Complications of low compared to standard pneumoperitoneum pressures in laparoscopic surgery for benign gynecologic pathology: a systematic review protocol

The methodology of this protocol follows the recommendations of the Cochrane Collaboration
group 9] and is reported according to the preferred reporting items for systematic reviews
and meta-analyses (PRISMA) group 10] (see Additional file 1).

Eligibility and criteria

All randomised controlled trials comparing at least two different intra-peritoneal
pressure levels during gynecologic laparoscopy and reporting data on any complication,
length of surgery, or length of hospital stay will be eligible (Additional file 2). No restriction based on language or publication date will be applied. Considering
the need for more invasive surgery during the staging of potentially metastatic diseases
and the obvious increased morbidity associated with these diseases, trials including
more than 20 % of malignant indications for laparoscopy will be excluded. A malignant
indication is defined as a potentially metastatic disease.

Information sources

The search will be conducted throughout Medline (via PubMed), Embase, and the Cochrane
Library from their inception up to a maximum of 6 months before the submission of
the manuscript for publication. The reference lists of the selected publications will
be sought to possibly identify more trials and systematic reviews on the subject.
We will also search randomised controlled trials registries (www.isrctn.com, clinicaltrials.gov) in order to identify additional studies.

Search strategy

An information specialist was consulted to elaborate the search strategy, and all
authors reviewed it. The strategy used for Medline is available in Additional file
3. Keywords and index terms related to laparoscopy and artificial peritoneal pressures
were used. Validated filters 11], 12] are used to identify randomised controlled trials. We will use EndNote X7.3 to manage
the references and eliminate duplicates.

Study selection

Two authors will independently screen all references according to the eligibility
criteria. The first selection step will consist in a review of titles and abstracts.
Full text of remaining references will subsequently be revised to assess eligibility.
When necessary, a third reviewer will be involved to solve any disagreement. A PRISMA
flow diagram will be provided.

Data collection process

Two authors will independently use a standardised data extraction sheet. A pilot version
is available (see Additional file 4). We will test it with two (2) randomly selected included publications and refine
it accordingly. Authors will be contacted if relevant data is missing. When necessary,
a third reviewer will be involved to resolve any disagreement. In case of double publication
of the same study, data will be combined to consider as one study.

Data items

Data that will be extracted from each trial include characteristics of the trial study
design, characteristics of the participants (age, body mass index, surgery indication),
characteristics of the intervention (type of surgery, intra-peritoneal pressures,
positioning of the patient, emergency, or elective surgery), and outcomes (blood loss,
hemorrhage, organ injury, open surgery conversion, anesthetic complications, re-intervention,
re-admission, mortality, length of surgery, length of hospital stay, post-operative
pain, conversion from low to standard pressure, and any other complication reported
by authors). Intra-peritoneal pressures will be classified as low (12 mm Hg) or standard
(?12 mm Hg).

Risk of bias assessment

To assess the risk of bias in individual studies, the Cochrane Collaboration’s tool
for assessing risk of bias 13] will be used regarding the primary outcome (as shown in the data collection sheet,
Additional file 4). Two reviewers will independently apply the tool to each study included in the review,
and a third reviewer will be involved if needed to solve any disagreement. We will
consider a study at high risk of bias when one criterion or more is of “unclear risk”
or “high risk”. Since blinding of personnel is not possible intra-operatively, this
aspect will not be considered in the overall risk of bias. A sensitivity analysis
will be performed to compare the low risk of bias to those associated with a “high
risk” or “unclear risk” of bias.

Outcomes

As any type of complications reflect an uncertainty regarding the safety of an intervention
and assuming variations in the reporting of complications, we considered the frequency
of any intra- or post-operative complications as our primary outcome. This outcome
variable will be a composite measure of the number of patients with any complications
or else with at least one specific complication reported by authors. The secondary
outcomes will be specific-patient-oriented complications, i.e., hemorrhage, organ
injury, open surgery conversion, anesthetic complications (such as arrhythmia, severe
hypotension, severe oxygen desaturation, etc.), mortality, blood loss, post-operative
pain, post-operative re-intervention, re-admission after discharge from hospital,
and length of surgery, length of hospital stay, quality of exposure, and need for
rising intra-peritoneal pressure level.

Summary measures

Frequencies of complications as a composite issue according to intra-peritoneal pressure
levels will be compared through risks ratios and their 95 % confidence intervals (CI).
We will also compute risk ratios for each type of complication (including hemorrhage,
organ injury, open surgery conversion, anesthetic complications, re-intervention,
re-admission, mortality, other reported complications). The continuous variables such
as length of surgery, length of hospital stay, and blood loss will be evaluated with
mean differences and their 95 % CI. Post-operative pain being possibly measured on
various scales, standardised mean differences, and their 95 % CI will be used to evaluate
this outcome.

Synthesis of results

Analyses will be conducted in an intent-to-treat approach. Meta-analyses will be performed
using random-effects models within the Cochrane statistical package RevMan5 (Computer
program, Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration,
2014). Risk ratios and their 95 % CI will be pooled using Mantel-Haenszel method.
Mean differences and standardised mean differences will be pooled with inverse variance
method. In case of rare events, as expected for mortality, analysis will be done using
Peto odds ratios. Heterogeneity between included trials will be assessed with the
I²14].

Additional analysis

After discussions with field experts, we plan to conduct sensitivity and subgroup
analysis determined a priori, to explore sources of statistical heterogeneity. Sensitivity
analyses according to studies with low risk of bias will be conducted. It is known
and understood that the Trendelenburg position, obesity, and the pneumoperitoneum
can affect the anesthetic management during abdominal laparoscopy 15]–17]. The Trendelenburg position is usually used during gynecologic surgery so that bowels
move toward the head of the patient, thus out of the pelvis, allowing better visualisation
and security. For this reason it is possible that all studies found report surgeries
performed in that position, but if not, it could influence the effect measure. Also,
because of the often vital situation in which they take place, emergency surgeries
are more prone to complications and higher pressure levels might be more beneficial
than in elective surgery. In addition, as mini-laparoscopy uses smaller trocarts and
instruments, it might have an influence on the association between pressure levels
and complications. Therefore, we also plan to conduct subgroup analyses, if sufficient
data are available, according to Trendelenburg position vs. other position, mini-laparoscopy
vs. other, emergency vs. elective surgery, body mass index 30 vs. body mass index
?30. Furthermore, additional analyses are planned to compare high intra-peritoneal
pressure level (?15 mm Hg) vs. standard (12 to 14 mm Hg) and high intra-peritoneal
pressure level (?15 mm Hg) vs. low (12 mm Hg), if data are available.

Risk of bias across studies

We plan to use the GRADE approach to assess the quality of evidence of this review.
The GRADEprofiler (Computer program on www.gradepro.org. McMaster University, 2014) will help us judge the strength of our recommendations.