Description of the condition
Gastrointestinal (GI) disorders are the most common complications after abdominal surgery, contributing to a prolonged duration of hospital stay in 50% of patients (Guerrero 1999). Postoperative GI disorders mainly manifest as abdominal distension, inability to tolerate an enteral diet (either by mouth or via a feeding tube), nausea, vomiting, postoperative ileus, anastomotic leakage and wound infection (Mythen 2009). These disorders, regardless of whether they involve minor GI discomfort or are more serious, can lead to increased mortality, prolonged hospital stay and increased medical costs (Lang 2001). Therefore, it is important to employ effective perioperative interventions to prevent the occurrence of postoperative GI disorders.
Description of the intervention
The goal-directed approach to fluid management involves determining the optimum haemodynamic parameters for an individual patient, instituting a series of interventions (typically an intravenous fluid bolus) and then measuring the response in the patient to determine the appropriate course of action. This strategy is repeated until haemodynamic parameters have reached the goals set ( Table 1). For example, after a patient is given a 200mL bolus of intravenous fluid over 10 minutes, a rapidly increased stroke volume (SV) of > 10% indicates a low preload. This suggests that the relationship between preload and SV is in the ascending segment of the left ventricular performance curve in accordance with the classic Frank-Starling law of the heart (Lopes 2007). (The Frank-Starling Law states that SV of the heart increases in response to an increase in the volume of blood filling the heart when all other factors remain constant.) In such a case, repeated fluid boluses can be given until the SV increases by < 10%, indicating that the relationship between preload and SV is close to or at the plateau of the left ventricular performance curve. Bolus therapy can then be terminated because optimal SV has been achieved through increases in preload.
The key feature of this management approach is individualization of fluid management according to patient responsiveness to fluid administration during the perioperative period, rather than according to a predetermined formula.This approach effectively protects against both volume depletion and overload during the perioperative period and contrasts with traditional or non－goal-directed fluid therapy, which involves perioperative fluid administration for maintenance, replacement for deficits and “third space” fluid loss. In some cases, traditional fluid therapy can be guided by the use of routine monitoring indicators such as mean arterial pressure (MAP), central venous pressure (CVP) and urinary output (Miller 2009). These traditional approaches to fluid replacement in patients undergoing major operative procedures, on the other hand, do not take into account the specific surgical procedures or anaesthetic techniques and anaesthetic agents used.
How the intervention might work
Goal-directed fluid management is designed to improve the microcirculation and increase tissue perfusion by optimizing the fluid load in the circulatory system and stabilizing haemodynamic parameters, thereby optimizing the patient's systemic oxygen delivery (Donati 2007; Wakeling 2005). This approach uses the concept of the Starling mechanism, whereby the response in haemodynamic parameters to a fluid bolus will vary according to the adequacy of fluid loading. Seeking the optimum SV point of the left ventricular performance curve is an approach designed to effectively protect against both volume depletion and overload during the perioperative period. For patients undergoing GI surgery, accurate evaluation of fluid status by standard approaches is difficult to achieve because of pathophysiological changes of the disease itself, preoperative bowel preparation, intraoperative fluid loss and the stress response.
The pathogenesis of GI tract complications is multifactorial, and gut hypoperfusion plays a key role. Although the body can normally compensate for a 25% to 30% decrease in blood volume without changes in heart rate (HR) and blood pressure (BP), compromise in GI tract perfusion can nevertheless occur with as little as a 10% to 15% reduction in intravascular volume (Hamilton-Davies 1997).
Therefore, hypoperfusion of the GI tract cannot be reliably determined through routine monitoring such as HR and BP, and the GI tract may be injured during prolonged periods of hypoperfusion that may not be readily apparent. Goal-directed fluid therapy is designed to avoid this situation and reduce the incidence of significant GI complications following surgery and anaesthesia.
Why it is important to do this review
Fluid therapy is an important measure in stabilizing vital signs and avoiding tissue hypoperfusion during the perioperative period, especially in patients undergoing GI surgery. Intraoperative tissue hypoperfusion increases the incidence of postoperative complications and mortality and should be avoided wherever possible (Davies 2004). However, individualized, timely and accurate fluid therapy programs for patients undergoing GI surgery have yet to achieve wide recognition. A number of studies suggest that goal-directed fluid management can stabilize a patient's haemodynamics, improve tissue oxygenation and facilitate both GI recovery and a favourable postoperative outcome (Donati 2007; Gan 2002; Hiltebrand 2009). Unfortunately, these studies included a relatively small number of patients, resulting in a lack of high-quality evidence for the effectiveness of goal-directed fluid management.
Some attempts have been made to summarize the evidence in this area, but none have been comprehensive. Giglio 2009 published a systematic evaluation of goal-directed fluid therapy for patients undergoing major surgery that focused only on liver and GI complications. These authors did not evaluate other outcomes such as pulmonary oedema and the amount of fluid infused. Hamilton 2011 published a systematic evaluation of preemptive haemodynamic fluid therapy; however, these authors included participants who were not undergoing major abdominal surgery, and the outcomes did not include postoperative GI complications.
To assess whether goal-directed fluid management reduces GI complications in adult patients after major abdominal surgery .
Criteria for considering studies for this review
Types of studies
We will include randomized controlled clinical trials (RCTs), quasi-randomized trials (e.g. sequence generated by date of birth or clinical record number) and cluster-randomized trials.
Types of participants
We will include participants aged 18 years and older undergoing major open abdominal surgery(including bowel resection, gastric resection, liver resection, oesophageal resection, Whipple procedures and similar significant GI procedures).
We will exclude participants who are subject to emergency surgery, low-risk surgery and surgery restricted to management of minor problems and injuries.
Types of interventions
We will include trials that compare goal-directed fluid management with standard fluid management during the intraoperative period.
We will define goal-directed fluid management as the determination of the optimum haemodynamic parameters for an individual participant, followed by the institution of a series of interventions (typically an intravenous fluid bolus) and measurement of response to determine the subsequent appropriate course of action. This strategy is repeated until haemodynamic parameters have reached set goals.
We will consider traditional or non－goal-directed fluid management to be based on the use of routine monitoring indicators, including HR, MAP and CVP.
Types of outcome measures
- Proportion of participants with any one of the following GI complications: postoperative ileus lasting >24 hours; anastomotic leak requiring reoperation; postoperative nausea and vomiting (PONV) requiring two or more than two antiemetic interventions.
- Mortality－all causes.
We will divide into short-term, medium-term (90 days after postop) and long-term (six months after postop) follow-up data.
- Total volume of intravenous fluid (ml).
- Proportion of participants with a significant rate of adverse events of therapy (myocardial infarction, cerebrovascular accident or central pontine myelinolysis).
- Proportion of participants with pulmonary oedema requiring specific therapy.
- Proportion of participants with acute kidney failure defined as an abrupt decline in glomerular filtration rate (GFR) resulting from Ischaemic or toxic injury to the kidney requiring specific therapy.
We will divide into short-term, medium-term (90 days after postop) and long-term (six months after postop) follow-up data.
Search methods for identification of studies
We will search the following databases: the current issue of the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, see Appendix 1); MEDLINE (1966 to date, see Appendix 2); EMBASE (1980 to date); CINAHL (1982 to date) and the Web of Science (1945 to date). We will focus on RCTs and quasi-randomized trials while comparing goal-directed fluid management with standard (routine) fluid management in perioperative adult participants undergoing major abdominal surgery.
We will not impose a language restriction.
Searching other resources
We will make efforts to contact experts in the field to identify unpublished research and any current trials. We will collect data from unpublished and ongoing trials by contacting the trial investigators.
We will search the following databases for ongoing trials:
- Meta Register of Controlled Trials;
- The National Research Register; and
Data collection and analysis
Selection of studies
We (HG, H-CW) will independently screen the titles and abstracts of all publications identified by the searches to assess their eligibility. We (HG, H-CW) will document the reason(s) for excluding any publications. We will resolve between－review author disputes by consulting with a third review author (HZ). We (HG) will contact the first author of all selected publications to obtain detailed information about publications. We will collect, process and code these data from each publication using a unique identifier on a Form for Eligible Trials for those included (see Appendix 3).
Data extraction and management
We (J-RY and Y-NY) will independently extract data from the included trials and then will record the data on a Data Extraction Form (see Appendix 3). We will conduct cross-checking to ensure the accuracy of the extracted data.
Assessment of risk of bias in included studies
We (HG and H-CW) will assess the risk of bias of each included study independently according to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will resolve disputes by consulting with a third review author (HZ). We will assess the risk of bias in included studies on the basis of the following domains. If all of the following criteria are assessed as adequate, we will consider a study as having a low risk of bias. If one or more of the following criteria is not assessed as adequate, we will consider a study as having a low risk of bias.
Random sequence generation
We will consider random sequence generation to be adequate if the sequence is generated using a computer or a random number table algorithm. We will consider assignment according to birthday or participant hospital number to be inadequate. If the methods of randomization used are not mentioned in the study, we will consider the method of randomization to be unclear.
We will consider allocation of concealment to be adequate if participants and observers are unaware of the allocation of the next participant to be enrolled in the study. Acceptable systems of allocation include central randomization (including telephone-, network- and pharmacy-controlled randomization), sealed opaque envelopes and an on-site locked computer. If the patients or the participants are aware of the grouping (e.g. alternate medical record numbers, reference to case record numbers or date of birth, open allocation sequence or unsealed envelopes), we will consider there to be a high risk of bias. If the means of concealment are not mentioned in the study or are described in insufficient detail to allow for assessment of the effectiveness of the allocation concealment, we will consider the method of concealment to be unclear.
We will consider blinding to be adequate if both participant and observer are blind, unclear if the authors say nothing and inadequate if the authors say that no blinding of participant or observer occurred. We will complete a risk of bias table for each eligible study and outcome using the categories low, high and unclear risk of bias. For each outcome, we will present the summary risk of bias assessments within domains in risk of bias graphs or figures and across domains in the summary of findings table.
Measures of treatment effect
We will use relative risks (RRs) with 95% confidence intervals (CIs) for dichotomous data. Additionally, we will use mean difference (MD) with 95% CI for continuous data where we find data provided as mean and standard deviation.
Unit of analysis issues
For cluster-randomized trials, we will conduct the analysis at the same level as the allocation, using a summary measurement from each cluster. For multiple treatments, we will ensure that the numbers of participants, not the numbers of treatment attempts, are randomly assigned.
Dealing with missing data
Whenever possible, we will use an intention-to-treat (ITT) analysis. If it is unclear whether the missing data are the result of participant withdrawal or dropout before data collection, we will contact the authors to seek the missing information. In cases with unresolved missing data, we will perform a sensitivity analysis using a best-case and worst-case approach to assess their relative effect in our analyses.
Assessment of heterogeneity
We will assess the clinical heterogeneity of included studies as clinical diversity (e.g. different types of surgical procedures, different types of interventions, different types of participant characteristics).
We will assess statistical heterogeneity for included studies. We will use standard the I
Assessment of reporting biases
We will assess reporting biases in a qualitative manner using a funnel plot if more than ten studies are included in the meta-analysis. Funnel plot asymmetry may exist for the following reasons: publication bias, language bias, time lag bias, multiple (duplicate) publication bias, location bias, outcome reporting bias, etc.
If no important clinical heterogeneity is noted, we will synthesize the results in a meta-analysis. We will synthesize data according to major open abdominal surgery (e.g. intestinal resection, gastric resection, liver resection, stomach resection, oesophageal resection, Whipple) and types of interventions (e.g. goal-directed fluid management, standard fluid management). We will use a fixed-effect model if an I
Subgroup analysis and investigation of heterogeneity
When appropriate, after consideration of statistical and clinical heterogeneity, we will perform subgroup analysis based on the following:
- Surgery type;
- Different types of goal-directed fluid management; and
- Short-term, medium-term (90 days after postop) and long-term (six months after postop) follow-up data.
We will consider a sensitivity analysis for study quality by assessing the effect on our overall estimate of the removal of trials deemed to be at high risk of bias. Where necessary, we will also perform sensitivity analysis to assess the impact of missing data using best-case and worst-case imputation for the missing data.
We will exclude and then will include any study that appears to have a large effect size (often the largest or the earliest study) to assess its impact on the meta-analysis.
We will exclude any studies of low quality or high weight. If the data of the results are not changed, we will consider that the evidence is robust.
Summary of findings
We will use the principles of the GRADE system (Guyatt 2008) to assess the quality of the body of evidence associated with the following specific outcomes in our review:
- Proportion of participants with any one of the following GI complications:postoperative ileus, anastomotic leakage, PONV and wound infection;
- Mortality－all causes;
- Total volume of intravenous fluid (ml);
- Case rate of adverse events (myocardial infarction, cerebrovascular accident and central pontine myelinolysis);
- Case rate of pulmonary oedema; and
- Case rate of acute kidney failure (acute renal failure (ARF) in the critical care setting is defined as an abrupt decline in glomerular filtration rate (GFR) resulting from ischaemics or toxic injury to the kidney).
We will construct a summary of findings (SoF) table using the GRADE software. The GRADE approach appraises the quality of a body of evidence based on the extent to which one can be confident that an estimate of effect or association reflects the item being assessed. Determining the quality of a body of evidence takes into consideration within-study risk of bias (methodological quality), directness of the evidence, heterogeneity of the data, precision of effect estimates and risk of publication bias.
We thank the Cochrane Anaesthesia Review Group for their kind technical assistance with the protocol for the review.
We would like to thank Mike Bennett (content editor), Marialena Trivella (statistical editor) and TJ Gan, Richard L Nelson and Luzius Hiltebrand (peer reviewers) for their help and editorial advice during the preparation of this protocol for the systematic review.
Appendix 1. CENTRAL search strategy
#1 goal-direct* or (goal direct*) or GDT or ((fluid or hemodynamic) near management)
#2 MeSH descriptor Digestive System Surgical Procedures explode all trees
#3 MeSH descriptor Gastrointestinal Diseases/ explode all trees
#4 ((gastro* or abdomen*) near (surg* or operat* or complicat*))
#5 (#2 OR #3 OR #4)
#6 (#1 AND #5)
Appendix 2. OVID MEDLINE search strategy
1. goal-direct*.af. or (goal direct*).af. or GDT.mp. or ((fluid or hemodynamic) adj3 management).mp.
2. exp Digestive System Surgical Procedures/ or Gastrointestinal Diseases/ or ((gastro* or abdomen*) adj3 (surg* or operat* or complicat*)).mp.
3. 1 and 2
Appendix 3. Data collection form
Goal-directed fluid management for reduction of gastrointestinal complications in major surgery
EXCLUDE: Reason for exclusion
DO NOT PROCEED IF STUDY EXCLUDED FROM REVIEW
Will the NICOM monitor be included in review?, 9 October 2014
Are you including the NICOM monitor for non-invasive cardiac output?
The NICOM™ (Cheetah Medical, Washington, DE) is a noninvasive cardiac output monitoring device based on chest bioreactance that has been validated in clinical practice. The NICOM requires the connection of 4 double electrode stickers symmetrically placed on the thorax. The upper electrode pair delivers a small alternating current, and the lower pair analyses the variation in the frequency spectra of the delivered current (bioreactance). The time delay between the applied current and the measured voltage is correlated with cardiac stroke volume and allows the monitoring of cardiac output.
We will include, in our review, trials that compare goal-directed fluid management with standard fluid management during the intraoperative period. We define goal-directed fluid management as the determination of the optimum haemodynamic parameters for an individual participant, followed by the institution of a series of interventions (typically an intravenous fluid bolus) and measurement of response to determine the subsequent appropriate course of action.
The NICOM monitor could be used as the monitoring technique for a goal-directed fluid management protocol. Any such devices are included in our stated methods. As such, any RCT using the NICOM for this purpose would be included in our review
Feedback posted by: firstname.lastname@example.org
I certify that I have no affiliations with or involvement in any organization or entity with a financial interest in the subject matter of my feedback.
Reply: Hai Guo (author of Cochrane protocol)
Contributions of authors
Conceiving the review: Hong Zheng (HZ).
Co-ordinating the review: HZ.
Undertaking manual searches: Yi-Ning Yang (Y-NY).
Screening search results: Hai Guo (HG), Hui-Cai Wang (H-CW), HZ.
Organizing retrieval of papers: HG.
Screening retrieved papers against inclusion criteria: HG, H-CW, HZ.
Appraising quality of papers: HG, H-CW, HZ.
Abstracting data from papers: Jian-Rong Ye (J-RY), Y-NY.
Writing to authors of papers for additional information: HG, Lin Chen (LH).
Providing additional data about papers: HG, LH.
Obtaining and screening data on unpublished studies: H-CW, HG.
Data management for the review: J-RY, Y-NY.
Entering data into Review Manager (RevMan 5.1): H-CW, J-RY, Y-NY.
RevMan statistical data: H-CW, J-RY, Y-NY.
Other statistical analysis not using RevMan: H-CW, J-RY, Y-NY.
Interpretation of data: HG, H-CW, J-RY, Y-NY.
Statistical inferences: HZ, HG.
Writing the review: HG, Lena S Sun (LSS).
Securing funding for the review: HG.
Performing previous work that was the foundation of the present study: HZ.
Guarantor for the review (one author): HZ.
Person responsible for reading and checking review before submission: HZ, LSS.
Declarations of interest
Hong Zheng: none known.
Hai Guo: none known.
Lena S Sun: none known.
Jian-Rong Ye. none known.
Lin Chen: none known.
Hui-Cai Wang: none known.
Yi-Ning Yang: none known.
Sources of support
- No sources of support supplied
- The Doctoral Program of Higher Education,number:20126517110005, China.