Regime for Bowel Preparation in Patients Scheduled to Colonoscopy: Low-Residue Diet or Clear Liquid Diet? Evidence From Systematic Review With Power Analysis

Abstract

Clear liquid diet (CLD) is used to perform bowel preparation before colonoscopy traditionally, but several clinical studies indicated that low-residue diet (LRD) generates equal effects to CLD and a conclusive conclusion has not yet been yielded. The systematic review was performed to address this conflict and facilitate informed decision-making eventually. To capture randomized controlled trials (RCTs) comparing LRD with CLD in terms of bowel preparation, a search was performed in PubMed, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL), Science Direct, recent conference abstracts, Google Scholar, and Clinicaltrials.gov through May 2015. We performed all meta-analyses based on fixed- or random-effects model, which is generated from clinical characteristics and methodology. Moreover, the G*Power software was adopted to achieve statistical power for each outcome. In total, we captured 109 potential citations at initial search stage and 2 topic-related articles were included through other sources. After critical appraisal, 7 RCTs were eligible for our inclusion criteria. Meta-analyses generated similar effects in bowel preparation quality, efficacy of colon cleansing, and compliance with recommended dietary regime when LRD versus CLD regime, but patients who were prescribed to receive LRD have slightly better tolerance (RR, 1.06; 95% CI, 1.02-1.11) and tended to repeat the same preparation regime in future (RR, 1.17; 95% CI, 1.09-1.26) relative to patients in CLD. Importantly, both regimes resulted in similar adverse events (AEs). With the best available evidence, LRD could be recommended to be as standard regime for bowel preparation prior to colonoscopy.

Conflict of interest statement

The authors have no funding or conflicts of interest to disclose.

Figures

FIGURE 1

Flow diagram of literature retrieval and selection: 109 potential citations and additional 2 records were initially obtained and eventually 7 eligible studies were incorporated into this systematic review and meta-analysis.

FIGURE 2

Assessment of risk of bias: (A) risk of bias graph, (B) risk of bias summary; the whole quality of all eligible studies was good because most of risk of bias indices were rated as low risk and no index was graded as high risk.

FIGURE 3

Meta-analysis on the quality of bowel preparation (excellent—good preparation): 5 eligible studies including 1171 participants were included and no significant difference for this given outcome was identified based on a fixed-effect model.

FIGURE 4

Meta-analysis on the efficacy of colon cleansing: subgroup analyses according to OBPS and BBPS were not statistically significant. BBPS = Boston Bowel Preparation Scale, OBPS = Ottawa Bowel Preparation Scale.

FIGURE 5

Meta-analysis on patient tolerance to the same diet regime in future: 4 eligible studies which enrolled 995 participants reported this given outcome and patients in LRD diet regime group reported better tolerance based on a random-effect model. LRD = low-residue diet.

FIGURE 6

Meta-analysis on willings to repeat the same diet regime: 3 studies including 921 eligible participants were incorporated and the patients were prescribed to use the LRD diet regime; they were allowed to select the same diet regime if a bowel preparation is needed in future according to the synthesis analysis based on a fixed-effect model. LRD = low-residue diet.

FIGURE 7

Meta-analysis on compliance with recommended diet regime: 3 eligible studies in which 489 participants were recruited provided the essential information on this outcome and the pooled result based on a random-effect model indicated no statistically significant difference.

FIGURE 8

Meta-analysis on overall AEs: 3 studies including 905 were eligible for the inclusion criteria prespecified in our study and the synthesis analysis with fixed-effect model did not indicate statistically significant difference.