British Journal of Anaesthesia, 132 (3): 491e506 (2024)

doi: 10.1016/j.bja.2023.11.050
Advance Access Publication Date: 6 January 2024
Review Article

CRITICAL CARE

Pharmacological agents for procedural sedation and analgesia in the

emergency department and intensive care unit: a systematic review
and network meta-analysis of randomised trials
Sameer Sharif1,2,3,* , Jasmine Kang4 , Behnam Sadeghirad3,5 , Fayyaz Rizvi4, Ben Forestell1 ,
1,2 1,2 6,7 8
Alisha Greer , Mark Hewitt , Shannon M. Fernando , Sangeeta Mehta ,
9 3,10
Mohamed Eltorki , Reed Siemieniuk , Mark Duffett , Maala Bhatt12, Lisa Burry8,13,
11

Jeffrey J. Perry6 , Andrew Petrosoniak14 , Pratik Pandharipande15, Michelle Welsford1 and

Bram Rochwerg2,3
1
Department of Medicine, Division of Emergency Medicine, McMaster University, Hamilton, ON, Canada, 2Department of
Medicine, Division of Critical Care, McMaster University, Hamilton, ON, Canada, 3Department of Health Research
Methods, Evidence and Impact, McMaster University, Hamilton, ON, Canada, 4Michael G. DeGroote School of Medicine,
McMaster University, Hamilton, ON, Canada, 5Department of Anesthesia, McMaster University, Hamilton, ON,
Canada, 6Department of Emergency Medicine, University of Ottawa, Ottawa, ON, Canada, 7Division of Critical Care,
Department of Medicine, University of Ottawa, Ottawa, ON, Canada, 8Department of Medicine, Sinai Health System;
Interdepartmental Division of Critical Care Medicine, University of Toronto, Toronto, ON, Canada, 9Department of
10
Pediatrics, Division of Pediatric Emergency Medicine, McMaster University, Ottawa, ON, Canada, Department of
11
Medicine, McMaster University, Hamilton, ON, Canada, Department of Pediatrics, McMaster University, Hamilton, ON,
12
Canada, Department of Medicine, Sinai Health System, Interdepartmental Division of Critical Care Medicine, Toronto,
13
ON, Canada, Department of Pharmacy, Sinai Health System, Leslie Dan Faculty of Pharmacy, University of Toronto,
14
Toronto, ON, Canada, Department of Medicine, Division of Emergency Medicine, University of Toronto, Toronto, ON,
15
Canada and Department of Anesthesiology, Division of Critical Care, Vanderbilt University School of Medicine,
Nashville, TN, USA

*Corresponding author. E-mail: [email protected]

Abstract
Background: We aimed to evaluate the comparative effectiveness and safety of various i.v. pharmacologic agents used
for procedural sedation and analgesia (PSA) in the emergency department (ED) and ICU. We performed a systematic
review and network meta-analysis to enable direct and indirect comparisons between available medications.
Methods: We searched Medline, EMBASE, Cochrane, and PubMed from inception to 2 March 2023 for RCTs comparing two
or more procedural sedation and analgesia medications in all patients (adults and children >30 days of age) requiring
emergent procedures in the ED or ICU. We focused on the outcomes of sedation recovery time, patient satisfaction, and
adverse events (AEs). We performed frequentist random-effects model network meta-analysis and used the Grading of
Recommendations, Assessment, Development, and Evaluation (GRADE) approach to rate certainty in estimates.
Results: We included 82 RCTs (8105 patients, 78 conducted in the ED and four in the ICU) of which 52 studies included
adults, 23 included children, and seven included both. Compared with midazolam-opioids, recovery time was shorter
with propofol (mean difference 16.3 min, 95% confidence interval [CI] 8.4e24.3 fewer minutes; high certainty), and patient
satisfaction was better with ketamine-propofol (mean difference 1.5 points, 95% CI 0.3e2.6 points, high certainty).
Regarding AEs, compared with midazolam-opioids, respiratory AEs were less frequent with ketamine (relative risk [RR]
0.55, 95% CI 0.32e0.96; high certainty), gastrointestinal AEs were more common with ketamine-midazolam (RR 3.08, 95%

Received: 7 August 2023; Accepted: 30 November 2023

© 2023 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved.
For Permissions, please email: [email protected]

491
 492 - Sharif et al.

CI 1.15e8.27; high certainty), and neurological AEs were more common with ketamine-propofol (RR 3.68, 95% CI
1.08e12.53; high certainty).
Conclusion: When considering procedural sedation and analgesia in the ED and ICU, compared with midazolam-opioids,
sedation recovery time is shorter with propofol, patient satisfaction is better with ketamine-propofol, and respiratory
adverse events are less common with ketamine.

Keywords: analgesia; critical care; emergency; network meta-analysis; procedural sedation; systematic review

the purposes of PSA is to ensure that patients have adequate

analgesia and anaesthesia for their painful procedure;
Editor’s key points
furthermore, sedation recovery time is a useful measure to

Despite many large RCTs comparing procedural ensure that we find which medications are most likely to save
sedation and analgesia medications, uncertainty the healthcare resource of monitoring time in the department.
persists regarding the optimal medication or combi-
nation of medications considering both safety and
efficacy. Methods

In this systematic review with network meta-
analysis, the authors highlight the importance of an We followed the Preferred Reporting Items for Systematic
individualised approach to procedural sedation and Review and Meta-Analysis (PRISMA) statement extension for
analgesia based upon patient and procedure network meta-analysis (Supplementary Appendix).8,9 We
characteristics. registered the protocol with the Center for Open Science

These findings provide a current, comprehensive (https://osf.io/vrkns/wiki/home/). There was no external
summary of evidence to guide clinical practice funding.
for procedural sedation and analgesia. The specific
regimes represented by smaller nodes in the analysis
would benefit from more randomised clinical trial Data sources and strategy
data. We searched four databases (Medline, EMBASE, Cochrane, and
PubMed) from inception to May 2021. The search was updated
on 2 March 2023. The search strategy was developed by an
Procedural sedation and analgesia (PSA) refers to the admin- expert health sciences librarian and peer-reviewed
istration of medications with sedative, analgesic, or dissocia- (Supplementary Appendices 1 and 2). To search for unpub-
tive properties with the goal of suppressing a patient’s lished studies, we reviewed conference proceedings from the
consciousness to facilitate care or to perform procedures.1 PSA following organisations for 2020 and beyond: Society of Critical
is commonly performed in-hospital, particularly in the emer- Care Medicine, American Thoracic Society, American College
gency department (ED) and ICU to facilitate procedures such as of Emergency Physicians, Canadian Association of Emergency
bronchoscopy, tracheostomy, emergent endoscopy,2 ortho- Physicians, European Society of Intensive Care Medicine, and
paedic manipulation, abscess incision and drainage, and the American Academy of Pediatrics.
electrical cardioversion.3 There are a variety of medications
that can be selected for PSA with propofol, fentanyl, and
midazolam being the most commonly used4; however, eto- Study selection
midate, ketamine, and dexmedetomidine have seen increased
Screening of titles and abstracts was performed independently
use of late.4
and in duplicate by pairs of reviewers using Covidence soft-
Despite the large number of RCTs comparing these medi-
ware (Melbourne, Australia). The same pairs of reviewers
cations, uncertainty persists regarding the optimal medication
assessed the eligibility of full texts of those citations deemed
or combination of medications considering both safety and
potentially eligible at title and abstract review, independently
efficacy, as there have been numerous randomised trials since
and in duplicate. We resolved disagreements at full text
the last review on this subject was published.1,5,6 Also, previ-
through discussion and consensus. We included published full
ous systematic reviews and meta-analyses have been limited
text or conference abstracts of RCTs, without language re-
to head-to-head pairwise comparisons between two drug re-
striction (Supplementary Appendices 1 and 2).
gimes.1 The objective of this study was to perform a system-
atic review and network meta-analysis of patients (adult or
paediatric >30 days of age) undergoing PSA for emergent
Inclusion criteria
procedures in the ED and ICU in an effort to compare the ef-
ficacy and safety of various i.v. PSA medications. From a safety We used the following eligibility criteria to include studies
perspective, we will focus on reporting adverse events (AEs). that: (i) enrolled adults or children (>30 days of age); (ii)
From an efficacy perspective, we will focus on patient satis- compared at least two different i.v. PSA medication regimes-
faction and sedation recovery time. All these outcomes were these may have included single or combined medications used
selected with patient importance in mind as per Grading of for procedural sedation; (iii) examined sedation in patients for
Recommendations, Assessment, Development, and Evalua- a specific procedure performed in the ED or ICU; (iv) evaluated
tion (GRADE) guidance.7 We chose these outcomes as one of at least one of the outcomes of interest.
 Procedural Sedation in the ED and ICU - 493

Exclusion criteria residual missing data were incorporated into RoB and GRADE
certainty assessments. When the number of events was zero,
We excluded RCTs that used PSA in the following contexts: (i)
we used the continuity correction as recommended by
noninvasive positive pressure ventilation (NIPPV); (ii) as part of
Sweeting and colleagues.16 We assessed the feasibility of
a strategy that included general anaesthesia; (iii) for tracheal
performing network meta-analysis for each outcome by
intubation; (iv) in combination with neuromuscular block; (v)
checking network connectivity, ensuring the availability of
restraining and controlling aggression or delirium; (vi) pro-
more trials than number of intervention nodes, and having at
cedures exceeding a duration longer than 1 h, as procedures of
least 10 trials for each outcome network. When appropriate to
this length are not frequently undertaken in the ED or ICU.
perform network meta-analysis, we calculated direct effect
We included the following outcomes of interest: sedation
estimates using the DerSimonian and Laird random-effects
recovery time (defined as time from procedure completion
model, for all comparisons with two RCTs or more.17
until return to baseline mental status, or as defined by study
We performed frequentist random-effects network meta-
authors), patient satisfaction (defined as patient perception of
analysis using multivariate meta-analysis assuming a com-
procedure success based on any scale used by study authors),
mon heterogeneity parameter.18,19 We assessed the transi-
and AEs related to PSA medications (as defined by study au-
tivity assumption by comparing the distribution of important
thors). Studies that examined non-synthetic opioids, such as
characteristics of trial populations, interventions, and co-
morphine, as part of PSA were analysed as separate notes in
interventions, and the methodological characteristics of the
the analysis. Synthetic opioids, such as alfentanil, remifenta-
studies across treatment comparisons. We identified issues of
nil, and fentanyl, are highly lipid-soluble with a far more rapid
incoherence by comparing direct evidence with indirect evi-
onset of action than morphine10; therefore, combining these
dence using the side-splitting method.20 We also confirmed
classes of opioids would introduce a high degree of clinical
the coherence assumption in the entire network using the
heterogeneity.
‘design-by-treatment’ model.21 At the request of reviewers, we
performed a post hoc sensitivity analysis without ICU studies.
Data extraction and risk of bias assessment After display of the rank probabilities using rankogram, we
used the surface under the cumulative ranking (SUCRA) to aid
Using a pre-designed data extraction form, two investigators in interpretation of relative effect of the interventions. All
extracted the following data: author names, study inclusion analyses were performed using the ‘network’ suite in Stata
and exclusion criteria, number of patients enrolled and rand- (version 17.0, StataCorp., College Station, TX, USA).22
omised, patient age and setting, procedure type and length
categorisation, and outcomes data. Pairs of investigators
independently collected all study data in duplicate and Assessment of certainty of evidence
assessed risk of bias (RoB) of the included studies using the We used the Grading of Recommendations, Assessment,
modified Cochrane RoB 2.0 tool.11 Although the published Development, and Evaluation (GRADE) approach to assess the
protocol describes using the Cochrane RoB 1.0 tool to assess certainty of evidence for each outcome.23 First, we assessed
individual study RoB, we used the modified Cochrane RoB 2.0 certainty of evidence in direct estimates using the traditional
tool. We believe this RoB tool is optimal as it eliminates the GRADE framework incorporating RoB, consistency, directness,
‘unclear’ category found in the original RoB tool and instead and publication bias. Next, we rated certainty of indirect es-
rates RoB across domains as either low, probably low, probably timates using the lowest certainty of the highest order loop
high, or high.12 The RoB examines the bias from the following while also considering issues related to intransitivity. We then
domains: randomisation process, deviations from the inten- rated the certainty in network estimates based on the higher
ded interventions, missing outcome data, measurement of the certainty between the direct and indirect estimates, while also
outcome, selection of the reported results. We resolved dis- considering issues of incoherence and imprecision at the
agreements in data extraction and RoB assessments through network estimate level.23 We used a minimally contextualised
discussion. approach to evaluate certainty in effect estimates24 using the
null as the threshold for all outcomes except sedation recovery
time. For sedation recovery time, we used 5 min as the
Data synthesis and analysis
threshold for clinically important effect. The GRADE approach
For continuous outcomes such as sedation recovery time, we was used by two authors with extensive experience (SS and
calculated the mean difference (MD) and corresponding 95% BR) to rate certainty of evidence. These ratings were discussed
confidence intervals (CIs); this includes the analysis of patient amongst the authorship group to ensure consensus. In keep-
satisfaction as a continuous outcome. Specifically, for patient ing with GRADE methods, ratings have been provided along
satisfaction as a continuous outcome, when studies used in- with transparent description and rationale to inform readers.
struments that measured the same construct, we used linear We used GRADE narrative statements to communicate the
transformation to convert measure to a 0e10 scale13 and used findings from the network meta-analysis (e.g. ‘probably’,
weighted MD for pooling study estimates. For dichotomous ‘may’, etc.).25
outcomes such as AEs, we calculated the relative risk (RR) and
the corresponding 95% CIs including when patient satisfaction
Subgroup analyses
was reported as a dichotomous outcome. We assessed statis-
tical heterogeneity between trials using visual inspection of We performed subgroup analysis using a network meta-
the forest plots, the I2 statistic and the c2 test. When only regression model for the following a priori defined subgroups:
median was reported for a continuous outcome, we converted (i) adults (aged 18 yr or older) vs paediatrics (under 18 yr of age);
this to a mean using the Hozo method.14 Moreover, we used (ii) short procedures (cardioversion, central line insertion,
the methods by Weir to calculate standard deviation when not incision and drainage, foreign body removal) vs long proced-
reported.15 No other imputation was performed and any ures (orthopaedic procedures, bronchoscopy, endoscopy,
 494 - Sharif et al.

tracheostomy, lumbar puncture, chest tube insertion); (iii) medications used in the included studies are provided in
patients admitted to the ICU vs those in the ED; (iv) high vs low Supplementary Appendix 3, (Supplementary Table S4). The
RoB studies; and (v) PSA with opioids vs without opioids. network maps for all the outcomes are in Supplementary
Appendix 5. The league tables and GRADE assessment of evi-
dence are also provided in the online supplementary
Results appendix.
Search results and study characteristics
We identified 15 341 citations (Fig. 1) in the search. Of these, 168
Sedation recovery time
underwent full-text review and we included 82 RCTs with a total
of 8105 patients. Characteristics of the included trials are in Compared with midazolam-opioid, sedation recovery time
Appendix 3, Supplementary material, Supplementary Table S1. was shorter with propofol (MD 16.3 min less, 95% CI 8.4e24.3
Seventy-eight studies were performed in the ED (n¼7822 min less; high certainty), and probably shorter with propofol-
patients)26e102 and four in the ICU (n¼283 patients).103e106 Nine- opioid (MD 13.6 min less, 95% CI 6.6e20.7 min less; moderate
teen were determined to be at overall high or probably high certainty), ketamine-propofol (MD 10.5 min less, 95% CI
RoB29,34,35,38,42,49,51e53,57,59,75,80,83,95,98,101,104,105 and 63 were found 3.4e17.6 min less; moderate certainty), etomidate-opioid (MD
to be at low or probably low RoB26e28,30e33,36,37,39e41,43e48,50,54e56,58 14.8 min less, 95% CI 3.5e26.0 min less; moderate certainty),
,60e74,76e79,81,82,84e94,96,97,99,100,102,103,106,107
(Supplementary Appen and opioids (MD 12.1 min less, 95% CI 25.4 min less to 1.3 min
dix 3, Supplementary Table S2). Fifty-two studies included adults more; moderate certainty) (Table 1; Supplementary Appendix
only (n¼4850 patients),26e28,30,32,33,36,38e40,47e50,52,54,57e60,65e75, 3, Supplementary Tables S5 and S13). Compared with
77e80,82e84,87,89e95,98,99,102e105,107
23 included paediatrics only midazolam-opioid, sedation recovery time may be longer with
(n¼2358 patients),34,35,42e46,51,53,55,61e64,76,85,86,88,96,97,101,103,106 and the use of ketamine-midazolam (MD 8.3 min more, 95% CI
seven included a mix of both populations (n¼897 1.1e15.5 min more; low certainty) (Table 1; Supplementary
patients).29,31,37,41,56,81,100 Appendix 3, Supplementary Tables S5 and S13).
The most common comparators were midazolam-opioid Compared with ketamine-propofol, recovery time may be
(n¼1188 patients), ketamine-propofol (n¼1497 patients), pro- shorter with propofol (MD 5.8 min less, 95% CI 12.01 min less to
pofol (n¼912 patients), and ketamine alone (n¼894 patients). 0.4 min more; low certainty) (Table 1; Supplementary
The opioids included were fentanyl, remifentanil, and alfen- Appendix 3, Supplementary Tables S5 and S14). Compared
tanil. The definitions of all AEs recorded from the 79 RCTs that with ketamine-propofol, there is probably no difference in
reported them are provided in Supplementary Appendix 3, sedation recovery time with the use of propofol-opioids (MD
(Supplementary Table S3). The dosing regimens of the PSA 3.1 min less, 95% CI 8.5 min less to 2.3 min more; moderate

Records identified through database searching

Identification

MEDLINE (n=3382)
EMBASE (n=6449)
Cochrane CCTR (n=5183)
PubMed (n=327)
Total (n=15 341)

Records after duplicates removed

Screening

(n=10 464)

Records screened (n=10 464) Records excluded (n=10 296)

Eligibility

Full-text articles Full-text articles excluded, with reasons (n=78)

assessed for eligibility (n=160) Wrong study design (n=7)
Wrong patient population (n=1)
Duplicate (n=16)
Studies included in qualitative
Wrong outcomes (n=2)
synthesis (n=82)
Trial registration (n=25)
Included

Wrong comparator (n=5)

Studies included in Wrong setting (n=5)
quantitative synthesis (meta-analysis) Wrong route of administration (n=11)
(n=82) Review article (n=6)

Fig 1. Study flowchart.

Table 1 Network estimates evaluating the efficacy of various procedural sedation and analgesia medication regimens for recovery time. CI, confidence interval; GRADE, Grading of
Recommendations, Assessment, Development, and Evaluation; MD, mean difference. *Imprecision only incorporated at network level, not at direct or indirect. yLowered for imprecision.
z
Lowered for inconsistency. ¶Lowered two levels for very serious imprecisions.

Comparison Direct estimate Indirect estimate Network estimate* GRADE Narrative summary
MD (95% CI) MD (95% CI) MD (95% CI)

Midazolam-opioids vs 21.7 (3.7e39.7) 14.7 (5.4e24.0) 16.3 (8.4e24.3) High Midazolam-opioids have a longer
propofol recovery time compared with propofol
y
Opioids vs midazolam- 5 (25.2 to 15.2 18.6 (37.3 to 0.2 12.1 (25.4 to 1.3 Moderate Opioids probably have a shorter recovery
opioids time compared with midazolam-
opioids
Etomidate-opioids vs 9.9 (27.0 to 7.1 18.5 (34.6 to 2.4 14.8 (26.0 to 3.5 Moderatey Etomidate-opioids probably have a
midazolam-opioids shorter recovery time compared with
midazolam-opioids
Midazolam-ketamine vs 2.7 (7.6 to 13.1 14.3 (3.9e24.8 8.3 (1.1e15.5 Lowy,z Midazolam-ketamine may have a longer
midazolam-opioids recovery time compared with
midazolam-opioids
Ketamine-propofol vs 6.9 (24.0 to 10.1) 11.3 (19.7 to 3.0) 10.5 (17.6 to 3.4) Moderatey Ketamine-propofol probably has a
midazolam-opioids shorter recovery time compared with
midazolam-opioids
Ketamine-propofol vs 8.4 (0.2 to 16.9) 2.1 (7.8 to 12.0) 5.8 (0.4 to 12.0) Lowz,¶ Ketamine-propofol may have a longer
propofol recovery time compared with propofol
Ketamine-propofol vs 0.5 (6.4 to 7.4) 11.0 (0.6e21.5) 3.1 (2.3 to 8.5) Moderatey Ketamine-propofol probably has no
propofol-opioids difference in recovery time compared
with propofol-opioids
Ketamine-propofol vs 3.6 (13.7 to 6.5) 5.0 (15.3 to 5.2) 3.6 (9.8 to 2.7) Lowy,z Ketamine-propofol may have no
ketamine difference in recovery time compared
with ketamine

Procedural Sedation in the ED and ICU

Ketamine vs propofol 10.1 (7.4 to 27.6) 10.0 (1.3e18.7) 9.4 (2.2e16.5) Moderatey Ketamine probably has a longer recovery
time compared with propofol
Ketamine vs propofol- 1.2 (15.5 to 17.9) 8.0 (0.1 to 16.0) 6.8 (0.5 to 13.8) Moderatey Ketamine probably has a longer recovery
opioids time compared with propofol-opioids
Ketamine vs etomidate- 4.9 (13.0 to 22.8) 10.9 (4.9 to 26.7) 7.8 (3.5 to 19.1) Moderatey Ketamine probably has a longer recovery
opioids time compared with etomidate-
opioids
Ketamine vs ketamine- 8.1 (18.4 to 2.1) 21.6 (32.1 to 11.1) 15.2 (22.4 to 8.1) High Ketamine has a shorter recovery time
midazolam compared with ketamine-midazolam
Etomidate vs ketamine 6.6 (6.0 to 19.2) 9.6 (23.7 to 4.6) 0.2 (9.6 to 9.1) Lowy Etomidate may have no difference in
recovery time compared with
ketamine

-
495
 496 - Sharif et al.

certainty) and may be no difference with the use of ketamine Compared with ketamine-propofol, patient satisfaction as
(MD 3.6 min more, 95% CI 2.7 min less to 9.8 min more; a dichotomous outcome was probably worse with the use of
low certainty) (Table 1; Supplementary Appendix 3, ketamine (RR 0.89, 95% CI 0.79e1.02; moderate certainty), and
Supplementary Tables S5 and S14). propofol-opioids (RR 0.93, 95% CI 0.83e1.05; moderate cer-
Compared with ketamine, recovery time is probably tainty), and may be worse with propofol (RR 0.94, 95% CI
shorter with propofol (MD 9.4 min less, 95% CI 2.2e16.5 min 0.82e1.07, low certainty) (Supplementary Appendix 3,
less; moderate certainty), propofol-opioids (MD 6.7 min less, Supplementary Tables S8, S14 and S17). Compared with ke-
95% CI 13.8 min less to 0.5 min more; moderate certainty), and tamine, there was probably no difference in patient satisfac-
etomidate-opioids (MD 7.8 min less, 95% CI 19.1 min less to 3.5 tion as a dichotomous outcome with propofol (RR 1.05, 95% CI
min more; moderate certainty) (Table 1; Supplementary 0.92e1.20; moderate certainty), midazolam-ketamine (RR 1.07,
Appendix 3, Supplementary Tables S5 and S15). Compared 95% CI 0.94e1.23; moderate certainty), and propofol-opioids
with ketamine, there was a longer recovery time with the use (RR 1.04, 95% CI 0.91e1.19; moderate certainty)
of midazolam-ketamine (MD 15.2 min more, 95% CI 8.1e22.4 (Supplementary Appendix 3, Supplementary Tables S8, S15
min more; high certainty) and may be no difference with and S17).
etomidate (MD 0.2 min less, 95% CI 9.6 min less to 9.1 min
more; low certainty) (Table 1; Supplementary Appendix 3,
Respiratory adverse events
Supplementary Tables S5 and S15).
Respiratory AEs were defined variably by the included studies
and included the following: apnoea, laryngospasm, bag-valve
Patient satisfaction
mask ventilation, oxygen desaturation, intubation, aspira-
Patient satisfaction was reported as a continuous outcome in 22 tion, hypoxia (as defined by the authors) amongst others
studies (involving 2126 patients) and measured as number of (Supplementary Appendix 3, Supplementary Table S3). The
patients satisfied with sedation/analgesia in 24 studies network diagram for this outcome is available in Figure 2.
(involving 2711 patients). With respect to the continuous scales, Compared to midazolam-opioids, there were fewer respiratory
a wide variety were used, including but not limited to scales AEs with the use of ketamine (RR 0.55, 95% CI 0.32e0.96; high
ranging from 1 to 5, 0 to 100, and 1 to 10 (Supplementary certainty), ketamine-midazolam (RR 0.57, 95% CI 0.37e0.86;
Appendix 3, Supplementary Table S6). Compared with high certainty), ketamine-propofol (RR 0.52, 95% CI 0.31e0.87;
midazolam-opioids, patient satisfaction was higher using high certainty), and may be fewer with the use of propofol (RR
ketamine-propofol (MD 1.5 points higher, 95% CI 0.3e2.6 points 0.71, 95% CI 0.43e1.16; low certainty) (Table 2, Fig. 3;
higher, high certainty), and may have been higher with dex- Supplementary Appendix 3, Supplementary Tables S7 and
medetomidine (MD 1.0 points higher, 95% CI 0.4 points lower to S11). Compared with midazolam-opioids, there may be no
2.4 points higher; low certainty) and propofol-opioids (MD 1.0 effect on respiratory AEs with the use propofol-opioids (RR
points higher, 95% CI 0.2 points lower to 2.2 points higher; low 1.05, 95% CI 0.61e1.81; low certainty), etomidate-opioids (RR
certainty) (Supplementary Appendix 3, Supplementary Tables 0.85, 95% CI 0.42e1.74; low certainty), midazolam (RR 0.49, 95%
S7, S13 and S16). Compared with midazolam-opioids, etomi- CI 0.14e1.67; low certainty), or dexmedetomidine-opioids (RR
date-opioids may have no impact on patient satisfaction (MD 0.84, 95% CI 0.15e4.83; low certainty) (Table 2, Fig 3;
0.01 points higher, 95% CI 1.2 points lower to 1.2 points higher; Supplementary Appendix 3, Supplementary Tables S9 and
low certainty) (Supplementary Appendix 3, Supplementary S13). Compared with midazolam-opioids, there may be more
Tables S7, S13 and S16) while opioids may result in decreased respiratory AEs with the use of opioids (RR 1.22, 95% CI
patient satisfaction (MD 0.7 points lower, 95% 2.2 points lower to 0.57e2.60; low certainty) (Table 2, Fig 3; Supplementary
0.8 points higher; low certainty) (Supplementary Appendix 3, Appendix 3, Supplementary Tables S9 and S13).
Supplementary Tables S7, S13 and S16). Compared with ketamine-propofol, there were more respi-
Compared with ketamine-propofol, patient satisfaction ratory AEs with the use of propofol-opioids (RR 2.03, 95% CI
may be lower with the use of propofol-opioids (MD 0.5 points 1.32e3.13; high certainty) and probably more with propofol (RR
lower, 95% CI 1.7 points lower to 0.7 points higher; low cer- 1.37, 95% CI 0.98e1.91; moderate certainty) (Table 2;
tainty), and may have no impact on satisfaction with the use of Supplementary Appendix 3, Supplementary Tables S9 and S14).
ketamine (MD 0.03 points higher, 95% CI 1.5 points lower to 1.6 Compared with ketamine-propofol, there was probably no dif-
points higher; low certainty) or propofol (MD 0.01 points lower, ference in respiratory AEs with the use of ketamine (RR 1.07, 95%
95% CI 1.1 points lower to 1.1 points higher; low certainty) CI 0.76e1.49; moderate certainty) (Table 2; Supplementary
(Supplementary Appendix 3, Supplementary Tables S7, S14 Appendix 3, Supplementary Tables S9 and S14).
and S16). Compared with ketamine, patient satisfaction was Compared with ketamine, there were more respiratory AEs
probably lower with the use of etomidate-opioids (MD 1.5 with the use of propofol-opioids (RR 1.90, 95% CI 1.15e3.15;
points lower, 95% CI 3.6 points lower to 0.6 points higher; high certainty), and may be more with etomidate (RR 1.43, 95%
moderate certainty) (Supplementary Appendix 3, CI 0.73e2.79; low certainty) or propofol (RR 1.29, 95% CI
Supplementary Tables S7, S15 and S16). Compared with 0.85e1.95; low certainty) (Table 2; Supplementary Appendix 3,
midazolam-opioids, there was probably no difference in pa- Supplementary Tables S9 and S15). Compared with ketamine,
tient satisfaction as a dichotomous outcome with the use of there may be no difference in respiratory AEs with the use of
opioids (RR 1.01, 95% CI 0.86e1.19; moderate certainty) or midazolam-ketamine (RR 1.03, 95% CI 0.63e1.67; low cer-
ketamine-midazolam (RR 1.01, 95% CI 0.90e1.14; moderate tainty), and dexmedetomidine-ketamine (RR 0.91, 95% CI
certainty) (Supplementary Appendix 3, Supplementary Tables 0.46e1.80; low certainty). There was an uncertain effect on
S8, S13 and S17). respiratory AEs with midazolam (RR 0.89, 95% CI 0.26e2.98;
 Procedural Sedation in the ED and ICU - 497

Ketamine-propofol-opioid
Ketamine-propofol
KPMFL combination
Ketamine-haloperidol Diazepam-opioid
Etomidate
Ketamine
Diazepam
Midazolam-lidocaine-opioid

Dexmedetomidine
Midazolam-flumazenil

Dexmedetomidine-ketamine
Midazolam-morphine

Propofol

Midazolam

Thiopental-opioid
Promethazine-chlorpromazine-opioid
Methohexital-opioid

Methohexital
Pentobarbital
Ketamine-midazolam
Propofol-morphine
Dexmedetomidine-opioid
Etomidate-opioid Propofol-opioid
Opioid
Midazolam-opioid

Fig 2. Network map for respiratory adverse events for x-node analysis. The size of the node corresponds to the number of patients
randomised to that intervention. The thickness of the line and the associated numbers correspond to the number of studies comparing the
two linked interventions. KPMFL, ketamine-propofol-midazolam-flumazenil-lidocaine.

very low certainty) (Table 2; Supplementary Appendix 3, with ketamine (RR 2.56, 95% CI 0.72e9.08; moderate certainty)
Supplementary Tables S9 and S15). (Table 3; Supplementary Appendix 3, Supplementary Tables
S10 and S14). Compared with ketamine, there was an uncer-
tain effect on cardiac AEs with the use of propofol (RR 1.95, 95%
Cardiac adverse events CI 0.44e8.67; very low certainty), propofol-opioids (RR 1.48,
95% CI 0.39e5.48; very low certainty), midazolam-ketamine
Cardiac AEs were defined differently amongst the included trials
(RR 0.82, 95% CI 0.14e4.81; very low certainty) or
but most of them included hypotension and bradycardia,
dexmedetomidine-ketamine (RR 0.92, 95% CI 0.16e5.48; very
whereas others also included dysrhythmias and the use of an
low certainty) (Table 3; Supplementary Appendix 3,
inotrope or vasoactive agent (Supplementary Appendix 3,
Supplementary Tables S10 and S15).
Supplementary Table S3). Compared with midazolam-opioids,
there may be fewer cardiac AEs with the use of ketamine-
propofol (RR 0.38, 95% CI 0.10e1.44; low certainty) and an un-
Gastrointestinal adverse events
certain effect on cardiac AEs with the use of ketamine-
midazolam (RR 0.83, 95% CI 0.25e2.81; very low certainty) Almost all the included studies defined gastrointestinal (GI) AEs
(Table 3; Supplementary Appendix 3, Supplementary Tables S10 as nausea, vomiting, or both (Supplementary Appendix 3,
and S13). Compared with midazolam-opioids, there was an Supplementary Table S3). Compared with midazolam-opioids,
uncertain effect on cardiac AEs with the use of opioids (RR 2.67, there were more GI AEs with ketamine-midazolam (RR 3.08,
95% CI 0.22e32.19; very low certainty), propofol (RR 1.89, 95% CI 95% CI 1.15e8.27; high certainty), and there may be more with
0.44e8.04; very low certainty), propofol-opioids (RR 1.44, 95% CI ketamine-propofol (RR 1.97, 95% CI 0.58e6.66; low certainty)
0.39e5.30; very low certainty) and dexmedetomidine-opioids (Supplementary Appendix 3, Supplementary Tables S11, S13,
(RR 4.02, 95% CI 0.37e43.87; very low certainty) (Table 3; and S18). Compared with midazolam-opioids, there were prob-
Supplementary Appendix 3, Supplementary Tables S10 and S13). ably fewer GI AEs with the use of dexmedetomidine-opioids (RR
Compared with ketamine-propofol, there were more car- 0.07, 95% CI 0.00e0.97; moderate certainty) and an uncertain
diac AEs with the use of propofol-opioids (RR 3.80, 95% CI effect with the use of opioids (RR 0.32, 95% CI 0.04e2.30; very low
2.02e7.16; high certainty) and propofol (RR 4.99, 95% CI certainty), etomidate-opioids (RR 1.35, 95% CI 0.44e4.15; very low
1.91e13.02; high certainty), and probably more cardiac AEs certainty) and propofol (RR 1.99, 95% CI 0.30e13.21; very low
Table 2 Network estimates evaluating the efficacy of various procedural sedation and analgesia medication regimens for respiratory adverse events. CI, confidence interval; GRADE,

498
Grading of Recommendations, Assessment, Development, and Evaluation; RR, relative risk. *Imprecision only incorporated at network level, not at direct or indirect. yLowered for
imprecision. zLowered two levels for very serious imprecisions. ¶Lowered thee levels for very serious imprecisions. xLowered for incoherence.

-
Sharif et al.
Comparison Direct estimate RR Indirect estimate Network estimate* GRADE Narrative summary
(95% CI) RR (95% CI) RR (95% CI)

Ketamine-propofol vs 0.22 (0.04e1.15) 0.56 (0.32e0.99) 0.52 (0.31e0.87) High Ketamine-propofol has fewer respiratory adverse
midazolam-opioids events compared with midazolam-opioids
Ketamine vs midazolam- 0.08 (0.01e0.7) 0.54 (0.3e0.97) 0.55 (0.32e0.96) High Ketamine has fewer respiratory adverse events
opioids compared with midazolam-opioids
Midazolam-ketamine vs 0.55 (0.32e0.93) 0.53 (0.23e1.26) 0.57 (0.37e0.86) High Midazolam-ketamine has fewer respiratory adverse
midazolam-opioids events compared with midazolam-opioids
Opioid-dexmedetomidine 1 (0.02e52.41) 0.84 (0.12e6.14) 0.84 (0.15e4.83) Lowz Opioid-dexmedetomidine may have no effect on
vs midazolam-opioids respiratory adverse events compared with
midazolam-opioids
Opioids vs midazolam- 0.88 (0.22e3.61) 1.5 (0.58e3.91) 1.22 (0.57e2.60) Lowz Opioids may have more respiratory adverse events
opioids compared with midazolam-opioids
Midazolam-opioids vs 0.69 (0.22e2.13) 1.71 (0.95e3.06) 1.41 (0.86e2.32) Lowz Midazolam-opioids may have more respiratory
propofol adverse events compared with propofol
Midazolam-opioids vs 0.81 (0.22e2.9) 0.85 (0.44e1.63) 0.95 (0.55e1.64) Lowz Midazolam-opioids may have no effect on respiratory
propofol-opioids adverse events compared with propofol-opioids
Midazolam vs midazolam- 0.09 (0.01e0.89) 0.7 (0.19e2.59) 0.49 (0.14e1.67) Lowz Midazolam may have no effect on respiratory adverse
opioids events compared midazolam-opioids
Ketamine-propofol vs 0.83 (0.53e1.29) 0.54 (0.29e0.99) 0.73 (0.52e1.02) Moderatey Ketamine-propofol probably has fewer respiratory
propofol adverse events compared with propofol
Ketamine-propofol vs 1.03 (0.6e1.77) 1.18 (0.59e2.34) 0.94 (0.67e1.31) Moderatey Ketamine-propofol probably has no difference in
ketamine respiratory adverse events compared with ketamine
Ketamine-propofol vs 0.32 (0.16e0.65) 0.53 (0.29e0.98) 0.49 (0.32e0.76) High Ketamine-propofol has fewer respiratory adverse
propofol-opioids events compared with propofol-opioids
Dexmedetomidine- 0.71 (0.15e3.39) 1.06 (0.47e2.43) 0.97 (0.49e1.93) Moderatey Dexmedetomidine-ketamine probably has no effect
ketamine vs ketamine- on respiratory adverse events when compared with
propofol ketamine-;propofol
Etomidate vs ketamine 4.84 (1.8e12.99) 0.84 (0.45e1.58) 1.43 (0.73e2.79) Lowy,x Etomidate may have more respiratory adverse events
compared with ketamine
Ketamine vs propofol 1.6 (0.89e2.87) 0.53 (0.33e0.85) 0.78 (0.51e1.18) Lowy,x Ketamine may have fewer respiratory adverse events
compared with propofol
Ketamine vs midazolam 0.96 (0.02e49.82) 0.97 (0.26e3.58) 1.13 (0.34e3.79) Very Low¶ Ketamine has uncertain effect on respiratory adverse
events compared with midazolam
Ketamine vs midazolam- 0.72 (0.37e1.42) 1.1 (0.5e2.44) 0.97 (0.60e1.58) Lowz Ketamine may have no difference in respiratory
ketamine adverse events compared with midazolam-
ketamine
Ketamine vs midazolam- 0.08 (0.01e0.7) 0.54 (0.3e0.97) 0.55 (0.32e0.96) High Ketamine has fewer respiratory adverse events
opioids compared with midazolam-opioids
Dexmedetomidine- 1.28 (0.25e6.42) 085 (0.38e1.88) 0.91 (0.46e1.80) Lowz Dexmedetomidine-ketamine may have no effect on
ketamine vs ketamine respiratory adverse events when compared with
ketamine
Ketamine vs propofol- 0.43 (0.25e0.75) 2.37 (0.60e9.40) 0.53 (0.32e0.87) High Ketamine has fewer respiratory adverse events
opioids compared with propofol-opioid
 Procedural Sedation in the ED and ICU - 499

a Respiratory adverse events b Cardiac adverse events

Relative risk (95% confidence interval)

Relative risk (95% confidence interval)

6 20
18
5
16
4 14
12
3 10
8
2
6
Relative risk 1.00 4
1
2
Relative risk 1.00
0 0
PFOL OPPF KTFL KTMN MDZM OPET OPID OPDX MZKT PFOL OPPF KTFL OPID OPDX MZKT

c Gastrointestinal adverse events

d Neurological adverse events
Relative risk (95% confidence interval)

Relative risk (95% confidence interval)

20 20
18 18
16 16
14 14
12 12
10 10
8 8
6 6
4 4
2 2
Relative risk 1.00 Relative risk 1.00
0 0
PFOL KTFL OPET OPID OPDX MZKT KTFL OPET OPID MZKT

e
Better than No difference with Worse than
midazolam + opioid midazolam + opioid midazolam + opioid

High or moderate Better than No more effective Less effective than

certainty evidence midazolam + opioid than midazolam + midazolam + opioid
opioid

Low or very low May be better than May be no more May be less effective
certainty evidence midazolam + opioid effective than than midazolam +
midazolam + opioid opioid

Fig 3. Network meta-analysis results based on Grading of Recommendations, Assessment, Development, and Evaluation (GRADE)
certainty of evidence and treatment effectiveness for the comparisons of active treatments vs midazolam-opioid for the outcome of
adverse events. (a) respiratory adverse events; (b) cardiac adverse events; (c) gastrointestinal adverse events; (d) neurological adverse
events; (E) GRADE certainty of evidence table and figure legend. KTFL, ketamine-propofol; KTMN, ketamine; MDZM, midazolam; MZKT,
midazolam-ketamine; OPDX, opioid-dexmedetomidine; OPET, opioid-etomidate; OPID, opioid; OPPF, opioid-propofol; PFOL, propofol.

certainty) (Supplementary Appendix 3, Supplementary Tables Neurological adverse events

S11, S13 and S18).
There was a lot of variation in how the included studies
Compared with ketamine-propofol, there were probably
defined neurological AEs; briefly, the included recovery agita-
more GI AEs with ketamine (RR 2.08, 95% CI 1.05e4.11; mod-
tion, fasciculations, hallucinations, myoclonus, and vertigo
erate certainty) and may be fewer with propofol-opioids (RR
(Supplementary Appendix 3, Supplementary Table S3).
0.66, 95% CI 0.32e1.37; low certainty) (Supplementary
Compared with midazolam-opioids, there were more neuro-
Appendix 3, Supplementary Tables S11, S14 and S18).
logical AEs with the use of ketamine-propofol (RR 3.68, 95% CI
Compared with ketamine-propofol, propofol has an uncertain
1.08e12.53; high certainty), etomidate-opioids (RR 5.88, 95% CI
effect on GI AEs (RR 1.01, 95% CI 0.17e5.86; very low certainty)
1.96e17.62; high certainty), and ketamine-midazolam (RR 5.97,
(Supplementary Appendix 3, Supplementary Tables S11, S14
95% CI 2.15e16.62; high certainty) (Supplementary Appendix 3,
and S18). Compared with ketamine, there were fewer GI AEs
Supplementary Tables S12, S13 and S18). Compared with
with the use of propofol-opioids (RR 0.32, 95% CI 0.13e0.74;
midazolam-opioids, there was an uncertain effect on neuro-
high certainty) (Supplementary Appendix 3, Supplementary
logical AEs with the use of opioids (RR 0.34, 95% CI 0.07e1.72;
Tables S11, S15 and S18). Compared with ketamine, there
very low certainty) (Supplementary Appendix 3,
may be no effect on GI AEs with the use of midazolam-
Supplementary Tables S12, S13 and S19).
ketamine (RR 0.75, 95% CI 0.35e1.59; low certainty) and an
Compared with ketamine-propofol, there were more
uncertain effect with propofol (RR 0.49, 95% CI 0.08e2.85; very
neurological AEs with ketamine (RR 2.38, 95% CI 1.33e4.23;
low certainty) (Supplementary Appendix 3, Supplementary
high certainty) (Supplementary Appendix 3, Supplementary
Tables S11, S15 and S18).
 500
-
Sharif et al.
Table 3 Network estimates evaluating the efficacy of various procedural sedation and analgesia medication regimens for cardiac adverse events. AE, adverse events; CI, confidence
interval; GRADE, Grading of Recommendations, Assessment, Development, and Evaluation; RR, relative risk. *Imprecision only incorporated at network level, not at direct or indirect.
y
Lowered for imprecision. zLowered two levels for very serious imprecisions. ¶Lowered thee levels for very serious imprecisions.

Comparison Direct estimate Indirect estimate Network estimate* GRADE Narrative summary
RR (95% CI) RR (95% CI) RR (95% CI)

Ketamine-propofol vs 0.333 (0.014e7.88) 0.387 (0.088e1.694) 0.38 (0.10e1.43) Lowz Ketamine-propofol may have fewer cardiac
midazolam-opioids AEs compared with midazolam-opioids
Midazolam-ketamine vs 0.967 (0.236e3.964) 0.552 (0.051e5.909) 0.80 (0.24e2.69) Very low¶ Midazolam-ketamine may have no effect on
midazolam-opioids cardiac AEs compared with midazolam-
opioids
Opioid-dexmedetomidine 7.001 (0.373e131.382) 1.334 (0.021e83.063) 4.02 (0.37e43.87) Very low¶ Opioid-dexmedetomidine has an uncertain
vs midazolam-opioids effect on cardiac AEs compared with
midazolam-opioids
Midazolam-opioids vs 3.77 (0.158e90.033) 0.305 (0.06e1.555) 0.53 (0.12e2.25) Very low¶ Midazolam-opioids has an uncertain effect
propofol on cardiac AEs compared with propofol
Midazolam-opioids vs 0.441 (0.06e3.249) 1.142 (0.202e6.473) 0.69 (0.19e2.55) Very low¶ Midazolam-opioids has an uncertain effect
propofol-opioids on cardiac AEs compared with propofol-
opioids
Ketamine-propofol vs 0.155 (0.05e0.483) 0.345 (0.056e2.117) 0.20 (0.08e0.52) High Ketamine-propofol has fewer cardiac AEs
propofol compared with propofol
Ketamine-propofol vs 0.167 (0.02e1.412) 0.305 (0.016e5.766) 0.39 (0.11e1.38) Moderatey Ketamine-propofol probably has fewer
ketamine cardiac AEs compared with ketamine
Ketamine-propofol vs 0.32 (0.159e0.643) 0.119 (0.017e0.83) 0.26 (0.14e0.50) High Ketamine-propofol has fewer cardiac AEs
propofol-opioids compared with propofol-opioids
Dexmedetomidine- 2.00 (0.19e20.93) 4.03 (0.04e425.10) 2.37 (0.34e16.34) Very low¶ Dexmedetomidine-ketamine has an
ketamine vs ketamine- uncertain effect on cardiac AEs when
propofol compared with ketamine-propofol
Ketamine vs propofol 1.063 (0.022e52.527) 0.912 (0.111e7.486) 0.51 (0.12e2.29) Very low¶ Ketamine has an uncertain effect on cardiac
AEs when compared with propofol
Ketamine vs midazolam- 0.969 (0.02e48.05) 2.986 (0.272e32.82) 1.21 (0.21e7.07) Very low¶ Ketamine has an uncertain effect on cardiac
ketamine AEs when compared with midazolam-
ketamine
Ketamine vs propofol- 0.16 (0.01e2.46) 1.00 (0.23e4.46) 0.67 (0.18e2.59) Very low¶ Ketamine has an uncertain effect on cardiac
opioids AEs when compared with propofol-
opioids
 Procedural Sedation in the ED and ICU - 501

Tables S12, S14 and S19) and may be no difference in neuro- may wish to avoid propofol and propofol-opioids, as they were
logical AEs with the use of propofol-opioids (RR 1.00, 95% CI associated with the most cardiac AEs. A plausible alternative
0.35e2.80; low certainty) or propofol (RR 0.79, 95% CI 0.38e1.63; in these circumstances would be using either midazolam-
low certainty) (Supplementary Appendix 3, Supplementary opioid or ketamine-propofol as they were associated with
Tables S12, S14 and S19). Compared with ketamine, there the fewest cardiac AEs. Although both opioids and benzodi-
were fewer neurological AEs with the use of propofol (RR 0.33, azepines can cause hypotension, the use of a combination has
95% CI 0.15e0.71; high certainty), and probably fewer with been shown to require lower doses of each individual drug,
propofol-opioids (RR 0.42, 95% CI 0.15e1.15; moderate cer- perhaps abrogating negative sequelae.108
tainty) and dexmedetomidine-ketamine (RR 0.37, 95% CI In clinical circumstances where patients with an altered
0.12e1.17; moderate certainty). Compared with ketamine, mental status need PSA, healthcare providers may wish to
there may be no difference in neurological AEs with the use of avoid ketamine and etomidate given they were associated
midazolam-ketamine (RR 0.68, 95% CI 0.32e1.45; low certainty) with the most neurological AEs. This is likely a result of the
(Supplementary Appendix 3, Supplementary Tables S12, S15 post-emergence phenomenon that is associated with keta-
and S19). mine use; it is characterised by euphoria, vivid dreams, illu-
sions, and hallucinations.109 However, etomidate is associated
Additional analyses with myoclonic jerks which can explain the increase in noted
neurological AEs.110
We explored the impact of age (adults vs paediatrics), duration
The time it takes for a patient to recover from PSA is
of procedure (long vs short), and RoB on network estimates
important from a resource utilisation perspective, as these
using network meta-regression but found no evidence of
patients must be monitored closely until they fully recover.
important subgroup effect in relative effects across outcomes
This time includes monitoring by the registered nurse, the
of interest (Supplementary Appendix 4, Supplementary Tables
respiratory therapist, and the most responsible physician.
S20eS56). We did not have sufficient studies or granularity in
This is particularly noteworthy when sedating for short pro-
data to perform subgroup analysis for the comparison of ICU
cedures such as electrical cardioversions. In these instances,
vs ED admission. Our post hoc sensitivity analysis without ICU
healthcare providers may wish to avoid using ketamine and
studies did not show a difference in conclusions for any of the
combination midazolam-ketamine, as they were associated
outcomes of interest (Supplementary Appendix). Ranking
with longest recovery time. Conversely, opioids, propofol,
probabilities and SUCRA values are provided in
propofol-opioids, and opioid-etomidate were associated with
Supplementary Appendix 3.
the shortest recovery time. From a satisfaction perspective,
patients prefer ketamine-propofol followed by propofol-
Discussion opioids. Opioids, propofol, and ketamine alone were associ-
This systematic review and network meta-analysis highlights ated with the lowest patient satisfaction. Although the abso-
the strengths and weaknesses of various PSA medications and lute differences in patient satisfaction were small, there is a
combinations. Specifically, this analysis demonstrates that consistent signal that combination drugs may be associated
compared with midazolam-opioids for PSA in the ED and ICU, with higher patient satisfaction, perhaps by optimising benefit
ketamine has fewer respiratory AEs. Furthermore, compared while minimising potential adverse effects associated with
with ketamine-propofol, propofol-opioids have more respira- higher doses.
tory and cardiac AEs, and may have fewer GI AEs. However, We did not identify any relative subgroup effect when
recovery time is shorter with propofol, and patient satisfaction comparing children vs adults (Supplementary Appendix 4). Of
is greater with ketamine-propofol. Moreover, compared with the 23 studies that focused on a paediatric population alone, 21
ketamine, propofol-opioids have fewer GI AEs and probably examined ketamine alone or in combination with another
fewer neurological AEs but have more respiratory AEs. drug. Ketamine has a good safety profile,111 and with many of
Patient and procedure characteristics often dictate the choice the studies in children including it as one of their arms, it may
of PSA medications used by healthcare providers. Based on this partly explain why no differences were found between the
analysis, ketamine and combination ketamine-propofol may be adult and paediatric populations. We could not perform a
the best choice for patients who have a tenuous airway status subgroup analysis comparing studies done in the ICU vs those
(i.e. those with lung pathology). In contrast, healthcare providers conducted in the ED because of a lack of data. There were four
may want to avoid propofol, propofol-opioids, and opioid- studies that examined PSA in critically ill patients with two in
midazolam in these patients given their association with more a paediatric population. One of the paediatric studies exam-
respiratory AEs. Healthcare providers providing PSA for patients ined sedation for the insertion of central venous catheters103
undergoing procedures such as emergent endoscopies may whereas the other examined sedation for procedures such as
want to use combination midazolam-opioids, as the analysis a lumbar puncture and bone marrow aspiration.106 Of the
found that this regimen had the fewest GI AEs. Conversely, ke- adult ICU studies, one examined the sedation of burn patients
tamine should perhaps be avoided in this clinical circumstance for the purpose of dressing changes104 and the other assessed
given it is associated with the most GI AEs. In circumstances sedating post-coronary artery bypass graft patients for
where healthcare providers want the benefit from ketamine’s synchronised cardioversion for atrial fibrillation.105
respiratory protective features, but want to avoid its GI AEs, us- Strengths of this review include a pre-registered protocol, a
ing propofol in combination with ketamine may be advisable as comprehensive literature search including unpublished sour-
this results in fewer GI AEs. ces, duplicate and independent screening and data abstrac-
Critically ill patients made up a smaller number of patients tion, network meta-analysis allowing for inclusion of both
included in this analysis. Amongst their complex clinical fac- direct and indirect evidence, and GRADE assessment of cer-
tors, many of them are often hypotensive as a result of shock tainty of evidence. These findings represent the most current,
of various aetiologies. In these instances, healthcare providers comprehensive summary of evidence to guide clinical practice
 502 - Sharif et al.

for PSA. Moreover, inclusion of studies in children allows for a References

more robust and generalisable understanding of the various
PSA medications used. 1. Foo TY, Mohd Noor N, Yazid MB, Fauzi MH, Abdull
Wahab SF, Ahmad MZ. Ketamine-propofol (Ketofol) for
procedural sedation and analgesia in children: a sys-
Limitations tematic review and meta-analysis. BMC Emerg Med 2020;
20: 81
First, there were only four ICU studies included and therefore
2. Sneyers B, Laterre P-F, Perreault MM, Wouters D,
conclusions regarding critically ill patients are less certain.
Spinewine A. Current practices and barriers impairing
Second, because many of the findings had low or very low
physicians’ and nurses’ adherence to analgo-sedation
certainty of evidence because of imprecision and wide CIs,
recommendations in the intensive care unit–a national
further RCTs are needed to improve certainty of findings.
survey. Crit Care 2014; 18: 655
Specifically, PSA regimens that are represented by smaller
3. Jagoda AS, Campbell M, Karas S Jr, et al. Clinical policy
nodes in the analysis (i.e. etomidate-opioids, dexmedetomi-
for procedural sedation and analgesia in the emergency
dine-opioids, dexmedetomidine alone) would benefit from
department. Ann Emerg Med 1998; 31: 663e77
more RCT data. Third, many of the included studies used
4. Raffay V, Fi
ser Z, Samara E, et al. Challenges in proce-
different definitions for AEs which introduced some hetero-
dural sedation and analgesia in the emergency depart-
geneity into the findings. These limitations were considered
ment. J Emerg Crit Care Med 2020; 4: 1e13
when using the GRADE approach assessing the certainty of
5. Bellolio MF, Gilani WI, Barrionuevo P, et al. Incidence
evidence. Fourth, given the clinical heterogeneity between
of adverse events in adults undergoing procedural
studies, indirect comparisons may have a degree of intransi-
sedation in the emergency department: a systematic
tivity, although we did not lower for this GRADE domain.
review and meta-analysis. Acad Emerg Med 2016; 23:
119e34
Conclusions 6. De Vries LJ, Veeger N, Van Roon EN, Lameijer H. Low-
dose ketamine or opioids combined with propofol for
Overall, these data illustrate that there is no perfect pharma-
procedural sedation in the emergency department: a
cological agent for procedural sedation and analgesia.
systematic review. Eur J Emerg Med 2023; 30: 244e51
Compared with midazolam-opioids for procedural sedation
7. Schünemann HJ, Mustafa RA, Brozek J, et al. GRADE
and analgesia in the acute care setting, ketamine was associ-
guidelines: 22. The GRADE approach for tests and
ated with fewer respiratory adverse events, sedation recovery
strategies-from test accuracy to patient-important out-
time is shortest with propofol, and patient satisfaction is
comes and recommendations. J Clin Epidemiol 2019; 111:
highest using a combination of ketamine-propofol. Compared
69e82
with ketamine-propofol, propofol-opioids may be associated
8. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020
with higher rates of respiratory and cardiac adverse events,
statement: an updated guideline for reporting systematic
and probably fewer gastrointestinal adverse events. As such,
reviews. BMJ 2021; 372: n71
our data highlights the importance of an individualised
9. Hutton B, Salanti G, Caldwell DM, et al. The PRISMA
approach based upon patient and procedure characteristics.
extension statement for reporting of systematic reviews
incorporating network meta-analyses of health care in-
terventions: checklist and explanations. Ann Intern Med
Authors’ contributions
2015; 162: 777e84
Designed the study: SS, BR 10. Pathan H, Williams J. Basic opioid pharmacology: an
Collected the data: JK, FR, BF, MH, AG, SS update. Br J Pain 2012; 6: 11e6
Analysed and interpreted the data: SS, BR, BS 11. Sterne JAC, Savovic  J, Page MJ, et al. RoB 2: a revised tool
Contributed to the writing of the manuscript: SS, BR, BS, SMF, for assessing risk of bias in randomised trials. BMJ 2019;
LM, ME, RS, MD, MB, LB, JP, AP, GM, PP, MW 366: l4898
12. Guyatt GBJ. Modification of Cochrane tool to assess risk of bias
in randomized trials 2021. Available from: https://www.
Acknowledgements evidencepartners.com/resources/methodological-
We thank Rachel Couban, medical librarian and information resources/. [Accessed 15 September 2023]
specialist, Faculty of Health Sciences, McMaster University, 13. Thorlund K, Walter SD, Johnston BC, Furukawa TA,
Hamilton, for her assistance in performing the comprehensive Guyatt GH. Pooling health-related quality of life out-
search of the databases. We acknowledge John Reynolds of the comes in meta-analysis-a tutorial and review of
University of Miami for peer-review of the search strategy. methods for enhancing interpretability. Res Synth
Methods 2011; 2: 188e203
14. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and
Declaration of interest variance from the median, range, and the size of a
sample. BMC Med Res Methodol 2005; 5: 13
SS holds a McMaster University Department of Medicine In- 15. Weir CJ, Butcher I, Assi V, et al. Dealing with missing
ternal Career Research Award. standard deviation and mean values in meta-analysis of
continuous outcomes: a systematic review. BMC Med Res
Methodol 2018; 18: 25
Appendix A. Supplementary data 16. Sweeting MJ, Sutton AJ, Lambert PC. What to add to
Supplementary data to this article can be found online at nothing? Use and avoidance of continuity corrections in
https://doi.org/10.1016/j.bja.2023.11.050. meta-analysis of sparse data. Stat Med 2004; 23: 1351e75
 Procedural Sedation in the ED and ICU - 503

17. DerSimonian R, Laird N. Meta-analysis in clinical trials. 33. Arhami Dolatabadi A, Mohammadian A, Kariman H.
Control Clin Trial. 1986; 7: 177e88 Lidocaine-midazolam-fentanyl combination in control-
18. White I. Network meta-analysis. Stata J 2015; 15: 951e85 ling pain for reduction of anterior shoulder dislocation; a
19. White IR, Barrett JK, Jackson D, Higgins JP. Consistency randomized clinical trial. Emerg (Tehran) 2018; 6: e24
and inconsistency in network meta-analysis: model 34. Barcelos A, Garcia PC, Portela JL, Piva JP, Garcia JP,
estimation using multivariate meta-regression. Res Synth Santana JC. Comparison of two analgesia protocols for
Methods 2012; 3: 111e25 the treatment of pediatric orthopedic emergencies. Rev
20. Dias S, Welton NJ, Caldwell DM, Ades AE. Checking Assoc Med Bras 2014; 61: 362e7
consistency in mixed treatment comparison meta-anal- 35. Bauman LA, Cannon ML, McCloskey J, et al. Unconscious
ysis. Stat Med 2010; 29: 932e44 sedation in children: a prospective multi-arm clinical
21. Higgins JP, Jackson D, Barrett JK, Lu G, Ades AE, White IR. trial. Paediatr Anaesth 2002; 12: 674e9
Consistency and inconsistency in network meta- 36. Burton JH, Bock AJ, Strout TD, Marcolini EG. Etomidate
analysis: concepts and models for multi-arm studies. and midazolam for reduction of anterior shoulder
Res Synth Methods 2012; 3: 98e110 dislocation: a randomized, controlled trial. Ann Emerg
22. Chaimani A, Higgins JP, Mavridis D, Spyridonos P, Med 2002; 40: 496e504
Salanti G. Graphical tools for network meta-analysis in 37. Cevik E, Bilgic S, Kilic E, et al. Comparison of ketamine-
STATA. PLoS One 2013; 8, e76654 low-dose midozolam with midazolam-fentanyl for or-
23. Brignardello-Petersen R, Bonner A, Alexander PE, et al. thopedic emergencies: a double-blind randomized trial.
Advances in the GRADE approach to rate the certainty in Am J Emerg Med 2013; 31: 108e13
estimates from a network meta-analysis. J Clin Epidemiol 38. Chan KKL, Ho HF. Etomidate and midazolam for proce-
2018; 93: 36e44 dural sedation in the emergency department of Queen
24. Brignardello-Petersen R, Florez ID, Izcovich A, et al. Elizabeth Hospital: a randomised controlled trial. Hong
GRADE approach to drawing conclusions from a network Kong J Emerg Med 2008; 15: 75e87
meta-analysis using a minimally contextualised frame- 39. Ozturk TC, Guneysel O, Akoglu H. Anterior shoulder
work. BMJ 2020; 371: m3900 dislocation reduction managed either with midazolam or
25. Santesso N, Glenton C, Dahm P, et al. GRADE guidelines propofol in combination with fentanyl. Hong Kong J Emerg
26: informative statements to communicate the findings Med 2014; 21: 346e53
of systematic reviews of interventions. J Clin Epidemiol 40. Coll-Vinent B, Sala X, Ferna  ndez C, et al. Sedation for
2020; 119: 126e35 cardioversion in the emergency department: analysis of
26. Abdolrazaghnejad ABM. Fentanyl-midazolam vs. effectiveness in four protocols. Ann Emerg Med 2003; 42:
midazolam-ketamine regarding patient sedation anal- 767e72
gesia for emergency orthopedic procedures. Bangladesh J 41. David H, Shipp J. A randomized controlled trial of keta-
Pharmacol 2017; 12 mine/propofol versus propofol alone for emergency
27. Afzalimoghaddam M, Khademi MF, Mirfazaelian H, department procedural sedation. Ann Emerg Med 2011;
Payandemehr P, Karimialavijeh E, Jalali A. Comparing 57: 435e41
diazepam plus fentanyl with midazolam plus fentanyl in 42. Del Pizzo JAB, Downes K, Mularoni P. Efficiency in seda-
the moderate procedural sedation of anterior shoulder tion for forearm fracture reduction in children: propofol
dislocations: a randomized clinical trial. J Emerg Med vs. ketamine-propofol vs. ketofol [abstract]. Pediatr Emerg
2021; 60: 1e7 Care 2011; 27: 999
28. Akhlaghi N, Payandemehr P, Yaseri M, Akhlaghi AA, 43. Derakhshanfar H, Bozorgi F, Hosseini A, et al. Comparing
Abdolrazaghnejad A. Premedication with midazolam or the effects of dexmedetomidine and midazolam on
haloperidol to prevent recovery agitation in adults un- sedation in children with head trauma to perform CT in
dergoing procedural sedation with ketamine: a ran- emergency department/UporeCivanje efekata deksme-
domized double-blind clinical trial. Ann Emerg Med 2019; detomidina i midazolama na sedaciju dece sa povredom
73: 462e9 glave radi snimanja CT-om na Odeljenju urgentne medi-
29. Amini A, Arhami Dolatabadi A, Kariman H, et al. Low- cine. Acta Facultatis Medicae Naissensis 2015; 32: 59e65
dose fentanyl, propofol, midazolam, ketamine and lido- 44. Di Liddo L, D’Angelo A, Nguyen B, Bailey B, Amre D,
caine combination vs. regular dose propofol and fentanyl Stanciu C. Etomidate versus midazolam for procedural
combination for deep sedation induction; a randomized sedation in pediatric outpatients: a randomized
clinical trial. Emerg (Tehran) 2018; 6: e57 controlled trial. Ann Emerg Med 2006; 48: 433e440, 40.e1
30. Aminiahidashti H, Shafiee S, Hosseininejad SM, et al. 45. Dilli D, Dallar Y, Sorgui NH. Intravenous ketamine plus
Propofol-fentanyl versus propofol-ketamine for proce- midazolam vs. intravenous ketamine for sedation in
dural sedation and analgesia in patients with trauma. lumbar puncture: a randomized controlled trial. Indian
Am J Emerg Med 2018; 36: 1766e70 Pediatr 2008; 45: 899e904
31. Andolfatto G, Abu-Laban RB, Zed PJ, et al. Ketamine- 46. Dis‚el NR, Yilmaz HL, Sertdemir Y, Yes‚ilag  aç H, Avci A.
propofol combination (ketofol) versus propofol alone for Etomidate versus ketamine: effective use in emergency
emergency department procedural sedation and anal- procedural sedation for pediatric orthopedic injuries.
gesia: a randomized double-blind trial. Ann Emerg Med Pediatr Emerg Care 2016; 32: 830e4
2012; 59: 504e512.e1e2 47. Dunn MJ, Mitchell R, DeSouza CI, Drummond GB, Waite A.
32. Arhami Dolatabadi A, Memary E, Shojaee M, Recovery from sedation with remifentanil and propofol,
Kamalifard H. Dexmedetomidine-fentanyl versus compared with morphine and midazolam, for reduction in
midazolam-fentanyl in pain management of distal anterior shoulder dislocation. Emerg Med J 2011; 28: 6e10
radius fractures reduction; a randomized clinical trial. 48. Ferguson I, Bell A, Treston G, New L, Ding M, Holdgate A.
Emerg (Tehran) 2018; 6: e10 Propofol or ketofol for procedural sedation and analgesia
 504 - Sharif et al.

in emergency medicine-the POKER study: a randomized of children for head and neck CT imaging. Pediatr Emerg
double-blind clinical trial. Ann Emerg Med 2016; 68: Care 2004; 20: 499e506
574e582.e1 64. Lee-Jayaram JJ, Green A, Siembieda J, et al. Ketamine/
49. Genzlinger MA, Salen P, Grossman M, Stehly C, midazolam versus etomidate/fentanyl: procedural seda-
Stoltzfus J. 145 “Put Me Out Doc”: ketamine versus eto- tion for pediatric orthopedic reductions. Pediatr Emerg
midate for the reduction of orthopedic dislocations. Ann Care 2010; 26: 408e12
Emerg Med 2012; 60: S52e3 65. Lemoel F, Contenti J, Giolito D, et al. Adverse events with
50. Gharavifard M, Tafakori A, Zamani Moghadam H. ketamine versus ketofol for procedural sedation on
Remifentanil versus fentanyl/midazolam in painless adults: a double-blind, randomized controlled trial. Acad
reduction of anterior shoulder dislocation; a randomized Emerg Med 2017; 24: 1441e9
clinical trial. Emerg (Tehran) 2016; 4: 92e6 66. Maltepe F, Kocaayan E, Ugurlu BS, Akdeniz B, Guneri S.
51. Godambe SA, Elliot V, Matheny D, Pershad J. Comparison Comparison of remifentanil and fentanyl in anaesthesia
of propofol/fentanyl versus ketamine/midazolam for for elective cardioversion. Anaesth Intensive Care 2006; 34:
brief orthopedic procedural sedation in a pediatric 353e7
emergency department. Pediatrics 2003; 112: 116e23 67. Messenger DW, Murray HE, Dungey PE, van Vlymen J,
52.  lu SN, Erkalp K, et al. The analgesic and
Gümüs‚ F, S‚inikog Sivilotti ML. Subdissociative-dose ketamine versus fen-
hemodynamic effects of dexmedetomidine and remi- tanyl for analgesia during propofol procedural sedation:
fentanil during chest tube removal. Türk G€ ogüs Kalp a randomized clinical trial. Acad Emerg Med 2008; 15:
Damar Cerrahisi Dergisi 2013; 21: 966e71 877e86
53. Hart LS, Berns SD, Houck CS, Boenning DA. The value of 68. Miner JR, Biros M, Krieg S, Johnson C, Heegaard W,
end-tidal CO2 monitoring when comparing three Plummer D. Randomized clinical trial of propofol versus
methods of conscious sedation for children undergoing methohexital for procedural sedation during fracture
painful procedures in the emergency department. Pediatr and dislocation reduction in the emergency department.
Emerg Care 1997; 13: 189e93 Acad Emerg Med 2003; 10: 931e7
54. Hatamabadi HR, Arhami Dolatabadi A, Derakhshanfar H, 69. Miner JR, Danahy M, Moch A, Biros M. Randomized
Younesian S, Ghaffari Shad E. Propofol versus mid- clinical trial of etomidate versus propofol for procedural
azolam for procedural sedation of anterior shoulder sedation in the emergency department. Ann Emerg Med
dislocation in emergency department: a randomized 2007; 49: 15e22
clinical trial. Trauma Mon 2015; 20, e13530 70. Miner JR, Driver BE, Moore JC, et al. Randomized clinical
55. Havel Jr CJ, Strait RT, Hennes H. A clinical trial of pro- trial of propofol versus alfentanil for moderate proce-
pofol vs midazolam for procedural sedation in a pedi- dural sedation in the emergency department. Am J Emerg
atric emergency department. Acad Emerg Med 1999; 6: Med 2017; 35: 1451e6
989e97 71. Miner JR, Gray RO, Bahr J, Patel R, McGill JW. Randomized
56. Hunt GS, Spencer MT, Hays DP. Etomidate and mid- clinical trial of propofol versus ketamine for procedural
azolam for procedural sedation: prospective, random- sedation in the emergency department. Acad Emerg Med
ized trial. Am J Emerg Med 2005; 23: 299e303 2010; 17: 604e11
57. Holger JS, Satterlee PA, Haugen S. Nursing use between 2 72. Miner JR, Gray RO, Stephens D, Biros MH. Randomized
methods of procedural sedation: midazolam versus clinical trial of propofol with and without alfentanil for
propofol. Am J Emerg Med 2005; 23: 248e52 deep procedural sedation in the emergency department.
58. Bahreini M, Talebi Garekani M, Sotoodehnia M, Rasooli F. Acad Emerg Med 2009; 16: 825e34
Comparison of the efficacy of ketamine- propofol versus 73. Miner JR, Moore JC, Austad EJ, Plummer D, Hubbard L,
sodium thiopental-fentanyl in sedation: a randomised Gray RO. Randomized, double-blinded, clinical trial of
clinical trial. Emerg Med J 2021; 38: 211e6 propofol, 1:1 propofol/ketamine, and 4:1 propofol/keta-
59. Masoumi K, Maleki SJ, Forouzan A, Delirrooyfard A, mine for deep procedural sedation in the emergency
Hesam S. Dexmedetomidine versus midazolam-fentanyl department. Ann Emerg Med 2015; 65: 479e488.e2
in procedural analgesia sedation for reduction of ante- 74. Miner JR, Moore JC, Plummer D, Gray RO, Patel S, Ho JD.
rior shoulder dislocation: a randomized clinical trial. Rev Randomized clinical trial of the effect of supplemental
Recent Clin Trial. 2019; 14: 269e74 opioids in procedural sedation with propofol on serum
60. Jamal SM, Fathil SM, Nidzwani MM, Ismail AK, Yatim FM. catecholamines. Acad Emerg Med 2013; 20: 330e7
Intravenous ketamine is as effective as midazolam/fen- 75. Monsef Kasmaee V, Zia Zibari SM, Aghajani Nargesi M.
tanyl for procedural sedation and analgesia in the Remifentanil versus propofol/fentanyl combination in
emergency department. Med J Malaysia 2011; 66: 231e3 procedural sedation for dislocated shoulder reduction; a
61. Kennedy RM, Porter FL, Miller JP, Jaffe DM. Comparison of clinical trial. Arch Acad Emerg Med 2019; 7: e10
fentanyl/midazolam with ketamine/midazolam for pe- 76. Moro-Sutherland DM, Algren JT, Louis PT, Kozinetz CA,
diatric orthopedic emergencies. Pediatrics 1998; 102: Shook JE. Comparison of intravenous midazolam with
956e63 pentobarbital for sedation for head computed tomogra-
62. Khutia SK, Mandal MC, Das S, Basu SR. Intravenous phy imaging. Acad Emerg Med 2000; 7: 1370e5
infusion of ketamine-propofol can be an alternative to 77. Nashibi M, Mottaghi K, Faraji M, Delavari A, Taghipour H,
intravenous infusion of fentanyl-propofol for deep Amiri M. Comparison of analgesic and sedative effects of
sedation and analgesia in paediatric patients undergoing ketamine-propofol (ketofol) and fentanyl-midazolam
emergency short surgical procedures. Indian J Anaesth (fentazolam) combinations in outpatient orthopedic
2012; 56: 145e50 procedures. Trauma Mon 2017; 22, e41315
63. Kienstra AJ, Ward MA, Sasan F, Hunter J, Morriss MC, 78. Nejati A, Moharari RS, Ashraf H, Labaf A, Golshani K.
Macias CG. Etomidate versus pentobarbital for sedation Ketamine/propofol versus midazolam/fentanyl for
 Procedural Sedation in the ED and ICU - 505

procedural sedation and analgesia in the emergency 92. Taylor DM, O’Brien D, Ritchie P, Pasco J, Cameron PA.
department: a randomized, prospective, double-blind Propofol versus midazolam/fentanyl for reduction of
trial. Acad Emerg Med 2011; 18: 800e6 anterior shoulder dislocation. Acad Emerg Med 2005; 12:
79. Parlak M, Parlak I, Erdur B, Ergin A, Sagiroglu E. Age effect 13e9
on efficacy and side effects of two sedation and analgesia 93. Uri O, Behrbalk E, Haim A, Kaufman E, Halpern P. Pro-
protocols on patients going through cardioversion: a cedural sedation with propofol for painful orthopaedic
randomized clinical trial. Acad Emerg Med 2006; 13: 493e9 manipulation in the emergency department expedites
80. Phillips W, Anderson A, Rosengreen M, Johnson J, patient management compared with a midazolam/ke-
Halpin J. Propofol versus propofol/ketamine for brief tamine regimen: a randomized prospective study. J Bone
painful procedures in the emergency department: clin- Jt Surg Am 2011; 93: 2255e62
ical and bispectral index scale comparison. J Pain Palliat 94. Vahidi E, Hemati R, Momeni M, Jahanshir A, Saeedi M.
Care Pharmacother 2010; 24: 349e55 Comparison of sedative effectiveness of thiopental
81. Rahman NH, Hashim A. The use of propofol for proce- versus midazolam in reduction of shoulder dislocation.
dural sedation and analgesia in the emergency depart- World J Emerg Med 2018; 9: 125e9
ment: a comparison with midazolam. Emerg Med J 2011; 95. Venkatakrishnan RT. Comparison of propofol/fentanyl
28: 861e5 vs ketamine/midazolam for procedural sedation &
82. Salen P, Grossman M, Grossman M, Milazzo A, analgesia in the emergency department. J Emerg Med
Stoltzfus J. A comparison of ketamine versus etomidate 2011; 41: P222
for procedural sedation for the reduction of large joint 96. Wathen JE, Roback MG, Mackenzie T, Bothner JP. Does
dislocations. Int J Crit Illn Inj Sci 2016; 6: 79e84 midazolam alter the clinical effects of intravenous ke-
83. Sawas AYS, Madsen TE, Davis VW. Combined ketamine tamine sedation in children? A double-blind, random-
and propofol sedation versus propofol sedation for ized, controlled, emergency department trial. Ann Emerg
emergency department procedures: a prospective ran- Med 2000; 36: 579e88
domized trial. Ann Emerg Med 2016; 62: S76e7 97. Weisz K, Bajaj L, Deakyne SJ, et al. Adverse events during
84. Sener S, Eken C, Schultz CH, Serinken M, Ozsarac M. a randomized trial of ketamine versus co-administration
Ketamine with and without midazolam for emergency of ketamine and propofol for procedural sedation in a
department sedation in adults: a randomized controlled pediatric emergency department. J Emerg Med 2017; 53:
trial. Ann Emerg Med 2011; 57: 109e114.e2 1e9
85. Seol TK, Lim JK, Yoo EK, Min SW, Kim CS, Hwang JY. 98. Wright SW, Chudnofsky CR, Dronen SC, et al. Compari-
Propofol-ketamine or propofol-remifentanil for deep son of midazolam and diazepam for conscious sedation
sedation and analgesia in pediatric patients undergoing in the emergency department. Ann Emerg Med 1993; 22:
burn dressing changes: a randomized clinical trial. Pae- 201e5
diatr Anaesth 2015; 25: 560e6 99. Azizkhani R, Kouhestani S, Heydari F, et al. Comparison
86. Shah A, Mosdossy G, McLeod S, Lehnhardt K, Peddle M, of the effects of dexmedetomidine and propofol in
Rieder M. A blinded, randomized controlled trial to reducing recovery agitation in pediatric patients after
evaluate ketamine/propofol versus ketamine alone for ketamine procedural sedation in emergency department.
procedural sedation in children. Ann Emerg Med 2011; 57: J Res Med Sci 2021; 26: 61
425e433.e2 100. Massaeli M, Nasouhi S, Bahrani H, et al. Comparison of
87. AL Sheikh Shihab, Ahmad MJ, Khan ZMTS, et al. Single- sedatives for the reduction of shoulder dislocation based
shot sub-dissociative dose ketofol versus ketamine alone on bispectral index system in emergency department: a
for emergency department procedural sedation and randomized, three-group, double-blinded clinical trial.
analgesia in adult. J Emerg Med Trauma Surg Care 2021; 8 J Adv Med Biomed Res 2022; 30: 407e16
88. Sherwin TS, Green SM, Khan A, Chapman DS, 101. Vardi A, Salem Y, Padeh S, Paret G, Barzilay Z. Is propofol
Dannenberg B. Does adjunctive midazolam reduce re- safe for procedural sedation in children? A prospective
covery agitation after ketamine sedation for pediatric evaluation of propofol versus ketamine in pediatric
procedures? A randomized, double-blind, placebo- critical care. Crit Care Med 2002; 30: 1231e6
controlled trial. Ann Emerg Med 2000; 35: 229e38 102. Gale DW, Grissom TE, Mirenda JV. Titration of intrave-
89. Soysal S, Karcioglu O, Demircan A, et al. Comparison of nous anesthetics for cardioversion: a comparison of
meperidine plus midazolam and fentanyl plus mid- propofol, methohexital, and midazolam. Crit Care Med
azolam in procedural sedation: a double-blind, random- 1993; 21: 1509e13
ized controlled trial. Adv Ther 2004; 21: 312e21 ~ o FV, Aguiar VE,
103. Lucas da Silva PS, Oliveira Iglesias SB, Lea
90. Stronati G, Capucci A, Dello Russo A, et al. Procedural Brunow de Carvalho W. Procedural sedation for insertion
sedation for direct current cardioversion: a feasibility of central venous catheters in children: comparison of
study between two management strategies in the midazolam/fentanyl with midazolam/ketamine. Paediatr
emergency department. BMC Cardiovasc Disord 2020; 20: Anaesth 2007; 17: 358e63
388 104. Yang ZB, Shen JY, Mi KD, Ma Q, Wu YS, Yao M. [Study on
91. Tajoddini S, Motaghi M. Sedative and analgesic effects of the application of dexmedetomidine combined with
propofoleketamine versus propofolefentanyl for emer- remifentanil in dressing change of conscious patients
gency department procedures. Hong Kong J Emerg Med with non-intubation in burn intensive care unit]. Zhong-
2022; 29: 212e9 hua Shao Shang Za Zhi 2018; 34: 707e13
 506 - Sharif et al.

105. Yildirim V, Doganci S, Bolcal C, et al. Combination 108. Lobb D, Clarke A, Lai H. Administration order of mid-
sedoanalgesia with remifentanil and propofol versus azolam/fentanyl for moderate dental sedation. J Dent
remifentanil and midazolam for elective cardioversion Anesth Pain Med 2018; 18: 47e56
after coronary artery bypass grafting. Adv Ther 2007; 24: 109. Aroke EN, Crawford SL, Dungan JR. Pharmacogenetics of
662e70 ketamine-induced emergence phenomena: a pilot study.
106. Yldzdas‚ D, Yapcoǧlu H, Ylmaz HL. The value of cap- Nurs Res 2017; 66: 105e14
nography during sedation or sedation/analgesia in pe- 110. Doenicke AW, Roizen MF, Kugler J, Kroll H, Foss J,
diatric minor procedures. Pediatr Emerg Care 2004; 20: Ostwald P. Reducing myoclonus after etomidate. Anes-
162e5 thesiology 1999; 90: 113e9
107. Mofidi M, Rouhi R, Mahshidfar B, et al. Propofol-ketamine 111. Green SM, Krauss B. Ketamine is a safe, effective, and
vs. propofol-fentanyl combinations in patients under- appropriate technique for emergency department
going closed reduction: a randomized, double-blind, paediatric procedural sedation. Emerg Med J 2004; 21:
clinical trial. Adv J Emerg Med 2018; 2: e44 271e2

Handling Editor: Jonathan Hardman