Journal of Clinical Anesthesia 41 (2017) 48–54

Contents lists available at ScienceDirect

Journal of Clinical Anesthesia

Laparoscopic cholecystectomy under neuraxial anesthesia compared

with general anesthesia: Systematic review and meta-analyses
Marcelo A. Longo, MD ⁎, Bárbara T. Cavalheiro, MD, Getúlio R. de Oliveira Filho, PhD
Department of Surgery, University Hospital, Federal University of Santa Catarina, Florianópolis, Santa Catarina, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Background: Pneumoperitoneum during laparoscopic cholecystectomy (LC) can cause hypercapnia, hypoxemia,
Received 30 April 2017 hemodynamic changes and shoulder pain. General anesthesia (GA) enables the control of intraoperative pain
Received in revised form 10 June 2017 and ventilation. The need for GA has been questioned by studies suggesting that neuraxial anesthesia (NA) is ad-
Accepted 16 June 2017 equate for LC.
Available online xxxx
Study objective: To quantify the prevalence of intraoperative pain and to verify whether evidence on the mainte-
nance of ventilation, circulation and surgical anesthesia during NA compared with GA is consistent.
Keywords:
Review, systematic
Design: Systematic review with meta-analyses.
Cholecystectomy, laparoscopic Setting: Anesthesia for laparoscopic cholecystectomy.
Anesthesia Patients: We searched Medline, Cochrane and EBSCO databases up to 2016 for randomized controlled trials that
Pain, referred compared LC in the two groups under study, neuraxial (subarachnoid or epidural) and general anesthesia.
Patient safety Measurements: The primary outcome was the prevalence of intraoperative pain referred to the shoulder in the NA
group. Hemodynamic and respiratory outcomes and adverse effects in both groups were also collected.
Main results: Eleven comparative studies were considered eligible. The pooled prevalence of shoulder pain was
25%. Intraoperative hypotension and bradycardia occurred more frequently in patients who received NA, with a
risk ratio of 4.61 (95% confidence interval [CI] 1.70–12.48, p = 0.003) and 6.67 (95% CI 2.02–21.96, p = 0.002), re-
spectively. Postoperative nausea and vomiting was more prevalent in patients who submitted to GA. The preva-
lence of postoperative urinary retention did not differ between the techniques. Postoperative headache was
more prevalent in patients who received NA, while the postoperative pain intensity was lower in this group.
Performing meta-analyses on hypertension, hypercapnia and hypoxemia was not possible.
Conclusions: NA as sole anesthetic technique, although feasible for LC, was associated with intraoperative pain re-
ferred to the shoulder, required anesthetic conversion in 3.4% of the cases and did not demonstrate evidence of re-
spiratory benefits for patients with normal pulmonary function.
© 2016 Elsevier Inc. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.1. Eligibility criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.2. Information sources and search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.3. Study selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.4. Data extraction process and data items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.5. Risk of bias in individual studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.6. Synthesis of results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.7. Risk of bias across studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.8. Quality of evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.1. Study selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

⁎ Corresponding author at: Department of Surgery, University Hospital, Federal University of Santa Catarina, Rua Professora Maria Flora Pausewang, s/n, Trindade, 88036-800
Florianópolis, Santa Catarina, Brazil.
E-mail address: marceloarent@hotmail.com (M.A. Longo).

http://dx.doi.org/10.1016/j.jclinane.2017.06.005
0952-8180/© 2016 Elsevier Inc. All rights reserved.
 M.A. Longo et al. / Journal of Clinical Anesthesia 41 (2017) 48–54 49

3.2. Study characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

3.3. Risk of bias within studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.4. Results of individual studies and synthesis of results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.5. Risk of bias across studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.6. Quality of evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Funding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

1. Introduction
Laparoscopic cholecystectomy (LC) has replaced the open technique
as the first choice for the surgical treatment of cholelithiasis and chole-
cystitis because of its less invasive approach and its association with
shorter hospital stay, faster return to usual activities, less probability of
complications in the surgical wound, and decreased postoperative pain
[1].
Pneumoperitoneum, which is required for the procedure, causes re-
spiratory changes, including an increase in the arterial pressure of CO2
in the arterial blood (PaCO2), a decrease in pulmonary compliance, an in-
crease in peak and plateau airway pressure, a reduction in vital capacity
and functional residual capacity, atelectasis, increased dead space and
ventilation/perfusion mismatch [2,3].
It also affects the cardiovascular system and causes a decrease in car-
diac output, an increase in afterload and systemic and pulmonary vascu-
lar resistance. Bradycardia may also occur because of vagal stimulation
during insufflation of the abdominal cavity [4,5].
The choice of anesthetic technique for LC is largely limited to general
anesthesia (GA) because it eliminates the discomfort caused by the
pneumoperitoneum and the changes in the patient's position on the sur-
gical table. In addition, GA enables the better control of ventilation and a
rigorous analysis of CO2 as well as tracheal intubation, which reduces the
risk of bronchoaspiration [2,6].
However, new studies have demonstrated the possibility of
performing neuraxial anesthesia (NA) for LC [7–9]. These studies indicate
that neuraxial block is associated with a low incidence of postoperative
pain, nausea and vomiting, short hospitalization time, low costs and ad-
equate surgical relaxation. Nevertheless, they point out the occurrence of
intraoperative pain, which is sometimes severe and requires conversion
to GA.
A recent systematic review, [10] which evaluated LC under spinal an-
esthesia and identified postoperative pain as the primary outcome and
duration of surgery and postoperative complications as secondary out-
comes, concluded that subarachnoid block is a viable and safe technique
for LC. However, it did not produce sufficient evidence on the occurrence
of hypoxemia, hypercarbia and other cardiovascular changes.
This systematic review with meta-analyses was conducted to quanti-
fy the prevalence of shoulder pain in the NA group and to verify whether
evidence on the maintenance of ventilation and circulation during LC
under NA compared with GA is consistent.
have described or allowed the extraction of data on the prevalence of in-
traoperative pain, hemodynamic and respiratory outcomes and adverse
effects. Studies that involved comparisons with combined anesthesia
(NA and GA) were not included.
2.2. Information sources and search strategy
We searched the MEDLINE, Cochrane Central Register of Controlled
Trials and EBSCO databases, without language restriction, from inception
to March 2016 using the following terms: “laparoscopic cholecystecto-
my” AND “general anesthesia” or “general anaesthesia” or general AND
“spinal anesthesia” or “spinal anaesthesia” or spinal or “epidural anes-
thesia” or “epidural anaesthesia” or epidural or “neuraxial anesthesia”
or “neuraxial anaesthesia” or neuraxial.
2.3. Study selection
A systematic search was conducted by the three authors indepen-
dently. Two authors (MAL and BTC) screened the abstracts of the re-
trieved articles and excluded reports that did not fulfill the inclusion
criteria. Any doubt concerning the inclusion of a trial was resolved by a
discussion with the third author (GROF).
The reference lists of included articles were screened for further rel-
evant articles. Unpublished reports and studies only published as con-
ference abstracts were not included. Authors were not contacted for
additional data.
2.4. Data extraction process and data items
The primary endpoint variable was the prevalence of intraoperative
pain referred to the shoulder in the NA group. The secondary endpoint
variables were the anesthetic conversion rate (NA to GA), prevalence
of intraoperative hypotension, hypertension, bradycardia and respirato-
ry outcomes (respiratory rate, hypercapnia and hypoxemia), postopera-
tive pain scores at 0, 2, 4, 6, 8, 12 and 24 h, and reported frequencies of
postoperative nausea and vomiting (PONV), urinary retention and head-
ache. The data were extracted by two authors (MAL and BTC) and con-
ferred by the third author (GROF).
2.5. Risk of bias in individual studies

2. Methods The methodological quality of each study was evaluated according to

This systematic review was performed according to the processes de-
scribed by the PRISMA guidelines [11], including the design, implemen-
tation of the steps, analysis and description of the results. The protocol
for this study was not registered.
2.1. Eligibility criteria
We included randomized controlled trials that compared LC in the
two groups under study, NA (subarachnoid or epidural) and GA, and
the Cochrane Collaboration's tool for assessing risk of bias [12].
2.6. Synthesis of results
The pooled prevalence of intraoperative pain referred to the shoulder
and the anesthetic conversion rate were estimated by the inverse vari-
ance method. As this method tends to underestimate or overestimate
the prevalence in situations in which studies show great heterogeneity,
the double arcsine transformation was applied to obtain stable estimates
of these measures [13]. The calculations and graphs on the prevalence of
50 M.A. Longo et al. / Journal of Clinical Anesthesia 41 (2017) 48–54
intraoperative pain and the anesthetic conversion rate were performed
and extracted through the MetaXL 3.0 software [14].
For the categorical variables, the results are presented as risk ratios
(RR) and their 95% confidence intervals (CI). For the continuous vari-
ables, the mean difference (MD) and its 95% CI were calculated.
A random effects model would be used if the data were heteroge-
neous (Cochrane Q test with a p b 0.10) and quantified by the I2 index.
If the data were homogenous (p ≥ 0.1), we would apply a fixed effect
model.
Sensitivity analyses were conducted to identify the source of hetero-
geneity, if present, by removing studies one by one and recalculating the
combined estimates. We also planned sensitivity analyses including only
trials with low risk of bias. Studies were considered to have high risk of
bias if one or more of the domains in the Cochrane Collaboration's tool
had a high or unclear risk of bias.
Visual analogue scale (VAS) pain scores were extracted as the mean
and standard deviation. Scores described only as median and amplitude
were transformed into the mean and standard deviation [15]. The scores
described in graphical form were digitally extracted using the Engauge
Digitizer 6.0 program [16].
Statistical analyses were conducted using the softwares Review Man-
ager 5.3 [17] and R version 3.4.0 (packages “meta” [18] and “metafor”
[19]).
2.7. Risk of bias across studies
Publication bias and small study effects were assessed by the inspec-
tion of funnel plots and by formal testing asymmetry with Egger's linear
regression test at the 5% level of significance (p b 0.05). Additionally,
Duval and Tweedie's trim and fill analysis was used to obtain the effect
estimate adjusted for publication bias. Only the endpoints described by
at least 10 studies were considered, as the statistical power was too low
to distinguish chance from true asymmetry when b10 studies are in-
cluded [20,21].
2.8. Quality of evidence
The Grading of Recommendations Assessment, Development and
Evaluation (GRADE) [22] methodology was used to evaluate the quality
of evidence for the individual outcomes of the review. The quality of ev-
idence for each outcome was graded as high, moderate, low, and very
low. We presented the results of quality of evidence for each outcome
through a summary of findings table.
3. Results
3.1. Study selection
The systematic search identified 191 potentially relevant citations.
Eleven trials fulfilled all the inclusion criteria and were included for re-
view and meta-analyses (Fig. 1).
3.2. Study characteristics
In total, 963 patients were included in the meta-analyses, that is, 482
under NA and 481 under GA (Table 1). All but two studies reported pa-
tients with an American Society of Anesthesiologists (ASA) physical sta-
tus I or II who submitted to elective LC as inclusion criterion. Two trials
included ASA III patients [23,24], but excluded if they had acute cholecys-
titis, contraindication for laparoscopic surgery or neuraxial anesthesia,
chronic obstructive pulmonary disease or currently taking chronic nar-
cotic pain management.
3.3. Risk of bias within studies
The risk of bias within the studies is summarized in the Appendix
(see Supplemental digital content [SDC], Figs. 2 and 3). Only three stud-
ies described the generation of an appropriate randomized sequence,
and only four reported an adequate method to conceal the allocation se-
quence. Performance and detection bias was present in almost all trials.
Five studies [24–28] described incomplete outcome data. The main rea-
son for the missing data was the anesthetic conversion due to severe
shoulder pain, which could be related to the outcomes. Three patients
were excluded (one withdrew and two had surgical conversion) in one
trial [28], but this exclusion was unlikely to affect the results.
3.4. Results of individual studies and synthesis of results
Intraoperative pain referred to the shoulder was reported in 10 stud-
ies and occurred in 25% (95% CI 22%–30%, I2 = 84%) of the patients (SDC,
Fig. 4). Two studies [24,29] were responsible for the observed heteroge-
neity. These studies presented modifications of the technique to maintain
the pneumoperitoneum or included instillation of the subdiaphragmatic
area with local anesthetics. In the sensitivity analysis on their removal,
the prevalence of pain reached 36% (95% CI 31%–42%, I2 = 10%).
In most patients, when right shoulder pain was present, administra-
tion of intravenous fentanyl and/or irrigation of the diaphragm with
local anesthetics was required for adequate relief.
In eight patients, shoulder pain was severe enough to require conver-
sion to general anesthesia. Other reasons for conversion were discomfort
on abdomen, bleeding and adhesions on Calot's triangle (two cases
each), as well as nausea/vomiting and anxiety (one case each). The anes-
thetic conversion rate was 3.4% (95% CI 1.97–5.25%, I2 = 9%).
In both groups, seven studies reported the occurrence of intraopera-
tive hypotension, which was more prevalent in patients who submitted
to NA (RR 4.61, 95% CI 1.70–12.48, p = 0.003, I2 = 56%) (SDC, Fig. 5).
Sensitivity analyses identified one of the studies [30] as the one respon-
sible for the heterogeneity. In removing this study, the RR of hypotension
in the NA group was 6.34 (95% CI 2.34–17.18, p = 0.0003, I2 = 35%).
The occurrence of intraoperative bradycardia was described in four
studies and was more prevalent in the NA group (RR 6.67, 95% CI 2.02–
21.96, p = 0.002, I2 = 33%) (SDC, Fig. 6).
Intraoperative hypotension and bradycardia were defined according
to the criteria of the respective studies. The former was managed with in-
fusion of fluids and vasopressor agents in all cases, while the latter was
treated with atropine. There were no reported episodes of hemodynamic
instability.
Postoperative nausea and vomiting occurred more frequently in pa-
tients who received GA (RR 0.40, 95% CI 0.27–0.60, p b 0.00001, I2 =
22%) (SDC, Fig. 7).
The occurrence of postoperative urinary retention did not differ be-
tween the two anesthetic techniques (RR 1.46, 95% CI 0.74–2.88, p =
0.27, I2 = 24%) (SDC, Fig. 8). Two trials [26,31] had no cases of urinary
retention and thus were not included in the calculation. One study [32]
was responsible for the absence of difference. In this study, 14 patients
in the GA group required postanesthesia care unit opioid administration,
as opposed to 2 patients in the NA group. The sensitivity analysis with its
exclusion showed a higher prevalence in the NA group, with an RR of
2.76 (95% CI 1.18–6.43, p = 0.02, I2 = 0%).
The occurrence of postoperative headache was reported in five stud-
ies and was higher in patients who submitted to NA (RR 4.08, 95% CI
1.17–14.30, p = 0.03, I2 = 0) (SDC, Fig. 9). The episodes were consistent
with post dural puncture headache after spinal anesthesia, using pencil
point 25-gauge or Whitacre 25-gauge needles, and were managed main-
ly with analgesics and IV fluids; one patient required an epidural blood
patch after failure of conservative management.
Postoperative pain intensity, measured on the VAS, was lower in pa-
tients who received NA between the immediate postoperative period
and the first 24 h, except at 6 h (SDC, Fig. 10). Sensitivity analyses
 M.A. Longo et al. / Journal of Clinical Anesthesia 41 (2017) 48–54
Fig. 1. Flowchart of study selection process.
identified a study [24] as the source of heterogeneity in two instances,
namely, at 6 h and 12 h after surgery. In removing the study, the patients
in the NA group had a MD on the VAS of 0.01 (95% CI −0.61 to 0.63; p =
0.98; I2 = 0%) and −2.44 (95% CI −2.67 to −2.21; p b 0,00001; I2 = 0%),
respectively.
Meta-analyses of hypertension, hypercapnia and hypoxemia were
not possible. None of the studies indicated if there were cases of
hypertension.
As for hypercapnia and hypoxemia, only one study [23] collected
blood gases to measure these respiratory variables. Higher pH and
lower PaCO2 values were observed in the GA group during the incision
and 5 min after insufflation but not 5 min after wound closure. In the
same group, PaO2 and SpO2 were higher at all times than in the epidural
anesthesia group. Despite these differences, PaO2 and PaCO2 remained
within normal limits in the patients who received NA; there were no re-
ported cases of respiratory difficulty.
The same study described the intraoperative monitoring of the end-
tidal CO2 pressure (PETCO2), with carbon dioxide sampling at the nares
in the NA group, but no difference was found between the groups.
Only two trials [31,32] described the respiratory rate variation, and a
significant increase was observed in the NA group compared with the GA
group in one of the studies [32], with oxygen saturation at 97%–98%
throughout. In the other study [31], PETCO2 was monitored postopera-
tively for 24 h and did not present any difference, but how such a mea-
surement was made was not explained.
Two studies [28,30] reported in their methods that the respiratory
rate, PaCO2 and acid-base balance would be monitored, but they did
not present the results for these variables.
Three trials [24,26,32] indicated the occurrence of hypoxemia (mea-
sured through the SpO2), which did not ensue in any of the patients, or
CO2 retention (measured through the expired air), which was observed
in eight patients of the GA group in one of the studies [26] but did not re-
port how the expired CO2 sampling was conducted.
The other four trials [25,27,29,33] did not report these respiratory pa-
rameters. Meta-analyses of hypercapnia and hypoxemia could not be
performed because of the lack of comparable measures between the
studies.
Sensitivity analyses including only low risk of bias studies were not
possible. Even considering the difficulty in the blinding of participants
and investigators, all studies presented a high or unclear risk of bias in
at least one of the other domains of the bias assessment tool.
51
3.5. Risk of bias across studies
Publication bias could be assessed for intraoperative pain referred to
the shoulder, postoperative nausea and vomiting and urinary retention.
The analysis of the funnel plots (SDC, Fig. 11) and Egger's linear regres-
sion test revealed no evidence of bias (intraoperative pain: p = 0.14,
95% CI −1.65 to 9.11; PONV: p = 0.06, 95% CI −3.38 to 0.11; urinary re-
tention: p = 0.32, 95% CI −3.08 to 8.12). The effect sizes adjusted with
Duval and Tweedie's trim and fill procedure for intraoperative pain
(five studies trimmed), postoperative nausea and vomiting (three stud-
ies trimmed) and urinary retention (one study trimmed) were 13.90%
(95% CI 11.24%–16.77%), 0.57 (95% CI 0.38–0.85) and 1.56 (95% CI
0.65–3.72), respectively.
3.6. Quality of evidence
The GRADE-rated quality of evidence was very low for most out-
comes, low for postoperative urinary retention and headache and mod-
erate for PONV (SDC, Table 2).
4. Discussion
All studies presented high risk of performance and detection bias. The
main reason for this result is the absence of blinding of patients and in-
vestigators, who knew which anesthesia, NA or GA, would be performed.
Another bias is due to the modification of the technique for estab-
lishing pneumoperitoneum, which occurred only in the NA group, in
some of the studies [23,25,26,29,30,33]. In these studies, the use of
low pressure and flow rate for intraperitoneal CO2 insufflation, as well
as the minimization of positional alterations and the infiltration of the
subdiaphragmatic space with local anesthetic, might have influenced
the outcomes, especially the occurrence of intraoperative pain referred
to the shoulder and the postoperative pain scores [34].
Intraoperative pain referred to the shoulder is caused by the irritation
of the subdiaphragmatic space induced by CO2. The innervation of this
region is supplied by the phrenic nerve, the roots of which originate in
the spinal segments C3–5 [35,36]. At conventional levels, NA does not
block this painful stimulus. In our study, this intraoperative pain was re-
ported by 25% of the patients who received NA and was the main reason
 52
Table 1
Characteristics of the clinical trials included in the meta-analyses.

Study Publication Patients randomized Neuraxial anesthesia General anesthesia Anesthetic Surgical Patients analyzed
year for NA/GA (n) conversion conversion (NA/GA)

Arati [29] 2010 50/50 SAa at L2–3 interspace (3 mL of hyperbaric bupivacaine 0.5%) Technique not described 2 – 50/50

M.A. Longo et al. / Journal of Clinical Anesthesia 41 (2017) 48–54

Bessa [33] 2010 30/30 SA at L2–3 interspace (3 mL of hyperbaric bupivacaine 0.5% and Induction (propofol, fentanyl, atracurium), maintenance – – 30/30
20 μg of fentanyl) (isoflurane, atracurium)
Bessa [25] 2012 90/90 SA at L2–3 interspace (3 mL of hyperbaric bupivacaine 0.5% and Induction (propofol, fentanyl, atracurium), maintenance 4 – 86/90
20 μg of fentanyl) (isoflurane, atracurium)
Ellakany [32] 2013 20/20 CSEb at the 10th thoracic interspace (1 mL of plain bupivacaine Induction (propofol, fentanyl, atracurium), maintenance – – 20/20
0.5% and 25 μg fentanyl) (sevoflurane, propofol)
Imbelloni [26] 2010 35/33 SA at L3–4 interspace (3 mL of hyperbaric bupivacaine and 20 μg Induction (propofol, fentanyl, rocuronium, lidocaine), 1 – 34/33
of fentanyl) maintenance (sevoflurane)
Kalaivani [27] 2014 25/25 SA at L2–3 interspace (3 mL of hyperbaric bupivacaine 0.5% and Induction (propofol, fentanyl, atracurium), maintenance 2 – 23/25
25 μg of fentanyl) (isoflurane)
Mehta [31] 2010 30/30 SA at L3–4 interspace (0.3 mg/kg of hyperbaric bupivacaine 0.5%) Technique not described – – 30/30
Ross [23] 2013 10/10 EAc at T4–6 interspace (3 ml of lidocaine 1.5% as test dose Induction (propofol, fentanyl, lidocaine, rocuronium), – – 10/10
+20–25 mL of lidocaine 2%, through an epidural catheter) maintenance (sevoflurane)
Tiwari [24] 2013 117/118 SA at L3–4 interspace (2.5–3.5 mL of hyperbaric bupivacaine 0.5%) Induction (propofol, rocuronium), maintenance 7d 7 110/114
(N2O, sevoflurane)
Turkstani [30] 2009 25/25 SA at L2–3 interspace (2 mL of bupivacaine 0.5% and 25 μg of fentanyl) Induction (propofol, fentanyl, atracurium), maintenance – – 25/25
(sevoflurane)
Tzovaras [28] 2008 50/50 SA at L2–3 interspace (3 mL of hyperbaric bupivacaine 0.5%, 20 μg Induction (propofol, fentanyl, atracurium), maintenance – 2 49e/48
of fentanyl and 0.25 mg of morphine) (sevoflurane, propofol)
a
Spinal anesthesia.
b
Combined spinal epidural.
c
Epidural anesthesia.
d
Three patients in the NA group required anesthetic and surgical conversion.
e
One patient withdrew informed consent.
 M.A. Longo et al. / Journal of Clinical Anesthesia 41 (2017) 48–54
for anesthetic conversion, so that 3.4% of the patients in the NA group re-
quired conversion to GA.
Regarding the respiratory variables, insufflation of the peritoneal cav-
ity with carbon dioxide increases intra-abdominal pressure, which can
have a significant effect on mechanical ventilation. It decreases diaphrag-
matic excursion and lung compliance, thus resulting in high peak airway
pressures and decreased end expiratory tidal volume, which may lead to
CO2 retention and acidosis [37].
Elevated end-tidal CO2 pressure values are unreliable indicators of
the arterial carbon dioxide tension because such values are usually
only slightly elevated, may be easily misinterpreted as acceptable values,
and may not indicate the true level of hypercarbia [37], which may per-
sist in the recovery room [3].
In our systematic review, most studies did not measure the intraop-
erative variation of respiratory parameters, and those who measured it
were subject to reporting bias for not describing adequately the monitor-
ing used or the results for the variables. Selective reporting occurred in
five studies, three of which described CO2 retention by expired air but
did not describe the CO2 sampling method [24,26,31]. The other two re-
ported the monitoring of these variables but did not present the results
[28,30].
Intraoperative hypotension during LC occurs mainly due to the com-
pression of the inferior vena cava and decreased venous return, thus
resulting in the reduction of cardiac output and hypotension [5].
Performing this procedure under NA leads to a greater concern because
of the effect of sympathetic tonus loss and peripheral vasodilation, thus
worsening the hypotension. In our study, intraoperative hypotension
was 4.6 times more prevalent in the NA group. None of the studies re-
ported systemic repercussion, so that pressure was adequately con-
trolled with fluid infusion and vasopressor agents in all cases.
Bradycardia during LC occurs as a result of a deep vagal response to
rapid peritoneal distension [4]. Decreased heart rate after a high neuraxial
block may occur as a result of the inhibition of T1–4 cardioaccelerator fi-
bers and the reduction of right atrial filling [38]. The prevalence of intra-
operative bradycardia was 6.6 times higher in the NA group in our study.
Postoperative nausea and vomiting after GA is more frequent in rela-
tion to NA because of the greater use of emetogenic agents, such as inha-
lational anesthetics and opioids [39]. In our study, it occurred more
frequently in patients who received GA at a 2.5-fold higher rate.
Postoperative urinary retention is related to the use of medications
and type of surgery [40]. Local anesthetics may slow the return of blad-
der function [41], and opioids promote the inhibition of urination re-
flexes, especially when administered intrathecally, because of the
suppression of detrusor contractility and the reduction of the sensation
of bladder urgency [42]. In our study, the prevalence of postoperative uri-
nary retention did not differ between the two anesthetic techniques.
Sensitivity analysis with the exclusion of a study responsible for the ab-
sence of difference demonstrated a 2.7-fold greater chance of urinary re-
tention in the NA.
The chance of postoperative headache was 4.1 times higher in the NA
group. The studies described it as a typical complication after dural
puncture.
Postoperative pain related to LC is multifactorial, with the predomi-
nance of visceral pain [43]. The NA reduces postoperative pain because
of the existing activity of anesthetics injected at the subarachnoid or epi-
dural space and may act pre-emptively in reducing central sensitization
to noxious stimulus [44]. In our meta-analysis, the postoperative pain in-
tensity was lower in patients receiving NA. Only postoperative pain at 6 h
did not show any difference between the groups. The results were limit-
ed because no trial described all the intervals of postoperative pain and
they presented great heterogeneity.
The review by Yu et al. [10] reported spinal anesthesia as a safe anes-
thetic technique for LC, associated with an intraoperative pain incidence
of 26,1%, less intense postoperative pain at 2–4 h and 6–8 h, but not at
24 h, lower incidence of PONV and higher of urinary retention. Our
study found similar results of intraoperative pain and postoperative
53
outcomes. In addition to these endpoints, our study demonstrated that
neuraxial anesthesia was associated with higher prevalence of intraoper-
ative hypotension and bradycardia, but without systemic repercussion.
Also, our results did not show evidence to support the benefit of NA for
the pulmonary function in patients who submitted to elective LC.
The main limitation of our review was the difficulty in comparing the
prevalence of intraoperative pain between the groups. Although patients
who receive GA did not complain of pain, we could not assume the ab-
sence of pain in the intraoperative period. Therefore we estimated the
pooled prevalence of intraoperative pain only in the NA group. Second,
meta-analyses of hypercapnia and hypoxemia could not be conducted
because of the lack of comparable measures. Also, we could not conduct
sensitivity analyses based on studies with low risk of bias and we could
not eliminate the heterogeneity for the postoperative pain.
In conclusion, this systematic review with meta-analyses showed
that NA as sole anesthetic technique, although feasible for LC, was asso-
ciated with intraoperative pain referred to the shoulder, required anes-
thetic conversion in 3.4% of the cases and did not demonstrate evidence
of respiratory benefits for patients with normal pulmonary function.
Funding
No funding was received.
Conflict of interest
The authors declare no conflicts of interest.
Acknowledgments
Preliminary data for this study were presented as an oral presenta-
tion at the 62nd Brazilian Congress of Anesthesiology, 14–18 November
2015, Brazil.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.jclinane.2017.06.005.
References
[1] Cunningham AJ, Brull SJ. Laparoscopic cholecystectomy: anesthetic implications.
Anesth Analg 1993;76:1120–33.
[2] Desmond J, Gordon RA. Ventilation in patients anaesthetized for laparoscopy. Can
Anaesth Soc J 1970;17:378–87.
[3] Iwasaka H, Miyakawa H, Yamamoto H, Kitano T, Taniguchi K, Honda N. Respiratory
mechanics and arterial blood gases during and after laparoscopic cholecystectomy.
Can J Anaesth 1996;43:129–33.
[4] Myles PS. Bradyarrhythmias and laparoscopy: a prospective study of heart rate
changes with laparoscopy. Aust N Z J Obstet Gynaecol 1991;31:171–3.
[5] Joris JL, Noirot DP, Legrand MJ, Jacquet NJ, Lamy ML. Hemodynamic changes during
laparoscopic cholecystectomy. Anesth Analg 1993;76:1067–71.
[6] Johnson A. Laparoscopic surgery. Lancet 1997;349:631–5.
[7] Yuksek YN, Akat AZ, Gozalan U, Daglar G, Pala Y, Canturk M, et al. Laparoscopic cho-
lecystectomy under spinal anesthesia. Am J Surg 2008;195:533–6.
[8] Sinha R, Gurwara AK, Gupta SC. Laparoscopic surgery using spinal anesthesia. JSLS
2008;12:133–8.
[9] Kumar A. Laparoscopic cholecystectomy under spinal anaesthesia: a prospective
study. Nepal Med Coll J 2014;16:139–43.
[10] Yu G, Wen Q, Qiu L, Bo L, Yu J. Laparoscopic cholecystectomy under spinal anaesthesia
vs. general anaesthesia: a meta-analysis of randomized controlled trials. BMC
Anesthesiol 2015;15:176.
[11] Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for sys-
tematic reviews and meta-analyses: the PRISMA statement. BMJ 2009;339:b2535.
[12] Higgins JP, Altman DG, Gotzsche PC, Juni P, Moher D, Oxman AD, et al. The Cochrane
Collaboration's tool for assessing risk of bias in randomised trials. BMJ 2011;343:
d5928.
[13] Barendregt JJ, Doi SA, Lee YY, Norman RE, Vos T. Meta-analysis of prevalence. J
Epidemiol Community Health 2013;67:974–8.
[14] EpiGear International Pty Ltd. MetaXL 3.0. [software]. Available from: http://www.
epigear.com/index_files/metaxl.html; 2015.
[15] Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median,
range, and the size of a sample. BMC Med Res Methodol 2005;5:13.
54 M.A. Longo et al. / Journal of Clinical Anesthesia 41 (2017) 48–54
[16] Mitchell M, Muftakhidinov B, Winchen T. Engauge digitizer 6.0. [software]. Available
from: http://markummitchell.github.io/engauge-digitizer; 2015.
[17] The Cochrane Collaboration. Review manager (RevMan) 5.3. [software]. Available
from http://community.cochrane.org/tools/review-production-tools/revman-5;
2014.
[18] Schwarzer G. Meta: an R package for meta-analysis. R News 2007;7:40–5.
[19] Viechtbauer W. Conducting meta-analyses in R with the metafor package. J Stat
Softw 2010;36:1–48.
[20] Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a
simple, graphical test. BMJ 1997;315:629–34.
[21] Sterne JA, Sutton AJ, Ioannidis JP, Terrin N, Jones DR, Lau J, et al. Recommendations
for examining and interpreting funnel plot asymmetry in meta-analyses of
randomised controlled trials. BMJ 2011;343:d4002.
[22] Schünemann HJ, Oxman AD, Vist GE, Higgins JP, Deeks JJ, Glasziou P, et al. Chapter 11:
presenting results and ‘Summary of findings’ tables. In: Higgins JP, Green S, editors.
Cochrane handbook for systematic reviews of interventions Version 5.1.0 [updated
March 2011]. The Cochrane Collaboration; 2011 Available from: www.handbook.
cochrane.org.
[23] Ross SB, Mangar D, Karlnoski R, Camporesi E, Downes K, Luberice K, et al. Laparo-en-
doscopic single-site (LESS) cholecystectomy with epidural vs. general anesthesia.
Surg Endosc 2013;27:1810–9.
[24] Tiwari S, Chauhan A, Chaterjee P, Alam MT. Laparoscopic cholecystectomy under spi-
nal anaesthesia: a prospective, randomised study. J Minim Access Surg 2013;9:65–71.
[25] Bessa SS, Katri KM, Abdel-Salam WN, El-Kayal el SA, Tawfik TA. Spinal versus general
anesthesia for day-case laparoscopic cholecystectomy: a prospective randomized
study. J Laparoendosc Adv Surg Tech A 2012;22:550–5.
[26] Imbelloni LE, Fornasari M, Fialho JC, Sant'Anna R, Cordeiro JA. General anesthesia ver-
sus spinal anesthesia for laparoscopic cholecystectomy. Rev Bras Anestesiol 2010;60:
217–27.
[27] V K, Pujari VS, R SM, Hiremath BV, Bevinaguddaiah Y. Laparoscopic cholecystectomy
under spinal anaesthesia vs. general anaesthesia: a prospective randomised study. J
Clin Diagn Res 2014;8 (NC01-4).
[28] Tzovaras G, Fafoulakis F, Pratsas K, Georgopoulou S, Stamatiou G, Hatzitheofilou C.
Spinal vs general anesthesia for laparoscopic cholecystectomy: interim analysis of
a controlled randomized trial. Arch Surg 2008;143:497–501.
[29] Arati S, Ashutosh N. Comparative analysis of spinal vs general anaesthesia for lapa-
roscopic cholecystectomy: a prospective randomized study. Int J Anesth 2010;24
(12p-p).
[30] Turkstani A, Ibraheim O, Khairy G, Alseif A, Khalil N. Spinal versus general anesthesia
for laparoscopic cholecystectomy: a comparative study of cost effectiveness and side
effects. Anaesth Pain Intensive Care 2009;13:9–14.
[31] Mehta PJ, Chavda HR, Wadhwana AP, Porecha MM. Comparative analysis of spinal
versus general anesthesia for laparoscopic cholecystectomy: a controlled, prospec-
tive, randomized trial. Anesth Essays Res 2010;4:91–5.
[32] Ellakany M. Comparative study between general and thoracic spinal anesthesia for
laparoscopic cholecystectomy. Egypt J Anaesth 2013;29:375–81.
[33] Bessa SS, El-Sayes IA, El-Saiedi MK, Abdel-Baki NA, Abdel-Maksoud MM. Laparo-
scopic cholecystectomy under spinal versus general anesthesia: a prospective, ran-
domized study. J Laparoendosc Adv Surg Tech A 2010;20:515–20.
[34] Gurusamy KS, Samraj K, Davidson BR. Low pressure versus standard pressure pneu-
moperitoneum in laparoscopic cholecystectomy. Cochrane Database Syst Rev 2009;
2:CD006930.
[35] Ciofolo MJ, Clergue F, Seebacher J, Lefebvre G, Viars P. Ventilatory effects of laparos-
copy under epidural anesthesia. Anesth Analg 1990;70:357–61.
[36] Pursnani KG, Bazza Y, Calleja M, Mughal MM. Laparoscopic cholecystectomy under
epidural anesthesia in patients with chronic respiratory disease. Surg Endosc 1998;
12:1082–4.
[37] Wittgen CM, Andrus CH, Fitzgerald SD, Baudendistel LJ, Dahms TE, Kaminski DL.
Analysis of the hemodynamic and ventilatory effects of laparoscopic cholecystecto-
my. Arch Surg 1991;126:997–1001.
[38] Pollard JB. Cardiac arrest during spinal anesthesia: common mechanisms and strat-
egies for prevention. Anesth Analg 2001;92:252–6.
[39] Apfel CC, Heidrich FM, Jukar-Rao S, Jalota L, Hornuss C, Whelan RP, et al. Evidence-
based analysis of risk factors for postoperative nausea and vomiting. Br J Anaesth
2012;109:742–53.
[40] Roehrborn CG. Acute urinary retention: risks and management. Rev Urol 2005;
7(Suppl. 4):S31–41.
[41] Matsuura S, Downie JW. Effect of anesthetics on reflex micturition in the chronic
cannula-implanted rat. NeurourolUrodyn 2000;19:87–99.
[42] O'Reilly PH. Postoperative urinary retention in men. BMJ 1991;302:864.
[43] Joris J, Thiry E, Paris P, Weerts J, Lamy M. Pain after laparoscopic cholecystectomy:
characteristics and effect of intraperitoneal bupivacaine. Anesth Analg 1995;81:
379–84.
[44] Kelly DJ, Ahmad M, Brull SJ. Preemptive analgesia I: physiological pathways and
pharmacological modalities. Can J Anaesth 2001;48:1000–10.