directlythreatening, throwing the food tray and yelling. He is paranoid aboutstaff poisoning him and appears to be responding to internal stimuli. Thepatient is unable to converse with staff at all. No data or history areavailable, but the patient clearly poses a risk to staff and self.

Two-thirds of the panel endorsed a diagnosis of a psychotic disor-der (either schizophrenia or psychosis not otherwise specified[NOS]), with a majority somewhat to very confident in the diagno-sis. The experts indicated that the diagnosis of a psychotic disorderwould determine or at least influence their treatment selection.Thus, when the experts are at least fairly confident they are dealingwith a psychotic disorder, they will treat more specifically.

Throughout the survey, for patients who appear to have a primarypsychiatric disturbance, the experts strongly endorsed beginningwith the least invasive interventions possible consistent with ensur-ing safety of patient and others (see algorithms, pp. 43 and 45). Eventhough the description specified that the patient posed a clear risk tostaff and self, the experts still endorsed beginning with a show offorce and trying verbal intervention before using involuntary med-ication or restraints. If restraints are needed, the panel did not rec-ommend using them alone but with concomitant medication.

We asked the experts to assume that Patient 1 was not able or will-ing to take oral medication and that, based on the initial assessment,they had decided to intervene with parenteral medication to treat theagitation before further medical intervention. In this situation, thepanel recommended beginning with I.M. haloperidol plus an I.M.BNZ, or I.M. olanzapine alone, but would also consider I.M.

ziprasidone alone or combined with a BNZ, or I.M. haloperidol oran I.M. BNZ alone. The experts clearly preferred to use an antipsy-chotic, either alone or combined with a BNZ, for a patient whoappears to have a psychotic disorder. We then asked the panel toassume that they had established that Patient 1 has a diagnosis ofschizophrenia, had prescribed what they consider the usual maxi-mum dose of an antipsychotic, and that the patient has calmed downsomewhat but is still moderately agitated. If additional medication isneeded to control the agitation and the patient is now willing to takeoral medication, the experts recommended giving an oral BNZ.Giving another dose of the same antipsychotic was a second-lineoption. The experts did not recommend using divalproex or a dif-ferent antipsychotic (both third line) as adjunctive oral agents forresidual agitation.

Patient 2: The patient, a 19-year-old college freshman, is brought to theemergency room by his parents because of agitation and unusual behavior.He has no psychiatric history, but there is a family history of major depres-sion. The patient is highly agitated, pacing, displaying pressured speechand marked grandiosity, and bragging about his artwork. The patientinsists that there is nothing wrong with him. The family reports that thepatient has not been sleeping. The patient refuses to follow directions andbegins to shout at staff. His thought processes are difficult to follow. Attimes he appears to be responding to internal stimuli and hallucinating.

The experts strongly endorsed (89%) a diagnosis of mania for thispatient, with the majority fairly or very confident in the diagnosis.As in the previous case, when the panelists are able to make a fairlyspecific diagnosis, they indicate that this will influence or determinetheir treatment selection. For a patient with presumed bipolarmania, the experts would always recommend including an antipsy-chotic in the initial treatment, preferably in combination with aBNZ, although they would also consider using the antipsychoticalone or in combination with a mood stabilizer. There was less sup-port for a mood stabilizer alone or combined with a BNZ as initialtreatment (both second line), presumably because of the need tocalm the patient as rapidly as possible. If it is possible to use an oralantipsychotic, the experts rated olanzapine and risperidone first linefor a patient with presumed bipolar mania, with quetiapine, ziprasi-done, and haloperidol high second line. These ratings may reflect thefact that olanzapine and risperidone were the first two SGAs toobtain an indication for bipolar mania as well as a desire to use anagent with more sedating properties such as olanzapine. If it is decid-ed to include a traditional mood stabilizer in the initial regimen,divalproex was agent of choice. This recommendation agrees withresponses from a survey of emergency psychiatrists, which foundthat, if a mood stabilizer is needed in this setting, 90% would usedivalproex/valproate, while only 8% would choose lithium and only2% another mood stabilizer.34 The experts supported using dival-proex loading doses in all types of manic episodes, probably reflect-ing the fact that lithium is not generally used in the emergencysetting and that loading doses of divalproex may help stabilize the



patient quickly.35 As in 2001, the panel favored strategies thatemploy higher doses of divalproex rather than usual titration (e.g.,beginning with 250 mg tid and titrating as tolerated). Whereas a 20mg/kg/day initiation strategy was preferred in 2001, there was aslight preference in the current survey for beginning with a loadingdose of 30 mg/kg/day for 2 days and then reducing the dose to 20mg/kg/day, despite inclusion of emergency physicians presumablyless familiar with this approach to mania. The experts also preferreddelayed- over extended-release divalproex for emergency use.

If an I.M. treatment is needed for a patient with presumed bipo-lar mania, the experts endorsed olanzapine alone or the traditionalcombination of haloperidol plus a BNZ. The preference for I.M.olanzapine over ziprasidone here may reflect the length of time olan-zapine has had an indication for mania, findings from the I.M. olan-zapine study in agitated patients with bipolar mania by Meehan etal.,36 and the desire to use a more sedating agent.

Patient 3: A 60-year-old man with a history of alcohol abuse and dia-betes presents to the emergency department with agitation and confusion.He is mildly diaphoretic with mild tachycardia but is afebrile. Thepatient is cheerful but tries to wander and is difficult to redirect. He asksthe same questions repeatedly and picks at his skin.

Two-thirds of the panel endorsed a diagnosis of delirium, with86% somewhat to very confident in this diagnosis. A third endorseda diagnosis of alcohol withdrawal, with 81% somewhat to fairly con-fident in this diagnosis. A majority indicated that these diagnoseswould determine their treatment selection, while nearly all the restsaid that the diagnosis would at least influence treatment selection.

If an older patient is delirious or suffering alcohol withdrawal, theexperts are much less likely to use show of force to avoid making thepatient feel threatened or increasing his or her confusion (see treat-ment algorithm p. 48). When an organic basis for the problem isstrongly suspected, the panel emphasized the need to gather as muchdata as possible from collateral sources before intervening with med-ication that might cloud the clinical picture. The experts did notgenerally endorse use of restraints in this situation. The panel rec-ommended monotherapy for this patient and rated a BNZ alone firstline. This choice probably reflects concern about the role of alcoholin the presentation and the limited evidence for haloperidol or otherantipsychotics in this type of patient. The experts’ ratings may alsoreflect warnings concerning potential for cerebrovascular accidents(CVAs) in elderly patients with dementia treated with SGAs,37–42

although recently published reports have indicated that this riskappears to be no different than in patients taking FGAs,43,44 and thereports involved chronic rather than acute treatment.

Patient 4: A 14-year-old boy with a history of behavior problems isbrought by police from school where he threatened a teacher. The boy isloud and profane in the emergency department. He is hostile to the nurs-es and demands immediate discharge. A member of the school staff whoaccompanies him reports that the school records show no medical history

beyond the usual childhood diseases and immunizations but that the boydoes have a history of many previous behavior problems

The panel strongly endorsed (85%) a diagnosis of conduct disorder,with 74% somewhat to very confident in the diagnosis, which amajority said would influence their treatment selection. The expertsclearly indicated a reluctance to medicate a child in the absence of aclear Axis I psychiatric disorder (see algorithm p. 50). They gave pri-ority to talking to the patient and family and using nonpharmacolog-ical means to try to calm the patient. Not surprisingly, given thereluctance to use medication, no agents were rated first line and aBNZ alone was the only medication rated high second line. Theexperts would avoid combining medications in this patient (all com-binations except I.M. haloperidol plus a BNZ were rated third line).Note that the panel was not recruited for expertise in child psychiatry.

Patient 5: A patient has been treated in the PES or acute inpatient serv-ice for a behavioral emergency characterized by significant agitation. Theunderlying condition has been adequately treated and the patient hascalmed down. The patient has a history of insomnia with restlessness andracing thoughts before sleep and it is decided to leave p.r.n. orders for thenursing staff in case the patient has difficulty at bedtime.

None of the sedating agents we asked about were rated first line,probably reflecting lack of data concerning the most appropriatesedatives for use in agitated patients. The panel gave high second-lineratings to lorazepam, zolpidem, trazodone, and temazepam. Thesupport for lorazepam may reflect the perception that the sleep prob-lems represent residual agitation. None of the antipsychotics, includ-ing quetiapine, received significant support in this situation.

Selecting Medications Based on an Established Diagnosis

In the next part of the survey, we eliminated any ambiguity aboutdiagnosis and asked the panel about treatments for agitation in thecontext of specific diagnoses, both if the patient was willing to takeoral medication or if it was necessary to use a parenteral agent.

General medical etiology. The experts would avoid using medica-tion if possible if the agitation appears due to a general medical con-dition (e.g., delirium, agitation in dementia), probably because ofconcern about inducing profound somnolence or clouding the sen-sorium and interfering with medical assessment. Available data sug-gest delirium should be managed according to its underlying etiologyif this can be identified. If a medication is needed to treat agitationin a patient with a medical condition such as delirium and thepatient is able and willing to take oral medication, the experts preferto use haloperidol or risperidone alone. If a parenteral medication isneeded, the experts would start with I.M. haloperidol alone andwould also consider I.M. olanzapine or I.M. ziprasidone alone.

It is interesting to examine changes in recommendations since thefirst survey. In choosing a medication for agitation related to a gen-eral medical etiology, oral or I.M. haloperidol alone, the highest



rated option in 2001, continued to be preferred in the current sur-vey. However, whereas oral or parenteral BNZs alone or combinedwith haloperidol were the next choice in 2001, oral or I.M. SGAswere the next choice in the current survey. Among the oral SGAs, theexperts preferred risperidone (43% first line in 2001 and 60% incurrent survey) and olanzapine (19% first line in 2001 and 44% incurrent survey) for agitation related to a general medical etiology. Ifa parenteral medication is needed, the next choices after haloperidolwere I.M. olanzapine or ziprasidone. These ratings reflect the factthat delirium due to a general medical etiology has usually beentreated with high-potency FGAs, although evidence to support thisis very limited. Breitbart et al.45 found that the FGAs were superiorto lorazepam in efficacy and side effects in a group of prospectivelydefined patients with AIDS delirium, although the generalizabilityof these findings to other types of delirium is questionable. The 1998Expert Consensus Guidelines for the Treatment of Agitation in OlderPersons with Dementia14 also recommended high potency FGAs fordelirium due to a general medical etiology (e.g., congestive heart fail-ure, urinary tract or upper respiratory infections) in patients withdementia, with risperidone high second line. The increased supportfor using an SGA rather than a BNZ alone or combined with anFGA in the current survey reflects the growing research base on theefficacy and safety of the SGAs. A number of recent studies havefound risperidone as effective in managing agitation in delirium ashaloperidol but with less risk of EPS.46–49 Similar but more limitedfindings have been reported for olanzapine.50,51 Risperidone was alsothe preferred antipsychotic for treatment of delirium in elderlypatients in the 2004 Expert Consensus Guidelines on UsingAntipsychotics in Older Patients.23 The warning concerning the possi-bility of CVAs in older patients now in the labeling for the SGAs38–42

does not seem to have influenced the experts’ choice of agents for usein acute emergency settings, probably because reports of CVAs haveall involved chronic rather than acute exposure and recent reportshave indicated that the risk appears no different than with FGAs.43,44

Substance or alcohol intoxication. Just as in the earlier survey, aBNZ alone is the preferred oral or parenteral agent for a patient withstimulant intoxication (and alcohol intoxication if a medicationmust be used). For stimulant intoxication, the next choice would bea BNZ plus a high-potency FGA. A report by van Harten et al.52

suggested that individuals who abuse stimulants may be more proneto EPS, which may be why the experts prefer BNZs in this situation(i.e., antipsychotics are not likely to have any special benefits for thispopulation but may be more likely to cause EPS). Cocaine toxicitymay also involve seizures, and the experts may prefer BNZs toantipsychotics for their protective effect in this situation.

In both surveys, the experts had no first-line recommendations fortreatment of agitation due to alcohol intoxication, but did rate a BNZalone high second line. The slight preference for BNZs for patientsintoxicated with alcohol may reflect the fact that a component ofwithdrawal may be contributing to the agitation for which the BNZ

may be specifically indicated. The American Psychiatric AssociationGuideline for the Treatment of Substance Use Disorders53 recommendsBNZs for alcohol withdrawal. A meta-analysis of 9 prospective con-trolled trials demonstrated that sedative-hypnotic agents are moreeffective than antipsychotics in reducing duration of alcohol with-drawal delirium and mortality.54 A 2001 bulletin from the Centers forMedicare and Medicaid Services (CMS [formerly the Health CareFinancing Administration])55 mentions use of BNZs for behavioraldisturbances associated with alcohol withdrawal as appropriate use ofmedication for purposes of treatment rather than chemical restraint.

Primary psychiatric disturbances. If the clinician can tentativelymake a more specific Axis I psychiatric diagnosis, the experts preferto provide treatment that is likely to be effective for the underlyingcondition and would generally use an antipsychotic. If the provi-sional diagnosis is schizophrenia or bipolar mania and the patientis willing to take oral medication, first-line options are olanzapine orrisperidone alone or risperidone or haloperidol plus a BNZ. Anotherfirst-line option for bipolar mania is divalproex plus an antipsy-chotic. Comparing recommendations from 2001 with the currentsurvey, it is clear that the general trend to choose an SGA as initialtreatment for psychotic and bipolar disorders22,24,56 is reflected inresponses to the current survey. In the first survey, the first-line rec-ommendation for acute oral treatment of a patient with a provision-al diagnosis of schizophrenia or bipolar mania was a BNZ combinedwith either an FGA or an SGA. Of the three SGAs available at thetime of the previous survey (olanzapine, quetiapine, and risperi-done), monotherapy with oral risperidone and olanzapine was onlyrated high second line for schizophrenia (risperidone 65%, olanzap-ine 49% first line) and mania (olanzapine 45%, risperidone 41%first line). In contrast, in the current survey, there was much moresupport for the use of monotherapy with SGAs, with oral olanzap-ine rated first line by 87% and oral risperidone rated first line by84% for a provisional diagnosis of schizophrenia, and oral olanzap-ine rated first line by 78% and oral risperidone rated first line by67% for mania. Quetiapine and ziprasidone received high second-line ratings for schizophrenia (rated first line by 58%) and quetiap-ine was also rated high second line for mania (49% first line).

None of the SGAs was available in a parenteral formulation at thetime of the earlier survey, but there was strong support for the SGAsin the current survey when a parenteral medication is needed. For apatient with a provisional diagnosis of schizophrenia who requiresparenteral medication, the panel gave first-line ratings to I.M. olan-zapine alone, I.M. haloperidol plus a BNZ, and I.M. ziprasidonealone or combined with a BNZ. For a patient who appears to havemania, the experts gave first-line ratings to I.M. olanzapine and toI.M. haloperidol plus a BNZ and high second-line ratings to I.M.ziprasidone alone or combined with a BNZ, possibly because theyare considering continuation treatment if needed.

In summary, there is increasing support for the use of SGAs, bothoral and parenteral, in emergency settings to treat patients with a



provisional diagnosis of schizophrenia or mania. Nevertheless, theexperts also gave more support to the use of haloperidol, usuallycombined with a BNZ, than is generally seen in other treatment set-tings, probably reflecting extensive experience with this agent inemergency care. However, use of haloperidol alone is rarely recom-mended, while there is less support for combining SGAs with BNZs.

For psychotic depression and personality disorder in a patientwilling to take oral medication, there was no consensus on first-lineagents in either survey, probably reflecting lack of data on emergencytreatment of these conditions. However, the panel in the current sur-vey was more likely to use an SGA alone to treat agitation in a patientwith psychotic depression or a personality disorder than the 2001panel, probably reflecting findings on the possible efficacy of SGAs inthese conditions in nonemergency treatment.22,57–61 If parenteralmedication is needed to manage agitation in a patient with a provi-sional diagnosis of psychotic depression, I.M. olanzapine alone wasfirst line, with ziprasidone or haloperidol alone or combined with aBNZ high second line. This contrasts with the first survey where thepanel gave first-line ratings to an FGA plus a BNZ, with a BNZ alonehigh second line. As in the first survey, an I.M. BNZ alone was pre-ferred for agitation associated with a personality disorder requiringI.M. medication, with I.M. haloperidol plus a BNZ high second line.In the current survey, I.M. olanzapine was also high second line.

Use of Emergency Medications: General Strategies

Most important factors in medication selection. The following fac-tors were rated most important in selecting initial emergency med-ication: acute (immediate) effect on behavioral symptoms, speed ofonset, availability of I.M., liquid, or rapidly dissolving formulation,patient’s history of response to the medication, limited liability forside effects, patient preference, and ease of administration (no needfor laboratory tests and simple dosing requirements). The panelwould also consider the medication’s safety in overdose and likeli-hood of promoting long-term adherence to treatment. Just as in thefirst survey,1 the experts placed less emphasis on continuity with thenext phase of treatment than on managing acute symptoms and didnot consider cost a significant factor in initial medication selection.

Medication characteristics. Treatment of agitation may be directedsolely at managing the current state or may also represent initiationof a continuing treatment. We thus asked the panel to assess medica-tions in terms of their perceptions of how calming (reducing agitationwithout inducing sleep) or sedating (reducing agitation by inducingsleep) they are and how likely they are to be effective for the underly-ing condition (e.g., a psychotic disorder). The following agents wererated most calming (without putting the patient to sleep): lorazepam,olanzapine, risperidone, and haloperidol, while chlorpromazine,lorazepam, quetiapine, and droperidol were rated most sedating. Interms of effectiveness for the underlying condition, risperidone, olan-zapine, ziprasidone, haloperidol, and perphenazine were first line.

Lorazepam is viewed as potentially calming or sedating but not aseffective for the underlying condition. This survey was conductedbefore publication of findings on perphenazine from the ClinicalAntipsychotic Trials of Intervention Effectiveness (CATIE).62

In both surveys, haloperidol and lorazepam were first line fordecreasing acute agitation. However, support for using SGAs to treatacute agitation has increased since 2001, when they were rated highsecond line. In the current survey, olanzapine and risperidone were firstline for being calming but not sedating, while risperidone, olanzapine,and ziprasidone were first line for being effective for the underlyingcondition. In the first survey, droperidol (an FDA-approved buty-rophenone available only for parenteral administration and used pri-marily in anesthesia) was rated most effective for decreasing agitationbut dropped to high second line in the current survey, probably due tothe black box warning concerning prolongation of the QTc intervaland risk of torsade de pointes.63 Although recent analyses failed to findevidence of a relationship between droperidol and cardiac events,64–66

there has been a dramatic decrease in its use and availability in emer-gency settings.67 Despite use as an antipsychotic in Europe, droperidolwas rated here as ineffective for treating an underlying psychosis,reflecting inexperience with its use in the United States.

Since the experts indicated that the preferred goal of emergencyintervention for acute agitation should be to calm the patient with-out sedation, we examined the rank ordering of the medications interms of combinations of characteristics. The highest combinedmean scores for being both calming and effective for the underlyingcondition went to risperidone, olanzapine, haloperidol, and ziprasi-done in that order. The highest combined mean score for beingcalming plus effective minus the mean rating for sedation went torisperidone, ziprasidone, olanzapine, and haloperidol, in that order.

Time to onset is also important. In managing agitated and poten-tially violent patients, faster onset may reduce the chance of injuriesand the need for, or time in, restraints. For speed of onset, the panelgave highest ratings to intravenous (I.V.) medication of any class;however, the panel in the first survey did not give strong support tomaking I.V. access available in the PES, probably because it wouldrequire a different staffing pattern. The next highest ratings (firstline) for speed of onset went to I.M. formulations of lorazepam,haloperidol, olanzapine, ziprasidone, and chlorpromazine. The mainchange between the surveys was that the two SGAs now available inI.M. formulations, olanzapine and ziprasidone, which were notincluded in the first survey, received first-line ratings for speed ofonset in the current survey. This is consistent with clinical trial datashowing that olanzapine separates from haloperidol at 15 and 30minutes after injection, after which haloperidol “catches up.”

The majority of the panel indicated they would not generally usethe following as initial treatment for a behavioral emergency: oralaripiprazole, chlorpromazine, and perphenazine and I.M. droperidolor perphenazine. Data have recently been presented concerning anI.M. formulation of aripiprazole, which was found to be similar tohaloperidol in clinical trials for agitation at a dose of 7.5 mg.68



Route of administration. We asked which formulations the panelpreferred to use if a patient is willing to take oral medication, takinginto account feasibility and convenience of use by nursing staff aswell as intrinsic properties. Rapidly dissolving formulations receivedthe highest ratings with strong support also for liquid concentrate ortablets. Although I.M. medications are clearly needed in some situa-tions, the panel in the first survey1 indicated that they prefer to useoral medication if possible, probably due to concern about possibleadverse effects of I.M. medication in this situation, such as mental orphysical trauma to patient, compromising the patient-physician rela-tionship, and effects on long-term adherence. The experts’ recom-mendations are consistent with responses to a survey ofapproximately 50 PES directors, in which the majority advocated useof oral medication whenever possible, with liquid formulations pre-ferred to tablets because of more rapid onset and ease of verifyingcompliance.34 In that survey, the medical directors estimated thatonly 1 in 10 patients in their PESs requires an injection. It has beenreported elsewhere that most agitated patients will assent to oralmedications.69 Although we did not repeat the question in the cur-rent survey, the earlier panel was asked what percentage of patientsin their services were likely to require emergency interventions(restraints, seclusion, parenteral medication), and the 19 whoresponded indicated that more than 80% of patients are managedwithout these interventions. CMS interim rules70 specify that“chemical restraint” be considered a last resort, suggesting that oralmedication should be offered to patients first if possible. These rec-ommendations are reinforced by findings from a consumer survey inwhich respondents expressed a strong preference for oral medica-tion.8 In that survey, the consumers also indicated that they preferredanti-anxiety medications (e.g., lorazepam) over combinations ofantipsychotics and anti-anxiety agents. Among the antipsychotics,the consumers appeared to prefer an SGA over an FGA. (Results ofthree questions from the consumer survey are presented on p. 108.)

Use of combination treatment. An ideal drug for emergency treat-ment of agitation would have a wide therapeutic index, be both safeand effective at high doses, and could be used nonspecifically in allconditions. However, medications that have historically been avail-able are not well tolerated in high doses. Two strategies to deal withthis problem have been used: repeated administration or use of drugcombinations. The most common medication strategy in the PES hasbeen combining I.M. haloperidol (usually 5 mg) and lorazepam (usu-ally 2 mg) in the same injection.71 The rationale for this widely usedcombination is that haloperidol is often poorly tolerated andlorazepam helps achieve adequate sedation while reducing the sideeffects that might occur with higher doses of haloperidol. Althoughthis strategy is generally considered safe and effective, research con-cerning it is very limited. Two randomized, controlled studies72,73

found the combination more effective than the component agentsalone early in treatment; however, the results were confounded byunequal dosing.73 Differences also tended to disappear in 2–4 hours,

perhaps because of additional doses given in the interval or passage oftime. The panel in the first survey indicated that the most importantpotential benefits of a BNZ and antipsychotic combination weregreater efficacy, faster onset of action, and reduced side effects. Thelimited literature is inconclusive as to whether combinations producethese benefits, but it appears to support the advantage of using lowerdoses of each of the component medications, thus reducing the lia-bility for side effects, especially from haloperidol.74

Given their recent introduction, even fewer data are available oncombining SGAs with BNZs in management of acute agitation. Twostudies75,76 compared haloperidol plus lorazepam with risperidoneplus lorazepam and found both combinations equally efficacious.However, because the SGAs have more complex mechanisms of actionand varying side effects, the risk-benefit ratio of combinations mayvary significantly from one SGA to another. To the extent that thenewer SGAs have more benign side-effect profiles than the FGAs, therationale of adding BNZs to reduce side effects by using a lowerantipsychotic dose may not apply. To the extent that the SGAs have arapid onset of action, the rationale for using a BNZ to achieve morerapid calming is eliminated. However, tolerability at high doses hasbeen demonstrated only for ziprasidone.77 We thus asked the panel torate the safety of combining different antipsychotics with lorazepamand for which antipsychotics there might be a therapeutic advantagein adding lorazepam instead of using higher doses of the antipsychot-ic alone (i.e., where clinical benefits of the combination outweighpotential adverse effects). The panel indicated they would base thesedecisions on the differing pharmacological properties of the variousantipsychotics and that, as elsewhere, they would consider each SGAseparately, rather than viewing these agents as a homogeneous class.

Haloperidol and perphenazine. The panel felt that the widely usedcombination haloperidol plus lorazepam is likely to be more benefi-cial than haloperidol monotherapy and gave high second-line ratingsto the therapeutic benefits of combining I.M. or oral perphenazinewith lorazepam. The majority felt these combinations are safe.

Risperidone, ziprasidone, and aripiprazole. Among the SGAs, thepanel felt there was therapeutic benefit in combining oral risperidonewith lorazepam and gave high second-line ratings to combining oralaripiprazole and oral or I.M. ziprasidone with lorazepam, probablybecause these agents are less sedating and have lower hypotensive lia-bility (although the majority would not use oral aripiprazole in emer-gency settings). The majority rated these combinations generally safe.

Chlorpromazine, olanzapine, and quetiapine. The panel suggestedcaution when combining oral or I.M. chlorpromazine, I.M. olanzap-ine, or oral quetiapine with lorazepam. The panel felt that it was gen-erally safe to combine oral olanzapine with lorazepam. In terms ofbenefit, the panel did not give strong support to combining oral orI.M. olanzapine or quetiapine with a BNZ, probably because theseagents are viewed as adequately sedating without the addition of aBNZ. The experts did not believe that there was therapeutic benefitin combining I.M. or oral chlorpromazine with lorazepam and wouldavoid this combination (third line). The ratings of chlorpromazine



and I.M. olanzapine probably reflect concern about the potential forexcessive sedation and cardiorespiratory depression when these agentsare combined with I.M. BNZs. The editors note that the current sur-vey was completed before the release of information about adverseevents associated with co-administration of I.M. olanzapine and I.M.lorazepam. As noted in the current labeling for olanzapine,“Concomitant administration of intramuscular olanzapine and par-enteral benzodiazepine has not been studied and is therefore not rec-ommended. If use of intramuscular olanzapine in combination withparenteral benzodiazepines is considered, careful evaluation of clini-cal status for excessive sedation and cardiorespiratory depression isrecommended.”39 The caution about quetiapine and BNZs probablyreflects similar concerns about overlapping side-effect profiles.

Strategies when no response to initial medication. Until the intro-duction of the SGAs, only a limited number of options were availableif the first treatment tried produced an inadequate response. Weasked the panel what strategy they would use if, after 60 minutes,they had not achieved an adequate response to monotherapy with anoral or I.M. antipsychotic or to the combination of an antipsychoticand lorazepam. The experts rejected switching to a different antipsy-chotic in any of these situations (rated third line in nearly every case).Instead they recommended giving another dose of the same antipsy-chotic at the same or higher dose, giving a dose of lorazepam, or using(or repeating if one started with it) a combination of the sameantipsychotic and lorazepam. The order in which they would trythese options reflects the recommendations for combinationsdescribed above. Thus, they are more likely to add a BNZ or give adose of the combination if they started with haloperidol, ziprasidone,or risperidone, but more likely to give another dose of the antipsy-chotic alone if they started with olanzapine or quetiapine. If theybegan with a combination, they would generally repeat the samecombination again and would also consider giving a single dose ofthe same antipsychotic or lorazepam. We did not ask about combi-nations of I.M. olanzapine and lorazepam because the package insertfor I.M. olanzapine indicates this combination is not recommended.

We asked the panel in the first survey1 how long they would rec-ommend waiting before changing medication strategies (e.g., switch-ing to a different agent or agents, using a combination of agents ifthey had started with monotherapy) if a patient was not responding,assuming that the goal is to get to the point where the patient is suf-ficiently improved to be able to converse with caregivers and takeoral medication. If the initial treatment was monotherapy, theexperts recommended changing strategies after 2 or more doses ofmedication had been totally ineffective or 3 or 4 doses had been onlypartially effective. If the initial treatment involved a combination ofmedications (e.g., an antipsychotic plus a BNZ), the experts recom-mended changing strategies after 3 or more doses of the combinationhad been totally ineffective or 4 or more doses had been only par-tially effective. These ratings reflect a willingness to continue treat-ment longer when the initial treatment involved a combination of

medications, reflecting the more limited options available at thispoint. It is interesting that even with more options available, thepanel seemed to assume that issues of poor response are betteraddressed by increasing dosage than switching compounds.

Dosage and frequency. The experts’ recommendations concerningdosing levels and intervals between doses are summarized inGuideline 7 (p. 42) and are generally consistent with data from thefew fixed dose studies available as well as recommendations in thedrugs’ labeling. The following medications were considered appro-priate for use in a behavioral emergency by 90% or more of the panelat the following minimum and maximum doses: oral lorazepam (1–3mg), I.M. lorazepam (0.5–3 mg), oral or I.M. haloperidol (2.5–10mg), oral risperidone (1–3 mg), oral olanzapine (5–20 mg), I.M.olanzapine (5–10 mg), and I.M. ziprasidone (10–20 mg).

Although few dose-response studies have been done to determineoptimal doses for emergency use, a few studies have compared differ-ent doses of medication for agitation. Three studies looked at dosingof haloperidol.78–80 Baldessarini et al.81 subsequently combined theresults of these studies and produced a dose-response curve. Theirresults suggest that a single dose of 7.5–10 mg of haloperidol mightbe expected to produce the most benefit possible with fewest sideeffects, and that higher doses associated with an increased incidenceof side effects are not likely to produce much additional benefit.These findings are consistent with our panel’s recommendations.

Brier et al.82 compared I.M. olanzapine 2.5, 5.0, 7.5, or 10.0 mgwith haloperidol 7.5 mg and placebo and found the 10 mg dosemost effective and equivalent to haloperidol. However, this study didnot indicate whether a higher dose might be more effective thanhaloperidol or whether it would have sufficient tolerability. Baker etal.83 compared a new approach to olanzapine dosing involving rapidinitial dose escalation (RIDE) up to 40 mg/day with standard dosingof olanzapine plus lorazepam and found that the RIDE strategy mayoffer superior efficacy in rapidly and effectively controlling agitationwithout causing excessive sedation. Clinicians are always seeking tostrike a compromise between effectiveness and tolerability. Since dataare limited and there is no way to predict the best dose for a specificpatient, clinicians have historically begun with a moderate dose.

There is some controversy in the literature, which is very limited,as to the most appropriate dose of BNZs with which to initiate treat-ment for agitation in a behavioral emergency, although most pub-lished studies have used a dose of 2.0 mg lorazepam. Bienek et al.73

discussed the use of a higher initial dose of 3–4 mg.

Agents Not Available in the United States

We did not include agents not available in the United States in thissurvey. Regional differences in management of agitation are quitedramatic even in the United States, where there are distinct clinicalcultures that favor different pharmacologic approaches to agitation,and this is even more true transnationally. A number of issues under-



lie this variation in practice, one of the most contentious of which isthe appropriate endpoint of the intervention. If one considers U.S.expert consensus, as presented in our 2001 guidelines1 and the cur-rent publication, and the NICE guidelines from the UnitedKingdom,5 it might appear that among large blocks of English-speaking psychiatrists, there is a consensus on goals (i.e., calmingwithout sedation) so as to permit patient participation in assessmentand care decisions. However, this view is not universal. It derivesfrom some tolerance for disturbed behavior and the value attached toa collaborative rather than authoritarian physician-patient relation-ship. However, published guidelines in some countries, the contin-ued use of sleep as an endpoint in some studies,84 the continuedpopularity of “cocktails,” and the unpublished opinions of manypractitioners reflect a preference for sleep as an endpoint. With thisin mind, there are a large number of sedatives available with a greatdeal of history and considerable anecdotal but very limited experi-mental support. Medications are also regulated and funded very dif-ferently around the world and economic reasons can lead to greatvariation in the medications available and preferred in differentregions. Levomepromazine, a phenothiazine not used by U.S. psy-chiatrists, is popular in Japan for its sedative effects.85 Loxapine ispopular in France.86 Chlorpromazine still has proponents.Barbiturates, despite their safety limitations, are still used in someareas. Promethazine is combined with haloperidol in Brazil andIndia.87 Zuclopenthixol, available as both an acute and a relativelyshort-acting depot preparation, is promoted in some Europeancountries although the depot formulation is not thought to beappropriate for acute agitation according to NICE. Flunitrazepam isa BNZ that was marketed in the United States as Rohypnol (knownhere as “roofies”).88 It is still used elsewhere but was removed fromthe U.S. market amid reports of date rape.89

To establish a more global consensus on best practices for themanagement of behavioral emergencies will require more conver-gence on the relative value of participation versus sedation and, asHuf et al.87 noted in a recent Cochrane Review, cost will still imposesignificant differences between developed and developing countries.

Transition to Oral Continuation Medication

Nearly all the experts (> 90%) agreed that the ideal medication formanaging a behavioral emergency may not be ideal for continuingcare and that there are circumstances when it may be advantageousto switch to an oral formulation of a different antipsychotic for long-term treatment. Over 80% of the panel did not believe that there isany significant disadvantage to switching within 24 hours to an oralformulation of a different antipsychotic; however, 44% of the panelindicated that there may be some minor disadvantages in makingsuch a switch. The panel indicated that the most important factorsthat would lead them to switch to a different medication for ongo-ing treatment (from one that had successfully managed a patient’sagitation early in an episode of care) were that the initial medication

was unlikely to be effective for long-term symptomatic treatment orwas likely to cause significant EPS or if the patient expressed a pref-erence for a different medication. They would also consider thepotential for substance abuse or side effects (e.g., weight gain, seda-tion, nausea, or dizziness) in deciding whether to switch to a differ-ent agent.

We asked the experts which agent they would use if continuationtreatment with an oral antipsychotic is indicated following initialtreatment with an I.M. antipsychotic. If the initial medication wasI.M. haloperidol (or perphenazine or droperidol), the experts rec-ommended using oral risperidone, with one of the other SGAs ahigh second-line option. If the initial medication was I.M. ziprasi-done or olanzapine, the experts’ treatment of choice was to continuewith the same agent in an oral formulation. Switching from I.M.olanzapine to oral risperidone was another first-line option.

Research Findings Concerning Choice of Medications

First generation antipsychotics. The literature provides little evidenceof differential effectiveness among the FGAs that is not accounted forby dosage levels or pharmacokinetics. The majority of studies havebeen done with haloperidol,78,79 although some studies that havelooked at other antipsychotics (e.g., thiothixene,90 molindone,91

loxapine92,93) have found comparable effectiveness with haloperidol.Although chlorpromazine is often mentioned for behavioral emergen-cies because of its sedative side effects, haloperidol has been foundsuperior at usual doses.80,94

Second generation antipsychotics. Although SGAs are recommend-ed as first-line agents for treatment of schizophrenia and bipolar dis-order,15,22,95,96 they have not up to now been as widely used as FGAsin emergency settings. Their advantages in schizophrenia have beenthought to be related to differences in their relative activity at sero-tonin and dopamine receptors. One of the putative advantages of theSGAs has been thought to be reductions in hostility relative to theFGAs. For example, in the 3-year Intercontinental SchizophreniaOutpatient Health Outcomes (IC-SOHO) study of 7,655 schizo-phrenia patients who initiated or changed to monotherapy withclozapine, olanzapine, quetiapine, risperidone, or haloperidol, theresearchers assessed the presence or absence of verbal or physical hos-tility/aggression retrospectively for the 6 months before enrollmentand prospectively in the first 6 months of the study.97 They foundthat hostile/aggressive behavior was reduced during treatment, witholanzapine and risperidone significantly superior to haloperidol andclozapine in reducing aggression, even when results were adjusted forbaseline imbalances in age, gender, age of onset, and substance abuse.However, it is not clear that this effect is relevant to conditions otherthan schizophrenia or in the time frame of an emergency, and acuteserotonin antagonism could be problematic. The relative tolerabilityof the SGAs is also dependent on doses and titration schedules thatare not pertinent to emergency settings. Until very recently, none of



the SGAs was available in an I.M. formulation. Unlike studies ofFGAs and other agents which were performed in heterogeneous pop-ulations using a “pain and fever” approach, recent studies of I.M.antipsychotics have examined agitation in particular conditions, thuslimiting their generalizability.

Oral risperidone. A study compared the relative efficacy, safety, andtolerability of oral risperidone plus oral lorazepam (2 mg of each givento patients willing to take oral medication) with 5 mg of I.M.haloperidol plus 2 mg of I.M. lorazepam (given to patients whorefused oral medication).75 The subjects were a convenience sample of60 psychotic patients (30 in each treatment group) admitted to a largePES who required emergency medication to control agitation and/orviolence. Both treatment groups showed improvement over time, withno significant differences between the groups. One patient in thehaloperidol group developed a dystonic reaction; there were noadverse reactions in the risperidone group. In a more recent random-ized study conducted at 24 sites in the United States, Currier et al.76

compared the same two combinations and doses in 162 patientsexhibiting agitation associated with active psychosis and found that asingle dose of oral risperidone plus lorazepam (n = 83) was as effectiveas I.M. haloperidol plus lorazepam (n = 79) for rapid control of agita-tion and psychosis and that both treatments were well tolerated. Thus,these studies indicate that oral treatment with risperidone pluslorazepam is a tolerable and comparable alternative to the traditionalcombination of I.M. haloperidol and lorazepam for short-term treat-ment of agitated psychotic patients in the emergency setting who arewilling to take oral medications.

I.M. ziprasidone. Several studies have looked at the use of I.M.ziprasidone in hospital settings. In a randomized, open-label study,Brook et al.98 compared the efficacy and tolerability of I.M. ziprasi-done (n = 90) and I.M. haloperidol (n = 42) for 3 days followed byoral treatment to day 7 in hospitalized patients with acute psychoticagitation. I.M. ziprasidone was significantly more effective in reduc-ing the symptoms of acute psychosis and was associated with a lowerincidence of EPS than I.M. haloperidol. The transition from I.M. tooral ziprasidone was effective and well tolerated. A post hoc analysisof data from the Brook et al. study demonstrated that ziprasidone orhaloperidol combined with lorazepam were similarly effective butziprasidone alone was superior to haloperidol alone.99 In a 24-hour,double-blind, fixed dose trial, Lesem et al.100 evaluated the efficacyand tolerability of I.M. ziprasidone in inpatients with psychosis andacute agitation. Patients received up to 4 injections (q 2 hr p.r.n.) of2 mg (n = 54) or 10 mg (n = 63) of I.M. ziprasidone, with 2 mg usedas a control dose. I.M. ziprasidone 10 mg rapidly reduced acute agi-tation and was significantly more effective than the 2 mg dose.Patients were calmed but not excessively sedated and no acute dys-tonia or behavioral disinhibition was reported; one patient in the 10mg group developed akathisia. Daniel et al.101 performed a very sim-ilar 24-hour, double-blind, fixed dose study comparing I.M. ziprasi-

done 2 mg (n = 38) and 20 mg (n = 41) in the acute management ofagitated psychotic patients. The 20 mg dose of I.M. ziprasidone sub-stantially and significantly reduced symptoms of acute agitation inpatients with psychotic disorders, was well tolerated, and producedno dystonia or akathisia.

It is usually not possible to perform controlled trials in the type ofseverely agitated patients seen in emergency settings because of theneed to obtain informed consent. However, Preval et al.102 recentlyexamined use of I.M. ziprasidone (n = 77) in a naturalistic study ina PES. Subjects were all sufficiently agitated to require restraint asopposed to the majority of previous clinical trial subjects. Theyfound that I.M. ziprasidone was well tolerated and appeared effectivein decreasing agitation in patients with severe agitation, includingagitation associated with alcohol or substance intoxication.

A review of clinical experience based on the first 2 years of post-marketing experience with I.M. ziprasidone was recently pub-lished.103 As of July 2005, there had been no published reports ofsudden cardiac death or death from any cause associated with ziprasi-done and changes in QTc interval associated with I.M. ziprasidoneat peak serum concentrations were modest and comparable to thoseseen with I.M. haloperidol. One case of sudden cardiac death waspublished in March 2005,104 which was not considered attributableto drug treatment. Serial EKGs for this patient showed some increas-es in QTc interval but always within normal limits (baseline of 377ms, highest level during treatment 429 ms). The patient also hadmultiple cardiac risk factors (low high density lipoprotein, over-weight, heavy smoker) and autopsy revealed significant occlusion ofall three coronary arteries. There is also one case report in press105 ofa brief, asymptomatic episode of torsades de pointe associated withziprasidone in a patient with multiple medical problems.

I.M. olanzapine. A number of studies have been done with I.M.olanzapine. In a double-blind, randomized, 24-hour study, Meehan etal.36 compared the efficacy and safety of one to three injections of I.M.olanzapine (10 mg for the first two injections, 5 mg for the third),lorazepam (2 mg for the first two injections, 1 mg for the third), andplacebo (placebo for the first two injections, olanzapine 10 mg for thethird) in 201 agitated patients with bipolar mania and found that I.M.olanzapine was safe and effective in reducing acute agitation in thesepatients. At 2 hours after the first injection, I.M. olanzapine producedsignificantly greater reduction in agitation than lorazepam or placebo;there was no statistical difference between lorazepam and olanzapine24 hours after the first injection but both were superior to placebo.Another double-blind, randomized study by Meehan et al.106 com-pared the efficacy and safety of I.M. olanzapine (2.5 or 5.0 mg), I.M.lorazepam (1.0 mg), and placebo in acutely agitated patients withdementia and found that both active treatments produced significantimprovement in agitation compared with placebo but did not differfrom placebo in sedation or adverse effects, including orthostasis. In adouble-blind, placebo-controlled, dose-response study, Brier et al.82

compared one to three injections of I.M. olanzapine (2.5, 5.0, 7.5, or



10.0 mg), haloperidol (7.5 mg), or placebo in the treatment of acuteagitation in 270 recently hospitalized patients with schizophrenia.They found that olanzapine exhibited a dose-response relationship inreducing agitation and had a lower incidence of treatment-emergentparkinsonism than I.M. haloperidol. Wright et al.107 performed a 24-hour, double-blind study comparing one to three injections of I.M.olanzapine (10 mg), I.M. haloperidol (7.5 mg), and I.M. placebo inthe treatment of acutely agitated hospitalized patients with schizo-phrenia. I.M. olanzapine reduced agitation significantly more thanI.M. haloperidol at 15, 30, and 45 minutes after the first injection;however, both I.M. olanzapine and haloperidol reduced agitation sig-nificantly more than placebo and were similar to each other at 2 and24 hours after the first injection, indicating that olanzapine starts outfaster but that haloperidol subsequently catches up. No patients in theI.M. olanzapine group experienced acute dystonia, while 7% of thosein the I.M. haloperidol group did. In a follow-up double-blind, ran-domized study, Wright et al.108 assessed the efficacy and safety of olan-zapine (n = 131) and haloperidol (n = 126) versus placebo (n = 54)during a 4-day transition from I.M. to oral therapy and found thatreductions in agitation were sustained with both study drugs through-out the oral treatment period, but patients in the haloperidol groupexperienced significantly more dystonia (4.5% vs. 0%) and akathisia(4.3% vs. 0%) than those in the olanzapine group. Although theSGAs are distinguished from the FGAs by reduced EPS, other sideeffects are also relevant. There are no published reports of vital signmeasurements with I.M. olanzapine, but an FDA Psychophar-macological Drugs Advisory Committee briefing document cites aprevalence of 11.9% for a 20 mm Hg drop in systolic blood pressurein clinical trial subjects.109 Monitoring of vital signs for orthostasis isrecommended if repeated administration of olanzapine is contemplat-ed. Concomitant use of I.M. olanzapine with BNZs has not beenstudied and is not recommended by the manufacturer.39

Safety findings based on the first year of post-marketing experiencewith I.M. olanzapine were recently presented,110 which indicated that,worldwide, patient exposure to I.M. olanzapine in the first year was278,600 with 91 spontaneous adverse events reported and 15 fatali-ties. Of the 62 adverse event reports in which patients were known tohave received concomitant medications, 38 received BNZs in proxim-ity to treatment with I.M. olanzapine. Most of the fatalities involvedmultiple concomitant medications, mainly BNZs or other antipsy-chotics. Of the 15 fatalities, 8 were treated with BNZs within 24hours of death. Other contributing factors identified in the reports offatalities were age (4/15 were � 60 years of age), overweight or obesi-ty (5/15), metabolic issues, such as diabetes, hyperlipidemia, hyper-glycemia (3/15), cardiac comorbidities (3/15), respiratory conditions(5/15), and vascular conditions (4/15).

Benzodiazepines. Studies concerning the use of BNZs in psychiatricemergencies suggest that they are at least as effective as haloperidolalone. Most of the studies have been done with lorazepam,72–74,111–113

but controlled data have also been published concerning midazo-

lam,114,115 clonazepam,116 and flunitrazepam.88 Studies comparing 5mg of haloperidol with 2 mg of lorazepam found that the 2 agentswere equal on some measures,72,74,112 but that 2 mg of lorazepam wassuperior on measures of aggression74 and clinical global improve-ment.112 Flunitrazepam 1 mg was compared with haloperidol 5 mgand found to be superior using the Overt Aggression Scale as a meas-ure of outcome.88 Midazolam 5 mg was reported to be superior tohaloperidol 10 mg in its effect on a measure of motor agitation.114

Battaglia et al.72 found lorazepam used alone to be more sedating thanhaloperidol used alone. A recent randomized, prospective double-blind study by Nobay et al.115 compared midazolam with lorazepamand haloperidol in the management of 111 violent and severely agi-tated patients in the emergency department and found that midazo-lam had a significantly shorter time to onset of sedation and a morerapid time to arousal than lorazepam or haloperidol, with all threedrugs appearing to have similar efficacy. These studies suggest thatBNZs used at doses typical in emergency settings may be more effec-tive than haloperidol.

Selecting Medications in Special Situations

Medical comorbidity. The panel gave first-line ratings to lorazepamfor patients with cardiac arrhythmias or conduction defects, to risperi-done for patients with delirium or dementia, to lorazepam, haloperi-dol, or ziprasidone for patients with diabetes or hyperglycemia, andlorazepam for obese patients. They did not give first-line ratings to anyagent for patients with alcohol intoxication, but recommended usinga BNZ rather than an antipsychotic if a medication is needed, proba-bly reflecting concern about withdrawal syndromes and the risk ofseizures. Note that BNZs may be initiated even while alcohol is stillpresent in the patient’s system. Although concerns have been raised ontheoretical grounds about the risk of respiratory depression whenBNZs are used in combination with alcohol or other sedatives andabout the possibility of behavioral disinhibition with BNZs, theseconcerns are not reflected in the high ratings the BNZs generallyreceived throughout the survey nor are they supported by researchfindings.117 The experts did not recommend the use of droperidol orchlorpromazine as treatments for behavioral emergencies and thusthey received third-line ratings in every situation, even when theseagents would not be expected to pose any particular safety risks. Inkeeping with findings concerning the different agents’ adverse-effectprofiles, the experts would avoid olanzapine in patients who have dia-betes or hyperglycemia or who are obese and would avoid ziprasidonein patients with cardiac arrhythmias or a conduction defect. Theresults of the CATIE study,62 which compared olanzapine, per-phenazine, quetiapine, risperidone, and ziprasidone, were publishedafter the survey was completed. The CATIE findings confirmed thatolanzapine was associated with greater weight gain and increases inmeasures of glucose and lipids than other agents in the study but didnot find any QTc problems with ziprasidone. There was little supportfor using BNZs in patients with delirium or dementia, probably



because of concern about sedation, gait problems, and the possibilityof falls. Risperidone is preferred for delirious patients, perhaps becausesome of the other newer antipsychotics have anticholinergic propertiesthat might increase confusion and cause excessive sedation.

Elderly patients. The panel preferred to avoid antipsychotics or BNZsin frail elderly patients. If medication is needed, risperidone, haloperi-dol, and olanzapine were preferred (high second line). The preferencefor risperidone and olanzapine may reflect the larger number of stud-ies done with these agents in geriatric populations (in agitated patientswith dementia and elderly patients with psychotic disorders) com-pared with other newer SGAs as well as the recently published ExpertConsensus Guidelines on use of antipsychotics in elderly patients23

and patients with dementia.25 SGAs may be associated with anincreased risk of death (due to cardiovascular or infectious causes) inchronic treatment of elderly patients with dementia-related psychosisand should be used cautiously in elderly patients with dementia or atrisk for cerebrovascular disease,37–41 although recent studies indicatethat this risk appears no higher than with FGAs.43,44

Children. The panel would prefer to avoid using antipsychotics orBNZs in children. However, if medication is needed for a child, risperi-done and lorazepam are preferred (high second line), probably reflect-ing the larger number of studies with risperidone in this population.118

Mental retardation. If the patient has mental retardation, risperidoneis the preferred agent, reflecting the fact that the largest number ofstudies with antipsychotics in this population have been done with thisdrug.119,120 The preference for SGAs in patients with mental retarda-tion agrees with recommendations in the Expert Consensus Guidelineson the Treatment of Psychiatric and Behavioral Problems in MentalRetardation, in which SGAs were strongly preferred over FGAs fortreatment of agitation, aggression, or self-injurious behavior.18,120

Pregnancy. We did not repeat questions about pregnant women in thecurrent survey. In 2001, when asked about medication strategies for apregnant woman who is agitated, psychotic, and unresponsive to direc-tion and for whom immediate medical intervention is judged necessaryto prevent harm to the mother or fetus or to reduce the deleteriouseffects that the stress of agitation may have on the maternal-fetal sys-tem, the panel preferred a high-potency FGA (rated first line by 76%),1

probably reflecting the much larger database concerning and lack ofteratogenicity reported for high-potency FGAs.121 These recommenda-tions agree with those for treating psychotic depression in pregnantwomen in the Expert Consensus Guidelines on the Treatment ofDepression in Women 2001.19 While the FDA rates the FGAs and SGAssimilarly in their Use-in-Pregnancy ratings (category C: “risk cannot beruled out”), the 2001 expert panel was less willing to endorse the useof SGAs, presumably because of less experience with these agents inpregnant women. For a discussion of psychotropic agents in pregnan-cy, readers are referred to a recent review by Jain and Lacy.122

Other complicating problems. We asked the panel which medicationswere most appropriate for an agitated aggressive patient with a historyof various complicating problems. Their recommendations reflect find-ings concerning the adverse-effect profiles of the different agents. Theywould avoid using BNZs if possible in patients with a history of drugabuse or dependence, while BNZs are preferred for patients with a his-tory of seizures (e.g., because of substance or alcohol abuse). If a patienthas a history of akathisia or other EPS, tardive dyskinesia, or neu-roleptic malignant syndrome, the panel recommended choosing agentswith the least liability for causing EPS (e.g., a BNZ, an SGA other thanrisperidone) and would generally avoid antipsychotics that are morelikely to cause EPS (e.g., risperidone, especially at higher doses, andhigh potency FGAs). If a patient has a history of amenorrhea or galac-torrhea, the experts would choose a prolactin-sparing medication.

Summary of Medication Recommendations

Although the differences are slight and in many cases not statisticallysignificant, it is of some interest to compare the available drugs, par-ticularly in terms of changes since the last survey.

Benzodiazepines alone are still favored in situations when there is noinformation, for which no specific treatment is available (e.g., person-ality disorders), or when they may have specific benefits (e.g., intoxi-cation). BNZs are not recommended for use alone in psychoticconditions.

Haloperidol. Although haloperidol alone is usually thought to beappropriate, in the current survey it received second-line ratings foragitation in several circumstances. Its use alone is favored over otherchoices only in medically compromised patients and, even in that con-text, risperidone received very similar ratings. The traditional combi-nation of haloperidol plus a BNZ was never the most highly ratedalternative, but it was a first-line option for schizophrenia, mania, andpsychosis NOS (although not for psychotic depression).

Risperidone. Among the SGAs, risperidone, alone or combined with aBNZ, received strong support and was the second or third choice inalmost every situation. In the current survey, risperidone alone receivedhigher ratings than haloperidol alone except for the medically compro-mised patient, and the combination of risperidone plus a BNZ receivedhigher ratings than haloperidol plus a BNZ for acute treatment of agi-tation related to primary psychiatric disturbances. For example, formania, risperidone plus a BNZ was rated first line by the highest per-centage of experts (80%), although the combination of divalproex andan antipsychotic was most frequently rated treatment of choice (42%).Oral risperidone is also recommended following parenteral use ofhaloperidol, perphenazine, or droperidol. Taken together, the ratingssuggest that risperidone may now be viewed somewhat as haloperidolwas in the past—as a safe utility drug, alone or combined with a BNZ,in many situations.



Oral olanzapine, consistent with findings in other settings, receivedhigh ratings for agitation in a variety of conditions. It generally receivedratings very similarly to risperidone, with slightly higher ratings inareas where it has been tested (e.g., schizophrenia, mania) and slightlylower ratings in situations where it has not been studied and safety maybe an issue. Even before publicity concerning potential problems withcombinations of olanzapine plus a BNZ, such combinations receivedless support. Thus, even when olanzapine alone is top rated, olanzap-ine plus a BNZ was rated second line.

Oral ziprasidone and quetiapine generally received similar second-lineratings. Quetiapine received the strongest support in personality disor-ders, where it was ranked fourth. The combination of oral ziprasidoneplus a BNZ sometimes outranked ziprasidone alone (e.g., in mania); incontrast, the combination of quetiapine plus a BNZ was consistentlyrated lower than monotherapy with quetiapine.

Parenteral formulations. The biggest change since the first survey wasthe advent of new parenteral agents. In primary psychiatric conditions(schizophrenia, mania, psychotic depression, personality disorder),I.M. olanzapine alone and I.M. ziprasidone alone were more highlyrated than I.M. haloperidol alone. I.M. olanzapine received slightlymore support than I.M. ziprasidone, even though the latter has beenavailable longer. I.M. olanzapine was first line and top rated in schizo-phrenia, mania, and psychotic depression. I.M. ziprasidone alone wasgenerally rated slightly higher than I.M. ziprasidone plus a BNZ, butboth are first line in schizophrenia. The experts made a greater distinc-tion between using I.M. olanzapine alone and combined with a BNZ.Thus, although I.M. olanzapine alone usually received somewhat high-er ratings than I.M. ziprasidone, I.M. ziprasidone plus a BNZ was gen-erally rated more highly than I.M. olanzapine plus a BNZ. Neither ofthe new parenteral formulations received as much support as tradi-tional agents (I.M. haloperidol, I.M. BNZ) when there is no diagnosisor the diagnosis involves a medical condition or intoxication.

Other agents. As in the first survey, chlorpromazine received consis-tently low ratings from the current panel. Although there is controver-sy over its role, droperidol was also rated third line throughout. Positiveresults for perphenazine in the CATIE trial62 had not been publishedat the time of the survey and perphenazine generally received lower rat-ings than SGAs except aripiprazole. Data concerning I.M. aripiprazolemay influence perceptions of this agent in the future.

Use of Restraints

The concept of chemical restraint. According to the CMS interimfinal rules,55,70 the distinction between what is considered a chemi-cal restraint and a treatment appears to hinge on whether medicationis given as part of a plan of care or merely to control the patient’sbehavior. The panel in the first survey indicated that a brief assess-ment leading to determination of a general category of presentation

(e.g., intoxication, psychosis) is usually adequate to create a plan ofcare (a more comprehensive assessment leading to a specific diagno-sis was also supported but may be impractical for various reasons).

Use of physical restraints. The current survey did not repeat ques-tions about the use of restraints, since no significant change inresponses was anticipated. The 2001 panel considered use of physi-cal restraints a last resort, agreeing with the CMS interim rules,70

which specify that use of restraints for “managing behavioral emer-gencies is allowed only when all less restrictive measures have failedand unanticipated severely aggressive or destructive behavior placesthe patient or others in imminent danger…” They consideredrestraints extremely or usually appropriate when patients pose anacute danger to other patients, bystanders, staff, or themselves andsometimes appropriate to prevent an involuntary patient from leav-ing prior to assessment or transfer to a locked facility, but did notconsider use of physical restraints appropriate in other situations,such as a patient who has a history of previous self-injury or aggres-sion but does not appear to pose any immediate risk at the moment,when adequate resources are not available to supervise the patientadequately, to maintain an orderly treatment environment, or to pre-vent a voluntary patient from leaving prior to assessment.

Use of medications in conjunction with restraints. If a patientbecomes calmer when put in restraints, the experts were dividedbetween using no medication or only oral medication, but did notgenerally recommend parenteral medication in this situation.However, if a patient continues to be violent and extremely agitatedwhile in restraints, the experts strongly supported the use of par-enteral medication in combination with the restraints, but wouldalso consider oral medication in this situation. They did not consid-er it appropriate to leave such a patient unmedicated in restraints.Overall, the responses indicate that the goal in this situation is to usemedication to reduce time in and complications of restraints.

Feasibility of reducing use of restraints. In the current survey, wedid ask the panel if they consider elimination of the use of physicalrestraints a feasible goal. Only 36% of the experts believed thatrestraint is a form of treatment. These respondents agreed that it maybe overutilized but did not feel that it should be eliminated. Another40% did not consider restraint a form of treatment but believed thatits advantages as a security measure outweigh its disadvantages. Theremaining 24% of the experts felt that the disadvantages of restraintoutweigh the advantages, but were divided as to whether it would bepossible to reduce or eliminate the use of restraint in their treatmentsettings.

Minimizing Negative Impact and Preventing Future Crises

The majority of the experts felt that involuntary treatment is likelyto have a negative impact on a patient’s willingness to seek care in the



future, although a third did not think it would have an effect, andabout 10% felt that it would have a positive effect. These responsesreinforce the results of a consumer survey which found that con-sumers of emergency care put great emphasis on the importance ofbeing allowed to express a preference concerning their treatment.8

We asked the experts which interventions were most importantfor staff to undertake after a crisis but before the person leaves theemergency setting to minimize negative impact on the therapeuticalliance or the person’s willingness to seek out mental health care inthe future, and more than 60% said is is essential to encouragepatients to discuss the experience and ask questions, provide medicalmonitoring as needed, and give patients (and family) informationabout the medication the person is being discharged on and its sideeffects. These ratings reflect findings that emergency interventionsare less likely to be effective and promote positive long-term out-comes if patients do not understand what happened and why theyreceived the treatment they were given. We then asked about inter-ventions to reduce risk of future psychiatric crises. More than 80%of the panel said it is essential to arrange an outpatient follow-upvisit before discharge. This reflects findings that merely telling theperson to call for an appointment is ineffective and that the best out-comes are obtained when an appointment soon after the emergencyintervention is scheduled for the patient before discharge.123

CONCLUSIONS AND DIRECTIONS FOR FUTURERESEARCH

Even as tremendous strides are made in the treatment of psychiatricillness, behavioral emergencies continue to pose a challenge becausethey tend to occur in the context of incomplete information. Thisremains a difficult and controversial area of practice because itinvolves limitations on patient autonomy and control, although it ishoped that better practices will lead to improvement in this area.This guideline, like the previous one, is dedicated to a new climateof increased respect and an effort to move from control to care.

Other core problems in a behavioral emergency are the perceivedneed to do something immediately and the lack of agreementbetween the individual at the center of the emergency and thoseresponsible for managing it. This guideline is an effort to reach agree-ment among providers. But how do we facilitate agreement betweenpatients and providers? Strategies with a narrow focus on the techni-cal issues that determine short-term outcome may do so at the costof relationship issues that influence long-term outcome. Althoughthe attitudes and behaviors that foster autonomy and respect are dif-ficult to incorporate into guidelines, it is important to take intoaccount available information concerning consumer preferences inchoosing the most appropriate interventions.8

Available data suggest that consumers understand the need foremergency interventions but often feel frightened and abandoned inthe midst of them. In this respect, recent regulations that stress con-tinuing contact with the patient during the episode and debriefing

afterward may bring improvement in this area. The inclusion of con-sumer perspectives in the training of providers should help to sensi-tize providers and peer counselors. Advocates and families can alsogive providers proxy data that might help avoid or shorten episodes.

LIMITATIONS AND ADVANTAGES OF EXPERTCONSENSUS GUIDELINES

These guidelines can be viewed as an expert consultation, to beweighed in conjunction with other information and in the contextof each individual patient-physician relationship. The recommenda-tions do not replace clinical judgment, which must be tailored to theparticular needs of each clinical situation. We describe groups ofpatients and make suggestions intended to apply to the averagepatient in each group. However, individual patients will differ great-ly in their treatment preferences and capacities, history of responseto previous treatments, family history of treatment response, and tol-erance for different side effects. Therefore, the experts’ first-line rec-ommendations certainly will not be appropriate in all circumstances.

We remind readers of several other limitations of these guidelines:1. The guidelines are based on a synthesis of the opinions of a large

group of experts. From question to question, some of the indi-vidual experts would differ with the consensus view.

2. We have relied on expert opinion precisely because we are askingcrucial questions that are not yet well answered by the literature.One thing that the history of medicine teaches us is that expertopinion at any given time can be very wrong. Accumulatingresearch will ultimately reveal better and clearer answers.Clinicians should therefore stay abreast of the literature for devel-opments that would make at least some of our recommendationsobsolete. We hope to revise the guidelines periodically based onnew research information and on reassessment of expert opinionto keep them up-to-date.

3. The guidelines are financially sponsored by the pharmaceuticalindustry, which could possibly introduce biases. Because of this,we have made every step in guideline development transparent,reported all results, and taken little or no editorial liberty.

4. These guidelines are comprehensive but not exhaustive; becauseof the nature of our method, we omit some interesting topics onwhich we did not query the expert panel.

Despite the limitations, these guidelines represent a significantadvance because of their specificity, ease of use, and the credibilitythat comes from achieving a very high response rate from a largesample of the leading experts in the field.

FINAL WORD

Advances in public health do not always require technological break-throughs or long periods of waiting for new data. Immediate gainscan be made by increasing the speed with which best practices areimplemented. Guidelines offer a rapid means for communicating a



distillate of expert opinion. When reaching a clinical decision point,practitioners and patients can use guidelines to generate a menu ofreasonable choices and then select the option that is judged best foreach individual. This process drives the next round of expert opinionand the next round of empirical studies.

References

1. Allen MH, Currier GW, Hughes DH, et al. The expert consensus guidelineseries: Treatment of behavioral emergencies. Postgrad Med Special Report2001; May:1–88.

2. Battaglia J. Pharmacological management of acute agitation. Drugs2005;65:1207–22.

3. Lindenmayer JP. The pathophysiology of agitation. J Clin Psychiatry. 2000;61(Suppl 14):5–10.

4. Laughren T. Regulatory issues on behavioral and psychological symptoms ofdementia in the United States. International Psychogeriatrics 2000;12(Suppl1):331–6; discussion 337–40.

5. National Institute for Health and Clinical Excellence (NICE). ClinicalGuideline 1: Schizophrenia: Core interventions in the treatment and man-agement of schizophrenia in primary and secondary care. London: NationalInstitute for Clinical Excellence; December 2002 (www.nice.org.uk).

6. De Fruyt J, Demyttenaere K. Rapid tranquilization: New approaches in theemergency treatment of behavioral disturbances. Eur Psychiatry 2004;19:243–9.

7. Lukens TW, Wolf SJ, Edlow JA, et al. The American College of EmergencyPhysicians Clinical Policies Subcommittee on Critical Issues in the Diagnosisand Management of the Adult Psychiatric Patient in the EmergencyDepartment. Clinical policy: Critical issues in the diagnosis and managementof the adult psychiatric patient in the emergency department. Ann EmergMed, in press.

8. Allen MH, Carpenter D, Sheets JL, et al. What do consumers say they wantand need during a psychiatric emergency? J Psychiatr Pract 2003;9:39–58.

9. Djulbegovic B, Hadley T. Evaluating the quality of clinical guidelines:Linking decisions to medical evidence. Oncology 1998;12:310–4.

10. Shekelle PG, Kahan JP, Bernstein SJ, et al. The reproducibility of a methodto identify the overuse and underuse of medical procedures. N Engl J Med1998;338:1888–95.

11. Kahn DA, Carpenter D, Docherty JP, et al. The expert consensus guidelineseries: Treatment of bipolar disorder. J Clin Psychiatry 1996;57(suppl12a):1–88.

12. McEvoy JP, Weiden PJ, Smith TE, et al. The expert consensus guideline series:Treatment of schizophrenia. J Clin Psychiatry 1996;57(suppl 12b):1–58.

13. March JS, Frances A, Carpenter D, et al. The expert consensus guidelineseries: Treatment of obsessive-compulsive disorder. J Clin Psychiatry1997;58(suppl 4):1–72.

14. Alexopoulos GS, Silver JM, Kahn DA, et al. The expert consensus guidelineseries: Treatment of agitation in older persons with dementia. Postgrad MedSpecial Report 1998; April:1–88.

15. McEvoy JP, Scheifler PL, Frances A. The expert consensus guideline series:Treatment of schizophrenia 1999. J Clin Psychiatry 1999;60(suppl 11):1–80.

16. Foa EB, Davidson JRT, Frances A. The expert consensus guideline series:Treatment of posttraumatic stress disorder. J Clin Psychiatry 1999;60(suppl16):1–76.

17. Sachs GS, Printz DJ, Kahn DA, et al. The expert consensus guideline series:Medication treatment of bipolar disorder 2000. Postgrad Med Special Report2000;April:1–104.

18. Rush AJ, Frances A. Expert consensus guideline series: Treatment of psychi-atric and behavioral problems in mental retardation. AJMR 2000;105(3):159–228.

19. Altshuler LL, Cohen LS, Moline ML, et al, eds. The expert consensus guide-line series: Treatment of depression in women 2001. Postgrad Med SpecialReport 2001;March:1–116.

20. Conners CK, March JS, Frances A, et al. The expert consensus guidelineseries: Treatment of attention-deficit/hyperactivity disorder. J Atten Disord2001;4(suppl 1):S1–S128.

21. Alexopoulos GS, Katz IR, Reynolds CF, et al. The expert consensus guidelineseries: Treatment of depressive disorders in older patients. Postgrad MedSpecial Report 2001;October: 1–86.

22. Kane JM, Leucht S, Carpenter D, et al. The expert consensus guideline series:Optimizing pharmacologic treatment of psychotic disorders. J Clin Psychiatry2003;64(suppl 12):1–100.

23. Alexopoulos GS, Streim J, Carpenter D, et al. The expert consensus guidelineseries: Using antipsychotic agents in older patients. J Clin Psychiatry2004;65(suppl 2):1–105.

24. Keck PE, Perlis RH, Otto MW, et al. The expert consensus guideline series:Treatment of bipolar disorder 2004. Postgrad Med Special Report2004;December:1–120.

25. Alexopoulos GS, Jeste DV, Chung H, et al. The expert consensus guidelineseries: Treatment of dementia and its behavioral disturbances. Postgrad MedSpecial Report 2005;January: 1–111.

26. Kahn DA, Docherty JP, Carpenter D, et al. Consensus methods in practiceguideline development: A review and description of a new method.Psychopharmacol Bull 1997;33:631–9.

27. Brook RH, Chassin MR, Fink A, et al. A method for the detailed assessmentof the appropriateness of medical technologies. Int J Tech Assess Health Care1986;2:53–63.

28. Corrigan JD. Development of a scale for assessment of agitation followingtraumatic brain injury. J Clin Exp Neuropsychol 1989;11:261–77.

29. Folstein MF, Folstein SE, McHugh PR. “Mini-Mental State”: A practicalmethod for grading the cognitive state of patients for the clinician. J PsychiatrRes 1975;12:189–98.

30. Tuokko H, Hadjistavropoulos T, Miller JA, et al. The Clock Test:Administration and scoring manual. Toronto: Multi-Health Systems; 1995.

31. Inouye SK, van Dyck CH, Alessi CA, et al. Clarifying confusion: TheConfusion Assessment Method, a new method for detection of delirium. AnnIntern Med 1990;113:941–8.

32. Serper MR, Allen MH. Rapid screening for cognitive impairment in the psy-chiatric emergency service: I. Cognitive screening batteries. Psychiatr Serv2002;53:1527–9.

33. Copersino ML, Serper M, Allen MH. Emergency psychiatry: rapid screeningfor cognitive impairment in the psychiatric emergency service: II. A flexibletest strategy. Psychiatr Serv 2003;54:314–6.

34. Currier GW, Allen MH. American Association for Emergency Psychiatry sur-vey, I: Psychiatric emergency service structure and function. Presented at the51st American Psychiatric Association Institute on Psychiatric Services, NewOrleans, LA, October 29–November 2, 1999.

35. Hirschfeld RM, Allen MH, McEvoy JP, et al. Safety and tolerability of oralloading divalproex sodium in acutely manic bipolar patients. J Clin Psychiatry1999;60:815–8.

36. Meehan K, Zhang F, David S, et al. A double-blind, randomized comparisonof the efficacy and safety of intramuscular injections of olanzapine,lorazepam, or placebo in treating acutely agitated patients diagnosed withbipolar mania. J Clin Psychopharmacol 2001;21:389–97.

37. US FDA MedWatch. Dear Healthcare provider letter (April 16 2003).(www.fda.gov/medwatch/SAFETY/2003/risperdal.htm).

38. Abilify. U.S. prescribing information, updated April 2005 (www.bms.com/cgi-bin/anybin.pl?sql=select%20PPI%20from%20TB_PRODUCT_PPI%20where%20PPI_SEQ=101&key=PPI, accessed October 2005).

39. Zyprexa. U.S. prescribing information, updated May 26, 2005.(http://pi.lilly.com/us/zyprexa-pi.pdf, accessed October 2005).



40. Risperdal. U.S. prescribing information, updated April 2005. (www.risperdal.com/active/janus/en_US/assets/common/company/pi/risperdal.pdf, accessedOctober 2005).

41. Seroquel U.S. prescribing information, updated June 2005. (www.astrazeneca-us.com/pi/Seroquel.pdf, accessed October 2005).

42. Geodon U.S. prescribing information, updated May 2005. (www.pfizer.com/pfizer/download/uspi_geodon.pdf, accessed October 2005).

43. Gill SS, Rochon PA, Herrmann N, et al. Atypical antipsychotic drugs and riskof ischaemic stroke: Population based retrospective cohort study. BMJ. 2005;330:445.

44. Herrmann N, Mamdani M, Lanctot KL. Atypical antipsychotics and risk ofcerebrovascular accidents. Am J Psychiatry 2004;161:1113–5.

45. Breitbart W, Marotta R, Platt MM, et al. A double-blind trial of haloperidol,chlorpromazine, and lorazepam in the treatment of delirium in hospitalizedAIDS patients. Am J Psychiatry 1996;153:231–7.

46. Han CS, Kim YK. A double-blind trial of risperidone and haloperidol for thetreatment of delirium. Psychosomatics 2004;45:297–301.

47. Mittal D, Jimerson NA, Neely EP, et al. Risperidone in the treatment of delir-ium: Results from a prospective open-label trial. J Clin Psychiatry 2004;65:662–7.

48. Liu CY, Juang YY, Liang HY, et al. Efficacy of risperidone in treating thehyperactive symptoms of delirium. Int Clin Psychopharmacol 2004;19:165–8.

49. Parellada E, Baeza I, de Pablo J, et al. Risperidone in the treatment of patientswith delirium. J Clin Psychiatry 2004;65:348–53.

50. Skrobik YK, Bergeron N, Dumont M, et al. Olanzapine vs haloperidol:Treating delirium in a critical care setting. Intensive Care Med 2004;30:444–9.

51. Breitbart W, Tremblay A, Gibson C. An open trial of olanzapine for thetreatment of delirium in hospitalized cancer patients. Psychosomatics 2002;43:175–82.

52. van Harten PN, van Trier JCAM, Horwitz EH, et al. Cocaine as a risk factorfor neuroleptic-induced acute dystonia. J Clin Psychiatry 1998;59:128–30.

53. American Psychiatric Association. Practice guideline for the treatment ofpatients with substance use disorders: Alcohol, cocaine, opioids. Am JPsychiatry 1995;152(Suppl 11):1–59.

54. Mayo-Smith MF, Beecher LH, Fischer TL, et al. Management of alcoholwithdrawal delirium. An evidence-based practice guideline. Arch Intern Med2004;164:1405–12.

55. U.S. Department of Health and Human Services, Centers for Medicare andMedicaid Services (CMS; formerly Health Care Financing Administration).Qs and As on Hospital COP for Patients’ Rights. Baltimore: CMS, October10. 2001 (available at www.cms.hhs.gov/cop/2b1.asp).

56. Lehman AF, Lieberman JA, Dixon LB, et al. Practice guideline for the treat-ment of patients with schizophrenia, 2nd edition. Am J Psychiatry 2004;161(Suppl 2):1–56.

57. Kennedy SH, Lam RW. Enhancing outcomes in the management of treat-ment resistant depression: A focus on atypical antipsychotics. Bipolar Disord2003;5(Suppl 2):36–47.

58. Masan PS. Atypical antipsychotics in the treatment of affective symptoms: Areview. Ann Clin Psychiatry. 2004;16:3–13.

59. Grootens KP, Verkes RJ. Emerging evidence for the use of atypical antipsy-chotics in borderline personality disorder. Pharmacopsychiatry2005;38:20–3.

60. Koenigsberg HW, Reynolds D, Goodman M, et al. Risperidone in the treat-ment of schizotypal personality disorder. J Clin Psychiatry 2003;64:628–34.

61. Keshavan M, Shad M, Soloff P, et al. Efficacy and tolerability of olanzapinein the treatment of schizotypal personality disorder. Schizophr Res 2004;71:97–101.

62. Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychoticdrugs in patients with chronic schizophrenia. N Engl J Med 2005;353:1209–23.

63. Federal Drug Administration. FDA strengthens warnings for droperidol(available at www.fda.gov/bbs/topics/ANSWERS /2001/ANS01123.html)

64. Kao LW, Kirk MA, Evers SJ, et al. Droperidol, QT prolongation, and suddendeath: What is the evidence? Ann Emerg Med 2003;41:546–58.

65. Martel M, Miner J, Lashkowitz S, et al. QT prolongation and cardiacarrhythmias associated with droperidol use in critical emergency departmentpatients. Acad Emerg Med 2003;10:510–11.

66. Shale JH, Shale CM, Mastin WD. Safety of droperidol in behavioural emer-gencies. Expert Opin Drug Saf 2004;3:369–78.

67. Jacoby JL, Fulton J, Cesta M, et al. After the black box warning: Dramaticchanges in ED use of droperidol. Am J Emerg Med 2005;23:196 (letter).

68. Modell S, Wilber R, Marcus R, et al. Efficacy and safety of intramusculararipiprazole. European Psychiatry 2004; 19(Suppl 1):177–8 (abstract P104).

69, Currier GW. Atypical antipsychotic medications in the psychiatric emergencyservice. J Clin Psychiatry 2000;61(Suppl 14): 21–6.

70. U.S. Department of Health and Human Services, CMS. Hospital Conditionsof Participation for Patients’ Rights. 42CFR 482.13. Baltimore, MD: CMS;1999 (www.cms.hhs.gov /cop/2b.asp).

71. Binder RL, McNiel DE. Emergency psychiatry: Contemporary practices inmanaging acutely violent patients in 20 psychiatric emergency rooms.Psychiatr Serv 1999;50:1553–4.

72. Battaglia J, Moss S, Rush J, et al. Haloperidol, lorazepam, or both for psy-chotic agitation? A multicenter, prospective, double-blind, emergency depart-ment study. Am J Emerg Med 1997;15:335–40.

73. Bienek SA, Ownby RL, Penalver A, et al. A double-blind study of lorazepamversus the combination of haloperidol and lorazepam in managing agitation.Pharmacotherapy 1998;18:57–62.

74. Salzman C, Solomon D, Miyawaki E, et al. Parenteral lorazepam versus par-enteral haloperidol for the control of psychotic disruptive behavior. J ClinPsychiatry 1991;52:177–80.

75. Currier GW, Simpson GM. Risperidone liquid concentrate plus lorazepamversus intramuscular haloperidol plus lorazepam for treatment of psychoticagitation. J Clin Psychiatry 2001;62:153–7.

76. Currier GW, Chou JC, Feifel D, et al. Acute treatment of psychotic agitation:A randomized comparison of oral treatment with risperidone and lorazepamversus intramuscular treatment with haloperidol and lorazepam. J ClinPsychiatry 2004;65:386–94.

77. Daniel DG, Zimbroff DL, Swift RH, et al. The tolerability of intramuscularziprasidone and haloperidol treatment and the transition to oral therapy. IntClin Psychopharmacol 2004; 19:9–15.

78. Anderson WH, Kuehnle JC, Catanzano DM. Rapid treatment of acute psy-chosis. Am J Psychiatry 1976;133:1076–8.

79. Neborsky R, Janowsky D, Munson E, et al. Rapid treatment of acute psy-chotic symptoms with high- and low-dose haloperidol. Arch Gen Psychiatry1981;38:195–9.

80. Reschke RW. Parenteral haloperidol for rapid control of severe, disruptivesymptoms of acute schizophrenia. Dis Nerv Sys 1974;35:112–5.

81. Baldessarini RJ, Cohen BM, Teicher MH. Significance of neuroleptic doseand plasma level in the pharmacological treatment of psychoses. Arch GenPsychiatry 1988;45:79–91.

82. Breier A, Meehan K, Birkett M, et al. A double-blind, placebo-controlleddose-response comparison of intramuscular olanzapine and haloperidol in thetreatment of acute agitation in schizophrenia. Arch Gen Psychiatry 2002;59:441–8.

83. Baker RW, Kinon BJ, Maguire GA, et al. Effectiveness of rapid initial doseescalation of up to forty milligrams per day of oral olanzapine in acute agita-tion. J Clin Psychopharmacol 2003;23:342–8.

84. TREC Collaborative Group. Rapid tranquilisation for agitated patients inemergency psychiatric rooms: A randomised trial of midazolam versushaloperidol plus promethazine. BMJ 2003;327:708–13.

85. Hatta K. Emergency pharmacological management of aggression among agi-



tated psychotic patients in Japan. Psychiatr Serv 2002;53:1077–8.86. Bourdinaud V, Pochard F. [Survey of management methods for patients in a

state of agitation at admission and emergency departments in France] [Articlein French]. Encephale 2003;29:89–98.

87. Huf G, Alexander J, Allen MH. Haloperidol plus promethazine for psychosisinduced aggression. Cochrane Database Syst Rev 2005 Jan 25;(1):CD005146.

88. Dorevitch A, Katz N, Zemishlany Z, et al. Intramuscular flunitrazepam ver-sus intramuscular haloperidol in the emergency treatment of aggressive psy-chotic behavior. Am J Psychiatry 1999;156:142–4.

89. Rickert VI, Wiemann CM. Date rape among adolescents and young adults. JPediatr Adolesc Gynecol 1998;11:167–75.

90. Stotsky BA. Relative efficacy of parenteral haloperidol and thiothixene for theemergency treatment of acutely excited and agitated patients, Dis Nerv Syst1977;38:967–73.

91. Binder R, Glick I, Rice M. A comparative study of parenteral molindone andhaloperidol in the acutely psychotic patient. J Clin Psychiatry 1981;42: 203–6.

92. Paprocki J, Versiani M. A double-blind comparison between loxapine andhaloperidol by parenteral route in acute schizophrenia. Curr Ther Res ClinExp 1977;21:80–100.

93. Tuason VB. A comparison of parenteral loxapine and haloperidol in hostileand aggressive acutely schizophrenic patients. J Clin Psychiatry 1986;47:126–9.

94. Gerstenzang ML, Krulisky TV. Parenteral haloperidol in psychiatric emer-gencies: Double-blind comparison with chlorpromazine. Dis Nerv Syst 1977;38:581–3.

95. Davis JM, Chen N, Glick ID. A meta-analysis of the efficacy of second-gen-eration antipsychotics. Arch Gen Psychiatry 2003;60:553–64.

96. Suppes T, Dennehy EB, Hirschfeld RM, et al. The Texas implementation ofmedication algorithms: Update to the algorithms for treatment of bipolar Idisorder. J Clin Psychiatry 2005; 66:870–86.

97. Bitter I, Czobor P, Dossenbach M, Volavka J. Effectiveness of clozapine, olan-zapine, quetiapine, risperidone, and haloperidol monotherapy in reducinghostile and aggressive behavior in outpatients treated for schizophrenia: Aprospective naturalistic study (IC-SOHO). Eur Psychiatry 2005;20:403–8.

98. Brook S, Lucey JV, Gunn KP. Intramuscular ziprasidone compared with intra-muscular haloperidol in the treatment of acute psychosis. Ziprasidone IMStudy Group. J Clin Psychiatry 2000;61:933–41.

99. Daniel D, Brook S, Benattia I. Efficacy of intramuscular ziprasidone withoutadjunctive benzodiazepines. Presented at the 16th Congress of the EuropeanCollege of Neuropsychopharmacology, Prague, Czech Republic, September20–24, 2003.

100. Lesem MD, Zajecka JM, Swift RH, et al. Intramuscular ziprasidone, 2 mgversus 10 mg, in the short-term management of agitated psychotic patients. JClin Psychiatry 2001;62:12–8.

101. Daniel DG, Potkin SG, Reeves KR, et al. Intramuscular (IM) ziprasidone 20mg is effective in reducing acute agitation associated with psychosis: A dou-ble-blind, randomized trial. Psychopharmacology (Berl) 2001;155:128–34.

102. Preval H, Klotz SG, Southard R, et al. Rapid-acting IM ziprasidone in a psy-chiatric emergency service: A naturalistic study. Gen Hosp Psychiatry2005;27:140–4.

103. Zimbroff DL, Allen MH, Battaglia J, et al. Best clinical practice with ziprasi-done IM: Update after 2 years of experience. CNS Spectrums 2005;10 (suppl11):1-15.

104. Scahill L, Blair J, Leckman JF, et al. A. Sudden death in a patient withTourette syndrome during a clinical trial of ziprasidone. J Psychopharmacol2005;19:205–6.

105. Heinrich TW, Biblo LA, Schneider J. Torsades de pointes associated withziprasidone. Psychosomatics, in press

106. Meehan KM, Wang H, David SR, et al. Comparison of rapidly acting intra-muscular olanzapine, lorazepam, and placebo: A double-blind, randomizedstudy in acutely agitated patients with dementia. Neuropsychopharmacology2002;26:494–504.

107. Wright P, Birkett M, David SR, et al. Double-blind, placebo-controlled com-parison of intramuscular olanzapine and intramuscular haloperidol in thetreatment of acute agitation in schizophrenia. Am J Psychiatry 2001;158:1149–51.

108. Wright P, Meehan K, Birkett M, et al. A comparison of the efficacy and safe-ty of olanzapine versus haloperidol during transition from intramuscular tooral therapy. Clin Ther 2003;25:1420–58.

109. FDA Psychopharmacological Drugs Advisory Committee. Addendum toBriefing Document for Zyprexa® Intramuscular (Olanzapine for Injection).Eli Lilly and Company; February 14, 2001.

110. Sarsaburu S, Hornbuckle K, Blake DS, et al. The first year of safety experi-ence with post-marketing use of olanzapine's intramuscular formulation.Presented at the American College of Clinical Pharmacy, October 23--26,2005, San Francisco, CA.

111. Richards JR, Derlet RW, Duncan DR. Chemical restraint for the agitatedpatient in the emergency department: Lorazepam versus droperidol. J EmergMed 1998;16:567–73.

112. Foster S, Kessel J, Berman ME, et al. Efficacy of lorazepam and haloperidolfor rapid tranquilization in a psychiatric emergency room setting. Int ClinPsychopharmacol 1997;12:175–9.

113. Salzman C, Green A, Rodriguez-Villa F, et al. Benzodiazepines combinedwith neuroleptics for management of severe disruptive behavior. Psycho-somatics 1986;27(Suppl 1):17–22.

114. Wyant M, Diamond B, O’Neal E, et al. The use of midazolam in acutely agi-tated psychiatric patients. Psychopharmacol Bull 1990;26:126–9.

115. Nobay F, Simon BC, Levitt MA, et al. A prospective, double-blind, random-ized trial of midazolam versus haloperidol versus lorazepam in the chemicalrestraint of violent and severely agitated patients. Acad Emerg Med 2004;11:744–9.

116. Chouinard G, Annable L, Turnier L, et al. A double-blind randomized clini-cal trial of rapid tranquilization with I.M. clonazepam and I.M. haloperidolin agitated psychotic patients with manic symptoms. Can J Psychiatry1993;38(Suppl 4):S114–21.

117. Salzman C. Use of benzodiazepines to control disruptive behavior in inpa-tients. J Clin Psychiatry 1988;49(suppl):13–5.

118. Findling RL, Aman MG, Eerdekens M, et al. Risperidone DisruptiveBehavior Study Group. Long-term, open-label study of risperidone in chil-dren with severe disruptive behaviors and below-average IQ. Am J Psychiatry2004;161:677–84.

119. Aman MG, Gharabawi GM. Special Topic Advisory Panel on Transitioningto Risperidone Therapy in Patients With Mental Retardation andDevelopmental Disabilities. Treatment of behavior disorders in mental retar-dation: Report on transitioning to atypical antipsychotics, with an emphasison risperidone. J Clin Psychiatry 2004;65:1197–210.

120. Aman MG, Crismon ML, Frances A, et al.. Treatment of psychiatric andbehavior problems in individuals with mental retardation: An update of theExpert Consensus Guidelines for mental retardation/developmental disabilitypopulations. Englewood, CO: Postgraduate Institute for Medicine, 2004.

121. Altshuler LL, Cohen L, Szuba MP, et al. Pharmacologic management of psy-chiatric illness during pregnancy: Dilemmas and guidelines. Am J Psychiatry1996;153:592–606.

122. Jain AE, Lacy T. Psychotropic drugs in pregnancy and lactation. J PsychiatrPract 2005;11:177-91.

123. Allen MH. Level 1 psychiatric emergency services. The tools of the crisis sec-tor. Psychiatr Clin North Am 1999;22:713–34, vii.



Expert Consensus Guidelines on the Treatment of Behavioral

Documents

Transcript of Expert Consensus Guidelines on the Treatment of Behavioral