Miralax With Gatorade for Bowel Preparation: A Meta-Analysis of Randomized Controlled Trials

Transcription

1 1566 CLINICAL AND SYSTEMATIC S nature publishing group see related editorial on page x Miralax With for Bowel Preparation: A Meta-Analysis of Randomized Controlled Trials Sameer Siddique, MD 1, Kristi T. Lopez, MD 1, Alisha M. Hinds, DO 1, Dina S. Ahmad, MD 1, Douglas L. Nguyen, MD 2, Michelle L. Matteson-Kome, APN, PhD 1, Srinivas R. Puli, MD 3 and Matthew L. Bechtold, MD, FASGE, FACG 1 OBJECTIVES: METHODS: RESULTS: Polyethylene glycol () is a very popular bowel preparation for colonoscopy. However, its large volume may reduce patient compliance, resulting in suboptimal preparation. Recently, a combination of Miralax and has been studied in various randomized controlled trials (RCTs) as a lower volume and more palatable bowel preparation. However, results have varied. Therefore, we conducted a meta-analysis assessing the use of Miralax () vs. for bowel preparation before colonoscopy. Multiple databases were searched (January 2014). RCTs on adults comparing ( g in 1.9 l that is 64 fl oz) vs. (3.8 4 l) for bowel preparation before colonoscopy were included. The effects were analyzed by calculating pooled estimates of quality of bowel preparation (satisfactory, unsatisfactory, excellent), patient tolerance (nausea, cramping, bloating), and polyp detection by using odds ratio (OR) with fixed- and random-effects models. Five studies met inclusion criteria ( N = 1,418), with mean age ranging from 53.8 to 61.3 years. demonstrated statistically significantly fewer satisfactory bowel preparations as compared with (OR 0.65; 95 % confidence interval (CI): , P = 0.04) but more willingness to repeat preparation (OR 7.32; 95 % CI: , P < 0.01). Furthermore, no statistically significant differences in polyp detection ( P = 0.65) or side effects were apparent between the two preparations for nausea ( P = 0.71), cramping ( P = 0.84), or bloating ( P = 0.50). Subgroup analysis revealed similar results for split-dose vs. split-dose. CONCLUSIONS: for bowel preparation before colonoscopy was inferior to in bowel preparation quality while demonstrating no significant improvements in adverse effects or polyp detection. Therefore, appears superior to for bowel preparation before colonoscopy. Am J Gastroenterol 2014; 109: ; doi: /ajg ; published online 19 August 2014 INTRODUCTION Colonoscopy is an extremely important and essential tool in the management of colorectal disease and screening for colorectal cancer. However, the success of colonoscopy is dependent on the quality of bowel preparation before the procedure ( 1 ). To be considered a quality bowel preparation, it must be efficacious, tolerable, and safe ( 2,3 ). Although many bowel preparations have been studied and used, polyethylene glycol 3,350 with electrolytes () seems to be most widely used ( 4,5 ). solutions offer the advantage of minimal fluid and electrolyte shifts that can be seen with other preparations; however, a large volume is required to produce adequate mucosal cleansing, even with split dosing. This amount of volume over a short period of time is a deterrent for patients and may result in patient intolerance, poor compliance, poor colon preparation, missed lesions, and potentially a need for repeat colonoscopy owing to unsatisfactory preparations. Therefore, many studies and meta-analyses have evaluated how to maintain efficacy while improving the tolerability of solutions by a variety of mechanisms, such as adding adjuncts, splitting the dose, or combining the two ( 6 20 ). Most recently, a bowel preparation consisting of the -based product of Miralax (polyethylene glycol 3,350 without electrolytes; Merck, Whitehouse Station, NJ) combined with (a sports drink; PepsiCo, Chicago, IL) has been gaining popularity owing to its low volume and increased 1 Division of Gastroenterology, University of Missouri Columbia, Columbia, Missouri, USA; 2 Division of Gastroenterology, University of California Irvine, Irvine, California, USA; 3 Division of Gastroenterology, University of Illinois Peoria, Peoria, Arizona, USA. Correspondence: Matthew L. Bechtold, MD, FASGE, FACG, Division of Gastroenterology and Hepatology, University of Missouri Health Sciences Center, CE405, DC , Five Hospital Drive, Columbia, Missouri 65212, USA. bechtoldm@health.missouri.edu Received 9 February 2014; accepted 1 July 2014

2 Miralax and for Bowel Preparations 1567 palatability ( 21 ). Few studies have examined the use of the Miralax (M G) bowel preparation with variable results (22 26 ). Therefore, we conducted a meta-analysis assessing the use of bowel preparation vs. for bowel preparation before colonoscopy. METHODS Literature search A comprehensive literature search was performed from multiple databases using a three-point system. First, multiple databases, including MEDLINE / PubMed, Embase, Cochrane databases, CINAHL, Scopus, and Google Scholar, were searched in January Search terms included miralax and gatorade and polyethylene glycol and gatorade. Second, abstracts from major conferences (Digestive Disease Week (DDW) and American College of Gastroenterology (ACG) national meetings from 2003 to present) were searched. Third, references of the retrieved articles and reviews were manually searched for any additional articles. If data were incomplete, missing, or required clarification, authors were contacted. Data extraction All randomized controlled trials (RCTs), irrespective of language, on adult patients that compared ( g in 1.9 l or 64 fl oz) with 3350 with electrolytes (3.8 4 l) for bowel preparation were included in the analysis. Exclusion criteria were patients < 18 years old, non-rcts, abstract publications outside of DDW or ACG national meetings, and abstract publications before Two reviewers (S.S. and K.T.L.) independently assessed the trials and extracted the appropriate data to be included in the analysis. Any disagreements were evaluated and settled by consensus or a third party (M.L.B.). Study quality assessment Each study was assigned a Jadad score to assess the quality of the study (5 = excellent quality, 0 = poor quality) (27 ). The Jadad score is a statistical point system based upon ve components to evaluate the quality of studies: randomization, method of randomization being appropriate and described, double blinding, double blinding being appropriate and described, and description of withdrawal and dropouts ( 27 ). If the methods of double blinding or randomization were inappropriate, a point may be deducted for each inappropriate parameter ( 27 ). Statistical analysis Meta-analysis for the effects of the compared with for bowel preparation before colonoscopy were analyzed by calculating pooled estimates of the primary outcome of preparation quality (satisfactory, unsatisfactory, excellent) and secondary outcomes of polyp detection ( 1 polyp), patient willingness to repeat preparation, and side effects (nausea, cramping, bloating). Preparation quality was defined in each study based upon authors interpretation of the bowel preparation scale(s) used (Boston Bowel Preparation Scale and / or Ottawa Bowel Preparation Scale). Given the differences in administration that exists for these bowel preparations, subgroup analyses were performed in a similar manner for these studies using split dose (split-dose vs. split-dose ), full dose (full-dose vs. full-dose ), addition of bisacodyl as an adjunct ( with bisacodyl in split and full dose vs. ), and no addition of bisacodyl ( without bisacodyl in split and full dose vs. ). In studies with multiple experimental and control arms, groups with similar outcomes were combined to create a single pair-wise comparison in accordance with the Cochrane Handbook for Systematic Reviews of Interventions ( 28 ). The results were reported using odds ratio (OR) with Mantel Haenszel fixed- and random-effects models (random-effects model only used for those outcomes with statistically significant heterogeneity). Heterogeneity was analyzed by calculating the I 2 measure of inconsistency and was considered significant if P < 0.10 or I 2 > 50 %. If heterogeneity was statistically significant, a sensitivity analysis was used to examine for heterogeneity when certain studies were excluded from the analysis. Statistical analysis was performed using RevMan 5.2 (Review Manager (RevMan), Version 5.2, Copenhagen, Denmark: The Nordic Cochrane Centre, The Cochrane Collaboration, 2012). Publication bias was assessed by funnel plots. RESULTS Article search and identification The initial search identified 205 articles ( Figure 1 ). After evaluation of titles and abstracts, 186 articles were excluded. Of the 19 articles remaining, 5 articles met the inclusion criteria ( N = 1,418) with a mean age range from 53.8 to 61.3 years and published in the United States ( Table 1 ). Three studies examined two experimental arms each and compared with either two control groups ( 23 ) or one control group ( ). All studies evaluated the use of compared with for bowel preparation before colonoscopy. However, three studies examined groups using split dose for and ( 23,25,26 ), three studies examined groups using full dose ( ), and three studies examined the addition of bisacodyl as adjunct (24 26 ) while four studies did not (22 25 ). The details of the studies and preparation techniques are summarized in Table 2. Given that only randomized clinical trials were included in this meta-analysis, the Jadad scale was used for assessing the quality of the studies ( 27 ). All studies were of good quality, receiving Jadad scores 3 (Table 3 ). All studies Quality of preparation. Satisfactory bowel preparation was analyzed by four studies ( N = 1,196) ( 22,24 26 ). For preparation quality, the unweighted pooled rates for satisfactory bowel preparations were 89.7 % (662 / 738) of the group and 90.8 % (416 / 458) of the group. When pooled and weighted via meta-analysis, statistically significantly fewer satisfactory bowel preparations were noted for patients receiving the preparation as compared with those receiving (OR 0.65; 95 % confidence interval (CI): ; P = 0.04; Figure 2 ). Furthermore, in three studies ( 23,25,26 ) ( N = 714), preparations (143 / 402, 2014 by the American College of Gastroenterology

3 1568 Siddique et al % ) demonstrated a trend toward fewer excellent preparations as compared with (175 / 312, 56.1 % ) but did not quite reach statistical significance (OR 0.51; 95 % CI: ; P = 0.18). Despite the preparation quality, two studies ( 24,25 ) ( N = 902) examined polyp detection and found no statistically significant difference between those receiving the preparation (292 / 601, 48.6 % ) and those receiving (151 / 301, 50.2 % ) (OR 0.94; 95 % CI: ; P = 0.65). Side effects. Two studies ( 22,25 ) evaluated patients for side effects of the preparation ( N = 406). No serious adverse events were reported in the studies. No statistically significant difference was noted between the group and the group for nausea (24 / 251, 9.6 % vs. 17 / 155, 11 % ; OR 0.88; 95 % CI: ; P = 0.71), cramping (16 / 251, 6.4 % vs. 9 / 155, 5.8 % ; OR 1.09; 95 % CI: ; P = 0.84), or bloating (26 / 251, 10.4 % vs. 24 / 155, Potentially relevant articles (N =205) Potentially appropriate articles (N =19) Trials included in meta-analysis (N =5) Excluded N =186 Case reports/series Retrospective Reviews Pediatric Duplicates Excluded N =14 Different outcomes (N =8) Pre-manuscript Abstract (N =4) Not RCTs (N =2) Figure 1. Article search and identification. RCT, randomized controlled trial % ; OR 0.81; 95 % CI: ; P = 0.50), with no statistically significant heterogeneity ( Figure 3 ). Willingness to repeat preparation. Th ree studies ( 22,24,26 ) ( N = 894) examined the willingness of patients to repeat the bowel preparation. The group (unweighted 497 / 537, 92.6 % ) demonstrated statistically significantly higher willingness to repeat the bowel preparation as compared with the group (unweighted 238 / 357, 66.7 % ) (OR 7.32; 95 % CI: ; P < 0.01; Figure 4 ). Subgroup analysis Split-dose regimens. For preparation quality, the unweighted pooled rates for satisfactory bowel preparations were 82.3 % (237 / 288) for the group and 88.7 % (181 / 204) for the group ( Table 4 ). When pooled and weighted via meta-analysis, statistically significantly fewer satisfactory and excellent bowel preparations (124 / 342, 36.3 % vs. 164 / 255, 64.3 % ) were noted in the split-dose group as compared with the split-dose group ( P < 0.01 for both). No statistically significant differences were noted between the two split-dose preparations for polyp detection (90 / 201, 44.8 % vs. 52 / 101, 51.5 % ; P = 0.27) or side effects including nausea (18 /201, 9 % vs. 11 /101, 10.9 % ; P = 0.59), cramping (12 / 201, 6 % vs. 7 / 101, 6.9 % ; P = 0.75), or bloating (14 / 201, 7 % vs. 8 /101, 7.9 % ; P = 0.76). Full-dose regimens. No statistically significant difference was noted in the full-dose group as compared with the fulldose group for satisfactory preparation (425 / 450, 94.4 % vs. 235 /254, 92.5 % ; P = 0.70) or polyp detection (202 / 400, 50.5 % vs. 99 /200, 49.5 % ; P = 0.82). However, willingness to repeat preparation was statistically significantly higher in the full-dose group as compared with the full-dose group (414 / 450, 92 % vs. 153 / 254, 60.2 % ; P < 0.01). Table 1. Details of the studies included in the meta-analysis Study Design Blinding Location Patients ( N ) Enestvedt et al. ( 26 ) RCT Single Oregon ( 25 ) RCT Single Texas Bowel preparation scale Preparation regimen Bisacodyl BBPS Split dose + OPS Split dose and + Samarasena et al. ( 23 ) RCT Single California BBPS and OPS Split dose and full dose Gerard et al. ( 24 ) RCT Single Illinois McKenna et al. ( 22 ) RCT Single Pennsylvania BBPS and OPS Full dose and + BBPS Full dose BBPS, Boston Bowel Preparation Scale;, Miralax ; OPS, Ottawa Preparation Scale;, polyethylene glycol 3350 with electrolytes; RCT, randomized controlled trial.

4 Miralax and for Bowel Preparations 1569 Table 2. Details of the bowel preparations used for each study Study Dosing Diet on day before colonoscopy group group Enestvedt et al. ( 26 ) Split dose Clear liquids Miralax 238 g with 1.9 l + bisacodyl 20 mg 3 PM on day before: 0.95 l + bisacodyl 20 mg 4 h before leaving home: 0.95 l. ( 25 ) No bisacodyl. ( 25 ) With bisacodyl Samarasena et al. ( 23 ) Split dose Samarasena et al. ( 23 ) Full dose Gerard et al. ( 24 ) No bisacodyl Gerard et al. ( 24 ) With bisacodyl Split dose Clear liquids Miralax 255 g with 1.9 l 4 PM on day before: 0.95 l 5 h before colonoscopy: 0.95 l Split dose Clear liquids Miralax 255 g with 1.9 l + bisacodyl 10 mg 12 PM on day before: bisacodyl 10 mg 4 PM on day before: 0.95 l 5 h before colonoscopy: 0.95 l Split dose Clear liquids Miralax 255 g with 1.9 l 6 PM on day before: 0.95 l 5 h before colonoscopy: 0.95 l Full dose Clear liquids Miralax 255 g with 1.9 l 6 PM on day before: 1.9 l Full dose Clear liquids Miralax 255 g with 1.9 l 12 PM on day before: placebo folic acid 5 PM on day before: 1.9 l Full dose Clear liquids Miralax 255 g with 1.9 l + bisacodyl 10 mg 12 PM on day before: bisacodyl 10 mg 5 PM on day before: 1.9 l McKenna et al. ( 22 ) Full dose Clear liquids Miralax 238 g with 1.9 l 8 PM on day before: 1.9 l 4 l 3 PM on day before: 2 l 4 h before leaving home: 2 l 4 l 4 PM on day before: 2 l 5 h before colonoscopy: 2 l 4 l 4 PM on day before: 2 l 5 h before colonoscopy: 2 l 3.8 l 6 PM on day before: 1.9 l 5 h before colonoscopy: 1.9 l 3.8 l 6 PM on day before: 3.8 l 4 l 5 PM on day before: 4 l 4 l 5 PM on day before: 4 l 4 l 8 PM on day before: 4 l Satisfactory bowel preparations ( N satisfactory / N total ) 59 / / / / / 101 NA NA NA NA 193 / / / / / 54, Miralax ; NA, not available;, polyethylene glycol 3350 with electrolytes. M G with bisacodyl. For preparation quality, statistically significantly fewer excellent preparations were noted in the group with bisacodyl as compared with (68 / 188, 36.2 % vs. 122 / 204, 59.8 % ; P < 0.01); however, satisfactory preparation showed no statistical significance (336 / 388, 86.6 % vs. 370 / 404, 91.6 % ; P = 0.13). Furthermore, patients were more willing to repeat the with bisacodyl preparation as compared with (265 / 287, 92.3 % vs. 199 / 303, 65.7 % ; P < 0.01). No statistically significant difference was noted between the preparation with bisacodyl and preparation for polyp detection (145 / 301, 48.2 % vs. 151 / 301, 50.2 % ; P = 0.63). without bisacodyl. No statistically significant difference was noted between the preparation without bisacodyl and preparation for satisfactory preparation (326 / 350, 93.1 % vs. 331 / 355, 93.2 % ; P = 0.91), excellent preparation (75 / 214, 35.1 % vs. 103 / 209, 49.3 % ; P = 0.40), polyp detection (147 / 300, 49 % vs. 151 / 301, 50.2 % ; P = 0.77), nausea (16 / 150, 10.7 % vs. 17 / 155, 11 % ; P = 0.94), cramping (9 / 150, 6 % vs. 9 / 155, 5.8 % ; P = 0.94), or bloating (19 / 150, 12.7 % vs. 24 / 155, 15.5 % ; P = 0.51). Again, patients were more willing to repeat the preparation without bisacodyl than (232 / 250, 92.8 % vs. 153 / 254, 60.2 % ; P < 0.01). Publication bias. No publication was found for any of the outcomes based upon funnel plots ( Figure 5 ). Sensitivity analysis. Although most outcomes displayed no statistically significant heterogeneity, in a few outcomes (excellent preparation analyses for all studies, split dose, and with and without bisacodyl, as well as satisfactory preparation analysis for with bisacodyl), heterogeneity was discovered and a sensitivity analysis was performed on those with 3 studies. Upon sensitivity analysis, statistically significantly fewer 2014 by the American College of Gastroenterology

5 1570 Siddique et al. Table 3. Details of quality assessment using the Jadad scale Study Randomized Randomization method appropriate Double blinded Double blinding method appropriate Description of withdrawals and dropouts Enestvedt et al. ( 26 ) ( 25 ) Samarasena et al. ( 23 ) Gerard et al. ( 24 ) McKenna et al. ( 22 ) Jadad score 1=yes; 0=no. SATISFACTORY PREPARATION Miralax-gatorade Polyethylene glycol Study or Subgroup Weight Enestvedt et al % 0.45 (0.23, 0.88) Gererd et al % 1.40 (0.64, 3.07) % 0.40 (0.15, 1.09) McKenna et al % 0.79 (0.28, 2.24) (95% Cl) 738 events Heterogeneity: τ 2 = 5.83, df = 3 (P = 0.12); l 2 = 49% Test for overall effect: Z = 2.05 (P = 0.04) % 0.65 (0.43, 0.98) Favors EXCELLENT PREPARATION Miralax-gatorade Polyethylene glycol Study or Subgroup Weight Enestvedt et al % 0.53 (0.29, 0.96) % 0.22 (0.13, 0.37) Samarasena et al % 1.15 (0.68, 1.95) (95% Cl) % events Heterogeneity: τ 2 = 0.68; χ 2 = 19.28, df = 2 (P < ); l 2 = 90% Test for overall effect: Z = 1.34 (P = 0.18) 0.51 (0.19, 1.37) Favors POLYP DETECTION Miralax- Polyethylene Glycol Odds Ratio Study or Subgroup Weight Gerard et al % 1.04 (0.74, 1.46) % 0.76 (0.47, 1.23) Odds Ratio (95% Cl) 601 events Heterogeneity: χ 2 = 1.06, df = 1 (P = 0.30); l 2 = 6% Test for overall effect: Z = 0.45 (P = 0.65) % 0.94 (0.71, 1.24) Favors Figure 2. Forest plots for preparation quality and polyp detection between Miralax () preparation and polyethylene glycol () preparation for all studies. CI, confidence interval. satisfactory bowel preparations without heterogeneity were noted for with bisacodyl as compared with (OR 0.40; 95 % CI: ; P < 0.01; I 2 = 0 %, P = 0.53), with removal of one study ( 24 ). Furthermore, the sensitivity analysis revealed statistically significantly fewer excellent preparations without heterogeneity with compared with for split dosing (OR 0.57; 95 % CI: ; P = 0.03; I 2 = 0 %, P = 0.67) after removal of one study ( 25 ). Given < 3 studies, sensitivity analysis could not be performed on excellent preparations for with or without bisacodyl vs.. In addition, significant heterogeneity

6 Miralax and for Bowel Preparations 1571 NAUSEA Miralax-gatorade Polyetlrylene glycol Study or Subgroup McKenna et al Weight 72.4% 27.6% 0.80 (0.36, 1.78) 1.09 (0.33, 3.63) (95% Cl) 251 events Heterogeneity: χ 2 = 0.17, df = 1 (P = 0.68); l 2 = 0% Test for overall effect: Z = 0.37 (P = 0.71) % 0.88 (0.46, 1.72) M-H, fixed, 95% Cl 1 Favors CRAMPING Miralax-gatorade Polyetlrylene glycol Study or Subgroup McKenna et al Weight 83.2% 16.8% 0.85 (0.33, 2.24) 2.26 (0.40, 12.92) M-H, fixed, 95% Cl (95% Cl) 251 events 16 Heterogeneity: χ 2 = 0.92, df = 1 (P = 0.34); l 2 = 0% Test for overall effect: Z = 0.20 (P = 0.84) % 1.09 (0.47, 2.52) Favors BLOATING Miralax- Polyetlrylene Glycol Odds Ratio Study or Subgroup McKenna et al Weight 45.9% 54.1% 0.87 (0.35, 2.15) 0.75 (0.31, 1.80) Odds Ratio (95% Cl) 251 events Heterogeneity: χ 2 = 0.05, df = 1 (P = 0.82); l 2 = 0% Test for overal effect: Z = 0.68 (P = 0.50) % 0.81 (0.43, 1.51) Favors Figure 3. Forest plots for side effects between Miralax () preparation and polyethylene glycol () preparation for all studies. CI, confidence interval. Miralax-gatorade Polyethylene glycol Study or subgroup Enestvedt et al Gerard et al Mckenna et al Weight 19.5% 68.3% 12.2% M-H, fixed, 95% Cl 4.39 (1.43, 13.53) 8.39 (5.33, 13.20) 6.03 (1.63, 22.34) M-H, fixed, 95% Cl (95% Cl) 537 events Heterogeneity: χ 2 = 1.22, df = 2 (P = 0.54); l 2 = 0% Test for overall effect: Z = 9.62 (P < ) % 7.32 (4.88, 10.98) Favors Figure 4. Forest plot for willingness to repeat preparation between Miralax () preparation and polyethylene glycol () preparation for all studies. CI, confidence interval. remained despite sensitivity analysis for excellent preparation in all studies. DISCUSSION The quality of colonoscopy is dependent upon the quality of the bowel preparation ( 4 ). In the United States, ~ 14 million screening colonoscopies are performed annually ( 29 ). Given the large number of procedures, poor bowel preparation requiring a repeat colonoscopy may be very costly ( 30 ). One of the most common issues leading to poor preparation is patient intolerance to the large volume of preparation, requiring 4 l of over a short period of time ( 9,31 ). Owing to this intolerance and to improve overall bowel preparation, adding an adjunct (such as bisacodyl or ascorbic acid) to a smaller volume of (2 l) or split dosing the solution has been studied and used ( 6 20,32 34 ) by the American College of Gastroenterology

7 1572 Siddique et al. Table 4. Details of subgroup analysis Outcome Split-dose preparations Study groups ( N ) Patients ( N ) OR 95 % Confidence interval P value Heterogeneity ( I 2, P value) Satisfactory preparation < 0.01 a 0 %, 0.87 Excellent preparation a 71 %, 0.03 b Polyp detection NA c Nausea NA c Cramping NA c Bloating NA c Full-dose preparations Satisfactory preparation %, 0.39 Polyp detection NA c Willingness to repeat prep < 0.01 a 0 %, Bisacodyl Satisfactory preparation %, 0.09 b Excellent preparation < 0.01 a 65 %, 0.09 b Polyp detection %, 0.23 Willingness to repeat prep < 0.01 a 0 %, 0.39 No bisacodyl Satisfactory preparation %, 0.40 Excellent preparation %, < 0.01 b Polyp detection %, 0.56 Nausea %, 0.81 Cramping %, 0.28 Bloating %, 0.83 Willingness to repeat prep < 0.01 a 0 %, 0.55 NA, not applicable; OR, odds ratio. a Statistically signifi cant value. b Statistically signifi cant heterogeneity in which a random-effects model was used. c Not applicable given pooled arms from only one study. Recently, split-dose bowel preparation has been shown to be superior to traditional night-before bowel preparations ( 5 ). However, despite only having 2 l of the night before and 2 l on the day of colonoscopy with split dosing, the volume of is still 4 l total and is not tolerated by many patients. Therefore, the M G bowel preparation has been frequently used in the United States as a low-volume preparation ( 21 ). Despite its frequent use, the evidence in studies regarding the preparation has varied in its efficacy and tolerability ( ). On the basis of these differences, this meta-analysis was performed on the randomized clinical trials comparing with solutions in various dosing strategies. Bowel preparations must meet three basic criteria: efficacious, tolerable, and safe. In our meta-analysis, we examined efficacy and tolerability given that safety was not adequately examined by the RCTs and cannot be analyzed. First, with regard to prepara- tion quality, this meta-analysis demonstrated that bowel preparation is inferior to in satisfactory preparations. Across all studies, preparations appear to be 35 % less likely to result in a satisfactory preparation as compared with. Given that split dosing appears to be the optimal method of bowel preparation ( 5,18 ), subgroup analysis demonstrated that split-dose is inferior to split-dose in a number of satisfactory preparations, with the group being 57 % less likely to have a satisfactory preparation. Second, with respect to tolerability, no statistically significant improvement was noted for the preparation as compared with for side effects, even with split dosing. Furthermore, palatability seems to be significantly increased in patients taking the preparation as demonstrated by willingness to repeat preparation. However, the fact that preparation is less efficacious may lessen this benefit. Finally, although Miralax is approved by FDA (Food and Drug

8 Miralax and for Bowel Preparations 1573 SE(log[OR]) Figure 5. Funnel plot demonstrating no publication bias. OR, odds ratio. Administration) for the treatment of constipation at a dose of 17 g per day, the preparation is not FDA approved, with the dose being ~ 15 times the dose for constipation. Therefore, safety of the preparation needs further evaluation. The preparation has been found in a few reports to induce electrolyte disturbances, especially sodium. preparation is a hypotonic solution, not isotonic ( 35,36 ). Owing to this, a few case reports and one case series have identified significant hyponatremia in patients after taking preparation ( ). In contrast to the case reports, a retrospective study in 2011 showed no statistically significant difference in severe hyponatremia in patients who consumed preparation as compared with (0.13 % vs %, P = 0.10) ( 40 ). However, given that the incidence of severe hyponatremia in this study was extremely low, there is a high potential for inadequate statistical power to discover a statistical difference between the two groups. Furthermore, despite the low incidence of hyponatremia, a trend was identified toward preparationinduced hyponatremia, leading to the need for further prospective studies, preferably RCTs, to fully assess the safety of the preparation on electrolytes. Therefore, given all of these results in this meta-analysis, preparation must be considered, at best, a second-tier bowel preparation and used only if the patient cannot tolerate split-dose. The strengths of this meta-analysis include the use of only RCTs in varying populations throughout the United States. This article represents the first meta-analysis to date on the subject of bowel preparation with a very extensive search of multiple databases. Furthermore, this meta-analysis examined clinically relevant outcomes with an extensive subgroup analysis. With regard to quality, this meta-analysis consisted of only RCTs that were deemed high quality. Given that bowel preparation studies such as these cannot be adequately double blinded, the studies used in the meta-analysis were the highest quality possible based upon Jadad scoring. Finally, a sensitivity analysis was performed for all outcomes consisting of 3 studies with statistically significant heterogeneity. However, as with any study, limitations are present and are as follows. First, bowel preparations did vary slightly between studies with some using split-dose or full-dose preparations with or without bisacodyl. However, an extensive subgroup analysis was OR performed to compensate for these variations. Second, not all outcomes were available for each parameter of the subgroup analysis, but these were the only data available on the subject. Third, heterogeneity was noted for five outcomes, mostly involving excellent preparation quality. In response, a random-effects model was used to minimize the heterogeneity effect. In addition, a sensitivity analysis was performed for outcomes with 3 studies. Fourth, Gerard et al. ( 24 )instructed patients in both groups to take sodium phosphate enema on the morning of colonoscopy if the stools were not clear. Given that both groups received this instruction and no data are available on how many in each group used the enema, the effect, although present, is likely minimal. Fifth, safety of the preparation was not adequately analyzed in the RCTs included in this meta-analysis. Furthermore, RCTs on safety parameters would be beneficial to assess the full safety profile of this bowel preparation. Finally, a potential of the Hawthorne effect may be present, with the patients in the group possibly being more compliant with the preparation than those in the group, resulting in improved preparation quality. In conclusion, our meta-analysis shows that despite patients being more willing to repeat the preparation, is an inferior bowel preparation as compared with. Therefore, the role of seems to be as an alternative to only when the patient is unable to tolerate or is simply unwilling to take split-dose. CONFLICT OF INTEREST Guarantor of the article: Matthew L. Bechtold, MD, FASGE, FACG. Specific author contributions: Conception and design: Siddique, Lopez, and Bechtold; acquisition of data: Siddique, Lopez, Hinds, and Ahmad; analysis and interpretation of data: Nguyen, Matteson- Kome, Puli, and Bechtold; drafting of the manuscript: Siddique, Lopez, Hinds, and Ahmad; critical revision of the manuscript: Nguyen, Matteson, Puli, and Bechtold; statistical expertise: Matteson, Puli, and Bechtold; overall supervision of the project: Bechtold. Financial support: None. Potential competing interests: None. REFERENCES 1. Hi l lye r G C, B as ch C H, B as ch C E et al. Gastroenterologists perceived barriers to optimal pre-colonoscopy bowel preparation: results of a national survey. J Cancer Educ 2012 ;27 : McLachlan SA, Clements A, Austoker J. Patients experiences and reported barriers to colonoscopy in the screening context a systematic review of the literature. Patient Educ Couns 2012 ; 86 : Moon W. Optimal and safe bowel preparation for colonoscopy. Clin Endosc 2013 ;46 : Bechtold ML, Choudhary A. Bowel preparation prior to colonoscopy: a continual search for excellence. World J Gastroenterol 2013 ;19 : E ne st ve dt BK, Tof an i C, L ai ne L A et al. 4-Liter split-dose polyethylene glycol is superior to other bowel preparations, based on systematic review and meta-analysis. Clin Gastroenterol Hepatol 2012 ; 10 : Va l i ante F, B el lu mat A, D e B ona M et al. Bisacodyl plus split 2-L polyethylene glycol-citrate-simethicone improves quality of bowel preparation before screening colonoscopy. World J Gastroenterol 2013 ;19 : Ponchon T, B oust ière C, Here s b a ch D et al. A low-volume polyethylene glycol plus ascorbate solution for bowel cleansing prior to colonoscopy: the NORMO randomised clinical trial. Dig Liver Dis 2013 ; 45 : G ent i l e M, D e R o s a M, C est aro G et al. 2 L plus ascorbic acid versus 4 L plus simethicone for colonoscopy preparation: a randomized single-blind clinical trial. Surg Laparosc Endosc Percutan Tech 2013 ;23 : by the American College of Gastroenterology

9 1574 Siddique et al. 9. Ho oke y LC, D ep e w W T, Van ne r SJ. C ombine d l ow volume p olye t hy le ne glycol solution plus stimulant laxatives versus standard volume polyethylene glycol solution: a prospective, randomized study of colon cleansing before colonoscopy. Can J Gastroenterol 2006 ;20 : Shar ma V K, C ho cka l i ng ham SK, Ug he oke E A et al. Pro sp e c t ive, r and - omized, controlled comparison of the use of polyethylene glycol electrolyte lavage solution in four-liter versus two-liter volumes and pretreatment with either magnesium citrate or bisacodyl for colonoscopy preparation. Gastrointest Endosc 1998 ;47 : D ipa lma JA, Wol ff B G, Me ag he r A et al. Comparison of reduced volume versus four liters sulfate-free electrolyte lavage solutions for colonoscopy colon cleansing. Am J Gastroenterol 2003 ;98 : El Sayed AM, Kanafani ZA, Mourad FH et al. A randomized single-blind trial of whole versus split-dose polyethylene glycol-electrolyte solution for colonoscopy preparation. Gastrointest Endosc 2003 ;58 : Park J S, S oh n C I, Hwang SJ et al. Q u a l it y and effect of single dose versus split dose of polyethylene glycol bowel preparation for early-morning colonoscopy. Endoscopy 2007 ;39 : Aou n E, Ab du l - B a k i H, A z ar C et al. A randomized single-blind trial of split-dose -electrolyte solution without dietary restriction compared with whole dose -electrolyte solution with dietary restriction for colonoscopy preparation. Gastrointest Endosc 2005 ;62 : Ab du l- B a k i H, Hashash JG, E l hajj I I et al. A randomized, controlled, double-blind trial of the adjunct use of tegaserod in whole-dose or splitdose polyethylene glycol electrolyte solution for colonoscopy preparation. Gastrointest Endosc 2008 ;68 : Park S S, Sin n DH, Kim Y H et al. E ffi cacy and tolerability of split-dose magnesium citrate: low-volume (2 liters) polyethylene glycol vs. single- or split-dose polyethylene glycol bowel preparation for morning colonoscopy. Am J Gastroenterol 2010 ;105 : Marmo R, Rotondano G, Riccio G et al. E ffective bowel cleansing before colonoscopy: a randomized study of split-dosage versus non-split dosage regimens of high-volume versus low-volume polyethylene glycol solutions. Gastrointest Endosc 2010 ;72 : Ki l gore T W, Ab d i no or A A, S zar y NM et al. Bowel preparation with splitdose polyethylene glycol before colonoscopy: a meta-analysis of randomized controlled trials. Gastrointest Endosc 2011 ; 73 : G o d f re y J D, C l ark R E, C hou d har y A et al. Low-volume polyethylene glycol and ascorbic acid for bowel preparation prior to colonoscopy: a metaanalysis. World J Meta-Anal 2013 ;1 : C l ark R E, G o d f re y J D, C hou d har y A et al. Low-volume polyethylene glycol and bisacodyl for bowel preparation prior to colonoscopy: a meta-analysis. Ann Gastroenterol 2013 ;26 : Hi l lye r G C, L ebwoh l B, B as ch C H et al. Split dose and MiraLAX-based purgatives to enhance bowel preparation quality becoming common recommendations in the US. Therap Adv Gastroenterol 2013 ;6 : McKenna T, Macgill A, Porat G et al. Colonoscopy preparation: polyethylene glycol with is as safe and efficacious as four liters of polyethylene glycol with balanced electrolytes. Dig Dis Sci 2012 ;57 : Samarasena JB, Muthusamy VR, Jamal MM. Split-dosed MiraLAX/ is an effective, safe, and tolerable option for bowel preparation in low-risk patients: a randomized controlled study. Am J Gastroenterol 2012 ;107 : Gerard DP, Holden JL, Foster DB et al. Randomized trial of gatorade/polyethylene GLYCOL with or without bisacodyl and NuLYTELY for colonoscopy preparation. Clin Transl Gastroenterol 2012 ; 3 : e Hj el k re m M, Ste ngel J, L iu M et al. MiraLAX is not as effective as GoLytely in bowel cleansing before screening colonoscopies. Clin Gastroenterol Hepatol 2011 ; 9 : Enestvedt BK, Fennerty MB, Eisen GM. Randomised clinical trial: MiraLAX vs. Golytely - a controlled study of efficacy and patient tolerability in bowel preparation for colonoscopy. Aliment Pharmacol Ther 2011 ;33 : Jad a d A R, Mo ore R A, C ar rol l D et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996 ;17 : Higgins JPT, Green S. Cochrane handbook for systematic reviews of interventions. The Cochrane Collaboration S e e ff LC, R i chards T B, Shapi ro JA et al. How many endoscopies are performed for colorectal cancer screening? Results from CDC s survey of endoscopic capacity. Gastroenterology 2004 ; 127 : Ness R M, Manam R, Ho e n H et al. Predictors of inadequate bowel preparation for colonoscopy. Am J Gastroenterol 2001 ;96 : G olub RW, Ker ne r BA, Wis e W E Jr et al. Colonoscopic bowel preparations which one? A blinded prospective, randomized trial. Dis Colon Rectum 1995 ;38 : Pontone S, Angelini R, Standoli M et al. Low-volume plus ascorbic acid vs. high-volume plus simethicone bowel preparation before colonoscopy. World J Gastroenterol 2011 ;17 : Jans en S V, G o e d hard JG, Win kens B et al. Preparation before colonoscopy: a randomized controlled trial comparing different regimes. Eur J Gastroenterol Hepatol 2011 ; 23 : Corporaal S, Kleibeuker JH, Koornstra JJ. Low-volume plus ascorbic acid versus high-volume as bowel preparation for colonoscopy. Scand J Gastroenterol 2010 ;45 : Occhipinti KE, Di Palma JA. How to choose the best preparation for colonoscopy. Nat Rev Gastroenterol Hepatol 2009 ; 6 : Weber FH Jr. Optimizing colonic preparation: the solution is becoming clearer and clearer. Clin Gastroenterol Hepatol 2011 ; 9 : Lewis J, Schoefeld P. Severe hyponatremia and Miralax- bowel preparation. Am J Gastroenterol 2011 ;106 :S She r i d M, S amo S, Su l aiman S. S e ve re el e c t roly te s d istu rb anc e after regimen of Miralax with bisacodyl for bowel preparation. Am J Gastroenterol 2012 ;107 :S L e w is J J, E sw ar an SL, S cho en fel d P S. S e ve re hy p onat rem i a ass o c i ate d w it h Miralax- bowel prep prior to colonoscopy. Gastroenterology 2011 ;140 :S S cho en fel d P, E l l i ott E. Inc re as e d r isk of s e vere hy p onat rem i a w it h Miralax- vs standard bowel preparation. Am J Gastroenterol 2011 ;106 :S583.