Bowel preparation for colonoscopy
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
The primary aim for this review is to determine the colonic cleanliness by using among the various agents and their side effects.
Another aim for this study is to look for timing of bowel preparation which has any influence on colonic cleanliness.
Background
Since 1950s to date, colonoscopy is the gold standard method for assessing the colonic pathologies. With an advent of various tools used in colonoscopy, those help to do therapeutic procedures from simple polypectomy to extensive submucosal resections. Apart from the handling of instrument and its related tools, the main factor governing during the procedure is bowel cleanliness. Good bowel cleanliness will help not only to get accurate diagnostic findings and also to help in the therapeutic safeness. Theoretically, an ideal colonic bowel preparation is mandatory to see the colonic mucosa clearly without visual and histological changes. Bowel preparation must be safe and minimal unwanted effects to the patient including patient daily activities and lastly it should be economical as well. However, currently available bowel preparation agents are still need to meet above mentioned expectations and trials are still carried out to get the one nearest to ideal preparation (Yakut 2010; Beck 2010).
Description of the condition
Poor bowel preparation in colonoscopy can cause incomplete visualization of colon and may lead to missed lesions, procedure failure, prolonged procedure time, and increased procedure complications (Rex 2002). Studies using large database reveals that up to 23% of colonoscopies have sub optimal bowel preparation. There are many reasons for inadequate bowel preparation. Poor compliance from patient aspect likes incomplete consumption of colon preparation is an another strong influence in cleanliness (Hsu 1998). Nowadays there is increase demand for colonoscopy attributable to widespread colorectal cancer screening and surveillance coupled with the introduction of new bowel cleansing preparation.
Description of the intervention
The commonly used bowel cleansing preparations in hospital setting are polyethylene glycol (PEG); sodium phosphate (NaP); magnesium citrate (Mg‐citrate) and sodium pico sulphate, citric acid and magnesium oxide (SPS) containing preparations. There are many trials that compare the efficacy of two or more bowel cleansing preparations for colonoscopy. Study results vary depending on their limitations and strength.
Sodium phosphate, a low volume hyper‐osmotic solution, draws fluid into the intestinal lumen osmotically. It is known to be superior to the polyethylene glycol (PEG) solution, introduced in 1980 by Davis et al. (Davis 1980), both for acceptability by the patients and for the quality of bowel cleansing (Choi 2011). Because of the fluid and electrolytes shift, it is contraindicated in patient with kidney disease, sodium restricted diet, congestive heart failure and advanced liver disease. On the other hand, PEG is iso‐osmolar and it is better to use in patients where NaP cannot be applied.
PEG has been available as bowel preparation agent since 1980s. Preparation containing PEG are large volume, osmotically balanced solution that acts as purgative to cleanse the bowel though ingestion of non absorbable fluid. Its efficacy is better compare with older diet and cathartic preparation, however, it cannot compete with NaP in patients acceptability and tolerability (Bakun 2006). Because of the large volume agent, PEG is contraindicated in patient with ileus, gastric retention, bowel perforation, GI obstruction and severe colitis.
SPS contains sodium picosulphate and magnesium citrate. Sodium picosulphate decreases water and electrolyte absorption and increases intestinal motility and magnesium citrate acts as osmotic laxative. SPS enhances the gastrointestinal peristalsis, mainly used in radiological imaging test of colon. The combination of SPS or Mg citrate with PEG or NaP give well tolerated than using alone. It yields a high percentage of positive rating for efficacy (Love 2009) .There is a study shown that Combination of magnesium citrate with a single dose of sodium phosphate preparation reduce the adverse effect of NaP by reducing the dosage require for the preparation( Choi 2011; Bakun 2006).
How the intervention might work
Apart from the various agent used and patient compliance in bowel preparation, the timing of bowel preparation and colonoscopy is also play a important role in bowel cleanliness. Some of the studies show same day bowel preparation is better than traditional previous day preparation in view of cleanliness, detection of flat lesions and success rate in caecal intubation (Parra‐Blanco 2006).
Why it is important to do this review
Therefore, it is important to review all the currently available agents used in bowel preparation regarding cleanliness and adverse effects as well as to look for the timing of agent before colonoscopy which determines the quality of colonic cleansing and facilitate the procedure.
Objectives
The primary aim for this review is to determine the colonic cleanliness by using among the various agents and their side effects.
Another aim for this study is to look for timing of bowel preparation which has any influence on colonic cleanliness.
Methods
Criteria for considering studies for this review
Types of studies
All randomised controlled trials comparing PEG, NaP and SPS in adults having bowel preparation for colonoscopy within a period from January 1990 to December 2010 will be considered for inclusion. If the studies are not randomised and have allocation bias during randomizations, these studies will be excluded in this review. Again, those do not provide details in patient selection, allocation, study design, outcome and measurement methods will also be excluded. Moreover, studies presented in abstract form without publication of the full paper will not be included in the analysis. There will be no language restriction of identified trials.
Types of participants
All adult patients undergoing elective colonoscopy for a specific clinical indication will be evaluated. Patient who need emergency colonoscopy, pregnant women, patient with previous colectomies and children will be excluded in this review.
Types of interventions
1. Identified trials will be divided into four groups to ensure uniformity in comparison. The four groups are:
(1) PEG compared with NaP
(2) PEG compared with SPS
(3) NaP compared with SPS
(4) combinations of different agents
2. Timing of bowel preparation and effect on colonic cleansing during colonoscopy will be also involved as subgroup analysis.
(1) Same day bowel preparation compared with the day before preparation.
Types of outcome measures
Data on the following outcome measures will be sought. They include efficacy of bowel cleansing, completion of bowel preparation, timing of bowel preparation, adverse events and the biochemical derangement. Studies will be considered for inclusion if they provide information on at least one outcome parameters as listed above. Data from duplicate studies will be analysed once only.
Search methods for identification of studies
All randomised controlled trials will be searched from the MEDLINE and pubMed data bases and the Cochrane Central Register of Controlled Trials between January 1990 and July 2011, including Internet links to the related articles. Search data will include the following key words: bowel preparation, colonoscopy, polyethylene glycol, sodium phosphate, sodium picosulphate and timing of bowel preparation. In addition, a manual search will also be carried out using the references of the retrieved papers to identify additional relevant trials for potential inclusion.
Data collection and analysis
All studies eligible for inclusion will be assessed and appraised by two independent reviewers (Thant Zin and Kye Mon Min Swe). Each study will be evaluated on methodology and randomizations. Study quality will be assessed based on the following parameters: (1) well‐defined population (inclusion and exclusion criteria and sample size), (2) allocation concealment, (3) use of blinding, (4) comparable treatment groups, (5) clearly defined outcome variables and (6) analysis based on "intention‐to‐treat". All data and results of statistical tests will be extracted from the papers onto a data sheet. For particular outcomes that are evaluated, if the data is not specifically reported, it will be regarded as missing and no assumption will be made from the missing data.
Dichotomous data will be tested by calculating the rate difference with their 95% confidence interval. The effect measures will be presented using fixed effect models. Heterogeneity of results and publication bias will be tested for by evaluating the inter‐study variability, using a Funnel plot (Cochrane 2002, Egger 1997, RevMan 2008). As the studies become less precise (i.e. higher standard error), the results (presented in this review as a log odds ratio) of the studies become more variable, scattered to both sides of the more precise larger studies. In this situation, the Funnel plot becomes asymmetrical and does not resemble an inverted funnel. If the Funnel plot don't meet this criteria, we will use a random effects model instead. Statistical analyses and Forest plots will be performed using the build‐in MetaView in the Cochrane software Review Manager 5.1 (RevMan 2008), where the confidence interval (CI) for each study is represented by a horizontal line and the point estimate is represented by a square. The size of the square corresponds to the weight of the study in the meta‐analysis. The confidence interval for summary statistics is represented by a diamond.
If analysis of variables is not possible, for example due to the lack of uniformity of data presentation, presentation of data as median and range does not allow analysis by a Forest plot, as there could be a skewed distribution of data. Thus medians cannot be assumed to be the same as means. Unless raw data or graphs will be provided in the primary trials, no assumption will be made to designate an arbitrary equivalence. Instead, those data in median or range will be presented unchanged separately.
If we identify studies, in which results are presented as scores, we will not include these in the analysis.
