Level A recommendations

Generally accepted principles for patient care that reflect a high degree of clinical certainty (eg, based on evidence from one or more Class of Evidence I or multiple Class of Evidence II studies).

Level B recommendations

Recommendations for patient care that may identify a particular strategy or range of strategies that reflect moderate clinical certainty (eg, based on evidence from one or more Class of Evidence II studies or strong consensus of Class of Evidence III studies).

Level C recommendations

Recommendations for patient care that are based on evidence from Class of Evidence III studies or, in the absence of any adequate published literature, based on expert consensus. In instances where consensus recommendations are made, “consensus” is placed in parentheses at the end of the recommendation.

There are certain circumstances in which the recommendations stemming from a body of evidence should not be rated as highly as the individual studies on which they are based. Factors such as heterogeneity of results, uncertainty about effect magnitude and consequences, and publication bias, among others, might lead to such a downgrading of recommendations.

When possible, clinically oriented statistics (eg, likelihood ratios, number needed to treat) are presented to help the reader better understand how the results may be applied to the individual patient. For a definition of these statistical concepts, see Appendix C .

This policy is not intended to be a complete manual on the diagnosis and management of adult psychiatric patients in the ED but rather a focused examination of critical issues that have particular relevance to the current practice of emergency medicine.

It is the goal of the Clinical Policies Committee to provide an evidence-based recommendation when the medical literature provides enough quality information to answer a critical question. When the medical literature does not contain adequate empirical data to answer a critical question, the members of the Clinical Policies Committee believe that it is equally important to alert emergency physicians to this fact.

This clinical policy is not intended to represent a legal standard of care for emergency physicians. Recommendations offered in this policy are not intended to represent the only diagnostic or management options available to the emergency physician. ACEP recognizes the importance of the individual physician’s judgment and patient preferences. This guideline defines for the physician those strategies for which medical literature exists to provide support for answers to the critical questions addressed in this policy.

Scope of Application

This guideline is intended for physicians working in EDs.

Inclusion Criteria

This guideline applies to adult patients presenting to the ED with psychiatric symptoms. Critical question 4 includes patients with delirium.

Exclusion Criteria

This guideline is not intended to be used for pediatric patients. It is also not intended for patients with delirium in regard to critical questions 1, 2, and 3.

For potential benefits and harms of implementing the recommendations, see Appendix D .

Level A recommendations

None specified.

Level B recommendations

None specified.

Level C recommendations

Do not routinely order laboratory testing on patients with acute psychiatric symptoms. Use medical history, previous psychiatric diagnoses, and physician examination to guide testing.

Key words/phrases for literature searches: emergency services, hospital, psychiatric, mental disorders, physical examination, diagnostic tests, routine, and variations and combinations of the key words/phrases. Searches included January 1, 2005 to search date of September 3, 2015.

Study Selection : Ninety-five articles were identified in the searches. Nine articles were selected from the search results for further review, with 2 Class III studies included for this critical question.

In patients with acute behavioral emergencies, physicians are often asked to obtain routine laboratory testing in addition to obtaining a history and performing a physical examination. The previous clinical policy on this topic, published in 2006, made a Level B recommendation: “In adult ED patients with primary psychiatric complaints, diagnostic evaluation should be directed by the history and physical examination. Routine laboratory testing of all patients is of very low yield and need not be performed as part of the ED assessment”; and Level C recommendations: “Routine urine toxicologic screens for drugs of abuse in alert, awake, cooperative patients do not affect ED management and need not be performed as part of the ED assessment; urine toxicologic screens for drugs of abuse obtained in the ED for the use of the receiving psychiatric facility or service should not delay patient evaluation or transfer.” However, the articles supporting this 2006 recommendation were regraded by the methodologists using the committee’s current criteria and were determined to be either Class X or Class III.

For this revision, the authors of 2 Class III studies reached similar conclusions that laboratory testing after medical screening by an emergency physician rarely if ever changes ED management or disposition. Janiak and Atteberry performed a chart review on 502 consecutive admissions at a large academic center. In this center, routine laboratory tests were obtained for all patients admitted to the psychiatric service, regardless of whether directed testing was performed by the emergency physician. The authors reviewed each of these laboratory test results and noted that, with only one exception (0.19%), laboratory tests obtained by the psychiatric service would not have changed management.

In a similar study, Parmar et al obtained a convenience sample of 598 patients presenting to an ED. After medical screening by an emergency physician, which included laboratory tests in 155 patients, 44% of patients had additional laboratory tests obtained by the psychiatric service. With only one exception (0.5%), no patient had a laboratory value that led to a change in disposition.

According to these 2 studies, it would appear that laboratory studies ordered by the psychiatric service rarely change patient disposition from an ED point of view. However, because both studies investigated only patients admitted to an inpatient unit, it is unknown whether laboratory test results influence disposition for patients sent to other locations (for instance, a crisis residence or rehabilitation facility). It is also unknown what harms the patients might have experienced had a medical illness been undetected. Finally, none of the studies reviewed included all patients presenting to the ED with acute psychotic symptoms, meaning that it is unknown whether there are patients who are missed by current ED screening methods.

In summary, existing literature indicates that routine or ancillary laboratory testing for psychiatric patients has little or no use in the ED. It is likely that subsets of patients with higher rates of disease (eg, elderly, immunosuppressed, new-onset psychosis, substance abuse) may benefit from routine laboratory testing. In addition, although urine toxicologic screening has no benefit for the management or disposition of the patient in the ED, it may be helpful to obtain an objective understanding of the patient’s potential substance abuse on transfer to a psychiatric facility. When transfer to a psychiatric facility may be delayed for hours, it may be helpful to obtain a urine toxicologic screen in the ED, when feasible. To expedite the care of patients, agreement between the ED and local psychiatric facilities regarding minimal laboratory testing for psychiatric clearance should be mutually determined.

Level A recommendations

None specified.

Level B recommendations

None specified.

Level C recommendations

Use individual assessment of risk factors to guide brain imaging in the ED for patients with new-onset psychosis without focal neurologic deficit. (Consensus recommendation)

Key words/phrases for literature searches: emergency services, mental disorders, physical examination, diagnostic tests, routine, and variations and combinations of the key words/phrases. Searches included January 1, 2005 to search date of September 4, 2015.

Study Selection : Ninety-three articles were identified in the searches, and 13 articles were selected from the search results for further review. None of the 13 articles were classified as Class I, II, or III; therefore, zero studies were included for this critical question.

Historically, computed tomography (CT) of the brain has often been recommended in the evaluation of patients with new-onset psychosis without focal neurologic deficits to exclude medical pathology such as mass lesion as a cause for symptoms. Because psychosis, delirium, dementia, and encephalopathy may share similar presenting features, CT of the brain has been seen as a potentially important part of the diagnostic algorithm for new-onset psychosis.

The rate of reported neuroimaging abnormalities in the Class X studies reviewed varied widely, from 3% to 66.1%. In many studies, the reported abnormalities were either not described or were characterized as incidental or unrelated to the patient’s psychiatric condition. A study comparing neuroimaging in patients presenting with psychosis versus a control group found no difference in the frequency of clinically relevant findings (11.1% versus 11.8%). In the Class X studies that did categorize imaging abnormalities, the percentage of imaging findings described as clinically relevant, influencing clinical management, or altering diagnosis ranged from 0% to approximately 5%. Definitions of “altering treatment” or “altering diagnosis” were not strictly described and may be difficult to apply to a retrospective chart review or lack external validity. Interpretation of the effect of imaging on diagnosis and treatment also may depend on blinding, which was not applied in all studies. In addition, as described below, poor study methodology may result in underreporting of abnormalities by systematic exclusion of patients with abnormal findings. Consequently, the results of these Class X studies should be applied with caution and attention given to patient-specific risk factors for central nervous system disease.

It is difficult to ascertain an accurate estimation of significant abnormal neuroimaging findings based on the current studies in the literature. There are a number of ways in which the rate of abnormal neuroimaging study results in patients with new-onset psychosis in the ED can be underestimated. Poorly described methods may mask biases such as low-quality chart abstraction, lack of blinding, or absence of strictly defined variables. Studies that retrospectively identified patients based on final diagnostic codes for “new psychosis” may not have included all patients who presented initially with altered mental status, such as those who later received alternative diagnoses such as encephalopathy, central nervous system mass, hydrocephalus, or stroke. In some studies, inpatient psychiatric patients were included; these patients are typically more homogeneous than patients in the ED and have fewer acute comorbidities, placing them at a lower risk than the undifferentiated patient in the ED.

Conversely, there are factors that may falsely elevate the rate of abnormal neuroimaging study results that were reported. To identify patients as having no focal neurologic deficits, a comprehensive structural examination would be required. None of these studies provided a detailed description of neurologic examination performed; it is not clear whether patients underwent a thorough examination to exclude deficits or if patients were included when there were no deficits documented in the medical record. Given that many acutely psychotic patients may not be able to cooperate with a comprehensive neurologic examination, emergency physicians may have a lower threshold to obtain neuroimaging in these patients.

The timeframe for imaging and the definition of abnormal were also not clearly defined among these studies, making their application to patients in the ED unclear.

Level A recommendations

None specified.

Level B recommendations

None specified.

Level C recommendations

In patients presenting to the ED with suicidal ideation, physicians should not use currently available risk-assessment tools in isolation to identify low-risk patients who are safe for discharge. The best approach to determine risk is an appropriate psychiatric assessment and good clinical judgment, taking patient, family, and community factors into account.

Key words/phrases for literature searches: suicidal ideation, ED, emergency services, hospital, risk assessment, patient discharge, and variations and combinations of the key words/phrases. Searches included January 1, 1990 to search dates of September 4, 2015, and November 5, 2015.

Study Selection : Eighty-five articles were identified in the searches. Nineteen articles were selected from the search results for further review, with 4 Class III studies included for this critical question.

Emergency physicians, as well as mental health professionals, are frequently called on to determine the suicide risk in a patient who presents with depression or suicidal ideation. There are many tools to screen for suicidal ideation, although few that determine the level of risk for the patient. There is a need to determine whether a patient has high, moderate, or low risk of suicide to help decide whether a patient should be hospitalized or discharged. An objective tool for patients’ risk determination, such as the Pulmonary Embolism Rule-out Criteria (PERC) rule for pulmonary embolism, National Emergency X-radiography Utilization Study (NEXUS) criteria for cervical spine radiographs, or Thrombolysis in Myocardial Infarction (TIMI) scores for coronary artery disease, would be helpful but has eluded current investigations. For example, if a good tool existed for psychiatric patients, it might classify patients as high risk (patient needs inpatient psychiatric care), moderate risk (patient needs further evaluation and treatment from a mental health professional), and low risk (patient may only need outpatient follow-up).

That the discovery of a tool to determine the level of risk for suicidal patients has eluded medical science is not surprising because suicide is a complex disease process with many persons who present with suicidal ideation, fewer patients with suicide attempts, and lesser number who complete a suicide. This determination process is made more challenging by the waxing and waning of suicidal thoughts over time, changes in psychiatric condition, social circumstances, and contribution from substance use and stressors.

A number of studies were reviewed for inclusion in this clinical policy. However, few studies examined tools used in the ED setting that would predict suicide within a short time period. The studies varied by technique, subject enrollment, end point, and length of follow-up. Four Class III studies were identified that investigated whether risk assessment can identify patients who are at risk for future self-harm.

Posner et al used a tool developed by Columbia University, the University of Pennsylvania, and the University of Pittsburgh, the Columbia–Suicide Severity Rating Scale (C-SSRS). The tool was used to distinguish suicidal ideation from suicidal behavior. The 4 constructs measured in this tool were severity, intensity of ideation, suicidal behavior subscale, and a lethality subscale. In adult patients with psychiatric problems, the C-SSRS had 100% sensitivity (95% confidence interval 98% to 100%) and 100% specificity (95% confidence interval 94% to 100%) for identifying lifetime actual attempts that were recorded on the Columbia Suicide Form. The study was limited by low prevalence and convenient outcome measurement, and the incremental predictive validity of C-SSRS could not be estimated. Unfortunately, the risk of lifetime suicide attempts does not help the emergency physician in the disposition of a patient presenting to the ED with suicidal ideation.

Tran et al examined a large retrospective electronic medical record database of patients with at least one suicide risk assessment to develop a prediction model; the investigators then compared performance of the electronic medical record–based model with an 18-item checklist used by clinicians to estimate suicide risk. The goal was to differentiate low-, moderate-, and high-risk suicidal behaviors. Although the predictive performance of the electronic medical record–based model was inadequate as a decision support tool (sensitivity=28% for high risk), the model did perform better than clinician assessment using the 18-item checklist (sensitivity=8% for high risk) based on 90-day outcomes.

Bilen et al examined a list of factors that could predict repeated deliberate self-harm or suicide. Risk factors associated with deliberate self-harm were female sex, self-injury and whether the self-injury required a surgical procedure, current psychiatric or antidepressant treatment, substance use disorder, personality disorder, and not having children younger than 6 years. Using these factors, patients could be stratified into low-, moderate-, and high-risk categories. Although deliberate self-harm has a close association to suicide, this study focused on deliberate self-harm rather than suicide.

Randall et al used a number of psychiatric scales and found modest performance according to receiver operating characteristic curve analysis and predictive values. They used several questionnaires such as the Beck Hopelessness Scale, the Barrett Impulsiveness Scan, the Brief Symptom Inventory, the Drug Abuse Screening Test, and the Cut down, Annoyed, Guilt, Eye-opener (CAGE) assessment to determine which of these scales were valuable in differentiating individuals at risk for self-harm within 3 months. This study was limited by selection bias and attrition. Clinicians’ predictive power was poor for high risk, as was the electronic medical record model. None of the tools were considered strong enough and the diagnostic usefulness is limited.

Although these studies were rated as Class III, the study designs were problematic. The studies need to be able to separate the tools that best predict suicide completion in an at-risk population with a low prevalence rate. The ideal tool would be useful in all age groups and have a greater than 90% sensitivity and specificity for high-risk in the next 30 days with co-occurring mental illness and substance use disorder.

This clinical policy review demonstrates that there is no tool currently available that can be solely used to predict the risk of suicide among patients in the ED who have suicidal ideation.

Level A recommendations

None specified.

Level B recommendations

None specified.

Level C recommendations

Ketamine is one option for immediate sedation of the severely agitated patient who may be violent or aggressive. (Consensus recommendation)

Key words/phrases for literature searches: ketamine sedation for agitation in ED, acute agitation, agitation, psychomotor agitation, ketamine, emergency service, hospital, emergency department, and variations and combinations of the key words/phrases. Searches included January 1, 2005 to search dates of September 4, 2015, and November 5, 2015.

Study Selection : One hundred thirty-three articles were identified in the searches, and 11 articles were selected from the search results for further review. None of the 11 articles were classified as Class I, II, or III studies; therefore, zero studies were included for this critical question.

At this time, there is a lack of Class I, II, or III studies establishing the safety and efficacy of ketamine to control acute agitation in the ED. However, there are a number of studies in the out-of-hospital literature describing its use for this indication, and there are 2 studies (Class X) addressing its use in the ED under the immediate direction of a physician.

Management of acutely agitated patients in the ED remains a critical issue. Most of these patients can be safely sedated with antipsychotics and/or benzodiazepines, but there remains a subset of extremely agitated patients for whom this approach will not be effective. Although small in number, these patients have a significant effect on the ED staff in terms of time and dedicated resources in order to maintain a safe environment for the patient and others in the ED. Multiple agents, including ketamine, have been suggested as rescue agents in instances when antipsychotics and/or benzodiazepines fail.

Ketamine has been used as a drug for sedation, anesthesia, and induction for many years. It functions through antagonism of the glutamate N -methyl- d -aspartate receptors, resulting in a dissociated state with analgesia and amnesia. Its rapid onset of action, achieving the dissociated state within 1 to 2 minutes by the intravenous route and within approximately 3 minutes by the intramuscular route, and short half-life make it useful for procedural sedation and pretreatment of intubation in the ED. Other benefits include few effects on vital signs, with typically protected respiratory drive and rare negative effects on systolic blood pressure.

Ketamine also has several potentially serious adverse effects, most notably tachycardia and hypertension in already agitated patients. Ketamine is associated with emergence phenomenon, laryngospasm, hypersalivation, and vomiting. Its duration of action is short; thus, patients may require readministration of medications. There are also concerns that it may worsen symptoms in psychiatric patients who are acutely psychotic.

Although no high-level studies currently describe its use in the ED, 2 Class X studies recently addressed its use for patients with acute agitation in the ED. In a retrospective review of 27 patients who received ketamine for acute agitation in the ED, none became hypoxic. Sixty-two percent of patients required additional sedating medications. The dosing range was wide in this group of patients, from 40 to 400 mg of intravenous or intramuscular ketamine. The median dose was 200 mg. A 2016 study described administration of ketamine as a rescue drug in ED patients after droperidol or droperidol and benzodiazepines failed. Forty-nine patients received intramuscular ketamine, with dosing of 4 to 6 mg/kg. Of these patients, 90% were adequately sedated within 1 hour; only one had hypoxia less than 90% responding immediately to oxygen administration. There is also a description of its use for aeromedical retrieval of patients with acute psychiatric complaints who required critical care monitoring and transport by physicians during a prolonged period. In these instances, no patients required intubation after ketamine administration. It was also thought that ketamine administration did not result in worsening agitation or psychosis in the group of patients who required aeromedical transport. Both groups of patients required additional sedating medications, and the group receiving aeromedical retrieval continued to receive ketamine throughout their transport time. This is to be expected, given the duration of action of ketamine and the fact that it did not treat the underlying disorder that resulted in agitation. A common dosing strategy is intramuscular 4 to 6 mg/kg, and in cases in which additional sedation is required after ketamine, low dosing is a likely culprit.

There are multiple studies describing the use of ketamine for the agitated patient in the out-of-hospital literature. These describe the known adverse effects of laryngospasm, hypersalivation, vomiting, and emergence reaction. In out-of-hospital situations in which a physician was not administering the drug, respiratory depression was not uncommon and required escalation of care ranging from airway positioning to intubation in as many as 29% of patients. It is unclear whether this was a result of improper dosing or concomitant sedatives either ingested by the patient or given by emergency medical services personnel; however, the authors pointed out that there was a significant difference in the dose of ketamine for patients who were intubated versus those who were not (6.16 mg/kg [SD 1.62] versus 4.90 mg/kg [SD 1.54]; P =.02). In most cases, emergence reaction can be easily treated with benzodiazepines.

Although there is limited literature for guidance, the skill set of emergency physicians and their familiarity with the use of ketamine make it a reasonable choice when immediate control of an acutely agitated patient is required for patient and/or staff safety.