Meta-analysis: Randomized controlled trials of 4-L polyethylene glycol and sodium phosphate solution as bowel preparation for colonoscopy

R. Juluri, G. Eckert, T. F. Imperiale

Research output: Contribution to journalArticle

34 Scopus citations

Abstract

Background Randomized controlled trials (RCTs) comparing polyethylene glycol (PEG) with sodium phosphate (NaP) are inconsistent. Aim To compare the efficacy of and tolerance to PEG vs. NaP for bowel preparation. Methods We used MEDLINE and EMBASE to identify English-language RCTs published between 1990 and 2008 comparing 4-L PEG with two 45 mL doses of NaP in adults undergoing elective colonoscopy. We calculated the pooled odds ratios (ORs) for preparation quality and proportion of subjects completing the preparation. Results From 18 trials (n = 2792), subjects receiving NaP were more likely to have an excellent or good quality preparation than those receiving PEG (82% vs. 77%; OR = 1.43; 95% CI, 1.01-2.00). Among a subgroup of 10 trials in which prep quality was reported in greater detail, there were no differences in the proportions of excellent, good, fair or poor preparation quality. Among nine trials that assessed preparation completion rates, patients receiving NaP were more likely to complete the preparation than patients receiving 4-L PEG (3.9% vs. 9.8% respectively did not complete the preparation; OR = 0.40; CI, 0.17-0.88). Conclusion Among 18 head-to-head RCTs of NaP vs. 4-L PEG, NaP was more likely to be completed and to result in an excellent or good quality preparation.

Original languageEnglish (US)
Pages (from-to)171-181
Number of pages11
JournalAlimentary Pharmacology and Therapeutics
Volume32
Issue number2
DOIs
StatePublished - Jul 15 2010

ASJC Scopus subject areas

  • Hepatology
  • Gastroenterology
  • Pharmacology (medical)

Fingerprint Dive into the research topics of 'Meta-analysis: Randomized controlled trials of 4-L polyethylene glycol and sodium phosphate solution as bowel preparation for colonoscopy'. Together they form a unique fingerprint.

  • Cite this