Psychiatric Quarterly (2022) 93:915–933

https://doi.org/10.1007/s11126-022-10001-y

ORIGINAL PAPER

The Utilization of a Rapid Agitation Scale and Treatment

Protocol for Patient and Staff Safety in an Inpatient
Psychiatric Setting

Tessa Manning1 · Sarah Beth Bell1 · Drew Dawson2 · Krista Kezbers1,3 ·

Micheal Crockett1 · Ondria Gleason1

Accepted: 5 August 2022 / Published online: 25 August 2022

© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022

Abstract
Agitation is a common and potentially dangerous condition requiring rapid recognition and
treatment in acute psychiatric units. Prompt intervention can prevent a patient with agita-
tion from harming themselves, harming others, or needing restraints or seclusion. After the
review of numerous guidelines, the Modified Agitation Severity Scale (MASS) agitation
treatment protocol was developed to identify and manage agitation in an inpatient adult
psychiatric setting. This protocol involved modifying an existing agitation scale and pair-
ing scores with a treatment algorithm to indicate which behavioral and medication inter-
ventions would be most appropriate. All scoring and interventions were recorded in the
electronic medical record (EMR). Three months of data were collected before and after the
protocol was implemented. The new, modified scale had high reliability and correlated well
with another validated agitation scale. Perceived patient safety was high during both study
phases. Nurses’ perceptions of safety trended upward after the protocol was implemented,
though these differences were not significant, likely due to insufficient power. Although
there was no decrease in seclusion events after implementation of the treatment protocol,
there was a 44% decrease in restraint events and average restraint minutes per incident.
Despite a potential increase in workload for nursing staff, implementation of the proto-
col did not increase burnout scores. Physicians continued to order the protocol for 55% of
patients after the study period ended. These findings suggest that including a rapid agita-
tion assessment and protocol within the EMR potentially improves nurses’ perceptions of
unit safety, helps assess treatment response, reduces time patients spend restrained, and
supports decision making for nurses.

Keywords (4–6): agitation · Psychiatric medication · Quality improvement · Inpatient ·

Agitation severity scale · MASS

* Tessa Manning
Tessa-Manning@ouhsc.edu
1
School of Community Medicine, The University of Oklahoma, Tulsa, OK, USA
2
Oklahoma City Indian Clinic, 5208 W Reno Ave, Oklahoma City, OK, USA
3
Health Promotion Research Center, The University of Oklahoma Health Sciences Center,
Stephenson Cancer Center, Oklahoma City, OK, USA

13
Vol.:(0123456789)
916 Psychiatric Quarterly (2022) 93:915–933

Introduction

Patients with acute agitation are common in the inpatient psychiatric setting [1–4]. Agita-
tion, which is characterized by excessive or inappropriate motor or verbal activity, signals
the beginning of a behavioral emergency and is associated with several psychiatric con-
ditions [3]. When unmanaged, agitation can progress quickly from mild symptoms, such
as restlessness and anxiety, to severe symptoms, such as assault of staff or other patients.
This progression not only decreases feelings of safety, but also increases staff burnout [5].
Early recognition, appropriate assessment, and effective treatment of agitation can reduce
the risk for escalation to violence and injuries [6, 7].
Higher rates of patient agitation affect healthcare in terms of longer lengths of stay, more
readmissions, higher medication use, higher workers’ compensation costs, staff time off
from work due to injuries, and increased administration costs [6, 7]. In most circumstances,
non-pharmacological methods such as verbal de-escalation are helpful to decrease agita-
tion [1, 8–10]. However, when initial measures fail, medication has been shown to effec-
tively calm agitated behaviors and treat the underlying causes of agitation, such as anxi-
ety, mania, and psychosis [1, 9, 11–13]. In cases where agitation leads to violent behavior,
more coercive responses, including involuntary medication administration, seclusion, and
restraints, could be necessary [13]. However, these methods should generally be avoided
if possible, given evidence they can lead to physical injury and psychological trauma [14].
For this reason, efforts to improve early identification and reduction of agitation in hospi-
tals have become an integral component of quality assessment surveys [15, 16].
Multiple guidelines have been written to direct the assessment and treatment of agita-
tion in patients with acute psychiatric disorders [1, 4, 13, 17, 18]. These guidelines gener-
ally recommend that the physician evaluate the patient and attempt to determine the etiol-
ogy of agitation, that symptoms be assessed with the use of a standardized rating scale, and
that treatment for agitation be specific to the underlying cause [13, 19, 20]. If there is no
routine procedure for assessment of agitation, there could be delays in identification and
treatment, leading to escalation of more coercive management responses or resulting in
staff and patient injuries [13]. Having a standard agitation treatment protocol could prevent
these delays and provide more direction about when certain interventions are appropriate.
While several evidence-based assessment scales for agitation exist, many do not have
strong enough clinical utility to guide rapid treatment decisions. A recent review summa-
rized rating scales of psychomotor agitation for use in various treatment settings [21]. In
this review, the scales that best fit inpatient psychiatric settings and had strong psychomet-
ric properties took too long to administer to be used practically. The shorter scales with
strong psychometric properties were either not intended for or not appropriate for in-the-
moment assessment of an inpatient psychiatric population, especially because they gener-
ally did not describe specific problematic behaviors [21].
In the present study, we modified an existing descriptive agitation scale, the Agitation
Severity Scale [22], for rapid clinical assessment. Though the original scale assesses for
various, specific agitated behaviors and demonstrates strong content validity and strong
interrater reliability, administration of the scale takes an average of 10 min. After modi-
fication (described in the methods section, below), the scale could be administered in one
minute. This modified scale was then linked to a treatment protocol to guide nursing staff
decisions regarding both behavioral interventions and medication distribution. The MASS
agitation treatment protocol was implemented in an adult inpatient psychiatric setting for
three months. Study results were compared with three months of baseline data before

13
Psychiatric Quarterly (2022) 93:915–933 917

implementation. Outcomes measured included patient perceptions of safety, nurse percep-

tions of safety, nurse burnout, and scale convergence validity with another commonly used
agitation measure. Additional information regarding seclusion and restraint rates was also
collected for 18 months before and after protocol implementation.

Methods

This study was conducted at a psychiatric hospital in the midwestern United States on a
high-acuity, adult psychiatric unit following approval by the hospital system’s Institutional
Review Board. A workgroup of experienced professionals, including psychiatric physicians
and nurses, pharmacists, research specialists, and information technology professionals,
convened to design and carry out this study. There were five phases of study design: selec-
tion and modification of an agitation measurement scale, development of a treatment proto-
col including behavioral and medication interventions, development of nursing and patient
safety surveys, training of physician and nursing staff, and data collection upon implemen-
tation into the electronic medical record (EMR).

Development of a Modified Version of the Agitation Severity Scale (MASS)

With permission from the author of the Agitation Severity Scale, a team of psychiatrists
collaborated with a PhD-level biostatistician and an experimental psychologist to revise the
scale. The original version of the Agitation Severity Scale [22] measured five domains of
agitation: nonverbal facial expressions, verbal behaviors, purposeful motor behaviors, non-
purposeful motor behaviors, and interpersonal behaviors. For each of these five domains,
4 to 6 specific behaviors were listed, for a total of 25 items. These behaviors were ranked
on Likert scales with scores ranging from 0 to 75 (see supplementary materials for original
Agitation Severity Scale).
In creating the Modified Agitation Severity Scale (MASS) (Table 1), the following
changes were made: 1) items were organized by severity level rather than domain; 2) each
item was ranked as either present or absent, rather than assessed by Likert scale; and 3)
scores on items were weighted. These modifications decreased the time for completion
from ten minutes to about one minute. Weighting scale items was the most significant
modification. On the original scale, a behavior like pacing held the same weight as did a
behavior like screaming. In the modified scale, psychiatrists on the research team relied on
clinical experience to categorize behaviors into severity weights (very mild behaviors = 1,
mild behaviors = 2, moderate behaviors = 3, and severe behaviors = 10) for a total score
ranging from 0–83.

Development of the MASS Agitation Treatment Protocol

Multiple guidelines for the behavioral and pharmacological treatment of acute agitation
were thoroughly reviewed by the research team and used to develop the study’s agitation
treatment protocol (Table 2) [1, 11–13, 20, 21]. A list of recommended non-pharmacologic
interventions was formulated based on these guidelines and with the help of nursing staff.
Medication use was encouraged if non-pharmacologic nursing measures were not suc-
cessful, with the purpose of calming the patient without causing over sedation. Based on
the most commonly recommended medications in recent agitation guidelines, the MASS

13
918 Psychiatric Quarterly (2022) 93:915–933

Table 1  Modified Agitation Severity Scale

Weighted Value Agitation Symptom Point Value
if Symptom
Present

1 Red in the Face 1

Grinding Teeth 1
Tapping, Clenching, Involuntary Movements of Hands 1
Disorganized Thinking 1
Emotionally Labile, Anxious, Nervous 1
Unable to Reason 1
2 Spitting 2
Getting Attention with Voice 2
Darting Eyes 2
Demanding 2
Speaking more Quickly than Baseline 2
Angry Tone of Voice 2
Persistent Disruptive Verbalizations 2
Restless 2
3 Yelling, Audibly Louder than Baseline 3
Unable to be Calmed 3
Inappropriate Behavior 3
“In Your Face” 3
Decreased Self-Control, Impulsiveness 3
“Puffed up”, Chest Out, Threatening Posture 3
Confrontational 3
10 Biting 10
Throwing Objects 10
Violating Self or Others 10
Physical Violence to Self or Others 10
TOTAL SCORE (Maximum score 83)
Agitation Severity: 1 to 4: Very mild agitation
4 to 6: Mild agitation
7 to 9: Moderate agitation
Greater than 9: Severe agitation and/or violent

Total score is a sum of each present symptom multiplied by the weighted value for the section

agitation treatment protocol incorporated four different pharmacologic tracks: a benzodiaz-

epine (lorazepam), a first-generation antipsychotic (haloperidol), and a second-generation
antipsychotic (olanzapine; with a low-dose option and a standard dose option). The patient’s
physician determined which medication track was ordered upon admission, based on the
most likely etiology of the patient’s agitation. An oral medication was preferred over the
intramuscular route if the patient was willing to accept it. Intramuscular medication was
only administered in the event the patient was a danger to themselves or others and was
preferred to be utilized before seclusion or restraint, which were only utilized as a last resort
and required notification of the physician.
The MASS was scored by the nurses at the time of admission and then twice daily as
long as the protocol was ordered. The MASS scores were entered directly into a flowsheet

13
 Table 2  MASS Agitation Treatment Protocol
Assess and record MASS now, then at 0900 and 2100 while awake
• Calculate MASS score based on behavior observed within the most recent 1 h
• If medication given, repeat MASS hourly until score is less than 4, then resume at 0900 and 2100 scoring schedule

Agitation Score Severity Possible Treatment Interventions

1–3 Very Mild Behavioral Interventions
4–6 Mild Behavioral Interventions Oral Medication
7–9 Moderate Behavioral Interventions Oral Medication Intramuscular Injection
>9 Severe and/or Violent Behavioral Interventions Oral Medication Intramuscular Injection Seclusion or
Restraint
Psychiatric Quarterly (2022) 93:915–933

Behavioral Interventions
Speak with patient about frustration Encourage patient to channel feelings into activity
Identify wants and feelings Bring in alternative staff person
Quiet room Staff directed time out
Encourage use of coping skills patient identifies as helpful Redirect attention and offer choices
Encourage patient to engage in relaxation techniques Offer fluids
Encourage self-time out Offer food
Remove provoking stimuli Offer nicotine replacement
Medication Interventions
Choose one of the following:
Ο Lorazepam
• 1 mg Oral six times daily prn for MASS score of 4 to 9; (Max 8 mg in 24 h)
• 1 mg Intramuscular six times daily prn for MASS score of 7 to 9 and patient unable/unwilling to accept oral medications (Max 8 mg in 24 h)
• 2 mg Oral four times daily prn for MASS score of 10 or above (Maximum 8 mg in 24 h)
• 2 mg Intramuscular four times daily prn if MASS score 10 or above and patient unable/unwilling to accept oral medications: (Max 8 mg in 24 h)
Ο Haloperidol
• 2 mg Oral six times daily prn for MASS score of 4 to 9; (Max 20 mg in 24 h)
• 2 mg Intramuscular six times daily prn for MASS score of 7 to 9 and patient unable/unwilling to accept oral medications (20 mg in 24 h)
• 5 mg Oral four times daily prn if MASS score 10 or above; (Max 20 mg in 24 h)

13
919
 Table 2  (continued)
920

Assess and record MASS now, then at 0900 and 2100 while awake
• Calculate MASS score based on behavior observed within the most recent 1 h

13
• If medication given, repeat MASS hourly until score is less than 4, then resume at 0900 and 2100 scoring schedule
• 5 mg Intramuscular four times daily prn if MASS score 10 or above and patient unable/unwilling to accept oral medications (Max 20 mg in 24 h)
Ο Olanzapine (Low Dose): Lower doses should be considered for the elderly, patients with low body weight, dehydration, and no previous exposure to antipsychotic medication
• 2.5 mg Oral three times daily prn for MASS score of 4 to 9; (Max 20 mg in 24 h)
• 2.5 mg Intramuscular three times daily prn for MASS score of 7 to 9 and patient unable/unwilling to accept oral medications (Max 20 mg in 24 h)
• 5 mg Oral three times daily prn for MASS score of 10 or above (Max 20 mg in 24 h)
• 5 mg Intramuscular three times daily prn if MASS score 10 or above and patient unable/unwilling to accept oral medications (Max 20 mg in 24 h)
Ο Olanzapine (Standard Dose)
• 5 mg Oral three times daily prn for MASS score of 4 to 9; (Max 30 mg in 24 h)
• 5 mg Intramuscular three times daily prn for MASS score of 7 to 9 and patient unable/unwilling to accept oral medications (Max 30 mg in 24 h)
• 10 mg Oral three times daily prn for MASS score of 10 or above (Maximum 30 mg in 24 h)
• 10 mg Intramuscular three times daily prn if MASS score 10 or above and patient unable/unwilling to accept oral medications (Max 30 mg in 24 h)
Notify Physician: Nurse to contact on-call physician if
• Maximum daily dose met for agitation medication,
• 3 or more doses of MASS protocol medication given in less than 4 h, or
• Concerns for acute muscle stiffness
Psychiatric Quarterly (2022) 93:915–933
Psychiatric Quarterly (2022) 93:915–933 921

within the EMR, with the score ranges serving as a decision tree for selecting various man-
agement strategies from the associated agitation treatment protocol (Table 2). For scores
of 0 to 3 (very mild), behavioral interventions such as speaking to the patient about their
frustrations, taking the patient to a quiet room, or offering nicotine replacement therapy
(for patients with a history of nicotine dependence) were recommended. For scores of 4
to 6 (mild), oral medication could be offered in addition to the previous recommendations.
For scores 7 to 9 (moderate), the nurse could utilize any of the previous recommendations,
as well as intramuscular injection of medication if the patient was acutely dangerous. For
scores 9 and above, seclusion or restraint could be administered if needed, with physician
notification.
Nursing staff recorded the treatment interventions in the EMR. Data related to any
involuntary treatments, such as medication administration, seclusion, and restraint, were
carefully documented and administered according to federal and state requirements. Nurses
were instructed to contact the on-call physician for acute muscle stiffness, three or more
MASS agitation treatment protocol medication doses administered in less than four hours,
or if the maximum daily dose was met for agitation medication.

Implementing the Modified Agitation Severity Scale (MASS) Protocol

Study data were collected in two blocks consisting of a baseline phase and an intervention
phase, both lasting three months (Fig. 1). At the beginning of the baseline phase, nurs-
ing staff were asked to voluntarily and anonymously complete a nursing safety survey
and burnout inventory. Next, three months of baseline data, including voluntary patient
safety surveys at discharge, were collected. After the baseline phase was complete, the new
MASS agitation scale and protocol were implemented as a physician order in the EMR.
Per the protocol, the nursing staff would routinely record agitation scores and medication
administered until patient discharge. As in the baseline phase, voluntary patient safety sur-
veys were administered at discharge.
After the intervention phase, nurses again completed a safety survey and burnout inven-
tory. Additional outcome measures collected during and after the study period included
total number of seclusion and restraint events. Demographic variables, urine drug screen
results, length of stay, and discharge psychiatric diagnoses were also collected through the
EMR for patients admitted during the study period.

Patient and Nurse Surveys

The patient and nursing safety surveys were developed by the research team and are avail-
able in the supplementary materials. All patients were asked if they were willing to vol-
untarily complete a patient safety survey on the day of discharge during both the baseline
and the intervention periods. This survey had eight items that were rated from 1 to 5, with
lower scores indicating positive perceptions of safety. Example items included “The unit
I stayed on was safe” and “Staff noticed my frustration as it was occurring or soon after.”
The nursing survey given at baseline had six items that were rated from 1 to 5, with
lower scores also indicating positive perceptions of safety. Example items included “I feel,
as a staff member, safe on the unit” and “I feel agitated patients get appropriately medi-
cated.” The post-intervention phase nursing survey had the original items plus four addi-
tional items: “A uniform agitation scale (MASS) is helpful,” “I understand our agitation

13
922 Psychiatric Quarterly (2022) 93:915–933

Fig. 1  Study Design

scale,” “I use the MASS to assess agitation,” and “I communicate MASS scores to the
team.”
The Oldenburg Burnout Inventory was used to assess nurse burnout at baseline and
post-intervention [23]. This scale has two dimensions, exhaustion and disengagement, as
well as a total score. Nurses were not followed over time, but rather were compared inde-
pendently due to the relative frequency of staff turnover.

Convergence Validity Measure

This study also correlated the MASS scale with another evidence-based agitation assess-
ment, the Clinical Global Impressions-Aggression Scale (CGI-A), which has been used in
emergency psychiatric settings [21]. The CGI-A consists of one item, “level of agitation,”
and is rated as none, slight, moderate, severe, and aggressive. Although a simple measure,
it does not describe individual symptoms or characterize agitation well for communication
or historical review. However, its brevity and clinical use made it a suitable scale to use as
a comparison measuring convergent validity in the revised MASS scale.

13
Psychiatric Quarterly (2022) 93:915–933 923

Statistical Analyses

Independent sample t-tests were used to compare nurse and patient survey responses at
baseline and following intervention, with one test comparing the total score and additional
tests comparing each item. The Oldenburg Burnout Inventory (total score and each dimen-
sion separately) was compared at baseline and following the intervention, also using an
independent samples t-test. Correlations between the CGI-A and the MASS were used to
determine how strong the relationship was between these two scales. These correlations
included the initial MASS score versus the initial CGI-A score and the average MASS
score versus the average CGI-A score throughout a patient’s stay. In addition, the CGI-A
was correlated with each individual MASS score given through the duration of each
patient’s stay (reported in the supplemental materials of this paper). The MASS scale was
correlated with age, gender, and length of stay. In addition, MASS scores were correlated
with patient diagnosis and urine drug screen results to measure potential sub-population
patterns.

Results

Demographics

This study included EMR data from 742 patients, with 353 in the baseline phase and 389
in the intervention phase (Table 3). Of the patients in the intervention phase, 243 received
the MASS agitation treatment protocol, leaving 110 patients for whom the protocol was
not ordered. Some patients had a very short length of stay and were discharged before the
scale could be utilized. For others, there may have been clinical reasons that the attending
physician did not feel the protocol was appropriate, such as for patients who were in active
alcohol withdrawal and were being treated with other protocols with similar medications.
Other patients were likely missed due to physician oversight.

Table 3  Demographics Age

Mean = 35.76 SD = 12.43

Gender
Male 50.1% (n = 372)
Female 49.9% (n = 370)
Race
Asian 0.7% (n = 5)
Black or African American 11.1% (n = 82)
Indigenous 8.6% (n = 64)
White 74.2% (n = 551)
Other 3.9% (n = 29)
Unknown 1.5% (n = 11)
Ethnicity
Hispanic 3.8% (n = 28)
Non-Hispanic 96.2% (n = 714)

13
924 Psychiatric Quarterly (2022) 93:915–933

There were no significant differences in any demographic variables in the baseline and
intervention periods of the study. The average length of stay was 6.51 days with a standard
deviation of 5.42 days. There were no significant differences in length of stay in the pre-
and post-periods of the study.

Nurse Survey

The Nurse Survey contained six individual items that were assessed at baseline and post-
intervention, with an additional four individual items added to the post-intervention survey.
Twenty-four nursing staff completed the baseline survey and the Oldenburg Burnout Scale.
Twenty-two nursing staff completed the post-intervention surveys. Figure 2 shows the per-
cent of nurses who “agreed” or “strongly agreed” with each item. There were no significant
differences between study phases on any of the nurse survey questions; however, a higher
percentage of nurses “agreed” or “strongly agreed” on each question after the intervention
period. Results from the second nurse survey showed that 73% of nurses reported they
either “agreed” or “strongly agreed” that “A uniform (MASS) agitation scale is helpful.”
There were no significant differences on the Oldenburg Burnout Scale at baseline and fol-
lowing the intervention (baseline mean = 36.17 (SD = 6.74), post intervention mean = 36.11
(SD = 8.29), t(35) = 0.03, p = 0.980).

Patient Survey

On the Patient Safety Survey, independent samples t-tests showed no significant pre- and
post- differences on the overall scale or any of the items (overall scale mean baseline 12.20
(SD = 6.38), mean post-intervention 13.25 (SD = 8.43), t(402) = -1.29, p = 0.199). Most patients
“agreed” or “strongly agreed” with each safety question (82.9% to 92.7% for each ques-
tion). The scale reliability as measured by internal consistency was high, with a Cronbach’s
alpha of 0.96 for the overall scale, 0.94 for the pre-period, and 0.97 for the post-period. 75.4%
of patients (266 of 353 patients) completed the survey in the baseline period, while only 36.8%
of patients (143 of 389 patients) completed the survey in the intervention period.

MASS (Modified Agitation Severity Scale)

The MASS has a score range from 0–83. Figure 3 shows the percent of patients who scored
in each severity category initially on the date of admission. The average initial MASS
score was 3.78 with a standard deviation of 7.49, indicating overdispersion in the data
due to a high amount of variability. The MASS did not significantly correlate with age,
r(241) = 0.02, p = 0.765. An independent samples t-test showed no gender differences on
MASS score, mean female 4.26 (SD = 9.00), mean male 3.23 (SD = 5.28) (no other genders
were in the EMR), t(241) = -1.07, p = 0.287. The MASS did not significantly correlate with
length of stay, r(241) = 0.03, p = 0.598.

MASS Correlation with the CGI‑A Agitation Scale

Average initial MASS scores were collected from 243 patients. Average initial CGI-A
scores were collected from 186 of those patients. Fewer CGI-A scores were collected
than MASS scores due to a technical problem with making the CGI-A a required item in

13
Psychiatric Quarterly (2022) 93:915–933 925

Fig. 2  Nursing Survey Results

the associated flowsheet of the EMR, though nursing staff were strongly encouraged to
complete both scales. The average initial MASS score had a strong correlation with the
CGI-A score, r(184) = 0.75, p < 0.001. The maximum number of times the MASS was
scored for a patient was 65. The minimum correlation between MASS and CGI-A was
0.620, and the maximum was 1.00. All correlations between the MASS score and the
CGI-A score except for 4 were significant (see supplementary materials).

13
926 Psychiatric Quarterly (2022) 93:915–933

Fig. 3  Initial MASS Scores

MASS Correlations with Psychiatric Diagnoses and Substance Use

Ten psychiatric diagnoses were assessed for correlation with the MASS (see supplementary
materials). Anxiety was negatively correlated with MASS scores, r(65) = -1.13, p = 0.044,
indicating a small to medium negative correlation. The remaining diagnoses that did not
correlate with MASS scores included major depressive disorder, bipolar disorder, psycho-
sis, post-traumatic stress disorder, traumatic brain injury, personality disorder, autism spec-
trum or neurodevelopmental disorder, substance use disorder, and substance withdrawal.

Seclusions and Restraints

Differences in seclusion and restraint events (both total incidents and minutes) were com-
pared in the 18 months before and after implementation of the MASS agitation treatment
protocol (see Fig. 4). Outliers were excluded if they were three standard deviations away
from the mean. These outliers included the following: a 145-min restraint and a 239-min
restraint before the MASS was implemented, a 143-min restraint after the MASS was
implemented, and a 580-min seclusion before the MASS was implemented.
With regard to restraints, there was a 44.1% nonsignificant decrease in total incidents,
with 68 incidents before the protocol was implemented and 38 incidents after implemen-
tation. There was also a 44.4% significant decrease (t(104) = 2.00, p = 0.047, effect size
d = 0.41 [medium]) in average restraint minutes per incident in the 18 months after the
MASS was implemented. Before the protocol, restraint incidents lasted an average of

13
Psychiatric Quarterly (2022) 93:915–933 927

Fig. 4  Average Restraint Minutes

per incident

18 min with a standard deviation of 22. After the protocol was implemented, the incidents
decreased to an average of 10 min with a standard deviation of 15 min. Figure 4 shows the
decreased variability in restraint minutes following MASS implementation.
There were no significant changes in seclusion incidents or minutes after MASS imple-
mentation. There were 22 seclusion incidents in the 18 months before protocol implemen-
tation and 28 seclusion incidents in the 18 months after protocol implementation. The aver-
age minutes for seclusion events in the 18 months before MASS implementation was 132
with a standard deviation of 141 min. After MASS implementation, the average minutes
for seclusion events was 137 min with a standard deviation of 97 min.

Discussion

In the inpatient psychiatric setting, most patient behavior throughout the stay will be
observed and documented by the nursing staff. This information is then typically reported
to the physician and other treatment team members, through either chart notations or inter-
disciplinary team meetings. However, it is common for these communications to include
only vague descriptions of behaviors and even for information to be lost during shift
changes, making this typical standard of care less helpful for treatment decision making.
A prominent benefit of this protocol was the ability for all team members to see patient
MASS scores in the EMR flowsheet, along with the behavioral and medication interven-
tions utilized. These scores and interventions could then be tracked over the course of
the patient stay to guide treatment and measure effectiveness. It is also common in the
inpatient psychiatric setting for the physician to place standing orders for medication to
be given for agitation as needed. However, there is often no further direction given to the
nursing staff about when medication administration is appropriate. This protocol offers
nursing staff guidance about when these specific interventions should be considered. It is
also possible for patients to become unexpectedly agitated on the unit after hours when the
physician has left the hospital. In these cases, the nurses may have to contact the physician
and wait on additional orders to be placed, causing delays in treatment, which can lead to

13
928 Psychiatric Quarterly (2022) 93:915–933

dangerous consequences. Having a pre-ordered protocol for agitation treatment could pre-
vent these delays.
After modifying the Agitation Severity Scale, the research team sought to determine
whether the new, more time efficient version would correlate with another well-validated
agitation scale. Results showed that the MASS had a very large correlation, high reliability,
and strong convergent validity with the CGI-A scale, which is commonly used in the emer-
gency psychiatric setting. This finding indicates that the MASS score is in agreement with
the one-item CGI-A score regarding the severity of agitation being measured. Therefore,
the MASS is able to capture level of agitation very similarly to the CGI-A, while providing
the detail needed for treatment decisions. Rates of coercive treatments, such as restraint
and seclusion events, were also carefully recorded to determine the possible efficacy of
the new protocol. Although there was no change in seclusion events, restraint frequency
and minutes sizably decreased in the 18 months following MASS agitation protocol imple-
mentation. The effect could be due to routine assessments leading to earlier recognition of
agitation and needed intervention so that more coercive measures were needed less often.
This finding could also reflect that nursing staff had easier access and more direction about
when medication intervention was appropriate. This decrease in restraint episodes likely
decreased patient suffering and increased feelings of safety for all involved. Restraint min-
utes were also much less variable after implementation of the MASS agitation treatment
protocol. This reduced variability may suggest that patients were being treated more sys-
tematically, and that this systematic treatment may have resulted in shorter and less vari-
able restraint time. It is unclear why there was no significant change in seclusion events.
However, it could be related to the relatively low frequency of these events at baseline
compared with restraint events. Therefore, a larger sample size might be needed to find any
differences.
Data were also collected to determine whether MASS scores were correlated with sev-
eral different psychiatric diagnoses. Results showed that MASS scores were negatively cor-
related with anxiety, but not with any other psychiatric disorder. This finding could reflect a
tendency for patients with high anxiety to be more withdrawn in this setting and less likely
to have aggressive behaviors. However, it was surprising that there were no correlations
with other disorders, given the large literature base that indicates that mental health disor-
ders, especially acute mania and psychosis, are associated with higher degrees of violence
[24, 25]. This lack of correlation may be related to the highly variable nature of the data,
with a standard deviation so variable that it was over dispersed. With this level of variabil-
ity, a larger sample size might be needed to detect potential differences.
This study also specifically assessed nursing staff feelings of safety, as they spend the
most face-to-face time with patients on the unit. Caring for patients displaying agitated
behaviors can be dangerous, stressful, and demanding, and prevents nurses from attend-
ing to the other numerous duties that are expected of them. Although there were no sig-
nificant differences found on the nursing staff safety survey, perceptions of safety after
the protocol was initiated trended in the positive direction on all questions (Fig. 2). In
the post-intervention survey, nursing staff reported overwhelmingly that they thought the
protocol was helpful. This is likely because the protocol provided more specific directions
about how to manage patients with agitation, whereas there was no specific guidance in
the baseline phase. Having standing orders for medication also likely improved feelings of
safety for nurses. Having to contact a physician after hours for medication orders can be
onerous for nursing staff and can delay the ability to provide treatment to a patient with
acute agitation, leading to multiple safety concerns.

13
Psychiatric Quarterly (2022) 93:915–933 929

The burden associated with caring for patients displaying agitated behaviors is well
known and includes nursing distress and burnout [5]. Agitated patient behaviors have been
linked with increased absenteeism, increased staffing turnover, and negative reactions
towards patients [5]. The present study’s findings showed no significant differences in nurs-
ing perception of burnout as measured on the Oldenburg Burnout Scale before and after
implementation of the protocol. This result is of particular interest given that the nursing
staff were asked to perform the additional daily task of scoring the MASS multiple times a
day for each patient, effectively increasing their workload. However, this extra work might
have been offset by the nurses having to do fewer emergent tasks, such as contacting a phy-
sician for medication orders after the patient had already become agitated, with potentially
dangerous delays in treatment.
Patients’ perceptions of safety were not significantly different before or during the study
intervention period. All patients were asked to complete this questionnaire voluntarily at
discharge, but many patients chose not to respond, leaving many potential responses unre-
corded, especially in the intervention period. This discrepancy is likely also the result of
the stress of the beginning of the COVID-19 pandemic during the last few weeks of the
intervention period. The nursing staff may not have offered the survey to every patient on
discharge due to an increase in workload related to changing pandemic-related require-
ments. On average, patients who did complete the survey answered that they either
“strongly agreed” or “agreed” on each of the seven items of the survey, indicating that
even before the protocol was implemented, they overwhelmingly felt safe. This suggests a
potential ceiling effect on patients’ perceptions of safety, although the voluntary nature of
the survey presents a risk of selection bias.
Physicians’ perceptions of the utility of the scale and protocol were not directly meas-
ured. Many of the physicians who work on the unit were involved in the development of
the study, and any survey of physicians would likely introduce additional bias. However,
post-study analysis of the EMR showed that in the 20 months after the study period ended,
physicians continued to order the MASS agitation treatment protocol for 55% of admit-
ted patients. It is likely that the protocol is being ordered specifically for patients who the
physician thinks may become agitated during their hospital stay. However, further research
is needed to determine in what cases and for what reasons physicians decide to order the
protocol.

Limitations and Future Directions

An important limitation of this study relates to comparing the amount of medicine given
for agitation in the baseline period and the intervention period. In the baseline period, phy-
sicians did not always use the term “agitation” or similar terms (i.e., aggression, violence)
when ordering medicine for these purposes. Instead, they sometimes used terms such as
“anxiety” or “mania.” Therefore, through EMR review it was not possible to differentiate
when as-needed medication was given for agitation-related reasons in the baseline period.
We did record as-needed protocol medications given per day during the intervention period
and could use these data in potential follow-up studies of this scale and protocol.
Another limitation was the underpowered nature of the nurse survey. Even though the
responses to each question trended in the positive direction, no statistically significant dif-
ferences were demonstrated. This survey was especially underpowered because individual
nurse responses were not tracked over time; instead, independent samples were used.

13
930 Psychiatric Quarterly (2022) 93:915–933

A final limitation is that data collection was stopped immediately before the COVID-19 pan-
demic began in March of 2020. This means that the post-intervention surveys were given dur-
ing the start of the COVID pandemic in early 2020. It is possible that perceptions of safety on
the unit were negatively affected by the threats of the virus, rather than patient agitation. Addi-
tional stressors during this time included the institution of COVID screening and testing, mask
availability, new isolation protocols, general fear of contracting the virus, concerns about loss
of income due to need to quarantine, and patient and staff refusal to wear masks. Interestingly,
the only study measure collected during the COVID-19 pandemic was the nurse survey, and
responses on every question in the nurse safety survey directionally improved. However, staff
turnover on this unit was relatively high during this period. This turnover could be an indicator
of burnout that was not captured in our survey.
Future studies of this protocol could include the evaluation of the response of underly-
ing psychiatric symptoms, such as the speed of resolution of mania, psychosis, or anxiety.
Items in the scale that were used less often could be removed in future iterations, stream-
lining this measure. In addition, future versions of this study could have an experimental
design allowing for accurate comparisons of the amount of medication given with the pro-
tocol. It could also be useful to pilot the protocol with different and more varied popula-
tions, including geriatric populations and emergency department populations, to examine
its utility in multiple medical settings.
After approximately two years of utilizing the MASS agitation treatment protocol, the
state health care authority asserted that intramuscular (IM) medication should be removed
from the protocol after a routine audit. A further discussion and revised protocol are pre-
sented in the supplementary materials.

Conclusions

The MASS agitation scale and corresponding treatment protocol were successfully imple-
mented over a three-month period in an adult inpatient psychiatric population. The scale
had high internal consistency and correlated well with another validated agitation scale.
This rapid assessment measure resulted in nurses regularly and systematically measur-
ing agitation symptoms in patients for the purpose of initiating behavioral and medica-
tion treatments as early as possible, with the goal of avoiding more severe symptoms and
coercive treatments. The protocol provided clear instructions to nursing staff about what
treatments could be administered and when they should be administered in the absence of
direct physician oversight, potentially preventing delays in treatment. Nurses did not feel
more burned out when they were asked to do the additional task of using the scale, and
the post intervention nurse survey indicated that the nurses found the protocol to be help-
ful. Following the study period, physicians continued to order the protocol for a major-
ity of patients, indicating they found it useful for treatment of patients at risk for agitated
behavior. Furthermore, there has been a clear trend of decreasing frequency and length
of restraint events since the protocol was implemented, thereby increasing unit safety for
patients and staff. These findings suggest that including a rapid agitation assessment such
as the MASS and a corresponding treatment protocol within the EMR allows for agitation
to be quickly detected and addressed, reduces restraint events, increases nurse feelings of
safety, and simplifies individual decision making for nurses with regard to managing agi-
tated patients.

13
Psychiatric Quarterly (2022) 93:915–933 931

Supplementary Information The online version contains supplementary material available at https://​doi.​
org/​10.​1007/​s11126-​022-​10001-y.

Acknowledgements We thank Dr. Jeffrey S. Grantham, MD, Kim Clark, Kassi Davis, Dr. Emily Gray,
PharmD, Dr. Jerrod Spring, MD, Renova Uwingabire, Jim Lightbody, and Teddi Beard for their contribu-
tions to this project.

Funding This research received no specific grant from any function agency in the public, commercial, or
not-for-profit sectors.

Declarations
Ethics Approval/Consent to Participate Both patients and nursing staff were provided with a written consent
form before completing surveys. Survey responders were not provided with any compensation and no identi-
fying information was collected. The authors assert that all procedures contributing to this work comply with
the ethical standards of the relevant national and institutional committees on human experimentation and with
the Helsinki Declaration of 1975, as revised in 2008. All procedures involving human subjects/patients were
approved by the Saint Francis Health System Institutional Review Board.

Conflict of Interest The authors have no relevant financial or non-financial interests to disclose.

References
1. Garriga M, et al. Assessment and management of agitation in psychiatry: Expert consensus. World J
Biol Psychiatry. 2016;17(2):86–128. https://​doi.​org/​10.​3109/​15622​975.​2015.​11320​07.
2. Caruso R, et al. Aggressive behavior and psychiatric inpatients: a narrative review of the literature
with a focus on the European experience. Curr Psychiatry Rep. 2021;23(5). https://​doi.​org/​10.​1007/​
s11920-​021-​01233-z.
3. Cornaggia CM, et al. Aggression in psychiatry wards: A systematic review. Psychiatry Res.
2011;189(1):10–20. https://​doi.​org/​10.​1016/j.​psych​res.​2010.​12.​024.
4. Hankin CS, Bronstone A, Koran LM. Agitation in the inpatient psychiatric setting: a review of clinical
presentation, burden, and treatment. J Psychiatr Pract. 2011;17(3):170–85. https://​doi.​org/​10.​1097/​01.​
pra.​00003​98410.​21374.​7d.
5. Adams AMN. Caring for patients displaying agitated behaviours in the intensive care unit - A mixed-
methods systematic review. Austrailian Critical Care. 2021;35(4):454–65. https://​doi.​org/​10.​1016/j.​
aucc.​2021.​05.​011.
6. Hunter M, Carmel H. The cost of staff injuries from inpatient violence. Hosp Community Psychiatry.
1992;43(6):586–8. https://​doi.​org/​10.​1176/​ps.​43.6.​586.
7. Rubio-Valera M, et al. Health service use and costs associated with aggressiveness or agitation
and containment in adult psychiatric care: a systematic review of the evidence. BMC Psychiatry.
2015;15:35. https://​doi.​org/​10.​1186/​s12888-​015-​0417-x.
8. Richmond J, et al. Verbal De-escalation of the agitated patient: consensus statement of the Ameri-
can association for emergency psychiatry project BETA De-escalation workgroup. West J Emerg Med.
2012;13(1):17–25. https://​doi.​org/​10.​5811/​westj​em.​2011.9.​6864.
9. Patel MX, et al. Joint BAP NAPICU evidence-based consensus guidelines for the clinical management
of acute disturbance: De-escalation and rapid tranquillisation. J Psychopharmacol. 2018;32(6):601–40.
https://​doi.​org/​10.​1177/​02698​81118​776738.
10. Martínez-Raga J, et al. 1st international experts’ meeting on agitation: conclusions regarding the cur-
rent and ideal management paradigm of agitation. Front Psych. 2018. https://​doi.​org/​10.​3389/​fpsyt.​
2018.​00054.
11. Bak M, et al. The pharmacological management of agitated and aggressive behaviour: A systematic
review and meta-analysis. Eur Psychiatry. 2019;57:78–100. https://​doi.​org/​10.​1016/j.​eurpsy.​2019.​01.​
014.
12. Battaglia J. Pharmacological management of acute agitation. Drugs. 2005;65(9):1207–22. https://​doi.​
org/​10.​2165/​00003​495-​20056​5090-​00003.

13
932 Psychiatric Quarterly (2022) 93:915–933

13. Wilson M, et al. The psychopharmacology of agitation: consensus statement of the american asso-
ciation for emergency psychiatry project BETA psychopharmacology workgroup. West J Emerg Med.
2012;13(1):26–34. https://​doi.​org/​10.​5811/​westj​em.​2011.9.​6866.
14. Wong AH, et al. Experiences of individuals who were physically restrained in the emergency depart-
ment. JAMA Netw Open. 2020;3(1):e1919381. https://​doi.​org/​10.​1001/​jaman​etwor​kopen.​2019.​19381
15. Knox D, Holloman G. Use and avoidance of seclusion and restraint: consensus statement of the amer-
ican association for emergency psychiatry project BETA seclusion and restraint workgroup. West J
Emerg Med. 2012;13(1):35–40. https://​doi.​org/​10.​5811/​westj​em.​2011.9.​6867.
16. Relias Media. New Regs on Restraint use Follow JCAHO’s Lead. 1999. https://​www.​relia​smedia.​com/​
artic​les/​43937-​jcaho-​restr​aint-​guide​lines-​follow-​other-​groups-​8217-​leads.
17. Holloman JGH, Zeller SL. Overview of Project BETA: Best practices in Evaluation and Treatment of
Agitation. West J Emerg Med. 2012;13(1):1–2. https://​doi.​org/​10.​5811/​westj​em.​2011.9.​6865.
18. Vieta E, et al. Protocol for the management of psychiatric patients with psychomotor agitation. BMC
Psychiatry. 2017;17(1). https://​doi.​org/​10.​1186/​s12888-​017-​1490-0.
19. Nordstrom K, et al. Medical evaluation and triage of the agitated patient: consensus statement of the
American association for emergency psychiatry project BETA medical evaluation workgroup. West J
Emerg Med. 2012;13(1):3–10. https://​doi.​org/​10.​5811/​westj​em.​2011.9.​6863.
20. Roppolo LP, et al. Improving the management of acutely agitated patients in the emergency department
through implementation of Project BETA (Best Practices in the Evaluation and Treatment of Agitation).
J Am Coll Emerg Physicians Open. 2020;1(5):898–907. https://​doi.​org/​10.​1002/​emp2.​12138.
21. Pompili M, et al. The management of psychomotor agitation associated with schizophrenia or bipolar
disorder: a brief review. Int J Environ Res Public Health. 2021;18(8):4368. https://​doi.​org/​10.​3390/​
ijerp​h1808​4368.
22. Strout TD. Development of an agitation rating scale for use with acute presentation behavioral man-
agement patients [Doctoral dissertation, Connell School of Nursing]. Boston College. 2011. http://​hdl.​
handle.​net/​2345/​1839.
23. Demerouti E, Bakker AB. The oldenburg burnout inventory: a good alternative to measure burnout
and engagement. In Jonathan R.B. Halbesleben (Ed.), Handbook of Stress and Burnout in Health Care.
2008;65–78. Nova Science Publishers.
24. Whiting D, Lichtenstein P, Fazel S. Violence and mental disorders: a structured review of associa-
tions by individual diagnoses, risk factors, and risk assessment. Lancet Psychiatry. 2021;8(2):150–61.
https://​doi.​org/​10.​1016/​S2215-​0366(20)​30262-5.
25. Volavka J. Violence in schizophrenia and bipolar disorder. Psychiatr Danub. 2013;25(1):24–33.

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.

Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the
author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is
solely governed by the terms of such publishing agreement and applicable law.

Dr. Tessa L. Manning, MD is an Assistant Professor at the University of Oklahoma School of Community
Medicine where she serves as the Medical Director of Consultation-Liaison Psychiatry and Associate Pro-
gram Director in the Psychiatry Department. She received her medical degree and completed Psychiatry
residency at the University of Texas Southwestern Medical School in Dallas. She is board certified in Psy-
chiatry by the American Board of Psychiatry and Neurology and in Addiction Medicine by the American
Board of Preventative Medicine. Previous publications are in the areas of Consultation-Liaison Psychiatry
and Psychiatry education. Her current clinical work involves the care of Adult Psychiatry and Addiction
patients in both inpatient and outpatient settings. Her interests include improving medical student and resi-
dent education, psychiatric care of the medically ill, and reducing stigma and inequalities for patients with
mental illness and addiction.

Dr. Sarah Beth Bell serves in a senior scientist role in the central research office at the School of Community
Medicine in the University of Oklahoma Health Sciences Center. She received her PhD in social neuro-
science from the University of Kentucky. Her research lines center around pain and rejection, aggressive
behavior, equity for diverse populations, and social determinants of health. She has a clinical background in
trauma informed care. Her interests include student-centered learning and advocacy for people experienc-
ing bias and discrimination.

13
Psychiatric Quarterly (2022) 93:915–933 933

Dr. Drew Dawson, MD, MBA is an outpatient child and adolescent psychiatrist at the Oklahoma City Indian
Clinic. He received his medical degree from the University of Oklahoma Health Sciences Center. Later
he went on to complete his psychiatry internship at the Medical University of South Carolina followed by
general psychiatry residency and child and adolescent psychiatry fellowship at the University of Oklahoma
School of Community Medicine. He is board certified by the American Board of Psychiatry and Neurol-
ogy in both general psychiatry as well as child and adolescent psychiatry. His interests include treatment of
minority groups including indigenous native populations.

Dr. Krista Kezbers is the Senior Program Manager for Dr. Michael Businelle’s lab at the Health Promotion
Research Center in the Stephenson Cancer Center at the University of Oklahoma Health Sciences Center.
She received her PhD from Oklahoma State University in Health and Human Performance and her Med
from the University of Texas in Kinesiology. Dr. Kezbers is a Diplomate for the American College of Life-
style Medicine and is also an Exercise is Medicine certified professional. She served as a statistician and
research development professional for the University of Oklahoma School of Community Medicine for four
years. Dr. Kezbers’ research interests are in the areas of lifestyle medicine as well as improving the health
of specialized, unique populations.

Micheal Crockett is a medical student at the University of Oklahoma College of Community Medicine.

13