Emergency Psychiatry

Annotated Abstracts of Journal Articles
2015, 3rd Quarter

Emergency Psychiatry

Annotations by Scott Simpson, MD, MPH
of the APM Emergency Psychiatry SIG

September 2015

  1. The diagnostic performance of the Richmond Agitation Sedation Scale for detecting delirium in older emergency department patients
  2. Changes in monthly unemployment rates may predict changes in the number of psychiatric presentations to emergency services in South Australia

PUBLICATION #1 — Emergency Psychiatry
The diagnostic performance of the Richmond Agitation Sedation Scale for detecting delirium in older emergency department patients

Han JH, Vasilevskis EE, Schnelle JF, et al
Acad Emerg Med 2015; 22(7):878-882

ANNOTATION (Scott Simpson)

The Finding: About 75% of cases of delirium in emergency departments (EDs) are missed. The investigators examined the use of the Richmond Agitation Scale (RASS) for detecting delirium among geriatric patients (65 years or older) compared to a full clinical assessment by a psychiatrist. A RASS score other than zero was 82-84% sensitive and 82-87% specific for delirium. Using a RASS score >+1 or <-1 as a cut-off enhanced specificity to 99% (though sensitivity was only 16-22%). Interrater reliability between the psychiatrist and research assistants with five minutes of training was moderate (kappa 0.63).

Strength and Weaknesses: Administering a RASS takes less than 10 seconds, making it well-suited to emergency department practice. Using both a psychiatrist and research assistant to report RASS scores allows the authors to not only report reliability but also anticipate the use of this tool by non-physician staff. The RASS is already a well-known instrument for evaluating level of sedation.
This study is limited to a convenience sample of patients 65 years or older. Each patient was only assessed once, which is easier to do than serial assessments but risks misdiagnosing patients, especially those who may have received sedating medications while in the emergency department. Patients with dementia were excluded—a group at particular risk for delirium.

Relevance: A non-zero RASS appears to be a faster, simpler, and comparably valid screener for delirium as other existing, multi-step screeners (e.g., the Brief Confusion Assessment Method or the Delirium Triage Screen). The RASS can also be administered by non-physician staff.


Objectives: Delirium is frequently missed in older emergency department (ED) patients. Brief (<2 minutes) delirium assessments have been validated for the ED, but some ED health care providers may consider them to be cumbersome. The Richmond Agitation Sedation Scale (RASS) is an observational scale that quantifies level of consciousness and takes less than 10 seconds to perform. The authors sought to explore the diagnostic accuracy of the RASS for delirium in older ED patients.

Methods: This was a preplanned analysis of a prospective observational study designed to validate brief delirium assessments for the ED. The study was conducted at an academic ED and enrolled patients who were 65 years or older. Patients who were non-English-speaking, deaf, blind, comatose or had end-stage dementia were excluded. A research assistant (RA) and a physician performed the RASS at the time of enrollment. Within 3 hours, a consultation-liaison psychiatrist performed his or her comprehensive reference standard assessment for delirium using Diagnostic and Statistical Manual of Mental Disorders Fourth Edition, Text Revision (DSM-IV-TR) criteria. Sensitivities, specificities, and likelihood ratios with their 95% confidence intervals (CIs) were calculated.

Results: Of 406 enrolled patients, 50 (12.3%) had delirium diagnosed by the consult-liaison psychiatrist reference rater. When performed by the RA, a RASS other than 0 (RASS > 0 or < 0) was 84.0% sensitive (95% CI = 73.8% to 94.2%) and 87.6% specific (95% CI = 84.2% to 91.1%) for delirium. When performed by physician, a RASS other than 0 was 82.0% sensitive (95% CI = 71.4% to 92.6%) and 85.1% specific (95% CI = 81.4% to 88.8%) for delirium. Using a RASS > +1 or < -1 as the cutoff, the specificity improved to approximately 99% for both raters at the expense of sensitivity; the sensitivities were 22.0% (95% CI = 10.5% to 33.5%) and 16.0% (95% CI = 5.8% to 25.2%) in the RAs and physician raters, respectively. The positive likelihood ratio was 19.6 (95% CI = 6.5 to 59.1) when performed by the RA and 57.0 (95% CI = 7.3 to 445.9) when performed by the physician, indicating that a RASS > +1 or < -1 strongly increased the likelihood of delirium. The weighted kappa was 0.63, indicating moderate interobserver reliability.

Conclusions: In older ED patients, a RASS other than 0 has very good sensitivity and specificity for delirium as diagnosed by a psychiatrist. A RASS > +1 or < -1 is nearly diagnostic for delirium, given the very high positive likelihood ratio.

Back to top of page

PUBLICATION #2 — Emergency Psychiatry
Changes in monthly unemployment rates may predict changes in the number of psychiatric presentations to emergency services in South Australia
Bidargaddi N, Bastiampillai T, Schrader G, et al
BMC Emerg Med 2015; 15:16

ANNOTATION (Scott Simpson)

The Finding: The authors conducted an ecological study of regional monthly unemployment rates and mental health presentations to emergency departments across South Australia. Over the study period (2005-2011), the regional unemployment rate ranged from 4.0–5.8%. This monthly change correlated with 69% of the monthly variation in emergency department mental health visits. Women appeared more strongly affected, and there was a 1-2 month lag from unemployment rate to emergency department presentation. The most common diagnoses were, by far, “neurotic, stress related, and somatoform disorders” followed by substance use disorders.

Strength and Weaknesses: Unemployment has been previously correlated with suicide rates, and this study suggests that unemployment increases the use of emergency psychiatric services more generally. That there is a lag from unemployment to emergency presentation is consistent with the course of personal crisis, in which the anxiety of a life stressor builds over 2-6 weeks. The authors suggest that administrators might adjust for near-term healthcare demand based on unemployment rates.
This study uses databases to examine a large amount of data, so the observed association remains correlative, not causative. (One might think that patients with mental illness are more vulnerable to unemployment during economic contracture.) The authors also cannot exclude confounding, and it is impossible to know if it is unemployed patients themselves contributing to increased visits, or family members, or someone else entirely. The effect of unemployment on patient’s social network and crisis resources is complex and difficult to capture with this methodology. Other than emergency department utilization, no other clinical outcomes are described. And, it is unclear if the pattern of observed diagnoses change with unemployment rates as well.

Relevance: Presentations for emergency mental health care reflect an interplay of psychiatric, medical, and social factors. Emergency psychiatrists’ biopsychosocial formulations should consider the impact of larger economic and social trends on the patient.


Background: To determine the extent to which variations in monthly Mental Health Emergency Department (MHED) presentations in South Australian Public Hospitals are associated with the Australian Bureau of Statistics (ABS) monthly unemployment rates.

Methods: Times series modelling of relationships between monthly MHED presentations to South Australian Public Hospitals derived from the Integrated South Australian Activity Collection (ISAAC) data base and the ABS monthly unemployment rates in South Australia between January 2004-June 2011.

Results: Time series modelling using monthly unemployment rates from ABS as a predictor variable explains 69% of the variation in monthlyMHED presentations across public hospitals in South Australia. Thirty-two percent of the variation in current month’s male MHED presentations can be predicted by using the 2 months’ prior male unemployment rate. Over 63% of the variation in monthly female MHED presentations can be predicted by either male or female prior monthly unemployment rates.

Conclusions: The findings of this study highlight that even with the relatively favourable economic conditions, small shifts in monthly unemployment rates can predict variations in monthly MHED presentations, particularly for women. Monthly ABS unemployment rates may be a useful metric for predicting demand for emergency mental health services.

Back to top of page