Sodium Phosphate Is Superior to Polyethylene Glycol in Constipated Patients Undergoing Colonoscopy: A Systematic Review and Meta-Analysis
Abstract
Background
Constipation is an important and highly prevalent predictor of inadequate bowel preparation during colonoscopy. In North America, between 2 and 28% of the general population suffer from constipation. Despite this high prevalence, to our knowledge, no meta-analysis on the optimal bowel preparation for constipated patients has been performed. We aimed to systematically review the literature to determine the ideal bowel preparation regimen for patients with chronic constipation.
Methods
A comprehensive search of electronic databases was performed to include studies assessing bowel preparation quality in constipated patients receiving different agents prior to colonoscopy. The primary outcome was colon cleanliness, with secondary outcomes including tolerability of the bowel preparation and serious adverse events.
Results
Preliminary database search yielded 1581 articles after duplicates were removed. After screening, eight studies meeting inclusion criteria were included for qualitative synthesis. Three randomized controlled trials with a total of 225 eligible constipated patients were included for meta-analysis. Of these, 107 patients received sodium phosphate (NaP) and 118 received polyethylene glycol (PEG). Patients receiving NaP before colonoscopy had a higher chance of a successful bowel preparation than patients receiving PEG (OR 1.87, CI 1.06 to 3.32, P = 0.003). Two studies found NaP had greater tolerability, while one study found PEG resulted in better tolerability.
Conclusions
In chronically constipated patients undergoing colonoscopy, the use of NaP may result in superior colonic cleanliness compared to PEG, although the quality of evidence was low. Further high-quality studies are needed to delineate the optimal bowel preparation for patients with constipation.
Introduction
Colonoscopy is a vital diagnostic and therapeutic tool for gastrointestinal disease and is recognized as the most effective screening method for pre-malignant and malignant colorectal lesions. The success of colonoscopy depends on the ability to visualize the colonic mucosa, which is affected by the quality of bowel preparation. If the mucosa cannot be adequately visualized, the colonoscopy may be incomplete, with reported rates as high as 20%. Poor visualization increases economic costs due to repeat procedures, prolongs procedural time, raises the risk of adverse events, and can increase the adenoma miss rate by around 8%.
Several predictors of poor bowel preparation have been identified, including late colonoscopy start time, non-adherence to preparation instructions, inpatient status, male gender, use of tricyclic antidepressants, chronic constipation, and a history of stroke, dementia, or cirrhosis. Given the variables associated with inadequate colonic cleansing and the associated costs, optimizing bowel preparation for high-risk patients is critical.
Patients with chronic constipation are particularly at risk for incomplete bowel preparation. The prevalence of constipation increases with age, coinciding with the age at which many patients undergo colonoscopy. Studies have shown constipated patients often have technically difficult colons and poorer bowel preparation. Although randomized controlled trials have tried to determine the optimal preparation, no meta-analysis focusing on constipated patients has been conducted. This review aimed to systematically analyze the literature to determine the ideal bowel preparation for patients with chronic constipation.
Materials and Methods
Search Criteria
We searched MEDLINE, EMBASE, SCOPUS, and Web of Science from January 1946 to January 2018 using terms related to colonoscopy and bowel preparation agents.
Selection Criteria
Two authors independently screened titles and abstracts. Inclusion criteria were adult patients (age ≥18 years), randomized controlled trial design, studies comparing two or more bowel preparations, inclusion of constipated patients, and assessment of preparation effectiveness. Exclusion criteria included non-English studies, non-human studies, and studies with fewer than 10 patients.
Data Extraction
Data on patient demographics, bowel preparation quality, dosing regimens, tolerability, and serious adverse events were collected. Definitions for successful bowel preparation were based on validated scales (Aronchick, BBPS, OBPS) or author-defined criteria when validated scales were not used.
Statistical Analysis
Categorical variables were presented as frequencies and percentages; continuous data as mean ± SD. Meta-analysis comparing NaP and PEG was conducted using a random-effects model, with statistical significance set at P = 0.05. Heterogeneity was assessed using the I² statistic.
Results
Study Selection
Of 1581 articles identified, eight met inclusion criteria after full-text review. Three studies with a total of 225 constipated patients were eligible for meta-analysis.
Study Characteristics
In total, 107 patients received NaP and 118 received PEG. The weighted mean age was 57.6 years; about 53.5% were female. Definitions of constipation varied, with most studies using versions of the Rome criteria, while some used self-report or other criteria.
Quality of Bowel Preparation
Four studies compared NaP to other bowel preparations. Three found NaP superior to PEG or other combinations. One study found bisacodyl improved preparation with both NaP and PEG, and NaP + bisacodyl was the best combination.
Other studies assessed adjuncts added to PEG, such as bisacodyl or lactulose, which sometimes improved bowel cleanliness.
Tolerability to Bowel Preparation
Most studies assessed tolerability by patient-reported symptoms like nausea, vomiting, abdominal pain, and bloating. Three studies reported NaP alone was better tolerated; one found PEG alone was better tolerated. Adjuncts sometimes increased adverse symptoms.
Meta-Analysis
Pooled data from three studies showed that patients receiving NaP had a higher chance of successful bowel preparation compared to those receiving PEG (OR 1.87, CI 1.06–3.32, P = 0.003). Heterogeneity was low (I² = 0%).
GRADE Certainty in Evidence
Certainty that NaP is superior was rated low due to risk of bias, inconsistency in dosing, use of adjuncts, and differences in definitions and measurement scales across studies.
Discussion
This is the first systematic review and meta-analysis focused on bowel preparation in constipated patients undergoing colonoscopy. Our findings suggest NaP may provide better bowel cleansing than PEG, with improved tolerability in some studies. Despite recognition of constipation as a risk factor for poor preparation, guidelines provide little specific advice for these patients.
PEG, usually in 4 L split dosing, is the current standard for bowel preparation. NaP, a hyperosmotic agent, may enhance peristalsis and bowel evacuation more effectively in constipated patients compared to the isosmotic PEG. However, NaP use has declined due to risks such as phosphate nephropathy, particularly in older patients or those with kidney disease. In otherwise healthy patients, adverse effects tend to be mild.
Limitations of our study include variation in definitions of constipation, differences in dosing and combinations of bowel agents, and lack of validated scales in some studies. High risk of bias and lack of blinding in several studies also reduce confidence in the findings.
Conclusion
Constipation is a common and significant predictor of inadequate bowel preparation. In chronically constipated patients undergoing colonoscopy, NaP may result in better colonic cleanliness compared to PEG. For otherwise healthy constipated patients without contraindications, NaP may be considered over PEG. However,Disodium Phosphate further high-quality research is needed to confirm the best regimen and dosing.