Level of agitation of psychiatric patients presenting to an emergency department

Level of Agitation of Patients Presenting to an ED Level of Agitation of Psychiatric Patients
Presenting to an Emergency Department
Leslie S. Zun, M.D., M.B.A.; and La Vonne A. Downey, Ph.D.
Received May 25, 2007; accepted Oct. 18, 2007. From the Department of Emergency Medicine, Rosalind Franklin University ofMedicine and Science/Chicago Medical School, and the Department of Objectives: The primary purpose of this study
Emergency Medicine, Mount Sinai Hospital (Dr. Zun); and the Schoolof Policy Studies, Roosevelt University (Dr. Downey), Chicago, Ill. was to determine the level of agitation that psy- Dr. Zun has been a speakers/advisory board member for Eli Lilly. chiatric patients exhibit upon arrival to the emer- Dr. Downey reports no financial or other relationship relevant to the gency department. The secondary purpose was to determine whether the level of agitation changed Corresponding author and reprints: Leslie S. Zun, M.D., M.B.A., over time depending upon whether the patient Department of Emergency Medicine, Mount Sinai Hospital, 15th andCalifornia, Chicago, IL 60608 (e-mail: [email protected]). Method: An observational study enrolling a
convenience sample of 100 patients presentingwith a psychiatric complaint was planned, in sychiatric patients frequently present to emergency order to obtain 50 chemically and/or physically Pdepartments (EDs) across the country.1 Many of
restrained and 50 unrestrained patients. The study these patients are agitated, necessitating treatment for their was performed in summer 2004 in a community, agitation in the ED. In an unselected ED sample, some inner-city, level 1 emergency department with45,000 visits per year. The level of patient agita- patients who exhibit agitation are intoxicated, delirious, tion was measured using the Agitated Behavior and/or otherwise impaired. This study focuses on patients Sedation Scale (RASS) upon arrival and every The treatment for these agitated patients frequently 30 minutes over a 3-hour period. The inclusion includes physical restraint, chemical treatment, and se- criteria allowed entry of any patient who pre-sented to the emergency department with a psy- clusion. The Joint Commission on Accreditation of Health- chiatric complaint thought to be unrelated to care Organizations, Centers for Medicare and Medicaid physical illness. Patients who were restrained Services, and many states have regulated the use of re- for nonbehavioral reasons or were medically There are few studies regarding the level of agitation Results: 101 patients were enrolled in the
study. Of that total, 53 patients were not re- of undifferentiated psychiatric patients presenting to EDs.
strained, 47 patients were restrained, and 1 had There is little information in the medical literature con- incomplete data. There were no differences in cerning the relationship between the level of patient agita- gender, race, or age between the 2 groups. Upon tion and restraint and seclusion use. Information exists for arrival, 2 of the 47 restrained patients were rated the use of restraint and seclusion of these agitated patients severely agitated on the ABS, and 13 of 47 re-strained patients were rated combative on the outside the ED. In a review of 13 published studies in adult RASS. There was a statistical difference (p = .01) inpatient psychiatric settings, a range of 1.9% to 66% of between the groups on both scales from time 0 patients had a need for seclusion and restraint.2 Another to time 90 minutes. Scores on the agitation scales study found an average of 2 restraints on 17% of patients decreased over time in both groups. One patient in an acute medical unit.3 In psychiatric emergency rooms, in the unrestrained group became unarousableduring treatment.
the percentage of patients restrained (20%–25%) was sig- Conclusion: This study demonstrated that
nificantly higher than in an inpatient facility (7%–20%).4–7 patients who were restrained were more agitated In other EDs, Lavoie et al.8 found that 25.2% of teaching than those who were not, and that agitation levels hospitals restrained at least 1 patient per day. An average in both groups decreased over time. Some re- of 3.7% of all ED patients needed restraint and seclusion, strained patients did not meet combativeness orsevere agitation criteria, suggesting either that use of other criteria is needed or that restraints were The relationship between the use of restraints and used inappropriately. Further study of the level of the level of agitation of psychiatric patients in the ED is agitation and the effects of restraints is needed.
unclear. It is thought that highly agitated patients are re- (Prim Care Companion J Clin Psychiatry 2008;10:108–113) strained in the ED to prevent further escalation and resul-tant violence. However, the relationship between the levelof agitation and restraint use needs further definition. In Prim Care Companion J Clin Psychiatry 2008;10(2) order to better understand the relationship between re- added together. Only 13 of the 14 items were used since straint use and the level of agitation, we proposed this the patients were not allowed to wander from the treat- study. The secondary purpose was to determine the ment area (item 7). The RASS is a 10-level scale based on change in the level of agitation of the psychiatric patients observation of the patient’s level of agitation or sedation, ranging from combativeness (+4) to unarousability (–5).
These scales were chosen because of their ease of use and variable measure of sedation and agitation.
The data were input into an SPSS program for analysis (Version 10; SPSS Inc.; Chicago, Ill.). To analyze the In order to determine if there was a significant differ- data, the investigators grouped the scores into broader ence between the groups, we planned to enroll 50 patients categories. The ABS scale was divided into no (< 22), who were restrained and 50 who were not. The inclusion mild (22–28), moderate (29–35), and severe (≥ 36) agita- criteria allowed entry of patients of any age who pre- tion, and the RASS was divided into agitated (+4 to +1), sented to the emergency department with a psychiatric alert and calm (0), and sedated (–1 to –5).13,14 The groups complaint thought to be unrelated to physical illness. Pa- were compared using the χ2, analysis of variance, and tients who were restrained for nonbehavioral reasons or were medically unstable were excluded from the study.
Basic demographic information was obtained on each One hundred one patients were enrolled in the study.
Of those patients, 53 were not restrained, and 47 were re- This observational study was performed in a commu- strained. Although various elements in the data set were nity, inner-city, level 1 teaching hospital ED with 45,000 not completed, only 1 patient had significant incomplete visits per year, located in Chicago, Ill. The city’s police data and was eliminated from consideration. There were department has designated the hospital as the referral site no differences between the 2 groups in gender (χ2 = 5.79, for psychiatric patients in the southwest side of the city.
df = 2, p = .12), race (χ2 = 7.22, df = 2, p = .30), age During the summer of 2004, a convenience sample of (χ2 = 2.73, df = 2, p = .59), or ED diagnosis (χ2 = 31.4, patients who presented with psychiatric complaints to the ED when a research fellow from the Department of Emer- All restrained patients were restrained within 15 min- gency Medicine was available in the ED were enrolled in utes of arrival to the ED. Of the restrained patients, 21 the study. None of the patients were restrained prior to were only physically restrained, 13 were chemically and arrival because neither the police nor paramedics are ca- physically restrained, and 13 were only chemically re- pable of behaviorally restraining patients. The emergency strained. Lorazepam was the most frequently used medi- physicians independently determined the need for physi- cation (12), followed by other agents (5), olanzapine (3), cal and/or chemical restraint without input from any and haloperidol and lorazepam (2). Among restrained pa- study personnel. Since patients did not receive a psychiat- tients with follow-up information, 15 of 27 patients were ric therapeutic plan in the ED, the use of medication admitted, and 12 of 27 went home. Mania was the most would be considered chemical restraint rather than be- frequent diagnosis (27 of 46 patients), followed by psy- havior modulation in the context of this study.7 Research chosis (9 of 46 patients) and depression (5 of 46 patients) fellows were responsible for completing an agitation (total Ns less than 47 due to missing data). The reason for checklist for each patient enrolled in the study. The psy- restraint was violent behavior in 28 of 44 patients and agi- chiatric diagnoses used for the study were provided by the emergency physicians and may not reflect the In the unrestrained group with follow-up information, DSM-IV criteria. Seclusion was not used at this hospital.
19 of the 36 patients went home, and 17 patients were ad- The study was institutional review board–approved as mitted. The leading diagnosis in this unrestrained group exempt from consent due to the observational nature of was manic-depressive illness (17 of 47 patients), followed the study. Data were collected without patient identifying by depression (13 of 47 patients) and psychotic illness (12 of 47 patients) (total Ns less than 53 due to missing data).
The patients were evaluated for their level of agitation There were no statistical differences found between the at arrival and every 30 minutes for 3 hours. We chose 2 groups for admission rates or diagnoses.
validated tests of agitation to determine the patients’ level There was a statistical difference between the groups of agitation in the ED: the Agitated Behavior Scale (ABS) on both scales from time 0 to time 90 minutes (ABS: and the Richmond Agitation-Sedation Scale (RASS).10–14 F = 18.4, df = 1, p = .01 [0 minutes] to F = 3.86, df = 1, The ABS is a scale with 14 items rated 1 (no agitation) p = .01 [90 minutes]; RASS: F = 10.4, df = 1, p = .01 [0 to 4 (highest level of agitation) and the individual scores minutes] to F = 5.74, df = 1, p = .01 [90 minutes]) (Tables Prim Care Companion J Clin Psychiatry 2008;10(2) Level of Agitation of Patients Presenting to an ED Table 1. Agitated Behavior Scale Scores Over Time for
Table 2. Richmond Agitation-Sedation Scale Scores Over
Unrestrained and Restrained Patientsa
Time for Unrestrained and Restrained Patientsa
aThe Agitated Behavior Scale is scored using a scale of 1 (no aThe Richmond Agitation-Sedation Scale is scored on a 10-level scale agitation) to 4 (highest level of agitation) for each of 13 items.
based on observation of the patient’s level of agitation or sedation, Category data were computed by grouping the total scores into the ranging from combativeness (+4) to unarousability (–5). The following categories: no (< 22), mild (22–28), moderate (29–35), category data were computed by grouping the total scores into the following categories: agitated (score of +4 to +1), alert and calm (score of 0), and sedated (score of –1 to –5).
1 and 2). The agitation scales decreased over time in bothgroups (ABS mean decreased from 16.5 to 14.0 in the un- Table 3. Agitated Behavior Scale: Change Over Time
restrained group and 26.5 to 15.0 in the restrained group, and RASS mean decreased from 0.7 to 0.1 in the unre- strained group and 2.9 to 0.1 in the restrained group) (Tables 1 and 2). Tables 3 and 4 show the statistical results for change over time in ABS and RASS scores.
The RASS was used to assess patients who became unarousable, and only 1 patient in the unrestrained group became unarousable, at 30 minutes. No other patient in ei- ther group became unarousable throughout the period ofobservation. One unrestrained patient was judged as com-bative on the RASS at 30 minutes, but no other un- numbers of patients who were in the no agitation ABS cat- restrained patients were found to be combative during the egory increased from time 0 to time 180 minutes, going evaluation period. Thirteen restrained patients were from 15 to 45 in the restrained group and 39 to 50 in the judged as combative upon presentation, 8 were judged as unrestrained group. The standard deviation and median combative at 30 minutes, and 1 was judged as combative values for the ABS (Table 5) had less variation within each time period for unrestrained patients as compared The ABS indicated that 2 patients in the restrained with restrained patients. The same difference in variation group and none in the unrestrained group reached the se- was not seen using the RASS (Table 6). This finding indi- verely agitated category at time 0, and 0 or 1 restrained cates that the ABS would have placed in the restrained patient was in the severely agitated category during the population some patients whom the RASS might have rest of the observation period. On the other hand, the placed in the unrestrained population.
Prim Care Companion J Clin Psychiatry 2008;10(2) Table 4. Richmond Agitation-Sedation Scale:
Table 6. Richmond Agitation-Sedation Scale Median and
Change Over Time
Standard Deviation Values in 30-Minute Increments
Table 5. Agitated Behavior Scale Median and Standard
Deviation Values in 30-Minute Increments
homicidal potential. Perhaps the emergency staff was us- ing other, unstudied criteria on which to base the decision to restrain a patient. As an example of such unstated crite- ria, an agitated patient brought to the emergency depart- ment by law enforcement in handcuffs for violent behav- ior would most likely be placed in restraints prior to assessment by the emergency physician.
Analogous to pain treatment, could the treatment of agitation using a measurement tool be more beneficial than the current “all-or-none” phenomenon, in which pa-tients either need or do not need restraints? Few studies The agitation levels of unrestrained patients started have measured the level of agitation a patient exhibits low and remained low throughout the study on both upon arrival to the emergency department. The natural scales. Restrained patients had higher agitation levels history of an agitated patient without treatment has not been evaluated and would be an interesting topic for Agitation levels remained significantly different be- tween restrained and unrestrained patients at each time Studies in the psychiatric literature found that restraint point during the first 2 of the 3-hour periods. The agita- and seclusion use reduced the level of agitation.16 The tion levels of both groups decreased over time.
medical literature offers limited information on the use of It is easy to understand the significance of the 15 pa- the agitation scales and testing in the emergency setting.
tients who were restrained at presentation and had no agi- We found no studies of the use of the RASS in emergency tation on the ABS scale. This finding was not sustained in medicine. The uses of ABS in emergency medicine in a the RASS, on which 13 of the restrained patients were selected population were examined in 1 study,17 and the found to be combative. Perhaps the patients were not Overt Aggression Scale was used in a study in a para- properly assessed. The difference could be explained by medic system.18 Battaglia and others17 used the ABS to as- the type of testing used. The RASS is a global rating with sess the differences seen with haloperidol, lorazepam, or an anchor that includes combativeness, and the ABS is both in the treatment of agitation. Patients had to score at composed of 14 items, of which only 2 involve anger or least 5 on the 11 psychosis/anxiety items on the Brief Psy- threats. However, it is concerning that a number of re- chiatric Rating Scale. The authors found that all treatment strained patients did not meet the criteria for combative- groups showed significant reduction in baseline scores ness or severe agitation on either scale.
over a 12-hour treatment phase. The scores began at a Possible explanations for this finding include that in- level of 40 for the patients to be enrolled in the study and appropriate patients were restrained, or that the scales do were at a level of approximately 20 by 2 hours of therapy not adequately reflect clinical decisions for restraints.
and continued at that level for 12 hours. The authors did The tools did not assess a patient’s level of suicidal or not examine the effect of physical restraint, nor did they Prim Care Companion J Clin Psychiatry 2008;10(2) Level of Agitation of Patients Presenting to an ED document how many patients received this intervention.
psychiatrists concerning the use of the terminology of In our study, few of the patients had an ABS score of 40 or chemical restraints. Emergency physicians, who do not de- velop therapeutic plans, use the term in reference to medi- The Overt Aggression Scale was used in a study by cation that quickly induces calm behavior. Psychiatrists do Mock et al.18 to measure the number of violence episodes not use this terminology; rather, medication is used as part encountered by emergency medical services personnel. In of a therapeutic plan. A comparison of the level of agita- the Mock et al. study, the tools were used to determine a tion found with the different treatment modalities would patient’s risk of violence, rather than the patient’s level of provide better guidance to determine the best technique for On the basis of our conclusions, an argument could be The tools chosen to measure agitation also limited this made that a scale or assessment of the need for restraint or study. Although the tools have been validated, their useful- seclusion that better matches the indications for restraint ness in the acute care setting has not. Modification of the and seclusion is needed. The chief indication for placing a ABS to 13 items may have biased the conclusions. Perhaps patient in restraint or seclusion is prevention of harm to there are other tests that would have provided better infor- the patient or staff. Such a scale would take into account mation than those used in this study. The raters’ agreement not only the level of agitation but also the probability of for each of the scales was not tested, and some of the vari- violence and elopement of patients with suicidal and ance may be attributable to lack of concordance among the raters. This study was limited by incomplete data collec- Many procedures are performed in the acute care set- tion for some of the patients on some of the inquiries. One ting. For most of these procedures, we have some under- serious potential bias of this study was observer bias, espe- standing of the effect of the procedure on the patient.19 cially in the cases in which patients were immediately re- This study demonstrated the level of agitation of patients strained upon presentation to the emergency department.
evaluated in the acute care setting with and without Another limitation was that the emergency physicians did restraints. In all procedures, one must understand the not utilize DSM-IV criteria in making their diagnoses. The indications and contraindications. The procedural steps groups were not homogeneous in terms of diagnoses or should be reviewed and technical aspects practiced. On indications for restraints. The treating emergency physi- the basis of the findings of this study, the procedural cians may have used other information, not identified in step to determine if a patient needs restraints is not well this study, to determine whether a patient needed to be In retrospect, the study could be improved if a greater In summary, the obvious conclusions of the study were number of patients were enrolled in order to determine if that patients who were restrained were more agitated and there was a difference between patients restrained chemi- that the use of restraints decreased agitation over time. Un- cally, physically, and both chemically and physically.
restrained patients were less agitated and became less so Stronger conclusions could be made if a protocol for over time. Dissecting the data further reveals that some pa- the initiation of restraints were used instead of physician tients who were restrained were not severely agitated, rais- ing the question of the relationship of restraint use to agita- Future study is needed in many areas on the basis of tion levels. Further study of the level of agitation and the the findings of this study. The best, most humane means effects of restraints, both chemical and physical, is needed.
of modulating agitated behavior in not only psychiatricpatients, but also demented or delirious patients, must be Drug names: haloperidol (Haldol and others), lorazepam (Ativan
and others), olanzapine (Zyprexa).
established. A multi-arm, randomized, prospective studyto examine these topics would be valuable, albeit difficult to accomplish in the acute care setting. A better under-standing of the rationale of treatment of the agitated 1. Kunen S, Niederhauser R, Smith PO, et al. Race disparities in psychiatric rates in emergency department. J Consult Clin Psychol 2005;73:116–126 patient is needed in order to determine whether these pa- 2. Soloff PG, Gutheil TG, Wexler DB. Seclusion and restraint in 1985: tients are being treated as part of a therapeutic plan or for a review and update. Hosp Community Psychiatry 1985;36:652–657 staff convenience. It would be valuable to study regula- 3. Robbins L, Boyko E, Lane J, et al. Binding the elderly: a prospective study of the use of mechanical restraints in an acute care hospital.
tory compliance with requirements for restraints, such as determining the use of alternatives prior to restraints and 4. Mion LC, Frengley JD, Jakovcic SA, et al. A further exploration of the use of physical restraints in hospitalized patients. J Am Geriatr Soc1989;37:949–956 5. Frengley DM. Incidence of physical restraints on acute medical wards.
This study did not separate out chemical from physical 6. Telintelo S, Kuhlman TL, Winger C. A study of the use of restraint in a psychiatric emergency room. Hosp Community Psychiatry 1983;34: restraint, somewhat limiting its usefulness. There is a dif- ference of opinion between emergency physicians and 7. Allen MH, Currier GW, Hughes DH, et al. The Expert Consensus Prim Care Companion J Clin Psychiatry 2008;10(2) Guideline Series: treatment of behavioral emergencies. Postgrad Med 14. Bogner JA, Corrigan JD, Bode RK, et al. Rating scale analysis of the Agitated Behavior Scale. J Head Trauma Rehabil 2000;15:656–669 8. Lavoie F, Carter GL, Danzl DF, et al. Emergency department violence 15. Sessler CN, Gosnell MS, Grap MJ, et al. The Richmond Agitation- in United States teaching hospitals. Ann Emerg Med 1988;17: Sedation Scale: validity and reliability in adult intensive care unit patients. Am J Respir Crit Care Med 2002;166:1338–1344 9. Lavoie FW. Consent, involuntary treatment, and the use of force in an 16. Fisher WA. Restraint and seclusion: a review of the literature.
urban emergency department. Ann Emerg Med 1992;21:25–32 10. Dubin WR. Evaluating and managing the violent patient. Ann Emerg 17. Battaglia J, Moss S, Rush J, et al. Haloperidol, lorazepam, or both for psychotic agitation? a multicenter, prospective, double blind, emergency 11. Corrigan JD, Mysiw WJ. Agitation following traumatic head injury.
department study. Am J Emerg Med 1997;15:335–340 18. Mock EF, Wrenn KD, Wright SW, et al. Prospective field study of 12. Corrigan JD. Development of a scale for assessment of agitation follow- violence in emergency medical services calls. Ann Emerg Med 1998;32: ing traumatic brain injury. J Clin Exp Neuropsychol 1989;11:261–277 13. Bogner JA, Corrigan JD, Stange M, et al. Reliability of the Agitated 19. Hsiao AK, Hedges JR. Competency and confidence: procedures in Behavior Scale. J Head Trauma Rehabil 1999;14:91–96 the emergency department. Ann Emerg Med 2001;37:686–687 Prim Care Companion J Clin Psychiatry 2008;10(2)

Source: http://www.mentalhealthacademy.net/journal_archive/ppp0815.pdf