Criteria for considering studies for this review
Types of studies
We plan to include both Randomized and non randomised Clinical Trials that compare the use of the planned relaparotomy versus on demand strategy in the management of secondary peritonitis. The non randomised clinical trials might be excluded if a sufficient number of RCT's to address the primary objective of this systematic review will be identified. If the non randomised clinical studies are included, these will be analysed separately from the RCT's. In addition, 30-day mortality rate will be a required item for inclusion of the non-RCT study. In the analysis we will include the studies in which temporary closure of the abdomen or the laparostomy with each technique has been made, provided the relaparotomy has been scheduled and not realized on the basis of clinical, laboratory and radiological findings and then compared with the on demand strategy. No language restriction will be applied.
Types of participants
Patients of any age affected by secondary peritonitis requiring emergency surgical intervention. Secondary peritonitis is defined as any intra-abdominal infection involving the peritoneum caused by the perforation, necrosis, ischemia or infection of any of the gastrointestinal organs or of any other intraabdominal organ included the peritonitis secondary to post-operative complications or peritonitis secondary to compartment syndrome or trauma. Exclusion criteria are patients without radical curative intention/approach at primary surgery and without minimum expected survival after primary surgery, or patients on continuous ambulatory peritoneal dialysis (CAPD).
Types of interventions
Planned strategy relaparotomies versus on-demand strategy.
Planned strategy relaparotomies is decided upon during or immediately after the first-index operation for peritonitis, when the consultant or a multidisciplinary team (van Ruler 2007) decides to re operate within 1-3 days, irrespective of the patient’s immediate postoperative course. The criteria for re-laparotomy in the on-demand strategy are the significant deterioration or lack of improvement and consist in the relaparotomy decision made assessing the worsening or persistence of any signs of sepsis, both clinical (including laboratory results) and radiological, including the unchanged or worsened scores of disease severity, after the primary laparotomy ( van Ruler 2007; Penninckx 1990; Koperna 2000). The following techniques were included as laparostomy - interventions that included the use of temporary mesh, plastic sheet, wound bags, “intraabdominal vacuum techniques” and cases where the abdomen was left open without any form of mesh or bags and studies that used a zipper or Velcro mesh technique to close the abdomen and packing of the intraperitoneal cavity. Important Co-Interventions with important impact on outcomes will be addressed and extracted: timing and modality of antibiotic therapy, fluid management.
Types of outcome measures
The primary outcome will be 30-day mortality
Secondary outcomes are postoperative in hospital complications classified using the Dindo grading system (Dindo 2004). Per patients and per events number of complications were calculated for each Dindo grade including specific abdominal complications of peritonitis-surgery like intra-abdominal abscesses, small bowel fistulae, ventilator-associated complications.Secondary outcomes include also the economic impact of planned vs on-demand strategy. If possible, results from included studies analysing patients with secondary peritonitis and treated by open abdomen techniques will be classified according to the open abdomen classification suggested by Björck 2009 in order to improve the inter-study and prognostic comparability.
Search methods for identification of studies
Searches will be carried out using medical subject headings (MeSH) and free text words in combination with the search strategy for randomised and non-randomized trials reported.
The proposed MEDLINE, EMBASE and Cochrane library strategies is given below and will be modified when applied to other databases.
[MP=title, abstract, original title, name of substance word, subject heading word, protocol supplementary concept, rare disease supplementary concept, unique identifier]
1. exp Laparotomy/
2. exp Abdomen/su [Surgery]
3. exp Peritoneal Lavage/
4. (etappenlavage or laparotom* or abdom* surger* or abdom* operation).mp.
5. 1 or 2 or 3 or 4
6. exp Reoperation/
7. exp Recurrence/
8. (relaparotom* or reoperation or repeat surgery or surgical revision).mp.
9. 6 or 7 or 8
10. exp Peritonitis/
11. exp Intestinal Perforation/co [Complications]
12. (peritonitis or peritone* or (abdom* adj3 infection*)).mp.
13. 10 or 11 or 12
14. 5 and 9 and 13
[mp=title, abstract, subject headings, heading word, drug trade name, original title, device manufacturer, drug manufacturer, device trade name, keyword]
1. exp laparotomy/
2. exp abdominal surgery/
3. exp peritoneum lavage/
4. (etappenlavage or laparotom* or abdom* surger* or abdom* operation).mp. [mp=title, abstract, subject headings, heading word, drug trade name, original title, device manufacturer, drug manufacturer, device trade name, keyword]
5. 1 or 2 or 3 or 4
6. exp reoperation/
7. exp recurrent disease/
8. (relaparotom* or reoperation or repeat surgery or surgical revision).mp. [mp=title, abstract, subject headings, heading word, drug trade name, original title, device manufacturer, drug manufacturer, device trade name, keyword]
9. 6 or 7 or 8
10. exp peritonitis/
11. exp intestine perforation/co [Complication]
12. (peritonitis or peritone* or (abdom* and infection*)).m_titl.
13. 10 or 11 or 12
14. 5 and 9 and 13
#1 MeSH descriptor Laparotomy explode all trees
#2 MeSH descriptor Abdomen explode all trees with qualifier: SU
#3 MeSH descriptor Peritoneal Lavage explode all trees
#4 (etappenlavage or laparotom* or abdom* surger* or abdom* operation):ti,ab,kw
#5 (#1 OR #2 OR #3 OR #4)
#6 MeSH descriptor Reoperation explode all trees
#7 MeSH descriptor Recurrence explode all trees
#8 (relaparotom* or reoperation or repeat surgery or surgical revision):ti,ab,kw
#9 (#6 OR #7 OR #8)
#10 (#5 AND #9)
#11 MeSH descriptor Peritonitis explode all trees
#12 MeSH descriptor Intestinal Perforation explode all trees with qualifier: CO
#13 (peritonitis or peritone* or (abdom* near3 infection*)):ti,ab,kw
#14 (#11 OR #12 OR #13)
#15 (#10 AND #14)
We will search for published and unpublished randomised controlled trials and non randomised controlled trials without language restrictions using the following electronic databases: Cochrane Central Register of Controlled Trials (CENTRAL), Cochrane Colorectal Cancer Group specialised register (September 2000), MEDLINE (1966 onwards), EMBASE (1980 onwards), CINAHL (1982 onwards), LILACS (Literatura Latinoamericana y del Caribe en Ciencias de la Salud), GoogleScholar (http://scholar.google.com), Web of Science and ClinicalTrials.gov.
Searching other resources
We will contact the authors of primary trials, pharmaceutical and device companies to identify further published and unpublished studies.
Data collection and analysis
Two reviewers (ST and RC) will independently assess the risk of bias for each included RCTs using “The Cochrane Collaboration’s Tool for Assessing Risk of Bias”. Items that will be assessed include: the generation of randomizations sequence, adequacy of allocation concealment, incomplete outcome data, selective outcome reporting and the adequacy of follow up. Given the difficulty of blinding of an invasive procedure, blinding for intervention will not considered in the assessment of bias risk.
Due to the fact that the primary outcome for this study is 30-day mortality and this represent an hard outcome, minimally affected by performance and detection bias, blinding of outcome assessment and data assessors will not be considered in the assessment of risk of bias. Each domain for each study will be described by "High", "Low" or "unclear" if a risk of bias is thought to be present, absent or uncertain respectively. Scoring approaches will not be used. The result of risk of bias assessment will be summarised in a Risk of Bias table.
Observational studies will be assessed for methodological quality using the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations. This is a checklist of 22 items useful in the assessment of methodological quality of cohort, case–control and cross-sectional studies (Vandenbroucke 2007).
Selection of studies
The titles and abstracts of all potentially relevant trials will be screened by two reviewer (ST and RC). The full text articles of relevant or potentially relevant studies will be obtained. Based on the full text article, two reviewers (ST and NG) will independently determine whether the study meets the inclusion/exclusion criteria. A third author (NG) will independently assess any discordant results. Discrepancies will be resolved through discussion. Agreement between the authors for selected studies will be reported.
Data extraction and management
Trial data will be extracted by each author (ST and CB) independently. Any discrepancies will be resolved by a third author (CR). For trials that include patients with multiple diagnoses, data will be extracted for patients with CRLMs. Primary authors will be contacted in order to clarify any issues or obtain additional or missing data, as required.
The following data will be extracted using standardised forms:
1. General study information
title, authors, contact address
year of publication
2. Trial characteristics
method of randomisation/study design
inclusion and exclusion criteria for patient recruitment
withdrawals and losses to follow-up
3. Study details
patient characteristics including age range, sex ratio, APACHE II score, MPI score
specific diagnosis leading to the surgical intervention
number of patients assigned to each treatment group
relaparotomy strategy for on demand and planned relaparotomy
criteria for deciding relaparotomy in the on demand group
details of intervention techniques
number of patients who died in hospital
definition and number of postoperative in hospital complications (retrieved separately per patients and per events)
length of hospital stay, of ICU and surgery department stay related to each intervention
number of patients death after the hospital discharge during the follow up time
data on quality of life after the hospital discharge using a validated scale
direct and indirect cost of two intervention
Assessment of risk of bias in included studies
The risk of bias in included RCTs will be assessed using the Cochrane Collaboration's tool Cochrane handbook chapter 8. This involves consideration of six features: sequence generation, allocation sequence concealment, blinding, incomplete outcome data, selective outcome reporting and ‘other’ potential sources of bias (Higgins 2011).
Measures of treatment effect
We will perform the analysis using Review Manager 5 statistical package recommended by the Cochrane Collaboration and using chapter 9 of Cochrane handbook for Systematic reviews of intervention as a guide (Higgins 2011).
Statistical analysis for dichotomous variables will be performed by using the odds ratio (OR) as the smmary statistics and for time to event analysis by calculating the hazard ratio (HR).
We will report study results with their associated 95% confidence intervals
Unit of analysis issues
All statistical analyses will be performed on an intention-to-treat basis. Analysis will be performed on a per patient and a per events basis. Statistical methods will include analysis using both relative risk and odds ratio. Fixed effects and the random effects models will both be tested. If these results are not significantly different, the fixed effects model will be reported.
Data will be displayed with Forest plots. Statistical significance will be set as P = 0.10 for testing heterogeneity, and P = 0.05 for all other analyses.
Dealing with missing data
The reference lists of all relevant studies found will be screened, and missing data will be sought by contacting the primary authors. Furthermore, pharmaceutical and device companies will be contacted to identify further published and unpublished studies.
Assessment of heterogeneity
The Q test will be used to test for heterogeneity between studies. Heterogeneity will be quantified also using I2 statistics.
Assessment of reporting biases
The publication bias will be assessed by using the Funnel plot.
Review Manager 5.1 will be used to analyse all primary and secondary outcome data. Meta-analysis will be only performed if all data will be comparable both clinically and methodologically. If included studies are not sufficiently homogeneous to combine in a meta-analysis, the results of included studies will be displayed in a forest plot.
Subgroup analysis and investigation of heterogeneity
In order to investigate the effect of treatment in patients with different degrees of severity of peritonitis we per formed a subgroup analysis considering separately each outcomes in patients with an APACHE II score less than 10 and greater than 10 and for the ones with APACHE II ranging from 10 to 19 and greater than 19. A subgroup analysis was performed also for considering the outcomes of interest in patients younger than 18 years and older then. Where possible a subgroup analysis considering the different closed and open abdomen groups will be realized for each considered outcomes.
Sensitivity analysis are planned to examine the influence of each individual trial and the influence of trial quality on the results of the meta-analysis. Publication bias will to be examined using funnel plots.