The First 500

American Journal of Emergency Medicine 33 (2015) 197201
Contents lists available at ScienceDirect
American Journal of Emergency Medicine

journal homepage: www.elsevier.com/locate/ajem
Original Contribution
The rst 500: initial experience with widespread use of low-dose

ketamine for acute pain management in the ED , , ,
Terence L. Ahern, MD a,, Andrew A. Herring, MD a,b, Erik S. Anderson, MD a, Virat A. Madia, MD a,
Jahan Fahimi, MD a,b, Bradley W. Frazee, MD a,b
a
b
Department of Emergency Medicine, Alameda Health System, Highland Hospital, Oakland CA

Department of Emergency Medicine, University of California, San Francisco, San Francisco CA
a r t i c l e
i n f o
Article history:
Received 17 September 2014
Received in revised form 5 November 2014
Accepted 7 November 2014
a b s t r a c t
Objectives: The objective of this study is to describe the clinical use and safety prole of low-dose ketamine (LDK)
(0.1-0.3 mg/kg) for pain management in the emergency department (ED).
Methods: This was a retrospective case series of consecutive patients given LDK for pain at a single urban ED between 2012 and 2013. Using a standardized data abstraction form, 2 physicians reviewed patient records to determine demographics, indication, dose, route, disposition, and occurrence of adverse events. Adverse events
were categorized as minor (emesis, psychomimetic or dysphoric reaction, and transient hypoxia) and serious
(apnea, laryngospasm, hypertensive emergency, and cardiac arrest). Additional parameters measured were
heart rate and systolic blood pressure.
Results: Five hundred thirty patients received LDK in the ED over a 2-year period. Indications for LDK were diverse. Median patient age was 41 years, 55% were women, and 63% were discharged. Route of administration
was intravenous in 93% and intramuscular in 7%. Most patients (92%) received a dose of 10 to 15 mg. Comorbid
diseases included hypertension (26%), psychiatric disorder (12%), obstructive airway disease (11%), and coronary
artery disease (4%). There was no signicant change in heart rate or systolic blood pressure. Thirty patients (6%)
met our criteria for adverse events. Eighteen patients (3.5%) experienced psychomimetic or dysphoric reactions.
Seven patients (1.5%) developed transient hypoxia. Five patients (1%) had emesis. There were no cases of serious
adverse events. Agreement between abstractors was almost perfect.
Conclusion: Use of LDK as an analgesic in a diverse ED patient population appears to be safe and feasible for the
treatment of many types of pain.
2014 Elsevier Inc. All rights reserved.
1. Introduction
The Institute of Medicine report, Relieving pain in America: a blueprint for transforming prevention, care, education, and research,
highlighted inadequate emergency department (ED) treatment of
pain as a major public health concern [1]. However, strategies to successfully manage acute pain in a safe and expeditious manner are the
source of considerable debate, and there is wide variation in clinical
Prior presentations: Accepted for presentation, American Society of Regional Anesthesia

and Pain Medicine, Annual Pain Medicine Meeting in San Francisco, CA, November 2014.
Conicts of interest: The authors report no conicts of interest.
Source of support: None.
Author contributions: TA, EA, VM, AH, JF, and BF conceived the study and designed
the trial. All authors participated in the conduct of the trial, data collection, and organized
the work of research assistants. TA, EA, and JF analyzed the data. TA drafted the manuscript, and all authors contributed substantially to its revision. TA takes responsibility for
the manuscript as a whole.
Corresponding author. Department of Emergency Medicine, Highland HospitalAlameda
Heath System, 1411 East 31st St, Oakland, CA 94602-1018. Tel.: +1 510 437 8497; fax: +1 510
437 8322.
E-mail address: terryahern@gmail.com (T.L. Ahern).
http://dx.doi.org/10.1016/j.ajem.2014.11.010
0735-6757/ 2014 Elsevier Inc. All rights reserved.
practice [2]. Current pharmacologic strategies in the ED rely heavily

on monotherapy with opioids; but adverse events such as sedation,
bradypnea, hypotension, and tolerance limit their utility in many patients [3-8]. In addition, the epidemic of opioid pain medication misuse
has become a nationally recognized problem, and emergency physicians have been tasked with carefully assessing opioid administration
and prescriptions [9]. More than ever, emergency physicians are considering alternative, complimentary medications, such as ketamine, that
can be combined with traditional drugs such as opioids and nonsteroidal antiinammatory drugs, to achieve multimodal analgesia in the
acute setting.
Ketamine has been used extensively in the ED for procedural sedation and rapid sequence intubation. An alternative, off-label, use of ketamine is for pain control, using subanesthetic dosingtypically 0.1 to 0.3
mg/kg. Research conducted over the last 15 years has demonstrated
that such low-dose ketamine (LDK) is safe, effective, and improves
pain management when combined with opioid analgesics [10-13].
Low-dose ketamine has been shown to potentiate the analgesic effect
of opioids, have opioid-sparing effects, and to attenuate development
of centralized chronic pain states [14-20]. For these reasons, LDK for analgesia has been widely adopted in the anesthesia, surgical, and
198
T.L. Ahern et al. / American Journal of Emergency Medicine 33 (2015) 197201
palliative care settings for the treatment of postoperative and chronic

cancer-related pain.
Emergency medicine has been slower to incorporate LDK for analgesia into routine practice, likely due to lack of familiarity with this indication as well as concerns over adverse effects, particularly
emergence phenomena. But a small yet growing body of evidence
has emerged over the last 10 years documenting the successful use
of LDK in the ED and prehospital environment [12,21-25]. These studies consistently show that the safety and side effect prole of LDK is
similar to that of opioids and that LDK causes few signicant
pyschomimetic reactions. In response, some institutions have begun
to routinely incorporate LDK into acute pain management as a complimentary and rescue analgesic. Two years ago, in collaboration with
emergency physicians, nursing, and pharmacy staff, we developed
an ED-specic LDK protocol to facilitate use for a broad array of painful
conditions in our department.
The aim of this study is to document the clinical use, safety, and side
effect prole of LDK for pain management in the ED.
2. Methods
2.1. Setting
This retrospective, consecutive case series was conducted in a
single ED at an urban trauma center. We obtained a database, derived from our electronic medical record (EMR) (Wellsoft Corporation, Sumerset, NJ), of all patients receiving ketamine in our ED
during a 2-year period from January 2012 to December 2013. This
2-year timeframe coincided with an increase in popularity and
awareness of ketamine on the part of ED providers after the creation
of an ED-specic LDK protocol in 2012. With broad inclusion criteria,
the protocol proposed LDK as an agent for analgesia in patients with
many types of acute or chronic pain, either alone or in combination
with additional pain relieving drugs. The protocol recommended
doses of 5 to 20 mg intravenous (IV) or 10 to 25 mg intramuscular
(IM). There were no absolute contraindications except for known allergy to ketamine. Relative contraindications included age younger
than 18 years, uncontrolled seizure activity, severe signs of elevated
intracranial pressure, renal and/or liver failure, and women who are
pregnant or breastfeeding. Patients were not specically excluded
for having abnormal vital signs (ie, hypertension, tachycardia, or
hypoxia), and the ultimate decision whether to order LDK was left
up to provider preference.
Our ED uses computerized drug storage units (Pyxis Corporation,
San Diego, CA) and EMRs that permit accurate tracking of department
drug ordering and administration, including dosage and route of administration. To facilitate ease of use and cut down on unnecessary
waste, our pharmacy began stocking preloaded syringes of 15 mg ketamine for IV administration, which were kept in the drug storage units.
Our hospital's institutional review board approved this retrospective
review.
2.2. Study population

We extracted data from electronic systems to include all ED patients for whom ketamine was ordered during the study period.
The data included medical record number, arrival date, age, sex, disposition, and chief complaint. Chief complaints were categorized
into 7 broad groups of indications for LDK before chart review. The
groups included muscular-skeletal pain, abdominal pain, chest
pain, skin and soft tissue infections, headache, back pain, and other.
For the purposes of this study, we dened LDK as a dose less than
or equal to 20 mg IV or 25 mg IM or roughly 0.1 to 0.3 mg/kg in the
average size adult.
2.3. Data abstraction

After a formal training period and data abstraction pilot trial, a standardized data abstraction form was used to review patient records independently by 2 authors. We abstracted ketamine dose (milligram),
route of administration (IV or IM) and systolic blood pressure (SBP),
and heart rate (HR) at 2 time points: at triage and within 1 hour of
LDK administration.
Detailed review of the clinical chart was done to ascertain the presence or absence of specic, predened adverse events with one hour of
LDK administration including cardiac arrest, apnea (respiratory rate
b 10 breaths per minute or need for jaw thrust and/or bag valve mask
ventilation), hypoxia (oxygen saturation, b 90% on room air or N5% decreased in oxygen saturation from baseline value if N90% at triage), hypertensive emergency (SBP, N 180 and the acute onset of chest pain,
shortness of breath, or severe headache), laryngospasm, emesis,
psychomimetic reaction (agitation, hallucinations, or unusual behavior
recorded by provider), and other (nurse or physician documentation
of specic problem related to LDK administration). Lastly, comorbidities
including a history of hypertension, coronary artery disease (CAD), psychiatric illness (schizophrenia, bipolar, and depression requiring medication), and chronic obstructive pulmonary disease (COPD) or asthma
were recorded. If not documented in the medical record, each of these
adverse events and comorbidities was assumed to be absent.
Frequent meetings were held between abstractors and study coordinators to answer questions, resolve disputes, and review identied adverse events. A random sample of 10% of charts reviewed was
duplicated to assess interrater reliability.
2.4. Data analysis
We report descriptive statistics and 95% condence intervals (CIs),
where appropriate. Interrater reliability was ascertained through the
Cohen statistic for route of administration and absence or presence
of any adverse events and the Spearman rank correlation for dose of administration. Statistical analysis was done using Stata version 11
(StataCorp, College Station, TX).
3. Results
We found almost perfect agreement between the 2 abstractors: =
0.98 for route of administration, = 0.90 for presence of adverse reaction, and r = 0.99 for dose of administration.
The initial database included 683 patients who received ketamine in
our ED over the study period. We excluded all cases of ketamine administration that did not meet our denition of LDK (20 mg IV or 25 mg
IM). Using this denition, 153 cases were excluded from the analysis.
The excluded cases primarily comprised ketamine used for conscious
sedation and rapid sequence intubation.
This series ultimately included 530 consecutive ED LDK administrations, of which 294 (55.5%) were female. The median age was 40 years,
and the distribution of patients was fairly even between the second to
fth decades of life (Table 1). Indications for LDK were diverse, and
many of the patients had substantial underlying illness including hypertension (26%), psychiatric disorder (12%), COPD (11%), and CAD (4%).
Ultimately, near two-thirds (63%) were discharged home from the ED.
Low-dose ketamine was administered IV in the vast majority of
cases (93%) and IM in the remaining cases. Most patient (92%) received
a single dose of either 10 or 15 mg, although the dose range was 5 to
25 mg IV/IM. There was no signicant change in SBP and HR within 1
hour of LDK administration, as compared with triage values. Mean triage SBP and HR was 141 (99% CI, 138-144) and 93 (99% CI, 91-95),
whereas SBP and HR within 1 hour of LDK administration were 138
(99% CI, 135-141) and 86 (99% CI, 84-88), respectively.
Of 530 LDK cases, only 30 (6%) met our criteria for an adverse event.
Each event is specically detailed in Table 2. There were 7 patients (1.5%)

Table 1
Patient characteristics and indications for LDK
Age (y)
19-29
30-39
40-49
50-59
60-69
N70
Sex
Male
Female
Disposition
Discharge
Admission
Indications
Musculoskeletal pain
Abdominal pain
Chest pain
Skin and soft
tissue infection
Back pain
Headache
Othera
Comorbidities
Hypertension
CAD
COPD
Psychiatric illness b
No. of patients
Percentage (%)
132
125
120
114
35
4
25
24
22
22
6
1
236
294
44
56
335
195
63
37
63
178
24
62
12
33
5
12
66
13
124
12
3
23
139
21
57
63
26
4
11
12
a
Included chronic pain, sickle cell crisis, genitourinary disorders, painful rashes, psychiatric complaints, and other miscellaneous painful complaints.
b
Included depression, bipolar disorder, and schizophrenia.
who developed transient hypoxia, 4 of whom had concurrently received

1 to 2 mg of hydromorphone. Most of the 7 had transient oxygen
desaturations to between 86% and 92%, and all but 1 patient responded
to 2 L oxygen by nasal cannula. Five patients (1%) had emesis, 3 received
ondansetron for symptomatic relief, and all 5 cleared their airway without assistance. There was no evidence of aspiration in any patient.
Eighteen patients (3.5%) experienced psychomimetic or dysphoric
reactions (hallucinations, agitation, unusual behavior, or provider documentation of LDK-related patient complaint), none of which were long
lasting or led to a change in ultimate disposition. Three patients were
given lorazepam for symptomatic anxiety or agitation. Most patients
improved without intervention or after reassurance by nurse or physician. There was 1 case of moderate-to-severe agitation in a 57-yearold man with metastatic cancer who was noted by the nurse to open
his eyes widely and scream while pulling at the gurney side rails after
receiving ketamine 15 mg IV and fentanyl 50 g IV; he was treated
with lorazepam resulting in resolution of his symptoms. Another patient was noted by the nurse to have a bad dream-like state and felt
like she might die in her dream, after receiving ketamine 10 mg IV.
Other notable patient quotations in the medical record included, I
feel like a zombie, if this is what people feel like on drugs, then I
don't want them, my pain is gone, but I feel crazy, I feel like I'm ying, and you all look like aliens.
4. Discussion
To our knowledge, this is largest series reported of LDK administration for pain in the ED. We found that LDK is feasible and safe for treatment of a wide variety of painful conditions. The adverse event rate was
6% overall, but the events were easily identied and dealt with by ED
staff. Furthermore, this adverse event rate is lower than that of opioid
medications in hospitalized patients, although a direct comparison is
problematic [26,27]. None of the adverse events caused harm or
changed disposition. Importantly, no patients experienced apnea,
laryngospasm, hypertensive emergency, or cardiac arrest.
199
Concerns over adverse psychomimetic affects, particularly emergence phenomena, have traditionally limited widespread use of LDK in
adult ED patients [28]. Our results conrm those of prior smaller studies
of LDK showing that psychomimetic reactions are mostly mild in nature
and rarely alter a patient's clinical course [10,12,21-24,29]. In our cohort,
18 patients (3.5%) had documented psychomimetic or dysphoric reactions within 1 hour of LDK administration. Although 3 patients required
lorazepam for sedation during the episode, most reactions were mild and
improved without intervention or with reassurance from ED staff.
It is now apparent that mild dysphoric effects of LDK occasional
occur with doses lower than what is traditionally considered the dissociative range (1-2 mg/kg IV), at which actual emergence phenomenon
can occur. The rate of such reactions in recent prospective studies
ranges from 16% to 26% [21,22,30]. It is important to note that the negative reactions are universally short lived and differ substantially from
emergence phenomenon. Our rate of mild dysphoric events is much
lower than described in previous prospective studies; but this is likely
due to the inherent limitations of retrospective chart review, reliance
on the medical records for documentation of events, and the sensitivity
of screening instruments for such events used in prospective studies.
Nonetheless, we suspect that some patients reported the effects as negative experiences primarily because they were taken by surprise. Based
upon our (the investigators) growing experience with LDK, we believe
that advising patients about the possibility of psychomimetic effects reduces the likelihood that the effect will be perceived as negative if it occurs. In addition, a prior prospective trial on LDK showed that the same
patient who reports very bothersome dissociative effects might report
high satisfaction at discharge [21]. It seems prudent that providers
who administer LDK should routinely coach patients just before administration, reassuring them that any dysphoric reaction will be short lived
and create as calm an environment as possible.
Other types of adverse events were infrequent. Seven patients (1.5%)
experienced transient oxygen desaturation within 1 hour of ketamine
administration. Of these patients, 4 were given concomitant opioids
with LDK, and all but 1 patient responded quickly with 2 to 4 L nasal
cannula oxygen. One patient required 2 hours of bilevel positive airway
pressure (bipap) support; but she had been hypoxic at triage, required
oxygen support via non-rebreather facemask and bipap before LDK,
and was already admitted for a COPD exacerbation. According to
provider's documentation, the indication for LDK in this case was to
treat chest pain but, perhaps, more importantly, to facilitate therapy
for hypoxia by way of providing anxiolysis and bronchodilation. Overall,
the rate of hypoxia is substantially less than reported in prior prospective research on opioid-based pain protocols in the ED [31]. For example, in the widely cited 1 + 1 hydromorphone titration protocol
study, Chang et al [31] found a 5% rate of hypoxia in patients receiving
hydromorphone. In addition, their study excluded patients with baseline oxygen saturation less than 95%. However, a direct comparison
with our heterogeneous cohort is not possible because some patients
may have been given LDK in spite of their hypoxia.
Similarly, we found a lower rate of emesis in our cohort than what
was reported in patients receiving hydromorphone in the study of
Chang et al [31] (1% vs 7%, respectively). Furthermore, most of our emesis cases were in patients who had experienced nausea and/or vomiting
before receiving LDK (reference, Table 2 for details), whereas such patients would have been excluded from the reporting of emesis in the
study of Chang et al [31].
We observed no signicant change in blood pressure or HR within 1
hour of administering LDK as compared with triage values. Patients who
were tachycardic, hypertensive, or hypoxic at triage remained so after
receiving LDK. This is not surprising given the well-established favorable hemodynamic prole of ketamine [32]. Although these ndings
suggest that LDK may be safe in patients who have abnormal vital
signs, there is much uncertainty in this patient population given the limitations of retrospective data. Furthermore, our LDK protocol does not
explicitly exclude patients with abnormal vital signs and allows for
200
Table 2
All 30 adverse events that occurred within 1 hour of LDK bolus, among 530 patients
Adverse eventa
Age Sex Disposition Indication
Dose Route Comorbiditiesb
Details
Hypoxia
41
Home
Abdominal
pain
15
IV
43
Admit
10
IV
Hypertension
43
Admit
Abdominal
pain
Abdominal
pain
15
IV
Depression,
hypertension, CAD
42
Admit
Abscess
20
IV
58
Home
Cancer
15
IV
48
Admit
COPD
15
IV
46
Admit
Trauma
20
IV
21
Admit
Abdominal
pain
15
IV
32
Home
Abdominal
pain
15
IV
Hypertension
51
Home
Back pain
15
IV
Hypertension
22
Home
15
IV
75
Home
Chronic
pain
Fracture
15
IV
Abdominal
pain
Abdominal
pain
Abdominal
pain
Abdominal
pain
Abdominal
pain
Abdominal
pain
10
IV
After LDK and 2 mg hydromorphone, patient reported, I feel dizzy.
IV
After LDK and 400 mg ibuprofen, patient reported, I feel dizzy.
15
IV
15
IV
15
IV
15
IV
Hypertension, COPD
LDK given for chronic pelvic pain. Hypoxia noted during LDK administration
requiring non-rebreather facemask, which resolved within 1 h.
Patient had been given 2 mg hydromorphone 45 minutes before LDK.
LDK given for abdominal pain. Placed on 2-L nasal cannula after LDK, although no
desaturation was noted.
LDK given for abdominal pain secondary to diabetic ketoacidosis. 45 min prior
received 2 mg hydromorphone. SpO2 dropped to 88%, transient
2-L NC applied.
LDK given for abscess drainage. 1 h after administration noted to have SpO2 of 88%
when asleep, which improved with elevation head of bed.
Patient may have had undiagnosed obstructive sleep apnea.
LDK and 2 mg hydromorphone given for back pain related to metastatic lesion. SpO2
dropped to 88%, transient 2-L NC applied.
Patient was admitted for COPD. SpO2 95% on 2-L NC before LDK but dropped to 80%
with increased work of breathing and lethargy obstructive
lung disease noted by MD afterward. Placed on bipap for next 2 h then improved.
LDK and 1 mg hydromorphone given for head laceration repair. SpO2 dropped to
90%, transient 2-L NC applied.
Emesis
Psychomimetic/dysphoricd
54 M
Home
55
Home
39
Home
67
Home
33
Home
43
Home
22
Admit
Abdominal
pain
20
IV
55
Admit
Abscess
10
IV
42
Home
Back pain
10
IV
40
Home
Back pain
15
IV
57
Admit
Cancer
15
IV
65
Admit
Chest pain
15
IV
67
Admit
Chest pain
15
IV
36
Home
Chest pain
15
IV
38
Home
Chest pain
10
IV
54
Home
Chest pain
15
IV
44
Home
Hematoma
15
IV
43
Home
Sickle cell pain
15
IV
Hypertension
LDK and 8 mg ondansetron given for nausea and vomiting in setting of

pyelonephritis. Patient had 2 small episodes of emesis afterward but
stated it's due to not eating.
LDK, 25 mg benedryl and 4 mg ondansetron given for nausea and vomiting in
setting of gastroparesis. Patient had large emesis 45 min
afterward, improved with 10 mg metoclopramide.
LDK, 30 mg ketorolac and 10 mg dexamethasone given for cauda equina syndrome.
Patient had small emesis 15 min afterward while lying at
for electrocardiogram.
LDK given for chronic abdominal pain and hyperemesis syndrome. Patient
complained of continued nausea and vomiting after LDK.
LDK and 1 mg hydromorphone given for humerus fracture. Patient had large emesis
10 min afterward, improved with 4 mg ondansetron.
COPD
After LDK and 1 mg hydromorphone, patient stated her pain is improved, but the
medicine made her feel like I'm going to die.
Hypertension
After LDK, noted that she does not want ketamine again for pain; that it made her
hallucinate.
CAD, COPD
After LDK and 4 mg morphine, patient reported pain gone but I feel crazy. Given
1 mg lorazepam.
After LDK and 4 mg morphine, patient noted by nurse to become unresponsive to
verbal stimuli. Awakened with sternal rub and began crying.
Vital signs normal except for HR of 110. Given 1 mg lorazepam.
After LDK, patient reported I feel dizzy, but the pain is gone. Later noted by RN to
patting the wall with hand repeatedly with eyes closed. She remained alert and
oriented but no explanation offered.
LDK and 25 g fentanyl given for abscess drainage. Noted to become anxious and was
crying because she didn't like the effect of the drug.
Hypertension
LDK and 2 mg hydromorphone given for back pain. Noted to be very anxious for
10 min afterward.
After LDK, patient became highly anxious and was crying. Reported I feel like a
zombie. Improved with reassurance by nurse.
After LDK and 50 g fentanyl, patient noted to have enlarged eyes and be screaming
in pain while pulling at side rails. Required 2 mg
lorazepam and was calmed by MD
COPD
After LDK, noted to be anxious and disoriented by nurse. Patient stated If this is
what people feel like on drugs, then I don't want them.
Feelings resolved spontaneously within 10 min.
Hypertension, coronary
Patient did not like feeling of LDK immediately, and the bolus was stopped before
completion.
Hypertension
Received LDK for asthma exacerbation. Afterward, noted to be more calm and stated
I feel like I'm ying, then I'm going to sleep.
During LDK administration, noted to have a bad dream-like state, and felt like she
was going to die in her dream.
Hypertension
After LDK noted, I feel weird. I feel funny What is wrong with me? Symptoms
resolved without intervention.
After LDK and 1 mg hydromorphone, patient reported pain is gone, but we all look
like aliens.
Depression, hypertension After LDK, patient became nauseated and ushed feeling. Improved with 25 mg
phenergan.
Abbreviations: M, male; F, female; SpO2, oxygen saturation as measured by pulse oximetry; NC, nasal cannula; MD, doctor of medicine; RN, registered nurse.
a
No patient experienced cardiac arrest, apnea, hypertensive emergency, or laryngospasm.
b
Signicant comorbities abstracted included history of hypertension, pyschiatric illness (depression, bipolar, and schizophrenia), CAD, and COPD.
c
Hypoxia was dened as oxygen saturation as measured by pulse oximetry less than 90% or decrease in oxygen saturation more than 5% from triage vital signs.
d
Pyschomimetic/dysphoric side effects were dened as hallucinations, agitation, unusual behavior, or registered nurse/doctor of medicine documentation of a specic problem
related to ketamine.
provider preference, so we cannot account for individual practice patterns and must assume some avoided LDK in these situations.
The favorable safety prole of LDK is especially notable given the
wide age distribution and prevalence of comorbidities in our cohort
(Table 1). To date, prior studies of LDK had rigorous inclusion and exclusion criteria and represented a tightly controlled cohort of patients. We
believe that our cohort represents a typical diverse, urban ED population, where many patients have chronic medical and psychiatric disease,
substance abuse, and lack of social support. In spite of this, our ndings
are consistent with those of a prior small retrospective study in a similar
setting [25] and recent prospective data [21,22,24,30], showing that LDK
is feasible, generally well tolerated, and very safe in the ED.
This study has the usual limitations inherent in a retrospective review. Quality of the data was dependent on that of the medical record,
particularly nursing documentation. To mitigate this, we focused on
data that were objective and not prone to interpretation or abstractor
bias using a standardized abstraction protocol based upon accepted
guidelines for chart review methodology [33]. Our EMRs include extensive documentation from nursing and physicians, so it is unlikely that
we missed any major adverse events (ie, cardiac arrest, apnea, hypoxia,
laryngospasm, and hypertensive emergency). Despite this, it is likely
our data underestimate minor adverse events, such as emesis or transient psychomimetic and dysphoric events.
Although emergency physicians should be encouraged by the safety
of LDK in this large and diverse cohort of ED patients, we emphasize that
data from prospective, randomized blinded trials are needed to denitively determine the efcacy, safety, and side effect prole of LDK compared with standard opioid analgesics and other opioid adjuncts.
5. Conclusion
Use of LDK alone or in combination with other pain medications as a
primary or rescue analgesic in a diverse ED patient population appears to
be safe and feasible for the treatment of many types of pain. Minor
psychomimetic side effects were observed but easily addressed by ED
personnel and did not alter disposition. Other side effects, including emesis and hypoxia, appear to be equally or less common than reported with
opioids. Prospective randomized trials are needed to determine the
efcacy and further elucidate the safety and side effect prole of LDK.
References
[1] Institute of Medicine Committee on Advancing Pain Research, Care, and Education.
Relieving pain in America: a blueprint for transforming prevention, care, education,
and research. Washington (DC): National Academies Press (US); 2011.
[2] Voscopoulos C, Lema M. When does acute pain become chronic? Br J Anaesth 2010;
105(Suppl. 1):i6985.
[3] Guant S, Taleb A, Borel-Khner J, Cauterman M, Raphael M, Nathan G, et al. Quality
of pain management in the emergency department: results of a multicentre prospective study. Eur J Anaesthesiol 2011;28(2):97105.
[4] Pyati S, Gan TJ. Perioperative pain management. CNS Drugs 2007;21(3):185211.
[5] Benyamin R, Trescot AM, Datta S, Buenaventura R, Adlaka R, Sehgal N, et al. Opioid
complications and side effects. Pain Physician 2008;11:S10520.
[6] Brennan MJ, Stanos S. Strategies to optimize pain management with opioids while
minimizing risk of abuse. PMR 2010;2(6):54458.
201
[7] Shah S, Mehta V. Controversies and advances in non-steroidal anti-inammatory drug

(NSAID) analgesia in chronic pain management. Postgrad Med J 2012;88(1036):738.
[8] Quinlan J, Carter K. Acute pain management in patients with persistent pain. Curr
Opin Support Palliat Care 2012;6(2):18893.
[9] Cantrill SV, Brown MD, Carlisle RJ, Delaney K, Hays D, Nelson L, Nelson L, et al.
American College Emergency Physicians Opioid Guideline Writing Panel. Clinical
policy: critical issues in the prescribing of opioids for adult patients in the emergency department. Ann Emerg Med 2012;60(4):499525.
[10] Cherry DA, Plummer JL, Gourlay GK, Coates K, Odgers C. Ketamine as an adjunct to
morphine in the treatment of pain. Pain 1995;62:11921.
[11] De Pinto M, Jalacic J, Edwards WT. Very-low-dose ketamine for the management of
pain and sedation in the ICU. J Opioid Manag 2008;4(1):546.
[12] Jennings PA, Cameron P, Bernard S. Ketamine as an analgesic in the pre-hospital setting: a systemic review. Acta Anaesthesiol Scand 2011;55:63843.
[13] Schmid RL, Sandler AN, Katz J. Use and efcacy of low-dose ketamine in the management of acute postoperative pain: a review of current techniques and outcomes.
Pain 1999;82:11125.
[14] Bell RF, Dahl JB, Moore RA, Kalso E. Perioperative ketamine for acute postoperative
pain. Cochrane Database Syst Rev 2006;1:CD004603.
[15] Bilgin H, Ozcan B, Bilgin T, Kerimolgu B, Uckunkaya N, Toker A, et al. The inuence of
timing of systemic ketamine administration on postoperative morphine consumption. J Clin Anesth 2005;17:5927.
[16] Childers Jr WE, Baudy RB. N-methyl-D-aspartate antagonists and neuropathic pain:
the search for relief. J Med Chem 2007;50(11):255762.
[17] Corchs F, Mercante JP, Guendler VZ, Mashruha M, Vieira D, Bernik M, et al. Sensitivity
to aversive stimulation, posttraumatic symptoms and migraines: what do they have
in common? Med Hypotheses 2011;77:5345.
[18] Dickenson AH. NMDA receptor antagonists: interactions with opioids. Acta
Anaesthesiol Scand 1997;41:1125.
[19] Fishman SM. Listening to pain: a physician's guide to improving pain management
through better communication. Oxford University Press; 2012.
[20] Koppert W, Dern SK, Sittl R, Albrecht S, Schuttler J, Schmelz M. A new model of electrically evoked pain and hyperalgesia in human skin: the effects of intravenous
alfentanil, S ketamine, and lidocaine. Anesthesiology 2001;95(2):395402.
[21] Ahern T, Herring A, Stone M, Frazee B. Effective analgesia with low-dose ketamine
and reduced dose hydromorphone in ED patients with severe pain. Am J Emerg
Med 2013;31(5):84751.
[22] Galinski M, Dolveck F, Combes X, Limoges V, Smail N, Pommier V, et al. Management
of severe acute pain in emergency settings: ketamine reduces morphine consumption. Am J Emerg Med 2007;25(4):38590.
[23] Jennings PA, Cameron P, Bernard S, Walker T, Jolley D, Fitzgerald M, et al. Morphine
and ketamine is superior to morphine alone for out-of-hospital trauma analgesia: a
randomized controlled trial. Ann Emerg Med 2012;59(6):497500.
[24] Johansson P, Kongstad P, Johansson A. The effect of combined treatment with morphine sulfate and low-dose ketamine in the pre-hospital setting. Scand J Trauma
Resusc Emerg Med 2009;17:616.
[25] Lester L, Braude DA, Niles C, Crandall CS. Low-dose ketamine for analgesia in the ED:
a retrospective case series. Am J Emerg Med 2010;28(7):8207.
[26] Davies EC, Green CF, Taylor S, Williamson PR, Mottram DR, Pirmohamed M. Adverse
drug reactions in hospital in-patients: a prospective analysis of 3695 patientepisodes. PLoS One 2009;4(2):e4439.
[27] Weiss AJ, Elixhauser A. Characteristics of adverse drug events originating during the
hospital stay, 2011. Healthcare Cost and Utilization Project. 2013. Statistical brief
#164. http://www.hcup-us.ahrq.gov/reports/statbriefs/sb164.jsp.
[28] Stoddard FJ, Sheridan RL, Saxe GN, King BS, King BH, Chedekel DS, et al. Treatment of
pain in acutely burned children. J Burn Care Rehabil 2002;23(2):13556.
[29] Porter K. Ketamine in prehospital care. Emerg Med J 2004;21:3514.
[30] Richards JR, Rockford RE. Low-dose ketamine analgesia: patient and physician experience in the ED. Am J Emerg Med 2013;31(2):3904.
[31] Chang AK, Bijur PE, Gallagher E. Randomized clinical trial comparing the safety and
efcacy of a hydromorphone titration protocol to usual care in the management of
adult emergency department patients with acute severe pain. Ann Emerg Med
2011;58(4):3529.
[32] Weinbraum AA. Non-opioid IV adjuvants in the perioperative period: pharmacological and clinical aspects of ketamine and gabapentinoids. Pharmacol Res 2012;65(4):
41129.
[33] Gilbert EH, Lowenstein SR, Koziol-Mclain J, Barta D, Steiner J. Chart reviews in emergency medicine research: where are the methods? Ann Emerg Med 1996;27(3):3058.
ORIGINAL RESEARCH
Safety of Peripheral Intravenous Administration of

Vasoactive Medication
Jose Cardenas-Garcia, MD1*, Karen F. Schaub, BS1, Yuly G. Belchikov, PharmD2, Mangala Narasimhan, DO1,
Seth J. Koenig, MD1, Paul H. Mayo, MD1
1
2
Division of Pulmonary, Critical Care and Sleep Medicine, Hofstra North ShoreLong Island Jewish School of Medicine, Hempstead, New York;
Clinical Pharmacy Services, Department of Pharmacy, Westchester Medical Center, Valhalla, New York.
BACKGROUND: Central venous access is commonly performed to administer vasoactive medication. The administration of vasoactive medication via peripheral intravenous
access is a potential method of reducing the need for central venous access. The aim of this study was to evaluate
the safety of vasoactive medication administered through
peripheral intravenous access.
METHODS: Over a 20-month period starting in September
2012, we monitored the use of vasoactive medication via
peripheral intravenous access in an 18-bed medical intensive
care unit. Norepinephrine, dopamine, and phenylephrine
were all approved for use through peripheral intravenous
access.
RESULTS: A total of 734 patients (age 72 6 15 years,
male/female 398/336, SAPS II score 75 6 15) received
vasoactive medication via peripheral intravenous access
783 times. Vasoactive medication used was norepinephrine (n 5 506), dopamine (n 5 101), and phenylephrine
(n 5 176). The duration of vasoactive medication via
Vasoactive medications (VMs) are often required to

improve hemodynamic function in patients with
shock. They are usually given through central venous
catheter (CVC) access, primarily out of concern that
extravasation of peripheral intravenous (PIV) access
may result in local tissue injury due to the vasoconstrictive effect of the VM. However, insertion of CVC
is associated with a variety of mechanical complications and risk of central lineassociated bacteremia.
To examine the feasibility and safety of using VM via
PIV access, we report on the administration of VM in
the form of norepinephrine, dopamine, and phenylephrine via PIV access, with the rationale that this
would be a method of reducing the need of CVC use.
Our hypotheses are that VM via PIV access is both
feasible and safe.
*Address for correspondence and reprint requests: Jose

Cardenas-Garcia, MD, Department of Medicine, Division of Pulmonary,
Critical Care and Sleep Medicine, HofstraNorth Shore LIJ School of
Medicine, 410 Lakeville Rd, Suite 107, New Hyde Park, NY 11042;
E-mail: jdecardenasg@gmail.com
Additional Supporting Information may be found in the online version of
this article.
Received: February 28, 2015; Revised: April 30, 2015; Accepted: April
30, 2015
2015 Society of Hospital Medicine DOI 10.1002/jhm.2394
Published online in Wiley Online Library (Wileyonlinelibrary.com).
An Official Publication of the Society of Hospital Medicine
peripheral intravenous access was 49 6 22 hours. Extravasation of the peripheral intravenous access during
administration of vasoactive medication occurred in 19
patients (2%) without any tissue injury following treatment,
with local phentolamine injection and application of local
nitroglycerin paste. There were 95 patients (13%) receiving vasoactive medication through peripheral intravenous
access who eventually required central intravenous
access.
CONCLUSIONS: Administration of norepinephrine, dopamine, or phenylephrine by peripheral intravenous access
was feasible and safe in this single-center medical intensive
care unit. Extravasation from the peripheral intravenous line
was uncommon, and phentolamine with nitroglycerin paste
were effective in preventing local ischemic injury. Clinicians
should not regard the use of vasoactive medication is an
automatic indication for central venous access. Journal of
C 2015 Society of
Hospital Medicine 2015;000:000000. V
Hospital Medicine
MATERIAL AND METHODS

Study Design
This was a single-arm, consecutive-patient study conducted from September 2012 to June 2014. The study
site was an 18-bed medical intensive care unit (MICU)
staffed by full-time attendings, fellows, and residents
at the Long Island Jewish Medical Center, which is an
827-bed tertiary care teaching hospital that is part of
the North ShoreLong Island Jewish Health System.
The primary outcome measure was the rate of local
tissue injury resulting from use of VM via PIV access.
The study was approved by the hospital institutional
review board (study #13583A), which waived
requirement for informed consent.
Protocol for Administration of VM via PIV Access
In cooperation with the Department of Pharmacy, medical and nursing staff developed a written protocol for
administration of VM via PIV access. The protocol was
reviewed and approved by the hospital pharmacy and
therapeutics committee and the MICU nursing leadership. The MICU nursing staff received in-service training before rollout of the protocol, which included
training on the recognition of PIV access extravasation
and the type of line that could be used. The MICU
housestaff teams were given specific instructions
concerning the protocol during their MICU rotations.
Journal of Hospital Medicine Vol 00 | No 00 | Month 2015
Cardenas-Garcia et al
Peripheral Administration of VM
TABLE 1. Summary of the Requirements for PIV
Access Used for Infusion of VM

Vein diameter >4 mm measured with ultrasonography
Position of PIV access documented to be in the vein with ultrasonography before starting
infusion of VM
Upper extremity only, contralateral to the blood pressure cuff
Intravenous line size 20 gauge or 18 gauge
No hand, wrist, or antecubital fossa PIV access position
Blood return from the PIV access prior to VM administration
Assessment of PIV access function every 2 hours as per nursing protocol
Immediate alert by nursing staff to the medical team if line extravasation, with prompt initiation of
local treatment
72 hours maximum duration of PIV access use
NOTE: Abbreviations: PIV, peripheral intravenous; VM, vasoactive medication.
A summary of the requirements for PIV access for VM

use is summarized in Table 1.
Patient Management
The decision to initiate treatment with VM was made
by the clinical management team. The standard concentrations of VM for use via PIV access were: norepinephrine 8 mg or 16 mg/250 mL normal saline,
dopamine 400 mg or 800 mg/250 mL D5W, and
phenylephrine 80 mg or 160 mg/500 mL normal
saline. If the attending or fellow in charge of the case
decided that VM should be administered through PIV
access, peripheral access was established that conformed to the requirements of the protocol, and VM
was administered via PIV access for as long as there
was clinical indication or until PIV access suitable for
VM administration was no longer feasible. If the
patient received VM via PIV access, a second PIV
access site was established in case of failure of the primary PIV site. If no PIV access could be inserted, the
patient received CVC access. The decision to use VM
via CVC access was made by the clinical management
team, as was the type, dose, and duration of the VM
use via PIV access or CVC access. Vasopressin was
not used via PIV access. Dobutamine was used via
PIV access but not recorded in our results, as it has
no a-mediated vasoconstrictor effect. Dobutamine
was not used concomitantly with other vasoactive
medication through the same PIV access. If PIV access
was not established using standard technique by nursing staff, medical residents or critical care fellows
inserted PIV access using real-time ultrasound guidance. The PIV access use for VM could also be used
for other medications providing they were compatible
with the VM. Only 1 type of VM was infused through
the PIV access.
As indicated in Table 1, the nursing staff examined
the PIV access site every 2 hours and checked that
blood could be aspirated from the line. The aspiration
of the line requires several seconds of discontinuation
of VM use, which we considered to have no clinical
relevance. If the nursing staff identified extravasation
2
of the PIV access site through which VM was infusing, they notified the medical housestaff, who
promptly initiated treatment with local injection of
phentolamine and local application of nitroglycerin
paste as described in Table 2. The extravasation site
was examined for tissue injury on a shift basis by the
nursing staff, and on bedside rounds by the attending
and fellow for at least 48 hours following PIV access
removal. Tissue injury was defined as any erythema,
blistering, skin breakdown, or necrosis in the site of
extravasation.
Data were collected prospectively by an investigator
(J.C.-G.) and entered into a standard data-collection
sheet for quality and safety assessment for the initial
13 months of the study. In the subsequent 7 months
of observation, data were collected from retrospective
chart review. The initial 13 months of data collection
were performed as an ongoing safety analysis project;
the subsequent 7-month review was performed as an
additional quality assessment project. The deidentified
data included patient demographics, patient disease
characteristics, use of VM, and VM via PIV access
complications.
Statistical Analysis
The statistical analysis was performed using SPSS 21
(Statistical Package for the Social Sciences; IBM,
Armonk, NY). Continuous variables are presented as
mean 6 standard deviation.
RESULTS
Characteristics of patients who received VM via PIV
access are presented in Table 3. During the study
period, there were 2462 admissions to the MICU, and
267 CVCs were inserted by the MICU team, 170 of
which were triple-lumen catheters and 97 were largegauge catheters for hemodialysis or plasmapheresis.
Of the total admissions, 953 cases received VM; 783/
953 (82%) received VM via PIV access, and 170/953
received VM via CVC access (18%). For VM use, an
18-gauge PIV catheter was used in 192/783 (25%), a
20-gauge catheter was used in 590/783 (75%), and a
22-gauge catheter was used in 1/783 of interventions.
TABLE 2. Treatment of VM via PIV Access
Extravasation
1. The VM via PIV infusion is stopped immediately.
2. Residual medication is aspirated through the PIV access, and the catheter is removed.
3. The extent of the extravasation is outlined to provide a baseline for monitoring.
4. Two vials, each containing 5 mg of phentolamine, are reconstituted with 5 mL of
normal saline per vial to yield a final concentration of 1 mg/mL.
5. The phentolamine solution is injected in 0.5- to 1-mL aliquots in 5 separate injections around
the leading edge of the extravasation, using separate 25-gauge or 27-gauge needles for each
injection.
6. Nitroglycerin paste (2.5 cm) is applied to the area of extravasation.
7. A medication occurrence report is filled out for review by the quality committee.
NOTE: Abbreviations: PIV, peripheral intravenous; VM, vasoactive medication.
TABLE 3. Demographic and Disease Characteristics
of the Study Population

Total Study Group
No. of patients
Age, y
Gender
Male
Female
SAPS II score
Patients on mechanical ventilation
Patients on hemodialysis
MICU mortality
Use of VM via PIV access
Extravasations of VM via PIV access
Total MICU admissions during study period
734
72 6 15
398 (54%)
336 (46%)
75 6 15
235 (32%)
90 (12%)
177 (23%)
783
19 (2%)
2,462
NOTE: Abbreviations: MICU, medical intensive care unit; PIV, peripheral intravenous; SAPS II, simplified
acute physiology score II; VM, vasoactive medication.
Catheter length was 30 mm, 45 mm, or 48 mm

depending on availability. The 22-gauge catheter,
which was a deviation from standard protocol, infiltrated shortly following insertion. We did not formally record the anatomic position of the PIV access
in the standard data-collection sheet; anecdotally, the
majority of PIV accesses were placed in the upper arm
basilic or cephalic vein. The duration of VM via PIV
access was 49 6 22 hours. Central intravenous access
was required in 95/734 (13%) of patients who initially had VM via PIV access. These catheters are
included in the 170 triple-lumen CVCs that were
inserted by the MICU team during the study period.
The type and highest dose of VM administered via
PIV access are presented in Table 4.
A total of 734 patients received VM via PIV access
during the 20-month study period; 49 of these
patients required 2 or more PIV access insertions, as
the initial and/or subsequent site timed out at 72
hours, resulting in a total of 783 separate interventions. Infiltration of the PIV access site occurred in
19/783 (2%) of interventions. All of them were identified by nursing staff with prompt response using local
injection of phentolamine and application of nitroglycerin paste at the site of the extravasation. There
was no tissue injury at the site of VM extravasation.
Sixteen of the extravasations occurred with norepinephrine infusions and 3 with dopamine infusions.
There were no infections of the PIV access sites used
for VM. Use of phentolamine and nitroglycerin paste
was not associated with hypotension, as defined as
mean arterial pressure less than 65 mm Hg.
DISCUSSION
Our study demonstrates that administration of VM
via PIV access is feasible, carries a low rate of complications, and offers an alternative to CVC access.
There are several elements that may have allowed safe
use of VM via PIV access. We developed a protocol
that involved a multidisciplinary team. The hospital
| Cardenas-Garcia et al
pharmacy performed an extensive literature search

and formulated the initial protocol with the MICU
attending staff. The protocol was then subjected to
iterative process improvement by a hospital committee
and nursing leadership in the MICU. Before program
rollout, the MICU nursing staff were educated and
trained to use the protocol. This was a key component
of the program, as the nurses were responsible for
many of the line insertions, line maintenance, and
identification of infiltration. Although we did not perform any formal measurement of the impact of PIV
access use on nursing workflow, we note that leadership and frontline nurses have been enthusiastic about
the implementation of VM via PIV access. The MICU
housestaff teams were given an in-service instruction
concerning the importance of prompt initiation of
local treatment in case of infiltration of the PIV access
site. Specific elements of the protocol that may have
improved safety were the use of ultrasonography to
insert difficult PIV access and confirmation of all PIV
access insertions using ultrasonography by the MICU
housestaff. The requirement for frequent checks of
PIV access function, prompt recognition of infiltration, and specific antidote to extravasation were
important elements of safety. The low rate of PIV
access extravasation (2%) may be related to the use
of ultrasonography to guide PIV access insertion in
patients who had challenging anatomy (eg, obesity,
edema, recreational drug use, history of multiple PIV
insertions), and ultrasonography was used to check
that the PIV access was well positioned before VM
infusion.
There were early literature reports that subcutaneous extravasation of catecholamines could result in
local ischemic injury both in human patients and animal models.15 Local phentolamine injection has been
identified as a specific antidote to block the local
ischemic injury.615 More recently, there have been
anecdotal reports showing that local application of
nitroglycerin paste blocks ischemic injury in the pediatric population.5,16,17 With this information, our
TABLE 4. Frequency, Highest Dose, and
Complications of Vasoactive Medication

Administered via PIV Access
Norepinephrine
Interventions
Dose, mg/kg/min, mean 6 SD
PIV access extravasations
Dopamine
Interventions
Phenylephrine
Interventions
506
0.70 6 0.23
16
101
12.7 6 5.23
3
176
3.25 6 1.69
0
NOTE: Abbreviations: PIV, peripheral intravenous; SD, standard deviation.
Cardenas-Garcia et al
protocol included the requirement of prompt treatment of local extravasation with phentolamine and
nitroglycerin paste at the site of VM via PIV access
extravasation. In theory, both phentolamine and nitroglycerin might cause hypotension. In our study,
administration of both phentolamine and nitroglycerin
paste was not associated with more hypotension nor
did it increase requirements for VM.
Multilumen small bore CVCs may be used for several reasons, some of which need to be reconsidered.
First, before introduction of the VM via PIV access
protocol, a common indication for triple lumen CVC
insertion in our MICU was the perception that VM
could only be administered through CVC access, for
fear of local tissue injury should extravasation of the
VM occur through the PIV access site. Our results
indicate that VM use is not an automatic indication
for CVC insertion. Second, a possible indication for
CVC insertion is to measure central venous pressure
for the purpose of guiding volume resuscitation in
patients with hemodynamic failure. As the utility of
central venous pressure monitoring has been called
into serious question,1820 we do not consider this
indication for CVC use to be valid. Third, CVC access
may be required due to anatomic constraints (ie, there
is no suitable PIV site). Fourth, there may be need for
such a large number of intravenous medications that
PIV access cannot support. Fifth, there is occasional
situation where the patient requires use of medications
where extravasation of PIV access would cause local
tissue injury without local antidote (eg, certain chemotherapeutic agents). The continued need for CVC
access in some patients is reflected in the finding that
13% of our study patients who received VM via PIV
access eventually required triple-lumen CVC insertion.
However, our results indicate that the rate of CVC
use may be reduced by using PIV access for VM
administration.
Our study has some methodological limitations.
Study design was single center and observational. The
focus of this study was to examine the safety of VM
via PIV access. We cannot comment on its effectiveness, indications, or influence on patient outcome nor
on why some patients required CVC insertion whereas
others did not. The decision to administer VM was
made by the clinical team, as was the route of its
administration and concentration, without any input
from the investigators. We did not collect data on
who performed the PIV access insertion (medical or
nursing staff), demographics, and disease characteristics of the CVC group, nor to what extent ultrasonography was used to guide PIV insertion. We did not
attempt to define whether there were any factors that
identified risk for PIV access extravasation, nor did
we evaluate for any differences between the PIV and
CVC group in terms of demographics and disease
characteristics. Lacking a control group, we cannot
say definitively that VM via PIV access is safer than
4
VM via CVC. Being a single-center study, it is not

possible to say that the results are transferable to
another clinical environment; this applies particularly
to the use of ultrasonography, which is a userdependent skill. We cannot determine which, if any,
component of the protocol was responsible for the
safe use of VM via PIV access. The rate of PIV access
extravasation was low, so it is possible that a larger
sample size is required to identify incidents of tissue
necrosis from extravasation of VM delivered via PIV
access despite the use of local antidote.
CONCLUSIONS
The delivery of VM via PIV access is safe and feasible.
Tto reduce the risk of extravasation leading to possible local tissue injury, we developed a protocol that
emphasized close cooperation between the nursing
and medical staff, routine use of ultrasonography,
rapid identification of extravasation of the PIV access,
and prompt response to local extravasation of VM
using phentolamine and nitroglycerin paste. This
approach offers a means of reducing CVC use, in
both intensive care unit (ICU) and non-ICU settings,
including hospital wards and emergency departments.
Clinicians should no longer consider administration of
norepinephrine, dopamine, or phenylephrine to be an
automatic indication for CVC access. This study
focused on the safety of VM administered via PIV
access, with emphasis on local complications related
to extravasation, and should be considered a preliminary single-center study that demonstrates that administration of certain vasoactive medications may not
universally require central venous access. A broader
study regarding assessment of safety and efficacy will
require a multicenter design.
Disclosures: J.C.-G., K.F.S., Y.G.B., M.N., S.J.K., and P.H.M. participated in the study design, statistical review, and manuscript writing.
J.C.-G. is the guarantor of the article, taking responsibility for the integrity of the work as a whole from inception to published article. This
work is original, and all authors meet the criteria for authorship, including acceptance of responsibility for the scientific content of the article.
This article is not under consideration in any other journal, and all of
the authors have read and approved the content of the article. No
potential conflict of interest exists with any companies or organizations
whose products or services are discussed in this article. This article has
not been funded by the National Institutes of Health, the Wellcome
Trust, or their agencies. All financial support of the study was derived
from the Division of Pulmonary, Critical Care and Sleep Medicine at
North ShoreLong Island Jewish Medical Center, New Hyde Park, New
York.
References
1. Close AS, Frackelton WH, Kory RC. Cutaneous necrosis due to
norepinephrine. II. Mechanism and prevention. Ann Surg. 1958;147:
4450.
2. Alexander CS, Sako Y, Mikulic E. Pedal gangrene associated with the
use of dopamine. N Engl J Med. 1975;293:591.
3. Julka NK, Nora JR. Gangrene aggravation after use of dopamine
[letter]. JAMA. 1976;235:2812.
4. Greene SI, Smith JW. Dopamine gangrene [letter]. N Engl J Med.
1976;294:114.
5. Chen JL, OShea M. Extravasation injury associated with low-dose
dopamine. Ann Pharmacother. 1998;32:545548.
6. Zucker G. Use of phenytolamine to prevent necrosis due to levarterenol.
JAMA. 1957;163:14771479.
7. Close AS. Phentolamine hydrochloride in prevention of cutaneous

necrosis due to levarterenol. JAMA. 1959;170:19161917.
8. Stetson JB, Reading GP. Avoidance of vascular complications associated with the use of dopamine. Can Anaesth Soc J. 1977;24:727733.
9. Zenk KE. Management of intravenous extravasations. Infusion. 1984;
6:7779.
10. Siwy BK, Sadove AM. Acute management of dopamine infiltration
injury with Regitine. Plast Reconstr Surg. 1987;80:610612.
11. Cooper BE. High dose phentolamine for extravasation of pressors [letter]. Clin Pharm. 1989;8:689.
12. Hinterberger JW, Kintzi HE. Phentolamine reversal of epinephrineinduced digital vasospasm. How to save an ischemic finger. Arch Fam
Med. 1994;3:193195.
13. Subhani M, Sridhar S, DeCristofaro JD. Phentolamine use in a neonate for the prevention of dermal necrosis caused by dopamine: a case
report. J Perinatol. 2001;21:324326.
14. Sinclair MD, Bailey MA, McAree BJ, Dewhurst D, Kent PJ. Images in
vascular medicine: rapid epinephrine reversal with phentolamine
15.
16.
17.
18.
19.
20.
| Cardenas-Garcia et al
following accidental autoinjector inoculation. Vasc Med. 2011;16:

215216.
Le A, Patel S. Extravasation of noncytotoxic drugs: a review of the literature. Ann Pharmacother. 2014 8;48:870886.
Denkler KA, Cohen BE. Reversal of dopamine extravasation injury with
topical nitroglycerin ointment. Plast Reconstr Surg. 1989;84:811813.
Wong AF, McCulloch LM. Treatment of peripheral tissue ischemia
with topical nitroglycerin ointment in neonates. J Pediatr. 1992;121:
980983.
Marik PE, Baram M, Vahid B. Does central venous pressure predict
fluid responsiveness? A systematic review of the literature and the tale
of seven mares. Chest. 2008;134:172178.
Marik PE, Cavallazzi R. Does the central venous pressure predict fluid
responsiveness? An updated meta-analysis and a plea for some common sense. Crit Care Med. 2013;41:17741781.
ProCESS Investigators, Yealy DM, Kellum JA, Huang DT, et al. A
randomized trial of protocol-based care for early septic shock. N Engl
J Med. 2014;370:16831693.
The Journal of Emergency Medicine, Vol. 49, No. 1, pp. 104117, 2015
Copyright ! 2015 Elsevier Inc.
Printed in the USA. All rights reserved
0736-4679/$ - see front matter
http://dx.doi.org/10.1016/j.jemermed.2014.12.041
Clinical
Review
EMERGENCY EVALUATION FOR PULMONARY EMBOLISM, PART 2:

DIAGNOSTIC APPROACH
Jeffrey A. Kline, MD* and Christopher Kabrhel, MD, MPH
*Department of Emergency Medicine and Department of Cellular and Integrative Physiology, Indiana University School of Medicine,
Indianapolis, Indiana and Department of Emergency Medicine, Center for Vascular Emergencies, Massachusetts General Hospital and
Department of Emergency Medicine, Harvard Medical School, Boston, Massachusetts
Reprint Address: Jeffrey A. Kline, MD, Department of Emergency Medicine and Department of Cellular and Integrative Physiology, Indiana
University School of Medicine, 720 Eskanazi Avenue, Indianapolis, IN 46202
, AbstractBackground: In part 1 of this two-part review, we discussed which risk factors, historical features,
and physical findings increase risk for pulmonary embolism
(PE) in symptomatic emergency department (ED) patients.
Objectives: Use published evidence to describe criteria
that a reasonable and prudent clinician can use to initiate
and guide the process of excluding and diagnosing PE. Discussion: The careful and diligent emergency physician can use
clinical criteria to safely obviate a formal evaluation of PE,
including the use of gestalt reasoning and the pulmonary embolism rule-out criteria (PERC rule, Table 2, part 1). We present published clinical and radiographic features of patients
with PE who eluded diagnosis in the ED. D-dimer can be
used to exclude PE in many patients, and employing agebased adjustments to the threshold to define an abnormal
value can further reduce patient exposure to pulmonary
vascular imaging. Moreover, we discuss benefits, limitations,
and potential harms of computed tomographic pulmonary
vascular imaging relevant to patients and the practice of emergency care. We present algorithms to guide exclusion and
diagnosis of PE in patients with suspected PE, including those
who are pregnant. Conclusions: Reasonable and prudent
emergency clinicians can exclude PE in symptomatic ED patients on clinical grounds alone in many patients, and many
more can have PE ruled out by use of the D-dimer. ! 2015
Elsevier Inc.
INTRODUCTION
This second part of a two-part review provides an
in-depth analysis of issues critical to deciding when to
initiate a formal diagnostic evaluation for pulmonary
embolism (PE) in emergency department (ED) patients,
and what diagnostic tests, if any, need to be ordered. We
explore evidence-based options for excluding PE to a
reasonable degree of diagnostic certainty but with
minimal exposure to radiation and iodinated contrast
material.
DISCUSSION
Decision to Initiate the Work-up and Empiric Treatment
Figure 1 presents an algorithm for the diagnostic evaluation of patients with possible PE. For PE to enter the
active differential diagnosis list for any patient, he or
she must have at least one possible physiologic manifestation of PE. The physiologic manifestation may be a
symptom (e.g., dyspnea, pleuritic chest pain, or new
fatigue) or a sign (e.g., heart rate > 100 beats/min or pulse
oximetry < 95% near sea level) that is not explained by
another cause. Other bedside physiological signs of PE
include a low (<30 mm Hg) end-tidal CO2, measured
by capnography, or signs of pulmonary hypertension on
12-lead electrocardiography, including T-wave inversion
, Keywordspulmonary embolism; medicolegal; defensive medicine; decision making; venous thromboembolism;

pregnancy; diagnosis; pregnancy
RECEIVED: 30 September 2014; FINAL SUBMISSION RECEIVED: 17 December 2014;

ACCEPTED: 21 December 2014
104
Emergency Evaluation for Pulmonary Embolism
105
Figure 1. Diagnostic algorithm for pulmonary embolism (PE) in patients who prompt enough clinical suspicion to warrant the documented consideration of PE. *Assumes a cutoff for abnormal of $500 ng/mL. Nondiagnostic ventilation-perfusion (V/Q) scan findings
require confirmation from results of another test, such as computed tomography pulmonary angiography (CTPA), if benefits outweigh
risks. Abbreviations: + = positive for PE; ! = negative for PE; Cr = creatinine; GFR = glomerular filtration rate; High = high probability
scan findings; LMWH = low-molecular-weight heparin; Nl = normal; Nondx = nondiagnostic (any reading other than normal or high
probability); PERC = pulmonary embolism rule-out criteria; quant = quantitative, sRGS = simplified revised Geneva score.
in leads V1 to V4, incomplete or complete right bundlebranch block, and the S1-Q3-T3 pattern (1,2).
Reasonable and prudent emergency care does not
dictate that all patients with a sign or symptom of PE
must be tested for PE. Nor does it dictate that a patient
with one or more risk factors for PE must undergo testing
for PE in the absence of a sign or symptom of PE. However, the authors believe that clinicians should consider
PE for patients with a sign or symptom of PE and a known
risk factor for PE (see Table 1, in part 1), and at least
mentally formulate an explanation why a work-up was
not pursued in the event that the patient had PE. If a
reasonable alternative disease explains the patients presentation, testing specifically directed at diagnosing PE
need not be ordered. The value of an alternative diagnosis
to obviate an evaluation for PE must be decided on a caseby-case basis, and is often a nuanced decision-making
process. For example, if an emergency physician cares

for a patient with long-standing dyspnea and tachycardia
with a known lung cancer and a large pleural effusion,
this does not mandate a computed tomographic pulmonary angiogram (CTPA). However, if the clinician was
aware that lung mass and effusion were radiographically
unchanged, but the patient recently developed new severe
dyspnea and tachycardia, this patient may warrant further
testing for PE.
The next step is to assess the pretest probability using
either gestalt or a validated scoring method, such as the
Wells score, or the revised Geneva score (RGS) or the
simplified RGS (Tables 3 and 4, part 1) (35). Gestalt
has the advantage of not requiring any memory aid, and
has similar diagnostic performance characteristics and
interobserver reliability as the Wells score and RGS
(3,6). If a patient has a high pretest probability (from
106
any method), the clinician should consider immediately

administering heparin or low-molecular-weight heparin
for patients with low bleeding risk. However, the benefits
of empiric anticoagulation remain unproven. One review suggested that the benefit of empiric systemic anticoagulation for 24 h exceeds the risks (bleeding and
heparin-induced thrombocytopenia) for any patient with
a pretest probability of PE of >20% (7). Several studies
have suggested that delay in administration of heparin
to patients with PE can increase mortality, but no study
has found that heparin administered prior to imaging improves morbidity or mortality (810).
Three studies have provided data on patients who
passed through the ED and were soon after diagnosed
with PE (8,1113). These patients can be categorized as
those admitted to the hospital and those discharged
home. Compared with patients who were promptly
diagnosed and treated for PE, patients admitted to the
hospital who went on to have delayed recognition of PE
tended to have a higher frequency of altered mental
status (either new or at baseline dementia) and
preexisting heart and lung disease (8,1113). Only one
study provided data on patients apparently discharged
with PE, and those patients were more likely to not have
dyspnea, have isolated pleuritic chest pain and
hemoptysis together with a pulmonary infiltrate on
imaging, and a lower D-dimer concentration with a small
distal clot seen on pulmonary vascular imaging (12). Coincidentally, in an analysis of PE(+) but pulmonary embolism rule-out criteria (PERC)(!) patients (see Table 2 in
part 1) in a large database, the presence of pleuritic chest
pain emerged as a common feature (14). Thus, it seems
that highly competent emergency physicians may miss
distal lung clots that produce pulmonary infarction and a
clinical picture of pneumonia. More evidence is needed
to determine if patients with these small distal clots, in
the absence of deep venous thrombosis (DVT), actually
benefit from systemic anticoagulation.
Exclusion of PE at the Bedside
About two-thirds of patients who are considered for testing
for PE in the United States have a low pretest probability,
regardless of the method used, and the prevalence of PE in
this group is <5% (15). Patients with a low gestalt pretest
probability (defined as a global estimate that the patient
has <15% probability of PE) are eligible to have PE ruled
out with the PERC rule (see Table 2 in part 1).
The authors suggest that ruling out PE requires a combination of pretest probability and diagnostic test results
that predict an outcome rate, or false negative rate
<2.0% for any one patient. This false negative rate,
synonymous with posttest probability, equals the product
of the likelihood ratio (LR) for a negative diagnostic test
J. A. Kline and C. Kabrhel
result (LR! = [1 ! sensitivity]/specificity) times the pretest odds (odds = probability/[1 ! probability]) (note that
odds are always higher than probability), which is then
converted from an odds value back to probability. Thus,
for a low-risk populationfor example, one with an underlying prevalence of venous thromboembolism (VTE)
of 45% defined by gestalt low clinical probability
the PERC rule, functioning as a diagnostic test, has an
LR! of about 0.2 or less, and therefore clearly can rule
out VTE, based upon a predefined posttest threshold of
2.0% (3,16,17):
Pretest probability = 4%.
Pretest odds = 0.04 (1!0.04) = 0.04/0.96 = 0.042.
Post-test odds = LR! * pretest odds = 0.2 *
0.042 = 0.0084.
Post-test probability = odds/(1 + odds) = 0.0084/
1.0084 = 0.0083 or 0.8%.
Here we refer to posttest under the assumption that
clinical criteria, namely the PERC rule, can function as
the diagnostic test. Importantly, a population of patients,
each with a pretest probability <2%, collectively has a
lower false negative rate. The combination of a low clinical gestalt impression plus a negative PERC rule reliably
predicts a probability of PE below 1% even in European
populations (1620). Use of the PERC rule after a low
pretest probability using other methods of pretest
probability assessment besides gestalt assessment has
not been validated (21). At a pretest probability <2%,
the risk of further testing outweighs the low probability
of failing to diagnose PE (22,23). Therefore, if all
criteria of the PERC rule are met in the setting of a
gestalt-based low pretest probability, not only is further
testing unnecessary, but it should be avoided if possible.
The PERC rule (Table 2, part 1) does not have 100%
sensitivity, and will be negative in the presence of small
PE at a rate of about 1 in 100 patients considered, and
even more rarely in the presence of larger PE (14,24).
In most cases, in a patient suspected of having PE if
any one of the eight criteria is not met, or the doctor
simply thinks a test is indicated, the patient should
undergo a diagnostic test for PE. Not all patients who
fail the PERC rule need an objective test for PE
ordered; the PERC rule provides only one set of criteria
to rule out PE, and other sets likely exist.
D-dimer Testing
Assuming PE cannot be ruled out with the PERC rule, the
next step is to determine which specific diagnostic test
makes sense in view of the patients pretest probability.
If the ED clinician has access to a quantitative D-dimer
assay, it should be strongly considered as a first diagnostic
test in patients for whom clinical suspicion is low or moderate based on either gestalt estimation, a Wells score of
#4, or an RGS #4 (see Tables 3 and 4 in part 1) (25,26).

Available data from the United States suggest that a
quantitative D-dimer at standard threshold produces a
false negative rate <1% even with a Wells score up to
6 (15). Most commercially available, automated,
quantitative D-dimer assays that employ either immunoturbidimetric latex agglutination or enzyme-linked immunoabsorbance colorimetry as the detection method have an
LR! of <0.15 (27,28). Different D-dimer assays have
variable thresholds for normal due to different capture
antibodies and optical methods of detection. Some
laboratories report results in D-dimer mass concentration
(e.g., nanograms per milliliter or micrograms per
milliliter) and others report fibrinogen equivalent units,
which are twice the mass concentration. The D-dimer
has a half-life in plasma of approximately 8 h, and extrapolating from humans and animal models of autologous PE,
the D-dimer level probably remains abnormally high for at
least 3 days after symptomatic PE (2932). However, as Ddimer may be continuously shed by unstable clot, it is
difficult to know exactly how long after an acute PE a Ddimer assay will remain positive.
The most common causes of false positive and false
negative D-dimer results are listed in Table 1 (29
31,33,34). Almost all risk factors for PE also elevate the
D-dimer concentration. The fact that D-dimer increases
with age has prompted numerous researchers to test if the
D-dimer can be adjusted upward for age and maintain
adequate exclusionary ability, mainly for suspected PE.
The most common formula studied is age * 10 ng/mL,
where a patient aged 80 years would have an ageadjusted threshold for abnormal at 800 ng/mL (3537). In
a large multicenter management study, this approach,
107
when used in conjunction with a Wells score #4 or a

simplified revised Geneva score #4, was associated with
a very low rate (0.3%) of PE diagnosis on 3-month
follow-up (see Tables 3 and 4, part 1) (37). Age adjustment
of D-dimer is also supported by previous meta-analyses of
other studies, as well as recent studies not yet aggregated
into a systematic review. It is our opinion that it is reasonable to use age-adjusted D-dimer values to rule out PE in
patients with low or moderate pretest probability (36).
All patients with a positive D-dimer result that cannot
be explained by another finding must undergo imaging
directed at discovering clots, and the choice of the
next test must be determined by a mix of patient and facility factors. As the physician becomes aware of new
information, PE may move up, down, or off the differential diagnosis list even after a D-dimer test result is
found to be positive. Removing VTE from the differential must be justified by the presence of a condition that
obviously explains the elevated D-dimer (e.g., one of the
causes listed in Table 1) together with a plausible explanation for the patients symptoms that is unlikely to coexist with PE (e.g., pneumothorax) (34). In particular,
clinicians should not assume that an elevated troponin
decreases the probability of PE (e.g., in favor of cardiac
ischemia), as troponin elevation occurs in about 20% of
patients with PE and is associated with worse outcomes
(3840). Similarly, emergency physicians must be
aware that about 45% of patients with PE have an
elevated brain natriuretic peptide concentration
(39,40). Clinicians should be aware that normalization
of initially abnormal vital signs has not been found to
reduce the probability of PE and should not be used to
justify cancelling a previously ordered test for PE (41).
Table 1. Factors that Cause Errors in D-dimer Measurements (3337))

False Positives
Patient factors:
" Increasing age: (6069 years [OR 2.6], 7079 years [OR 4.5],
$80 years [OR 10.5])
" Cocaine use (OR 2.0)
" Immobility: general (OR 2.3), limb (OR 2.8), or neurologic
(OR 3.0)
" Hemoptysis (OR 2.0)
" Hemodialysis (OR 2.2)
" Malignancy, active (OR 2.6)
" Rheumatoid arthritis (OR 2.8)
" Systemic lupus erythematosus (OR 2.1)
" Sickle cell disease (OR 24.2)
" Pregnancy and postpartum state: (2nd trimester [OR 7.3],
3rd trimester [OR 51.3], postpartum [OR 4.2])
" Surgery (<4 weeks prior): abdominal (OR 3.5), chest (OR
2.7), orthopedic (OR 2.2), other surgery (OR 3.2)
False Negatives*
Patient factors:
" Concomitant anticoagulation
" Symptoms lasting more than 5 days
" Subsegmental PE
" Isolated pulmonary infarction
" Chronic PE
System and machine issues:
" Wrong sample
" Severe lipemia or hemolysis
" Protein degradation by proteolysis that can occur with
prolonged time from sample draw to analysis
OR = odds ratio; PE = pulmonary embolism.

* Derived from case reports, experience and manufacturers information.
Theoretically, risk is greatest with vitamin K antagonists and dabigatran, as both inhibit active thrombin generation and therefore reduce
factor XIII generation, which could allow for non-cross-linked but insoluble clots. More likely, most PE diagnosed in patients on anticoagulation are simply chronic and thus liberate small amounts of D-dimer.
108
IMAGING
CT Pulmonary Angiography (CTPA)
A good quality computed tomography (CT) scan, which
requires about 200 Hounsfield units of contrast opacification in the main pulmonary artery, rules out PE at all pretest probabilities on the day of examination (42,43).
Chest CT does not rule out the possibility of future PE
from undiagnosed DVT. Chest CT angiography only
identifies filling defects in contrast-enhanced pulmonary
arteries. Most scanning protocols require the patient to
lie supine and hold his or her breath for a few seconds.
CT scanning requires injection of approximately
120 mL of contrast by a computer-controlled injection
device. The patient must have a peripheral intravenous
(i.v.) catheter (20 gauge or larger) or an approved
indwelling line to allow injection of the contrast. Equipment with multiple detector heads (e.g., 64-head scanners) allows better resolution so that filling defects can
be observed even in subsegmental pulmonary arteries
(44). The diagnostic sensitivity and specificity of a technically adequate CT scan, performed on a multidetector
CT scanner in an ED population independently of pretest
probability, are both about 90% (43). Interobserver agreement in identifying segmental or larger filling defects has
been consistently demonstrated to be very good, but interobserver agreement for subsegmental clots is poor (45).
Benefits of CT scanning include a binary positive or negative result and the ability to detect evidence supporting a
clinically significant alternative diagnosis (where pneumonia is most common, found in 822% of cases)
(33,4650).
Radiologists indicate presence of suboptimal image
quality in about 10% of their formal interpretations of
CTPA scans (45,51). Figure 2 (A and B) shows examples
of high-quality scan images and a degraded image. Image quality is most commonly degraded by low arterial
opacification, or motion artifact (e.g., from severe tachypnea) (52). Obesity increases risk of inadequate
CTPA imaging (53,54). Radiologists probably cannot
detect filling defects with <150 Hounsfield units of
opacification (42,52). In real practice, approximately
10% of CT scans yield technically inadequate images
secondary to motion artifact or poor pulmonary artery
opacification, most commonly in obese or very
tachypneic patients (53,54). Many diagnostic studies
of CTPA scanning exclude these scans from analysis,
but in the PIOPED II study, 11/51 patients with
indeterminate CTPA scans had PE on reference testing
(51). Patients with indeterminate CTPA results and
moderate or high pretest probability may need bridging
anticoagulation until a PE can be ruled out. This can be
done with a follow-up Ventilation Perfusion (V/Q) scan
that has homogeneous perfusion. Because the outcome

rate of PE after a negative quantitative D-dimer is
<1%, even in high pretest probability patients, the quantitative D-dimer, if negative (age adjustment allowed),
provides strong evidence to rule out PE in the setting
of a degraded image CTPA. However, as about 80
90% of high pretest probability patients will have a positive D-dimer, the usefulness of this approach is limited
(15,34). Alternatively, standard care also includes
bilateral lower-extremity venous ultrasonography, performed in the ED and again within 37 days (5558).
If results of this repeat examination are negative, VTE
can be ruled out in high-risk patients after indeterminate
CTPA scanning (27,37,51,5961).
The CT scan, despite its remarkable value as a diagnostic tool, poses risks to patients that emergency physicians must know. First among these is a 610% false
positive rate in low-risk populations, possibly leading to
over-diagnosis and unnecessary anticoagulation
(44,45,51,62). CTPA imparts approximately 10 to
20 mSv of radiation, with an estimated increased
lifetime risk of fatal cancer of at least 1 in 500 per chest
CT (63,64). This risk may be higher in young women
due to radiation to the breast (65). Furthermore, within
5 years afterward, more than one-third of patients who
have one CT angiography to rule out PE can be expected
to undergo subsequent CT pulmonary angiography,
incurring a second dose of radiation (66). Acute lifethreatening complications from CT scanning include
anaphylactoid reaction to contrast and pulmonary edema.
Other complications from CT scanning include contrast
extravasation into a limb that causes pain, or in severe
cases, compartment syndrome. Fortunately, extravasation
is rare, occurring in <1 in 500 patients (67).
For patients who report a history of previous immediate reaction to iodinated contrast (itching, urticarial,
wheezing, or full anaphylactoid reaction), their recurrence rate is approximately 615% with re-exposure,
compared to 1% for patients with no prior contrast reaction (67,68). The risk of breakthrough hypersensitivity
seems to be reduced by one-half with pretreatment with
parenteral corticosteroids (e.g., hydrocortisone, 200 mg,
i.v.) and antihistamines (e.g., chlorpheniramine 4 mg
i.v. or diphenhydramine 25 mg i.v.) (6973). In general,
patients with prior allergic diathesis (e.g., any allergy,
asthma or general atopy) have a 310-fold increased
risk of immediate contrast reaction, but data are mixed
as to whether a shellfish allergy increases a patients
risk of immediate contrast media reaction (74,75).
About 15% of patients undergoing contrast-enhanced
chest CT scanning go on to develop contrast nephropathy,
which, according to its minimal definition, comprises a
25% increase in the serum creatinine concentration,
measured within 27 days of the examination (76).
109
Figure 2. (A) Obvious saddle embolism (yellow arrowheads) in a main pulmonary artery with 329 Hounsfield units opacification
density. (B) Questionable filling defect (yellow arrowhead) within a left lower lobar artery associated with only 167 Hounsfield
units of opacification. Note that visual inspection of the quality contrast opacification can be misleading without on-screen region of interest measurement of the Hounsfield units. These images illustrate that computed tomography chest angiography can
range from highly certain to ambiguous and underscore the importance of talking to the radiologist about image quality.
Whether or not this laboratory finding represents clinically important kidney injury remains controversial, but
contrast nephropathy has been associated with worse outcomes (7679). At present, no specific prophylactic
measure beyond prehydration with intravenous saline
has demonstrated any beneficial effect to reduce the

incidence and significance of contrast nephropathy (80).
The increased resolution of multidetector-row
CT scanning has led to an increase in the detection of
isolated subsegmental PE. About one-quarter of all
110
contrast-enhanced chest CT scans read as positive for PE

have isolated subsegmental PE (8183). Isolated
subsegmental PE refers to a filling defect seen in one
small pulmonary artery, usually <3 mm in diameter, in
the absence of DVT. The problem with this diagnosis is
that when the same images are shown to a second
blinded radiologist, in about half of all cases, the
second radiologist finds no PE (45). This raises concern
that subsegmental PE may be a radiographic artifact
rather than a true disease. One survey found that most clinicians in Canada would opt not to treat these patients
without further testing (84). However, another study
found that the prognosis of patients with subsegmental
PE was not different than patients with segmental or
more proximal PE (85). No randomized trial has examined the safety of withholding anticoagulation for isolated subsegmental PE and perhaps as a result, one
clinical guideline recommends standard course anticoagulation (86,87). In general, most experts agree that
isolated subsegmental PE in patients with active cancer
should be treated, even if discovered incidentally
(84,8892). For patients without cancer, a second
international survey found that a majority thrombosis
experts recommend anticoagulation for most isolated
subsegmental PE (93). The authors recommend treatment
of isolated subsegmental PE if the patient has PE symptoms, a prior history of PE, an ongoing risk of PE (e.g.,
indwelling catheter or immobility), or an elevated
Ddimer (93). Patients with isolated subsegmental PE
and a high risk of bleeding (e.g., Registro Informatizado
de pacientes con Enfermedad TromboEmbolica [RIETE]
bleeding score > 1) probably should not be treated (94). In
these situations, and in any case where the risks and benefits of treatment are unclear, the patient should be
informed of the situation and help make the decision
about anticoagulation.
Ventilation Perfusion Scintillation (V/Q)
Ventilation-perfusion scintillation (V/Q) lung scanning
requires peripheral intravenous access, and for the patient
to sit upright during injection of a radioisotopic nuclide,
usually 99Tc macroaggregate, followed by positioning
the patient in front of a gamma camera to capture the
gamma emission from the radionuclide as it traverses
the pulmonary vasculature. Use of a central line to inject
isotope will often lead to inadequate images. If the perfusion lung scan shows a homogenous perfusion pattern
(i.e., a normal perfusion scan; Figure 3A), this is associated with a likelihood ratio negative of 0.05, and essentially rules out PE (95,96). Patients without PE and with
normal chest radiographs are far more likely to have
normal V/Q scan than patients with intrinsic lung
disease seen on chest radiograph. In patients with
nonnormal perfusion, most U.S. radiology departments

also perform the ventilation phase of the V/Q scan,
which requires the patient to inhale an aerosol
containing 99mTc diethylenetriamine-pentaacetic acid
or 133Xe. Although starting with the perfusion scan may
obviate the need for the ventilation scan, and thereby
reduce radiation exposure, image quality is best if the
ventilation phase is performed first because the background emission from the perfusion scan persists for
hours. To diagnose PE, the perfusion scan must show
two or more apex-central, wedge-shaped defects in perfusion pattern in a segmental or larger vascular distribution,
together with evidence of normal ventilation in the same
lung segments (Figure 3B). The primary technical limitations of V/Q scanning include the availability of
personnel to perform and interpret them and the availability of isotope. Some emergency clinicians may not be
aware that the availability of the 99Tc isotope depends
upon a cyclotron particle accelerator to manufacture
each day. The primary clinical limitation of V/Q scanning
is the fact that approximately two-thirds of scans are
neither normal nor diagnostic of PE, which requires patients to undergo further testing.
Bilateral Ultrasonography
Performing lower-extremity venous ultrasonography is
sensible due to its lack of ionizing radiation and the fact
that diagnosing DVT is tantamount to diagnosing PE
from the standpoint of the decision to administer heparin
anticoagulation in the ED. In fact, studies suggest that a
combination of D-dimer testing and lower-extremity ultrasound may be the most cost-effective approach to the
initial evaluation of PE (22,97). However, in the
absence of physical findings that suggest DVT, bilateral
ultrasonography is of limited usefulness for excluding
PE in the ED. Data from the largest study that
simultaneously performed bilateral leg ultrasonography
and performed pulmonary vascular imaging indicate
that a negative bilateral proximal lower-extremity venous
ultrasound has a sensitivity of 30% and a specificity of
57% (LR for a negative test = 1.22) for PE (98). Other
studies have yielded similar results, so all patients suspected of having PE for whom ultrasound findings are
negative require pulmonary vascular imaging (51,99).
Nonetheless, following the logic that discovery of DVT
is tantamount to diagnosis of PE in terms of treatment
action in the ED, pursuing DVT first makes sense for
patients with a positive D-dimer and for patients who
object to pulmonary vascular imaging.
Follow-up bilateral venous ultrasound is also the best
option to rule out PE in patients with high pretest probability, a positive D-dimer, and a negative CTPA scan with
any radiologist comment about degraded image quality.
111
Figure 3. (A) Normal ventilation-perfusion (V/Q) scan showing homogenous ventilation (top) and perfusion (bottom) images.
These images rule out pulmonary embolism. (B) V/Q lung scan series consistent with a high probability of acute pulmonary embolism using standard criteria. The top row of each panel shows the ventilation phase, conducted with 133Xe, which produces only
one planar image. The images labeled equilibrium or w/o represent washout images taken later. The second row of each
panel demonstrates perfusion phases of the examination, obtained with 99Tc macroaggregate. The black arrowheads point to
wedge-shaped defects in the perfusion images. Scans are read by looking for defects in the perfusion phase where the corresponding ventilation view shows relatively homogeneous scintillation activity in the anatomic segments that lack perfusion.
Management protocols using this approach find that 5%

of these high-risk patients will have a DVT at 37 days
follow-up (5558).
Pregnant Women
Figure 4 proposes an algorithm to rule out and diagnose PE
in pregnant patients. The algorithm starts with bilateral
lower-extremity venous ultrasound. If the bilateral ultrasound is positive, then treatment can be started. Otherwise,
the next step is determined by pretest probability assessment. To our knowledge, no pretest probability rules
have been validated in pregnant patients. It is clear that
over one-half of all VTEs diagnosed in pregnancy occur
in the third trimester (100). In the authors experience
(JAK and CK) and based on available patient-level data
112
Figure 4. Proposed algorithm for the exclusion and diagnosis of pulmonary embolism in pregnant patients with suspected pulmonary embolism (PE) in the emergency department setting. This algorithm has not been formally tested. Shared decision-making refers to discussion of the diagnostic options with the patient, including uncertain, but probably <5% risk of undiagnosed PE and the
potential risks of computed tomographic pulmonary angiogram (CTPA) or V/Q scanning to the fetus. Nonhigh pretest probability
(PTP) refers to absence of high PTP by gestalt, Wells or sRGS. See text for references. *D-dimer concentrations per trimester given
in ng/mL assuming a standard D-dimer threshold for abnormal of 500 ng/mL. Abbreviations: CXR = chest radiograph; Q = perfusion
lung scan; + = positive for PE; ! = negative for PE; Cr = creatinine; High = high-probability scan findings; LMWH = low-molecularweight heparin; Nl = normal; Nondx = nondiagnostic (any reading other than normal or high probability); PERC = pulmonary embolism rule-out criteria; quant = quantitative, sRGS = simplified revised Geneva score; V/Q = ventilation perfusion.
from pregnant ED patients, high Wells and Geneva scores,

the third trimester, or unexplained hypoxemia
(SaO2 < 95% breathing room air at sea level) predict a relatively higher pretest probability for PE (101).
The evaluation of possible PE in pregnancy challenges
clinicians, who must consider the epidemiological data
showing increased risk of PE, the potential catastrophe
of failing to protect the life of mother and child, and the
potential for increased lifetime risks from unnecessary radiation and contrast exposure to the mother and the fetus.
It is worth noting that most patients with pregnancy

selected by emergency physicians for PE work-up have
a low clinical probability (101). No firm guidelines exist
to guide the work-up of pregnant patients with suspected
PE (102,103). Efforts should be made to avoid fetal
exposure to radiation and iodinated contrast (104,105).
The proposed algorithm in Figure 4 draws from available
literature and expert opinion, and has been used informally by one of the authors for over 8 years, but has
not been formally tested (102). If the patient has a high
pretest probability by Wells or the revised Geneva score,

or is in the third trimester, or has unexplained hypoxemia,
then she should probably proceed to testing. Unfortunately, the D-dimer at standard threshold is not usually
helpful because almost all women in the third trimester
have a positive D-dimer when the standard positivity
threshold is used, and pulmonary vascular imaging is
frequently required (106). Because imaging with ionizing
radiation may be required, the authors recommend discussing the diagnostic approach with the patient prior to
initiating the work-up using a shared decision-making
approach, similar to that described by Hess et al. for
low-risk chest pain (107). When the risks and benefits
of testing are explained, some mothers will opt to proceed, whereas others will choose to avoid imaging at all
costs. To minimize radiation exposure, the authors propose a combined approach, where negative bilateral
lower-extremity venous ultrasonography is supported
by a negative PERC rule and a threshold-adjusted Ddimer assay. The exclusionary power of a single negative
bilateral leg ultrasound for PE per se has not been tested
in pregnancy, but seems to be similar to that of nonpregnant patients for excluding DVT (108,109). The D-dimer
threshold can be adjusted according to the trimester of
pregnancy, as follows: first trimester, 750 ng/mL;
second trimester, 1000 ng/mL; third trimester, 1250 ng/
mL (assuming a standard cutoff of 500 ng/mL)
(106,110,111). If the patient has a non-high-pretest probability, has no high-risk features, is PERC negative, and
the bilateral ultrasound is negative, and the D-dimer is
below the trimester-adjusted values, PE can be ruled
out to a reasonable degree of medical certainty. Note
that this recommendation does not state that the PERC
criteria can be used alone in pregnancy.
If the D-dimer is abnormal or the patient fails the
PERC criteria, then a pulmonary vascular imaging study
is warranted. The best choice of pulmonary vascular imaging is controversial and uncertain (102). Some evidence has suggested up to a 35% rate of inadequate
pulmonary vascular opacification with CPTA, especially
in the third trimester, resulting in a higher rate of nondiagnostic studies with CTPA than V/Q or Q-alone scanning
(112,113). Other data indicate that either CT pulmonary
angiography or V/Q scanning will produce adequate
images to rule out and diagnose PE in a pregnant
patient (103,114,115). The data used to estimate the
risk of fetal exposure to radiation for CT scanning vs.
V/Q scanning are both highly speculative. Shielding the
abdomen with a lead or bismuth-antimony apron during
CT scanning may reduce radiation based upon phantom
modeling (116). When available, tube voltage modulating technology may also serve to lower fetal radiation
exposure more than shielding (116). Magnetic resonance
imaging has not been adequately tested in pregnancy to
113
provide any basis for recommendation, but had too low

a sensitivity (78%) to rule out PE in nonpregnant patients
(117). As Figure 4 demonstrates, both CTPA and V/Q
scanning are equally justifiable when imaging a pregnant
patient is necessary. If a V/Q scan is chosen, the authors
suggest first performing a plain film chest radiograph, and
performing the V/Q scan only if the radiograph is normal.
Then, we suggest performing a perfusion-only (Q) scan
with half-dose 99Tc macroaggregate. Because 99Tc is
excreted in the urine, prehydration with 1 L of intravenous saline and insertion of a Foley catheter seems a
logical, but unproven, step to reduce fetal exposure to radiation. The risk of this approach is that if the perfusion
lung scan is nonnormal, and CT scanning is ultimately
required, the mother and fetus will be exposed to more radiation than if CTPA had been performed first. In patients
with an abnormal chest radiograph, we suggest performing a CTPA rather than a V/Q scan.
CONCLUSION
Acute pulmonary embolism can be ruled out on clinical
grounds without laboratory or radiographic imaging by
the combined use of gestalt pretest probability estimation
plus negative PERC rule. In the presence of non-highpretest probability by any method, including gestalt
assessment, a negative quantitative D-dimer rules out
PE at standard or age-adjusted threshold. However,
threshold adjustment is complicated by differing,
manufacturer-specific thresholds to define the cutoff for
an abnormal D-dimer. A good-quality CTPA scan rules
out PE. Patients with high pretest probability and a negative CTPA but with degraded image quality, can have PE
ruled out with a normal V/Q scan, a negative quantitative
D-dimer, or negative bilateral lower-extremity venous ultrasound performed in the ED and again 37 days later.
Exclusion of PE in pregnancy remains a controversial
subject, but a shared decision-making model that prioritizes testing without fetal radiation exposure may offer
the most effective and safe approach.
REFERENCES
1. Manara A, Dhoore W, Thys F. Capnography as a diagnostic tool
for pulmonary embolism: a meta-analysis. Ann Emerg Med
2013;62:58491.
2. Marchick MR, Courtney DM, Kabrhel C, et al. 12-Lead ECG
findings of pulmonary hypertension occur more frequently in
emergency department patients with pulmonary embolism than
in patients without pulmonary embolism. Ann Emerg Med
2010;55:3315.
3. Lucassen W, Geersing GJ, Erkens PM, et al. Clinical decision
rules for excluding pulmonary embolism: a meta-analysis. Ann
Intern Med 2011;155:44860.
4. Ceriani E, Combescure C, Le Gal G, et al. Clinical prediction
rules for pulmonary embolism: a systematic review and metaanalysis. J Thromb Haemost 2010;8:95770.
114
5. Klok FA, Mos IC, Nijkeuter M, et al. Simplification of the revised
Geneva score for assessing clinical probability of pulmonary embolism. Arch Intern Med 2008;168:21316.
6. Runyon MS, Webb WB, Jones AE, et al. Comparison of the unstructured clinician estimate of low clinical probability for pulmonary embolism to the Canadian score or the Charlotte rule.
Acad Emerg Med 2005;12:58793.
7. Hogg KE, Brown MD, Kline JA. Estimating the pretest probability to justify the empiric admininistration of heparin prior to pulmonary vascular imaging for pulmonary embolism. Thromb Res
2006;118:54753.
8. Smith SB, Geske JB, Maguire JM, et al. Early anticoagulation is
associated with reduced mortality for acute pulmonary embolism. Chest 2010;137:13829.
9. Jelinek GA, Ingarfield SL, Mountain D, et al. Emergency department diagnosis of pulmonary embolism is associated with significantly reduced mortality: a linked data population study. Emerg
Med Australas 2009;21:26976.
10. Kline JA, Marchick MR, Kabrhel C, et al. Prospective study of the
frequency and outcomes of patients with suspected pulmonary embolism administered heparin prior to confirmatory imaging.
Thromb Res 2012;129:e258.
11. Kline JA, Hernandez J, Jones AE, et al. Prospective study of the
clinical features and outcomes of emergency department patients
with delayed diagnosis of pulmonary embolism. Acad Emerg
Med 2007;14:5928.
12. Torres-Macho J, Mancebo-Plaza AB, Crespo-Gimenez A, et al.
Clinical features of patients inappropriately undiagnosed of pulmonary embolism. Am J Emerg Med 2013;31:164650.
13. den Exter PL, van den Hoven P, van der Hulle T, et al. Performance
of the revised Geneva score in patients with a delayed suspicion of
pulmonary embolism. Eur Respir J 2014;43:18014.
14. Kline JA, Slattery D, ONeil BJ, et al. Clinical features of patients
with pulmonary embolism and a negative PERC rule result. Ann
Emerg Med 2013;61:1224.
15. Kabrhel C, Courtney DM, Camargo CA Jr, et al. Potential impact
of adjusting the threshold of the quantitative D-dimer based upon
pretest probability of acute pulmonary embolism. Acad Emerg
Med 2009;16:32542.
16. Singh B, Mommer SK, Erwin PJ, et al. Pulmonary embolism ruleout criteria (PERC) in pulmonary embolismrevisited: a systematic review and meta-analysis. Emerg Med J 2013;30:7016.
17. Singh B, Parsaik AK, Agarwal D, Surana A, Mascarenhas SS,
Chandra S. Diagnostic accuracy of pulmonary embolism ruleout criteria: a systematic review and meta-analysis. Ann Emerg
Med 2012;59:51720. e1e4.
18. Fesmire FM, Brown MD, Espinosa JA, et al. Critical issues in the
evaluation and management of adult patients presenting to the
emergency department with suspected pulmonary embolism.
Ann Emerg Med 2011;57:62852.
19. Penaloza A, Verschuren F, Dambrine S, et al. Performance of the
pulmonary embolism rule-out criteria (the PERC rule) combined
with low clinical probability in high prevalence population.
Thromb Res 2012;129:e18993.
20. Bokobza J, Aubry A, Nakle N, et al. Pulmonary embolism rule-out
criteria vs D-dimer testing in low-risk patients for pulmonary embolism: a retrospective study. Am J Emerg Med 2014;32:60913.
21. Hugli O, Righini M, Le Gal G, et al. The pulmonary embolism
rule-out criteria (PERC) rule does not safely exclude pulmonary
embolism. J Thromb Haemost 2011;9:3004.
22. Lessler AL, Isserman JA, Agarwal R, et al. Testing low-risk patients for suspected pulmonary embolism: a decision analysis.
Ann Emerg Med 2010;55:31626.
23. Pauker SG, Kassirer JP. The threshold approach to clinical decision making. N Engl J Med 1980;302:110917.
24. Hennessey A, Setyono DA, Lau WB, et al. A patient with a large
pulmonary saddle embolus eluding both clinical gestalt and validated decision rules. Ann Emerg Med 2012;59:5213.
25. van Belle A, Buller HR, Huisman MV, et al. Effectiveness of managing suspected pulmonary embolism using an algorithm
combining clinical probability, D-dimer testing, and computed tomography. JAMA 2006;295:1729.

26. Bertoletti L, Le Gal G, Aujesky D, et al. Prognostic value of the
Geneva prediction rule in patients with pulmonary embolism.
Thromb Res 2013;132:326.
27. Stein PD, Hull RD, Patel KC, et al. D-dimer for the exclusion of
acute venous thrombosis and pulmonary embolism: a systematic
review. Ann Intern Med 2004;140:589602.
28. Brown MD, Lau J, Nelson RD, et al. Turbidimetric D-Dimer test in
the diagnosis of pulmonary embolism: a meta-analysis. Clin Chem
2003;49:184653.
29. Couturaud F, Kearon C, Bates SM, Ginsberg JS. Decrease in
sensitivity of D-dimer for acute venous thromboembolism after
starting anticoagulant therapy. Blood Coagul Fibrinolysis 2002;
13:2416.
30. Taira T, Taira BR, Carmen M, et al. Risk of venous thromboembolism in patients with borderline quantitative D-dimer levels. Am J
Emerg Med 2010;28:4503.
31. Kutinsky I, Blakley S, Roche V. Normal D-dimer levels in patients
with pulmonary embolism. Arch Intern Med 1999;159:156972.
32. Runyon MS, Gellar MA, Sanapareddy N, et al. Development and
comparison of a minimally-invasive model of autologous clot pulmonary embolism in Sprague-Dawley and Copenhagen rats.
Thromb J 2010;8:3.
33. Kline JA, Hogg MM, Courtney DM, et al. D-dimer threshold increase with pretest probability unlikely for pulmonary embolism
to decrease unnecessary computerized tomographic pulmonary
angiography. J Thromb Haemost 2012;10:57281.
34. Kabrhel C, Mark Courtney D, Camargo CA Jr, et al. Factors associated with positive D-dimer results in patients evaluated for pulmonary embolism. Acad Emerg Med 2010;17:58997.
35. Penaloza A, Roy PM, Kline J, et al. Performance of age-adjusted
D-dimer cut-off to rule out pulmonary embolism. J Thromb Haemost 2012;10:12916.
36. Adams D, Welch JL, Kline JA. Clinical utility of an age-adjusted
D-dimer in the diagnosis of venous thromboembolism. Ann Emerg
Med 2014;64:2324.
37. Righini M, van Es J, den Exter PL, et al. Age-adjusted D-dimer
cutoff levels to rule out pulmonary embolism: the ADJUST-PE
study. JAMA 2014;311:111724.
38. Becattini C, Vedovati MC, Agnelli G. Prognostic value of troponins in acute pulmonary embolism: a meta-analysis. Circulation
2007;116:42733.
39. Kline JA, Hernandez J, Rose G, et al. Surrogate markers for
adverse outcomes in normotensive patients with pulmonary embolism. Crit Care Med 2006;34:277380.
40. Jimenez D, Kopecna D, Tapson V, et al. Derivation and validation
of multimarker prognostication for normotensive patients with
acute symptomatic pulmonary embolism. Am J Respir Crit Care
Med 2014;189:71826.
41. Kline JA, Corredor DM, Hogg MM, et al. Normalization of vital
signs does not reduce the probability of acute pulmonary embolism in symptomatic emergency department patients. Acad Emerg
Med 2012;19:117.
42. Burnside P, Kline JA. Indirect computed tomography venography:
quality of vascular opacification and diagnostic implications.
Emerg Radiol 2010;17:195202.
43. Mos IC, Klok FA, Kroft LJ, et al. Safety of ruling out acute pulmonary embolism by normal computed tomography pulmonary angiography in patients with an indication for computed tomography:
systematic review and meta-analysis. J Thromb Haemost 2009;7:
14918.
44. Schissler AJ, Rozenshtein A, Kulon ME, et al. CT pulmonary angiography: increasingly diagnosing less severe pulmonary emboli.
PLoS One 2013;8:e65669.
45. Courtney DM, Miller CD, Smithline HA, et al. Prospective multicenter assessment of interobserver agreement for radiologist interpretation of multidetector CT angiography for pulmonary embolism. J Thromb Haemost 2010;8:53340.
46. Richman PB, Courtney DM, Kline JA. Prevalence and significance
of non-thromboembolic findings on chest computerized tomography angiography performed to rule-out pulmonary embolism-A
multi-center study of 1025 Emergency Department patients.
Acad Emerg Med 2004;11:6427.

47. Hall WB, Truitt SG, Scheunemann LP, et al. The prevalence of
clinically relevant incidental findings on chest computed tomographic angiograms ordered to diagnose pulmonary embolism.
Arch Intern Med 2009;169:19615.
48. van Es J, Douma RA, Schreuder SM, et al. Clinical impact of findings supporting an alternative diagnosis on CT pulmonary angiography in patients with suspected pulmonary embolism. Chest
2013;144:18939.
49. van Strijen MJ, Bloem JL, de Monye W, et al. Helical computed
tomography and alternative diagnosis in patients with excluded
pulmonary embolism. J Thromb Haemost 2005;3:244956.
50. Shujaat A, Shapiro JM, Eden E. Utilization of CT pulmonary angiography in suspected pulmonary embolism in a major urban emergency department. Pulm Med 2013;2013:915213.
51. Stein PD, Fowler SE, Goodman LR, et al. Multidetector computed
tomography for acute pulmonary embolism. N Engl J Med 2006;
354:231727.
52. Kay FU, Macedo AC, Chate RC, et al. Reduction of poor contrast
enhancement of the pulmonary artery in computed tomography
angiography using an alternative respiratory maneuver. J Thorac
Imaging 2014;29:10712.
53. Hawley PC, Hawley MP. Difficulties in diagnosing pulmonary embolism in the obese patient: a literature review. Vasc Med 2011;16:
44451.
54. Bae KT, Tao C, Gurel S, et al. Effect of patient weight and scanning duration on contrast enhancement during pulmonary multidetector CT angiography. Radiology 2007;242:5829.
55. Wells PS, Anderson D, Bormanis J. Value of assessment of pretest
probability of deep-vein thrombosis in clinical management. Lancet 1997;350:17958.
56. Heijboer H, Buller HR, Lensing AW, et al. A comparison of realtime compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic
outpatients. N Engl J Med 1993;329:13659.
57. Birdwell BG, Raskob GE, Whitsett TL, et al. The clinical validity of normal compression ultrasonography in outpatients suspected of having deep vein thrombosis. Ann Intern Med 1998;
128:17.
58. Wells P, Ginsberg J, Anderson D, et al. Use of a clinical model for
safe management of patients with suspected pulmonary embolism.
Ann Intern Med 1998;129:9971005.
59. Kline JA, Runyon MS, Webb WB, Jones AE, Mitchell AM. Prospective study of the diagnostic accuracy of the simplify D-dimer
assay for pulmonary embolism in emergency department patients.
Chest 2006;129:141723.
60. Runyon MS, Beam DM, King MC, et al. Comparison of the
simplify D-dimer assay performed at the bedside with a
laboratory-based quantitative D-dimer assay for the diagnosis of
pulmonary embolism in a low prevalence emergency department
population. Emerg Med J 2008;25:705.
61. Gottschalk A, Sostman HD, Coleman RE, et al. Ventilation-perfusion scintigraphy in the PIOPED study. Part II. Evaluation of the
scintigraphic criteria and interpretations. J Nucl Med 1993;34:
111926.
62. Hoffman JR, Cooper RJ. Overdiagnosis of disease: a modern
epidemic. Arch Intern Med 2012;172:11234.
63. Einstein AJ, Henzlova MJ, Rajagopalan S. Estimating risk of
cancer associated with radiation exposure from 64-slice
computed tomography coronary angiography. JAMA 2007;298:
31723.
64. Health Physics Society. Radiation exposure from medical diagnostic imaging procedures Health Physics Society fact sheet.
McLean, VA: Health Physics Society; 2008.
65. Parker MS, Hui FK, Camacho MA, et al. Female breast radiation
exposure during CT pulmonary angiography. AJR Am J Roentgenol 2005;185:122833.
66. Kline JA, Courtney DM, Beam DM, et al. Incidence and predictors
of repeated computed tomographic pulmonary angiography in
emergency department patients. Ann Emerg Med 2009;54:418.
67. Mitchell AM, Jones AE, Tumlin JA, et al. Immediate complications of intravenous contrast for computed tomography imaging
115
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
in the outpatient setting are rare. Acad Emerg Med 2011;18:

10059.
Kopp AF, Mortele KJ, Cho YD, et al. Prevalence of acute reactions
to iopromide: postmarketing surveillance study of 74,717 patients.
Acta Radiol 2008;49:90211.
Williams AN, Kelso JM. Radiocontrast-induced anaphylaxis
despite pretreatment and use of iso-osmolar contrast. Ann Allergy
Asthma Immunol 2007;99:4678.
Marshall GD Jr, Lieberman PL. Comparison of three pretreatment
protocols to prevent anaphylactoid reactions to radiocontrast media. Ann Allergy 1991;67:704.
Kim SH, Lee SH, Lee SM, et al. Outcomes of premedication for
non-ionic radio-contrast media hypersensitivity reactions in Korea. Eur J Radiol 2011;80:3637.
Bae YJ, Hwang YW, Yoon SY, et al. The effectiveness of automatic recommending system for premedication in reducing recurrent radiocontrast media hypersensitivity reactions. PLoS One
2013;8:e66014.
Delaney A, Carter A, Fisher M. The prevention of anaphylactoid
reactions to iodinated radiological contrast media: a systematic review. BMC Med Imaging 2006;6:2.
Wolf GL, Mishkin MM, Roux SG, et al. Comparison of the rates of
adverse drug reactions. Ionic contrast agents, ionic agents combined with steroids, and nonionic agents. Invest Radiol 1991;26:
40410.
Baig M, Farag A, Sajid J, et al. Shellfish allergy and relation to
iodinated contrast media: United Kingdom survey. World J Cardiol
2014;6:10711.
Sinert R, Brandler E, Subramanian RA, et al. Does the current definition of contrast-induced acute kidney injury reflect a true clinical
entity? Acad Emerg Med 2012;19:12617.
Mitchell AM, Jones AE, Tumlin JA, et al. One year outcomes
following contrast induced nephropathy. Am J Intern Med 2013;
1:16.
Mitchell AM, Jones AE, Tumlin JA, et al. Prospective study of the
incidence of contrast-induced nephropathy among patients evaluated for pulmonary embolism by contrast-enhanced computed tomography. Acad Emerg Med 2012;19:61825.
Mitchell AM, Kline JA. Contrast-induced nephropathy: doubts
and certainties. Acad Emerg Med 2012;19:12946.
Au TH, Bruckner A, Mohiuddin SM, Hilleman DE. The prevention of contrast-induced nephropathy. Ann Pharmacother 2014;
48:133242.
Stein PD, Goodman LR, Hull RD, et al. Diagnosis and management
of isolated subsegmental pulmonary embolism: review and assessment of the options. Clin Appl Thromb Hemost 2012;18:206.
Carrier M, Righini M, Le Gal G. Symptomatic subsegmental pulmonary embolism: what is the next step? J Thromb Haemost 2012;
10:148690.
Carrier M, Righini M, Wells PS, et al. Subsegmental pulmonary
embolism diagnosed by computed tomography: incidence and
clinical implications. A systematic review and meta-analysis of
the management outcome studies. J Thromb Haemost 2010;8:
171622.
Carrier M, Kimpton M, LE Gal G, et al. The management of a subsegmental pulmonary embolism: a cross-sectional survey of Canadian thrombosis physicians. J Thromb Haemost 2011;9:14125.
den Exter PL, van Es J, Klok FA, et al. Risk profile and clinical
outcome of symptomatic subsegmental acute pulmonary embolism. Blood 2013;122:11449.
Kearon C, Akl EA, Comerota AJ, et al. Antithrombotic therapy for
VTE disease: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians EvidenceBased Clinical Practice Guidelines. Chest 2012;141(2 Suppl):
e419S94.
Yoo HH, Queluz TH, El Dib R. Anticoagulant treatment for subsegmental pulmonary embolism. Cochrane Database Syst Rev
2014;(4):CD010222.
den Exter PL, Jimenez D, Kroft LJ, Huisman MV. Outcome of
incidentally diagnosed pulmonary embolism in patients with malignancy. Curr Opin Pulm Med 2012;18:399405.
116
89. Abdel-Razeq HN, Mansour AH, Ismael YM. Incidental pulmonary embolism in cancer patients: clinical characteristics and
outcomea comprehensive cancer center experience. Vasc Health
Risk Manag 2011;7:1538.
90. Douma RA, Kok MG, Verberne LM, et al. Incidental venous
thromboembolism in cancer patients: prevalence and consequence. Thromb Res 2010;125:e3069.
91. Dentali F, Ageno W, Becattini C, et al. Prevalence and clinical history of incidental, asymptomatic pulmonary embolism: a metaanalysis. Thromb Res 2010;125:51822.
92. Lim WY, Bozas G, Noble S, Hart S, Maraveyas A. Anticoagulating
the subsegmental pulmonary embolism in cancer patients: a survey
amongst different medical specialties. J Thromb Thrombolysis
2014 Oct 18. [Epub ahead of print].
93. Pesavento R, Casazza F, Filippi L, et al. An international survey on
isolated subsegmental pulmonary embolism. Thromb Res 2013;
131:1834.
94. Nieto JA, Solano R, Trapero IN, et al. Validation of a score for predicting fatal bleeding in patients receiving anticoagulation for
venous thromboembolism. Thromb Res 2013;132:1759.
95. The PIOPED Investigators. Value of the ventilation/perfusion scan
in acute pulmonary embolism. JAMA 1990;263:27539.
96. Worsley DF, Alavi A. Comprehensive analysis of the results of the
PIOPED study. J Nucl Med 1995;36:23807.
97. Duriseti RS, Brandeau ML. Cost-effectiveness of strategies for
diagnosing pulmonary embolism among emergency department
patients presenting with undifferentiated symptoms. Ann Emerg
Med 2010;56:32132.
98. Righini M, Le Gal G, Aujesky D, et al. Diagnosis of pulmonary
embolism by multidetector CT alone or combined with venous ultrasonography of the leg: a randomised non-inferiority trial. Lancet 2008;371:134352.
99. Daniel KR, Jackson RE, Kline JA. Utility of the lower extremity venous ultrasound in the diagnosis and exclusion of
pulmonary embolism in outpatients. Ann Emerg Med 2000;
35:54754.
100. Meng K, Hu X, Peng X, et al. Incidence of venous thromboembolism during pregnancy and the puerperium: a systematic review
and meta-analysis. J Matern Fetal Neonatal Med 2014 May 7;1
9. [Epub ahead of print].
101. Kline JA, Richardson DM, Than MP, et al. Systematic review and
meta-analysis of pregnant patients investigated for suspected pulmonary embolism in the emergency department. Acad Emerg Med
2014;21:94959.
102. Leung AN, Bull TM, Jaeschke R, et al. An official American
Thoracic Society/Society of Thoracic Radiology clinical practice
guideline: evaluation of suspected pulmonary embolism in pregnancy. Am J Respir Crit Care Med 2011;184:12008.

103. Tan M, Huisman MV. The diagnostic management of acute venous
thromboembolism during pregnancy: recent advancements and
unresolved issues. Thromb Res 2011;127(Suppl 3):S136.
104. Moon AJ, Katzberg RW, Sherman MP. Transplacental passage of
iohexol. J Pediatr 2000;136:5489.
105. Matthews S. Imaging pulmonary embolism in pregnancy: what is the
most appropriate imaging protocol? Br J Radiol 2006;79:4414.
106. Kline JA, Hambleton GW, Hernandez J. D-dimer concentrations in
normal pregnancy: new diagnostic thresholds are needed. Clin
Chem 2005;51:8259.
107. Hess EP, Knoedler MA, Shah ND, et al. The chest pain choice decision aid: a randomized trial. Circ Cardiovasc Qual Outcomes
2012;5:2519.
108. Ratiu A, Navolan D, Spatariu I, et al. Diagnostic value of a negative
single color duplex ultrasound in deep vein thrombosis suspicion
during pregnancy. Rev Med Chir Soc Med Nat Iasi 2010;114:4546.
109. Le Gal G, Kercret G, Ben Yahmed K, et al. Diagnostic value of single complete compression ultrasonography in pregnant and postpartum women with suspected deep vein thrombosis:
prospective study. BMJ 2012;344:e2635.
110. Chan WS, Lee A, Spencer FA, et al. D-dimer testing in pregnant
patients: towards determining the next level in the diagnosis
of deep vein thrombosis. J Thromb Haemost 2010;8:100411.
111. Chan WS, Ginsberg JS. Management of venous thromboembolism
in pregnancy. In: van Beek EJR, Buller HR, Oudkerk M, eds. Deep
vein thrombosis and pulmonary embolism. 1st edn. Chichester,
West Sussex: John Wiley & Sons; 2009:35371.
112. Ridge CA, McDermott S, Freyne BJ, et al. Pulmonary embolism in
pregnancy: comparison of pulmonary CT angiography and lung
scintigraphy. AJR Am J Roentgenol 2009;193:12237.
113. Cahill AG, Stout MJ, Macones GA, et al. Diagnosing pulmonary
embolism in pregnancy using computed-tomographic angiography
or ventilation-perfusion. Obstet Gynecol 2009;114:1249.
114. Nijkeuter M, Tan M, Middeldorp S, Kroft LJM, Huisman MV.
Safety of ruling out pulmonary embolism (PE) in pregnancy by
computed tomography pulmonary angiography (CTPA). J Thromb
Haemost 2013;11:130.
115. Shahir K, Goodman LR, Tali A, et al. Pulmonary embolism in
pregnancy: CT pulmonary angiography versus perfusion scanning.
AJR Am J Roentgenol 2010;195:W21420.
116. Chatterson LC, Leswick DA, Fladeland DA, et al. Lead versus
bismuth-antimony shield for fetal dose reduction at different gestational ages at CT pulmonary angiography. Radiology 2011;260:
5607.
117. Stein PD, Chenevert TL, Fowler SE, et al. Gadolinium-enhanced
magnetic resonance angiography for pulmonary embolism: a
multicenter prospective study (PIOPED III). Ann Intern Med
2010;152:43443.
ARTICLE SUMMARY
1. Why is this topic important?
Acute pulmonary embolism (PE) can cause sudden
death, and failure to diagnose PE can lead to devastating
patient outcomes and medicolegal allegations of negligence. However, overtesting and overdiagnosis for PE
pose a major threat to public health.
2. What does this review attempt to show?
With awareness of medicolegal implications, and a rich
literature base, this in-depth review considers current evidence to define a rational approach to the exclusion and
diagnosis of PE in the emergency department (ED) setting
that maximizes patient safety.
3. What are the key findings?
Not all patients with a sign, symptom, or risk factor for
PE require a formal evaluation for PE. In gestalt low-risk
patients, the pulmonary embolism rule-out criteria rule or
a D-dimer can be used to rule out PE. Emergency physicians should know the difference between a high-quality
and a low-quality computed tomographic pulmonary
angiogram scan. Algorithms are presented to guide the
process of evaluating possible PE in both nonpregnant
and pregnant patients.
4. How is patient care impacted?
Specific test modalities can rule out PE without patient
exposure to radiation and iodinated contrast material.
Although the exclusion and diagnosis of PE in symptomatic pregnant women remains controversial, a protocol
that includes a shared decision-making approach may be
a rational approach.
117

The First 500

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

The First 500

Hochgeladen von

Copyright:

Verfügbare Formate

American Journal of Emergency Medicine 33 (2015) 197201

Contents lists available at ScienceDirect

American Journal of Emergency Medicine

The rst 500: initial experience with widespread use of low-dose

Department of Emergency Medicine, Alameda Health System, Highland Hospital, Oakland CA

Prior presentations: Accepted for presentation, American Society of Regional Anesthesia

practice [2]. Current pharmacologic strategies in the ED rely heavily

T.L. Ahern et al. / American Journal of Emergency Medicine 33 (2015) 197201

palliative care settings for the treatment of postoperative and chronic

2.2. Study population

2.3. Data abstraction

T.L. Ahern et al. / American Journal of Emergency Medicine 33 (2015) 197201

who developed transient hypoxia, 4 of whom had concurrently received

T.L. Ahern et al. / American Journal of Emergency Medicine 33 (2015) 197201

Age Sex Disposition Indication

Dose Route Comorbiditiesb

After LDK and 2 mg hydromorphone, patient reported, I feel dizzy.

After LDK and 400 mg ibuprofen, patient reported, I feel dizzy.

Sickle cell pain

LDK and 8 mg ondansetron given for nausea and vomiting in setting of

T.L. Ahern et al. / American Journal of Emergency Medicine 33 (2015) 197201

[7] Shah S, Mehta V. Controversies and advances in non-steroidal anti-inammatory drug

Safety of Peripheral Intravenous Administration of

Vasoactive medications (VMs) are often required to

*Address for correspondence and reprint requests: Jose

An Official Publication of the Society of Hospital Medicine

MATERIAL AND METHODS

TABLE 1. Summary of the Requirements for PIV

Access Used for Infusion of VM

A summary of the requirements for PIV access for VM

An Official Publication of the Society of Hospital Medicine

Journal of Hospital Medicine Vol 00 | No 00 | Month 2015

TABLE 3. Demographic and Disease Characteristics

of the Study Population

Catheter length was 30 mm, 45 mm, or 48 mm

pharmacy performed an extensive literature search

Complications of Vasoactive Medication

NOTE: Abbreviations: PIV, peripheral intravenous; SD, standard deviation.

Journal of Hospital Medicine Vol 00 | No 00 | Month 2015

An Official Publication of the Society of Hospital Medicine

VM via CVC. Being a single-center study, it is not

Journal of Hospital Medicine Vol 00 | No 00 | Month 2015

7. Close AS. Phentolamine hydrochloride in prevention of cutaneous

An Official Publication of the Society of Hospital Medicine

following accidental autoinjector inoculation. Vasc Med. 2011;16:

Journal of Hospital Medicine Vol 00 | No 00 | Month 2015

EMERGENCY EVALUATION FOR PULMONARY EMBOLISM, PART 2:

, Keywordspulmonary embolism; medicolegal; defensive medicine; decision making; venous thromboembolism;

RECEIVED: 30 September 2014; FINAL SUBMISSION RECEIVED: 17 December 2014;

Emergency Evaluation for Pulmonary Embolism

process. For example, if an emergency physician cares

any method), the clinician should consider immediately

J. A. Kline and C. Kabrhel

Emergency Evaluation for Pulmonary Embolism

#4, or an RGS #4 (see Tables 3 and 4 in part 1) (25,26).

when used in conjunction with a Wells score #4 or a

Table 1. Factors that Cause Errors in D-dimer Measurements (3337))

OR = odds ratio; PE = pulmonary embolism.

J. A. Kline and C. Kabrhel

that has homogeneous perfusion. Because the outcome

Emergency Evaluation for Pulmonary Embolism

has demonstrated any beneficial effect to reduce the

contrast-enhanced chest CT scans read as positive for PE

J. A. Kline and C. Kabrhel