ORIGINAL ARTICLE
Postoperative Emergency Response Team
Activation at a Large Tertiary Medical Center
Toby N. Weingarten, MD; Sam J. Venus, MD; Francis X. Whalen, MD;
Brittany J. Lyne, SRNA; Holly A. Tempel, SRNA; Sarah A. Wilczewski, SRNA;
Bradly J. Narr, MD; David P. Martin, MD, PhD; Darrell R. Schroeder, MS; and
Juraj Sprung, MD, PhD
Abstract
Objective: To study characteristics and outcomes associated with emergency response team (ERT) activation in postsurgical patients discharged to regular wards after anesthesia.
Patients and Methods: We identified all ERT activations that occurred within 48 hours after surgery from June 1,
2008, through December 31, 2009, in patients discharged from the postanesthesia care unit to regular wards. For each
ERT case, up to 2 controls matched for age (⫾10 years), sex, and type of procedure were identified. A chart review was
performed to identify factors that may be associated with ERT activation.
Results: We identified 181 postoperative ERT calls, 113 (62%) of which occurred within 12 hours of discharge from
the postanesthesia care unit, for an incidence of 2 per 1000 anesthetic administrations (0.2%). Multiple logistic
regression analysis revealed the following factors to be associated with increased odds for postoperative ERT activation:
preoperative central nervous system comorbidity (odds ratio [OR], 2.53; 95% confidence interval [CI], 1.20-5.32;
P⫽.01), preoperative opioid use (OR, 2.00; 95% CI, 1.30-3.10; P⫽.002), intraoperative use of phenylephrine infusion
(OR, 3.05; 95% CI, 1.08-8.66; P⫽.04), and increased intraoperative fluid administration (per 500-mL increase, OR,
1.06; 95% CI, 1.01-1.12; P⫽.03). ERT patients had longer hospital stays, higher complication rates, and increased
30-day mortality compared with controls.
Conclusion: Preoperative opioid use, history of central neurologic disease, and intraoperative hemodynamic
instability are associated with postoperative decompensation requiring ERT intervention. Patients with these
clinical characteristics may benefit from discharge to progressive or intensive care units in the early postoperative
period.
© 2012 Mayo Foundation for Medical Education and Research 䡲 Mayo Clin Proc. 2012;87(1):41-49
E
mergency response teams (ERTs) have been
introduced by hospitals to evaluate and manage hospitalized patients whose condition is
acutely deteriorating. Patients assessed as clinically stable and able to be managed on regular
wards (ie, standard nursing wards) may experience acute deterioration,1,2 and the ERT is designed to promptly deliver care to these patients.
Identification of characteristics that can predict
postoperative adverse events would be desirable
because early intervention may prevent more severe complications.2 Critical analysis of types and
causes of ERT activation may provide clues to earlier identification and, potentially, prevention of
impending complications.
Because ERT systems are costly,3 some authors
have proposed preemptive triage of higher-acuity
patients to intensive care units (ICUs) or progressive
care units (eg, step-down units).4 The problem with
such an approach is that sudden postoperative adverse events can occur even in patients whose condition was stable in the postanesthesia recovery unit
(PACU) and who fulfilled discharge criteria for dis-
missal to regular wards.5 It is unknown whether
postoperative adverse events can be predicted from
patients’ comorbidities or other aspects of perioperative management.
Our objective was to examine factors associated
with the need for ERT activation after dismissal from
anesthetic care. Because we were primarily interested in perioperative factors, our study was limited
to ERTs activated within the first 48 postoperative
hours. Identification of factors associated with increased risk for these events may help to better
triage patients and minimize adverse postoperative outcomes.
From the Department of
Anesthesiology (T.N.W.,
F.X.W., B.J.L., H.A.T., S.A.W.,
B.J.N., D.P.M., J.S.) and Division of Biomedical Statistics
and Informatics (D.R.S.),
Mayo Clinic, Rochester, MN;
and Department of Critical
Care Medicine, Orlando Regional Medical Center, Orlando, FL (S.J.V.).
PATIENTS AND METHODS
The approval for review of medical records was obtained from the Mayo Clinic Institutional Review
Board in Rochester, MN. This study employed a retrospective case-control design that assessed potential factors associated with the need for an ERT activation after either surgery or diagnostic procedures
that required anesthesia.
Mayo Clin Proc. 䡲 January 2012;87(1):41-49 䡲 doi:10.1016/j.mayocp.2011.08.003 䡲 © 2012 Mayo Foundation for Medical Education and Research
www.mayoclinicproceedings.org
41
MAYO CLINIC PROCEEDINGS
The Department of Anesthesiology at Mayo
Clinic prospectively maintains a log of all ERT activations that occur at 2 Mayo Clinic-affiliated hospitals in Rochester: Saint Marys and Methodist. Using
this log, we identified adult patients who required
ERT activation within 48 hours of discharge from
the PACU to regular wards from June 1, 2008, to
December 31, 2009. (Regular ward refers to a standard nursing ward where patient vital signs are assessed at protocol-defined intervals as well as when
clinically indicated. Care may include continuous
pulse oximetry but not invasive monitoring as in an
ICU or progressive care unit setting.) By June 1,
2008, the ERT activation log system had been fully
implemented and was capturing 100% of events.
The 48-hour time window was selected to allow
identification of factors directly related to the intraoperative course. Surgical patients discharged to
monitored wards were excluded. (Monitored ward is
defined as an advanced patient care ward where patient vital signs are continually monitored as in an
ICU or other specialized patient care areas where
continual monitoring is indicated, eg, progressive
care unit.) Patients who underwent cardiac catheterization, bronchoscopy, or childbirth were excluded.
For each patient who required ERT activation, we
used the Mayo Clinic medical record database to
identify potential controls of the same sex and similar age (⫾10 years) who underwent the same procedure (as determined from International Classification of Diseases, Ninth Revision procedure codes)
during the study period and did not have ERT activation in the first 48 postoperative hours. From
these pools of potential controls, we randomly selected up to 2 controls for each ERT patient. For
cases in which fewer than 2 potential controls could
be identified, we did not select alternative controls.
Indications for ERT Activations
At our institution an ERT consists of either a rapid
response team (RRT) or code team. An RRT call can
be initiated by any health care team member concerned about the acutely deteriorating medical condition of a patient. Typically at our institution, RRT
calls are prompted as described by others,6 and indications include the following: decline in oxyhemoglobin saturation (assessed by either pulse oximetry or clinical assessment), bradypnea, tachypnea,
profound bradycardia or tachycardia, hypotension,
concern for possible heart attack (“chest pain”),
stroke (acute neurologic deficits), or acute mental
status changes (agitation, delirium). However, an
RRT call can be initiated for any other indication at
the discretion of the health care team member. The
RRT consists of an attending physician board-certified in critical care medicine, a critical care fellow or
senior anesthesia resident, a respiratory therapist,
42
and a critical care registered nurse. Code team activations are reserved for immediate life-threatening
events (cardiopulmonary arrest, severe respiratory
compromise that is assessed to require tracheal intubation and mechanical ventilation) or profoundly
unstable cardiac conditions (possibly requiring cardioversion, defibrillation). The code team is similar
in structure to the RRT but consists of an additional
critical care registered nurse, an internal medicine
resident, and a pharmacist. The level of ERT (RRT or
code team) activation is left to the discretion of the
individual health care team member (ie, nurse caring for an unstable patient); thus it is prone to subjectivity. Therefore, the level of activation occasionally is erroneous (making a “mistake on the safe
side” by activating the code team when the RRT
would have been more appropriate). Regardless, the
roles and capabilities of these 2 teams are closely
interrelated. Because our interest was to examine
patients who had an acute postoperative deterioration, we analyzed all ERT interventions, regardless
of the level of activation.
Data Abstraction
Electronic medical records were abstracted for demographics; comorbid conditions; preoperative, intraoperative, and postoperative variables; postoperative course and complications; and details of the
ERT activation. Comorbid conditions were defined
according to definitions used for numerous outcome studies at Mayo Clinic,7 including cardiovascular disease (coronary artery disease [myocardial
infarction, coronary stent placement, or cardiac bypass surgery], congestive heart failure or cardiomyopathy [ejection fraction ⬍40%], the potential for
cardiac dysrhythmia [atrial fibrillation or flutter, implanted pacemaker and/or automated defibrillator],
arterial hypertension [medically treated], peripheral
vascular disease), central neurologic disease (history
of seizures, dementia, stroke, or transient ischemic
attacks), pulmonary disease (asthma, chronic obstructive or restrictive pulmonary disease, pulmonary hypertension, obstructive sleep apnea), diabetes mellitus (medically treated), kidney disease, and
preoperative scheduled use of opioids and/or benzodiazepines. Overall physical status was assessed
from the American Society of Anesthesiologists
(ASA) Physical Status score.
The anesthetic record was reviewed for anesthesia duration, anesthetic method (general, regional),
urgency (elective, emergent), muscle relaxant use,
blood transfusion, fluid administration, and perioperative complications. Complications included hemodynamic instability or systemic arterial hypotension (inferred by use of vasopressor infusion as a
surrogate for recalcitrant hypotension), need for car-
Mayo Clin Proc. 䡲 January 2012;87(1):41-49 䡲 doi:10.1016/j.mayocp.2011.08.003
www.mayoclinicproceedings.org
ANESTHESIA AND EMERGENCY TEAM ACTIVATION
dioversion, adverse respiratory events, or other severe perioperative complication.
All data were collected and managed using
REDCap electronic data capture tools hosted at
Mayo Clinic.8 ERT notes were abstracted and supplemented by review of the medication administration record, ICU admission note, subsequent progress
notes, and discharge summaries. Data abstracted included the probable primary cause of the ERT activation: hypotension from hypovolemia or distributive
shock, respiratory cause, cardiac cause, hypertensive crisis, neurologic causes such as mental status
changes, uncontrolled pain, psychiatric reasons, or
drug reactions. The following interventions were recorded: respiratory (tracheal intubation, application of a noninvasive ventilatory device such as a
continuous positive airway pressure device, bronchodilator administration), cardiac (cardiopulmonary resuscitation, defibrillation or cardioversion,
administration of vasoactive drugs, nitroglycerin,
antiarrhythmic drugs, or diuretics), intravenous
fluid bolus or blood product administration, glucose administration, or administration of analgesics,
sedatives, naloxone, flumazenil, or antipsychotics.
Immediate outcome of the ERT was categorized as
follows: remained in the previous hospital setting
with or without intervention, transfer to a monitored ward, transfer to the operating room for exploration or treatment, or death.
Outcomes
Postoperative complications that occurred within
the first 30 postoperative days were reported. Information was obtained from the medical records from
the index hospitalization, rehospitalization, or outpatient visits. A 30-day mortality rate was calculated. Postoperative complications included myocardial infarction, cerebrovascular event, respiratory
failure requiring tracheal reintubation, acute kidney
injury (serum creatinine increase ⬎1 mg/dL and
above 1.5 mg/dL [to convert to ␮mol/L, multiply by
88.4]), thromboembolic event, sepsis or multiorgan
failure, blood transfusion requirement, or death.
Causes of death were recorded. Total days in the
ICU and hospital were recorded.
Statistical Analyses
Data are summarized using mean ⫾ SD or median
with interquartile range (IQR) for continuous variables and frequency percentage for nominal variables. To estimate the incidence of ERT activation, a
denominator was obtained using information provided by the revenue accounting office; it was based
on patient counts after excluding all surgical or procedural categories of patients expected to be admitted to monitored wards (eg, cardiac, major vascular,
thoracic). Of note, this denominator was not reduced to account for other planned or unplanned
admissions to monitored wards in patients undergoing operations that do not routinely require this
level of postoperative care. Two sets of analyses
comparing characteristics between ERT cases and
controls were performed. One analysis included all
ERT cases and compared characteristics between
cases and controls using the 2-sample t test for continuous variables and the Fisher exact test for categorical variables. The other analysis excluded ERT
cases for which no matched controls could be identified and was performed using conditional logistic
regression, taking into account the matched study
design. Characteristics found to have evidence
(P⬍.05) of an association in univariate analyses
were included as explanatory variables in a multiple
logistic regression model with ERT activation as the
dependent variable. In all cases, 2-tailed P values of
.05 or less were considered statistically significant.
Analyses were performed using SAS statistical software (Version 9.2, SAS Institute, Inc., Cary, NC).
RESULTS
During the study period approximately 95,000 patients underwent surgery or diagnostic procedures
requiring anesthesia and were discharged to a regular ward. Of those, 181 patients required ERT activation within 48 hours; therefore, the estimated rate
of ERT activation in this population was 2 per 1000
anesthetic administrations (0.2%). Of these events,
168 (93%) were RRT, and 13 (7%) were code team
activations. Of the code team activations, 6 met the
institutional definition of a code (5 patients received
cardiopulmonary resuscitation, and in one patient
the trachea was intubated to protect the airway in
the context of acute mental status changes). In the
other 7 patients, the code team was activated for
reasons other than cardiopulmonary collapse (eg,
transient syncope, hypotension). Of the entire cohort, 81 patients (45%) underwent general or urologic surgery; 62 (34%), orthopedic surgery; 19
(10%), gynecologic surgery; 11 (6%), otolaryngological surgery; 5 (3%), neurosurgery; and l3 (7%),
diagnostic procedures or minor operations requiring anesthetic care outside the operating room.
The majority (N⫽113; 62.4%) of ERT interventions occurred during the first 12 hours after surgery, and more than three-fourths (142; 78.5%) occurred within the first 24 hours (Figure). The mean
time from ERT activation to team arrival was 4⫾2
minutes. Table 1 describes the reasons for ERT activation and types of interventions. The most frequent
reasons for ERT activation were hypotension in 58
patients (32%), cardiac in 36 (20%), and pulmonary
in 31 (17%). In all cases a physician on the ERT
assessed the patient and made decisions regarding
Mayo Clin Proc. 䡲 January 2012;87(1):41-49 䡲 doi:10.1016/j.mayocp.2011.08.003
www.mayoclinicproceedings.org
43
MAYO CLINIC PROCEEDINGS
100
Percentage of ERT calls
80
60
40
20
0
0
6
12
24
30
36
18
Time from end of anesthesia (h)
42
48
FIGURE. Cumulative frequency of time to emergency response team
(ERT) activation.
treatment and disposition. Immediate outcomes of
ERT activation are summarized in Table 1.
The median ICU length of stay of patients
transferred to a monitored ward was 2 days (IQR,
2-4 days). Two patients died during code team
calls, both from massive saddle pulmonary emboli associated with a cardiac arrest. The administration of a fluid bolus (N⫽63; 35%) was the
most common overall intervention, followed by
naloxone administration (N⫽16; 9%) and opioid
administration (N⫽12; 7%) (Table 1).
Using previously described matching criteria,
we identified 318 controls for these 181 ERT patients. For 154 ERT patients (85%) we identified 2
matched controls; for 10 ERT patients (6%) only 1
control could be identified, and for 17 ERT patients
(9%) no controls could be identified because the
patient underwent an operation that was uncommon or that rarely requires anesthesia. Clinical and
demographic features of ERT cases and controls are
summarized in Table 2. ERT and control groups
were similar with the exceptions of higher rates of
central neurologic diseases and preoperative use of
opioid analgesics in ERT patients. Among 18 ERT
patients (10%) with preexisting central neurologic
diseases, 5 (28%) had an ERT activation for worsening neurologic status. Among 55 ERT patients
(30%) who used opioids, 7 (13%) received naloxone during the ERT intervention.
Surgical characteristics in the control and ERT
groups are shown in Table 3. The perioperative
course was similar between groups except that more
ERT patients had intraoperative hemodynamic in-
44
stability as reflected by an increased use of phenylephrine infusion and greater intravenous fluid totals. Of the 10 ERT patients (6%) who received
intraoperative phenylephrine infusion, 8 required
ERT intervention for later hemodynamic instability
(5, hypotension; 3, cardiac).
The intraoperative course in both groups was
devoid of any serious complications. The PACU
course was generally unremarkable, except for 1
ERT patient who required a transient phenylephrine
infusion and fluid bolus but was subsequently discharged to the regular nursing ward. No complications (tracheal intubation, aspiration, laryngospasm, pulmonary edema, or seizures) occurred in
the PACU, and all patients were transferred to regular wards. Most patients were discharged to regular
wards with supplemental oxygen (188 ERT patients
[61%] and 110 controls [50%]; P⫽.78).
From multivariate analysis, the following factors were found to be significantly associated with
ERT activation (Table 4): preoperative central nervous system comorbidity (odds ratio [OR], 2.53;
95% confidence interval [CI], 1.20-5.32; P⫽.01),
preoperative scheduled opioid use (OR, 2.00; 95%
CI, 1.30-3.10; P⫽.002), intraoperative use of phenylephrine infusion (OR, 3.05; 95% CI, 1.08-8.66;
P⫽.04), and greater intraoperative fluid administration (per 500-mL fluid bolus, OR, 1.06; 95%
CI, 1.01-1.12; P⫽.03). Similar results were obtained from the matched-set analysis that excluded the 17 ERT cases (9%) with no matched
controls (Table 4).
The length of stay was longer for patients who
required ERT activation (median, 4 days [IQR, 3-8
days] vs 3 days (IQR, 1-4 days); P⬍.001). Complication rates during hospitalization, including 30day mortality rates, were higher among patients
who required ERT activation compared with controls (Table 5).
DISCUSSION
Recovery from surgery or diagnostic procedures requiring anesthesia may be associated with complications. To avoid serious morbidities, it is important
to anticipate potential problems so that preemptive
measures can be implemented. Our study demonstrated that preoperative scheduled use of opioids,
history of central neurologic disease, and hemodynamic instability during surgery were characteristics
independently associated with unexpected deterioration within the first 48 postoperative hours. The
majority of ERT calls occurred within the first 24
postoperative hours, most often for hypotension,
mental status changes, or respiratory problems. Despite the fact that some patients in the ERT cohort
required minimal to no intervention, ERT patients
had more severe in-hospital complications.
Mayo Clin Proc. 䡲 January 2012;87(1):41-49 䡲 doi:10.1016/j.mayocp.2011.08.003
www.mayoclinicproceedings.org
ANESTHESIA AND EMERGENCY TEAM ACTIVATION
ERTs have been established to intervene quickly
to mitigate sudden deterioration of patients’ health.
Although this initiative intuitively appears effective,
ERT outcomes are difficult to assess. ERT intervention may reduce ICU resource utilization9 and immediate mortality after cardiac arrest.3,6,10-13 However, a large meta-analysis did not find evidence that
these initiatives reduce the overall in-hospital mortality.14 In contrast, introduction of ERTs was
found to increase long-term survival in surgical
patients.15 This discrepancy14,15 may be related
to differences in disease complexity between surgical and medical patients. Surgical critical events may
be associated with more reversible causes (bleeding,
oversedation, hypotension), whereas medical critical
events may be related to advanced or terminal conditions. If this is true, then implementation of ERTs, better triage of higher-risk patients, or both may improve
postoperative morbidity and mortality.
Characteristics used to predict ERT activation
have been reported only once.16 A small case-control study identified ASA Physical Status class 3 or
greater and after-hours surgery as predictors. As in
our study, hypotension and decreased level of consciousness were the main reasons for ERT activation.16 That study used ASA Physical Status as a
surrogate for comorbidity,16 but the small cohort
precluded examining the relationship between specific comorbidities and intraoperative events with
early postoperative complications. In our larger series, we identified 3 markers for increased postoperative ERT activation.
The first predictor was preoperative opioid therapy that led to respiratory depression and oversedation (as evidenced by a higher rate of naloxone administration). Other investigators also reported that
patient-controlled analgesia17 and intravenous morphine have been associated with higher rates of ERT
activation for respiratory depression.18 Furthermore, it is known that opioid-tolerant patients postoperatively report greater intensity of pain and use
more opioid analgesics than opioid-naïve patients.19 A retrospective case-control series found
that 47.8% of opioid-tolerant patients experienced
postoperative moderate to severe sedation compared with 18.5% of opioid-naïve patients.20 A
higher proportion of these patients in our study received naloxone (N⫽7 [13%] vs N⫽9 [7%]), further suggesting the need for increased vigilance for
signs of oversedation.
The second predictor was the use of phenylephrine infusions and greater intravenous fluid administration. At our institution we correct brief episodes
of hypotension with boluses of ephedrine or phenylephrine but typically initiate phenylephrine infusions, in addition to fluid boluses, when hypotension persists. Of note, these patients received more
TABLE 1. Causes for Emergency Response Team (ERT) Activation and Specific
Interventions During ERT Call
Characteristic
Patients (N⫽181)
Causes for ERT activation
Hypotension
58 (32)
Cardiac
36 (20)
Pulmonary
31 (17)
Neurologic
23 (13)
Pain/psychiatric issues
17 (9)
Drug interactions
12 (7)
Hypertension
3 (2)
Epistaxis
1 (1)
Specific interventions
Chest compressions
5 (3)
Defibrillation/cardioversion
2 (1)
Tracheal intubation
5 (3)
Noninvasive ventilation
9 (5)
Other interventions and medications administered
Intravenous fluid bolus
63 (35)
Naloxone
16 (9)
Opioids
12 (7)
Blood transfusion
10 (6)
Vasopressors (epinephrine, phenylephrine)
5 (3)
Bronchodilator
5 (3)
Benzodiazepine
3 (2)
Anticholinergic (atropine)
2 (1)
Flumazenil
2 (1)
Antiarrhythmic (amiodarone)
2 (1)
Diuretic (furosemide)
1 (1)
50% glucose intravenous bolus
2 (1)
Antihypertensive (labetalol, hydralazine)
1 (1)
Nitroglycerin
1 (1)
Immediate outcome
Death
Transfer to surgery
2 (1)
4 (2)
Transfer to monitored ward
71 (39)
Treatment on regular ward
57 (32)
Continued observation on regular ward
47 (26)
Data are presented as No. (percentage).
intraoperative fluids (crystalloids and colloids) compared with the average amounts given to the remainder of ERT patients (see footnote to Table 3). We
consequently used the initiation of “phenylephrine
infusion” as a surrogate for more pronounced hemodynamic instability. At our institution, phenylephrine
infusion is considered a temporizing measure to correct for factors associated with perioperative hypotension, and it is widely viewed as a benign intervention.
However, among ERT patients who required phenyl-
Mayo Clin Proc. 䡲 January 2012;87(1):41-49 䡲 doi:10.1016/j.mayocp.2011.08.003
www.mayoclinicproceedings.org
45
MAYO CLINIC PROCEEDINGS
TABLE 2. Demographics and Preoperative Comorbidities in Controls and in Patients Who Subsequently
Required Emergency Response Team (ERT) Activation
Control (N⫽318)a
ERT (N⫽181)
P valueb
Age (y), mean ⫾ SD
59.4⫾17.0
59.9⫾17.4
.77
Body mass index (kg/m2), mean ⫾ SD
29.0⫾6.3
28.3⫾6.4
.23
152 (48)
88 (49)
.93
1-2
198 (62)
99 (55)
3-4
120 (38)
82 (45)
1.2⫾2.3
1.1⫾0.7
.69
146 (46)
94 (52)
.23
141 (44)
82 (45)
Variable
Male sex
ASA Physical Status
.12
Preoperative creatinine (mg/dL), mean ⫾ SDc
Comorbidities
Cardiovasculard
Hypertension
Coronary artery disease
38 (12)
28 (15)
Atrial fibrillation or flutter
20 (6)
12 (7)
5 (2)
10 (6)
Peripheral vascular disease
Pacemaker/internal defibrillator
10 (3)
10 (6)
Congestive heart failure
12 (4)
9 (5)
Respiratory
d
60 (19)
45 (25)
Obstructive sleep apnea
31 (10)
21 (12)
Preoperative use of CPAP
18 (6)
4 (2)
Severe chronic lung diseasee
10 (3)
14 (8)
Asthma
24 (8)
16 (9)
14 (4)
18 (10)
10 (3)
8 (4)
2 (1)
6 (3)
Central nervous systemd
Stroke or transient ischemic attacks
Seizures
Dementia
2 (1)
6 (3)
Diabetes mellitus
55 (17)
29 (16)
.14
.02
.80
Preoperative scheduled medication use
Opioids
57 (18)
55 (30)
.002
Benzodiazepines
36 (11)
19 (10)
.88
Data are presented as No. (percentage) unless indicated otherwise. ASA ⫽ American Society of Anesthesiologists;
CPAP ⫽ continuous positive airway pressure.
b
P values are from t tests and Fisher exact test for categorical variables.
c
Data are missing for 59 patients. To convert mg/dL to ␮mol/L, multiply by 88.4.
d
Patients may have more than one comorbidity within the given category; therefore, the sum of the numbers within the category may
exceed the total for the overall category.
e
Includes chronic obstructive pulmonary disease, pulmonary fibrosis, and pulmonary hypertension.
a
ephrine infusion, postoperative hypotension was the
most common indication for ERT activation. Therefore, our perception that a phenylephrine infusion is
not a marker for subsequent instability, even after a
stable hemodynamic course in the PACU, needs to be
reexamined. Another study demonstrated that intraoperative hypotension predicts postoperative adverse
events.21
A third predictor was a history of central neurologic disease. Preexisting cognitive dysfunction is a
strong predictor of postoperative delirium.22 In our
patients, acute postoperative decline in neurologic
46
function was a frequent cause for ERT activation. Similarly, Lee et al16 reported a decline in mental status as
a common reason for ERT activation; however, they
did not comment on preoperative neurologic disease
as a risk factor. The relatively small number of patients in our study with this condition precludes
us from making a general recommendation regarding preemptive triage of these patients to a
higher level of postoperative care. Furthermore,
the hospital outcomes of these patients were
good; more than 50% did not require transfer to a
higher level of care after ERT activation.
Mayo Clin Proc. 䡲 January 2012;87(1):41-49 䡲 doi:10.1016/j.mayocp.2011.08.003
www.mayoclinicproceedings.org
ANESTHESIA AND EMERGENCY TEAM ACTIVATION
TABLE 3. Anesthetic and Intraoperative Characteristics in Controls and in Patients Who Required Emergency
Response Team (ERT) Activationa
Characteristic
Control (N⫽318)
ERT (N⫽181)
P value
16 (5)
7 (4)
.66
257 (81)
145 (80)
Neuroaxial
36 (11)
18 (10)
Peripheral nerve block ⫾ sedation
25 (8)
18 (10)
Endotracheal intubation
243 (76)
136 (75)
Laryngeal mask airway
14 (4)
9 (5)
Mask or nasal cannula
61 (19)
36 (20)
Emergency procedure
Type of anesthesia
.66
General
Airway management
.93
Intraoperative use
Nondepolarizing muscle blockers
135 (42)
87 (48)
.26
22 (7)
18 (10)
.24
2.2⫾1.5
2.6⫾2.0
.04
6 (2)
10 (6)b
.03
31 (10)
12 (7)
.23
Bronchospasm
6 (2)
6 (3)
.32
Hypoxemia
2 (1)
1 (1)
.92
3.3⫾1.7
3.4⫾2.1
.57
Blood transfusion
Crystalloids/colloids (L), mean ⫾ SD
Phenylephrine infusion
Antihypertensives
Adverse intraoperative events
Anesthetic duration (h), mean ⫾ SD
a
b
Data are presented as No. (percentage) unless indicated otherwise.
ERT patients (n⫽10) who had intraoperative phenylephrine infusion received 3.3⫾1.6 L of fluids intraoperatively.
The PACU course was generally uneventful in
our patients, which is not surprising because to be
included in our study, they needed to be in sufficiently stable condition to warrant discharge to
regular wards.5 A study of respiratory and cardiovascular events in the PACU demonstrated that
tachycardia and hypertension were predictive of un-
planned ICU admission and increased mortality,
demonstrating a relationship between early postoperative clinical deterioration and long-term outcomes.23 Furthermore, this study revealed a strong
tendency for patients to exhibit similar cardiovascular events intraoperatively and in the PACU.23 In
our study, patients with intraoperative hypotension
TABLE 4. Multivariate Analysis of Factors Associated With Emergency Response Team (ERT) Activationa
All ERT cases and controls
Matched-set analysis
Odds ratio
95% CI
P value
Odds ratio
95% CI
P value
Central nervous system comorbidityb
2.53
1.20-5.32
.01
2.44
1.07-5.58
.04
Preoperative scheduled opioid use
2.00
1.30-3.10
.002
1.95
1.18-3.24
.01
Phenylephrine infusion
3.05
1.08-8.66
.04
3.85
1.25-11.84
.02
Crystalloids/colloidsc
1.06
1.01-1.12
.03
1.12
1.04-1.20
.003
Preoperative
Intraoperative
a
Characteristics found to have a significant univariate association (see Tables 2 and 3) were included in the multiple logistic regression
model. In addition to an analysis that included all ERT cases and controls, a matched-set analysis was performed that excluded 17 ERT
cases that did not have any matched controls. The matched-set analysis was performed using conditional logistic regression, taking into
account the matched-set study design. CI ⫽ confidence interval.
b
Preoperative central nervous system comorbidities include cerebrovascular disease, seizures, and dementia.
c
Odds ratios are for a 500-mL increase.
Mayo Clin Proc. 䡲 January 2012;87(1):41-49 䡲 doi:10.1016/j.mayocp.2011.08.003
www.mayoclinicproceedings.org
47
MAYO CLINIC PROCEEDINGS
TABLE 5. Outcomes for Controls and for Patients Who Had Emergency Response Team (ERT) Intervention
During Hospitalizationa
Complicationsb
Control
(N⫽318)
ERT (N⫽181)
No. of patients
No. of patients
P value
.03
Myocardial infarction
1 (⬍1)
5 (3)
Stroke
1 (⬍1)
6 (3)
.01
Mechanical ventilationc
2 (1)
16 (9)
⬍.001
Renal failure
7 (2)
12 (7)
.03
Deep vein thrombosis
1 (⬍1)
3 (2)
.14
Pulmonary embolus
1 (⬍1)
2 (1)
.30
Sepsis/multiorgan failure
3 (1)
15 (8)
⬍.001
Need for blood transfusion
3 (1)
20 (11)
⬍.001
Alive
315 (99)
173 (96)
Dead
3 (1)
8 (4)
Mortality at 30 dd
.02
a
Data are presented as No. (percentage).
These complications represent all complications that occurred from the time of postanesthesia recovery unit discharge to hospital
discharge and are not limited to those occurring at the time of the ERT intervention.
c
Mechanical ventilation was implemented for patients who required ventilatory support for respiratory failure arising from cardiopulmonary arrest, acute lung injury, pneumonia, sepsis, or multisystem organ failure.
d
All deaths in the control group, and 1 death in the ERT group, occurred following hospital discharge. Causes of the 7 in-hospital deaths
in the ERT group included cardiac arrest during the time of the rapid response team activation in 2 patients, due to massive pulmonary
emboli as determined on autopsy; pneumonia in 1; multiorgan failure in 3; and acute hemorrhage in 1. The specific causes of the 4
out-of-hospital deaths were not determined. The 3 control patients were as follows: an 85-year-old woman with non-Hodgkin
lymphoma who had cardiac arrest after readmission following resection of a tongue lesion; a 57-year-old man with advanced
scleroderma (with cardiac and pulmonary involvement) who had undergone a muscle biopsy under monitored anesthetic care; a
62-year-old woman with renal cell carcinoma and multiple pulmonary emboli who died after placement of a pleural catheter for
malignant effusion under anesthesia. The ERT patient who died out of hospital was a 94-year-old woman who previously underwent
a left hemimandibulectomy.
b
had unremarkable PACU stays, and yet some experienced subsequent hemodynamic instability. This
suggests that the aggressive volume replacement
combined with intensive monitoring maintained
hemodynamic stability throughout the PACU stay.
However, the underlying comorbidities or conditions that led to subsequent instability after discharge to a regular ward were either not manifest or
not recognized in the PACU.
This is a retrospective study with all inherent
limitations. Because we limited our study to ERT
calls during the first 48 postoperative hours, we
could have missed significant complications that occurred outside that time frame but still were related
to anesthesia and surgery. Despite the existence of
well-defined clinical criteria for ERT activation, considerable subjectivity is involved. More specifically,
activation is triggered typically by allied health care
professionals with varied levels of training; therefore, some ERT activations were initiated for noncritical reasons (eg, epistaxis), which affects the ability to identify factors associated with increased odds
for more severe events. This is reflected in our data
48
in that 47 ERT activations (26%) resulted in no intervention by the team. This is also the reason why
we decided to include cases from both RRT and code
team activations; ie, the reasons why one team was
activated over the other are at times arbitrary. Regardless, we think that activation of either team represents an acute unexpected health decompensation
and thus is relevant for study. We believe our reported incidence represents an underestimate because invariably some patients were discharged to
the ICU for other reasons but were still included in
the denominator. However, our incidence of 0.2% is
similar to a previous report.16
CONCLUSION
Patients with preoperative central neurologic disease, scheduled preoperative opioid use, or intraoperative hemodynamic instability were found to be at
increased risk for ERT activation within 48 postoperative hours. More vigilant monitoring of such patients may be warranted in the immediate postoperative period.
Mayo Clin Proc. 䡲 January 2012;87(1):41-49 䡲 doi:10.1016/j.mayocp.2011.08.003
www.mayoclinicproceedings.org
ANESTHESIA AND EMERGENCY TEAM ACTIVATION
ACKNOWLEDGEMENT
We are thankful to Andrew Hanson for assistance
with data management and statistical analyses.
Grant Support: This project was supported by the Department of Anesthesiology, Mayo Clinic, Rochester, MN, and
National Institutes of Health/National Center for Research
Resources Clinical and Translational Science Awards Grant
Number UL1 RR024150.
Its contents are solely the responsibility of the authors and
do not necessarily represent the official views of the National Institutes of Health.
Correspondence: Address to Juraj Sprung, MD, PhD, Department of Anesthesiology, Mayo Clinic, 200 First St SW,
Rochester, MN 55905 ([email protected]).
10.
11.
12.
13.
14.
15.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
George AL Jr, Folk BP III, Crecelius PL, Campbell WB. Prearrest morbidity and other correlates of survival after inhospital cardiopulmonary arrest. Am J Med. 1989;87(1):28-34.
Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical
antecedents to in-hospital cardiopulmonary arrest. Chest.
1990;98(6):1388-1392.
Dacey MJ, Mirza ER, Wilcox V, et al. The effect of a rapid
response team on major clinical outcome measures in a community hospital. Crit Care Med. 2007;35(9):2076-2082.
Litvak E, Pronovost PJ. Rethinking rapid response teams. JAMA.
2010;304(12):1375-1376.
Aldrete JA. The post-anesthesia recovery score revisited. J Clin
Anesth. 1995;7(1):89-91.
Bellomo R, Goldsmith D, Uchino S, et al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;
32(4):916-921.
Hosking MP, Warner MA, Lobdell CM, Offord KP, Melton LJ.
Outcomes of surgery in patients 90 years of age and older.
JAMA. 1989;261(13):1909-1915.
Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG.
Research electronic data capture (REDCap)–a metadatadriven methodology and workflow process for providing
translational research informatics support. J Biomed Inform.
2009;42(2):377-381.
Baxter AD, Cardinal P, Hooper J, Patel R. Medical emergency
teams at The Ottawa Hospital: the first two years. Can J
Anaesth. 2008;55(4):223-231.
16.
17.
18.
19.
20.
21.
22.
23.
Bellomo R, Goldsmith D, Russell S, Uchino S. Postoperative
serious adverse events in a teaching hospital: a prospective
study. Med J Aust. 2002;176(5):216-218.
Bellomo R, Goldsmith D, Uchino S, et al. A prospective
before-and-after trial of a medical emergency team. Med J
Aust. 2003;179(6):283-287.
DeVita MA, Braithwaite RS, Mahidhara R, Stuart S, Foraida M,
Simmons RL. Use of medical emergency team responses to
reduce hospital cardiopulmonary arrests. Qual Saf Health Care.
2004;13(4):251-254.
Jones D, Bellomo R, Bates S, et al. Long term effect of a
medical emergency team on cardiac arrests in a teaching
hospital. Crit Care. 2005;9(6):R808-R815.
Chan PS, Jain R, Nallmothu BK, Berg RA, Sasson C. Rapid
response teams: a systematic review and meta-analysis. Arch
Intern Med. 2010;170(1):18-26.
Jones D, Opdam H, Egi M, et al. Long-term effect of a medical
emergency team on mortality in a teaching hospital. Resuscitation. 2007;74(2):235-241.
Lee A, Lum ME, O’Regan WJ, Hillman KM. Early postoperative
emergencies requiring an intensive care team intervention: the
role of ASA physical status and after-hours surgery. Anaesthesia. 1998;53(6):529-535.
Overdyk FJ, Carter R, Maddox RR, Callura J, Herrin AE, Henriquez C. Continuous oximetry/capnometry monitoring reveals frequent desaturation and bradypnea during patientcontrolled analgesia. Anesth Analg. 2007;105(2):412-418.
Fecho K, Joyner L, Pfeiffer DL. Opioids and code blue emergencies [abstract]. Anesthesiology. 2008;109:A34. http://www.
asaabstracts.com/strands/asaabstracts/abstract.htm;jsessionid⫽
DA80E5E3297E82419E075E6EC3F84D34?year⫽2008&index⫽
15&absnum⫽1555. Accessed September 1, 2011.
Ready LB. Acute pain: lessons learned from 25,000 patients.
Reg Anesth Pain Med. 1999;24(6):499-505.
Rapp SE, Ready LB, Nessly ML. Acute pain management in
patients with prior opioid consumption: a case-controlled retrospective review. Pain. 1995;61(2):195-201.
Kheterpal S, O’Reilly M, Englesbe MJ, et al. Preoperative and
intraoperative predictors of cardiac adverse events after general, vascular, and urological surgery. Anesthesiology. 2009;
110(1):58-66.
Robinson TN, Raeburn CD, Tran ZV, Angles EM, Brenner LA,
Moss M. Postoperative delirium in the elderly: risk factors and
outcomes. Ann Surg. 2009;249(1):173-178.
Rose DK, Cohen MM, DeBoer DP. Cardiovascular events in
the postanesthesia care unit: contribution of risk factors. Anesthesiology. 1996;84(4):772-781.
Mayo Clin Proc. 䡲 January 2012;87(1):41-49 䡲 doi:10.1016/j.mayocp.2011.08.003
www.mayoclinicproceedings.org
49