Colonoscopy is recommended for colorectal cancer screening

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2012;10:1225 1231 SYSTEMATIC REVIEWS AND META-ANALYSES 4-Liter Split-Dose Polyethylene Glycol Is Superior to Other Bowel Preparations, Based on Systematic Review and Meta-analysis Fasiha Kanwal, Section Editor BRINTHA K. ENESTVEDT,* CHRISTINA TOFANI, LOREN A. LAINE, ANN TIERNEY, and M. BRIAN FENNERTY *Division of Gastroenterology, Department of Medicine, Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, Philadelphia, Pennsylvania; Section of Digestive Diseases, Yale University, New Haven, Connecticut; and Division of Gastroenterology, Oregon Health & Science University, Portland, Oregon BACKGROUND & AIMS: Adequate bowel cleansing is an important determinant of the efficacy of screening colonoscopy. Polyethylene glycol (PEG)-based solutions are used commonly in bowel preparation, but their poor palatability and large volumes (4 L) influence compliance. Adjunct therapies, such as bisacodyl, splitdose regimens, and lower-volume regimens have been tested. We performed a meta-analysis to determine whether a 4-L split dose of PEG is better than others for bowel cleansing before colonoscopy. METHODS: We searched MEDLINE, the Cochrane Central Register of Controlled Trials and Database, recent abstracts from major conference proceedings, references from selected reviews and randomized trials (http://clinicaltrials.gov), and Google Scholar, through September 2011, for high-quality, randomized trials that compared 4-L split-dose PEG without adjunct therapy with other bowel preparation methods. Nine of 2477 trials considered were used in the analysis. We calculated pooled estimates of bowel preparation quality (primary outcome: excellent or good), preparation compliance, favorable overall experiences, willingness to repeat same preparation, and side effects. We calculated pooled estimates of odds ratios by fixed- and random-effects models. We also assessed heterogeneity among studies and publication bias. RESULTS: The overall pooled odds ratio for excellent or good bowel preparation quality for 4-L split-dose PEG was 3.46, compared with other methods (95% confidence interval, 2.45 4.89; P.01). Although there was significant heterogeneity in results among studies, 7 of 9 reported a significant benefit from the 4-L split-dose PEG preparation. There were no significant differences between PEG and others in preparation compliance, favorable overall experience, willingness to repeat the same preparation, abdominal cramping, nausea, or sleep disturbance. There was no significant publication bias based on funnel plot. CONCLUSIONS: A meta-analysis showed that 4-L splitdose PEG is better than other bowel preparation methods for colonoscopy. Significant heterogeneity among studies might result from differences in patient demographics and protocols. A 4-L split dose of PEG should be considered the standard with which new bowel preparation methods are compared. Keywords: Colorectal Cancer Screening; Visualization; Diagnostic Accuracy; Clinical Trial. Colonoscopy is recommended for colorectal cancer screening and surveillance, and several studies have suggested that colonoscopy may reduce the incidence of colorectal cancer mortality by approximately 50%. 1,2 Among the most important contributors to the effectiveness of colonoscopy as a cancer prevention tool is the quality of the bowel cleansing. Recent studies have shown an adenoma miss rate of 33% to 46% in those with inadequate bowel preparation at the time of their screening colonoscopy. 3,4 Even more worrisome is the advanced adenoma miss rate of 18% to 27% in the face of suboptimal bowel cleansing. Therefore, adequate preparation of the bowel is necessary for optimal visualization of the colonic mucosa and significantly influences diagnostic accuracy. 5,6 Polyethylene glycol (PEG)-based solutions are among the most commonly used and studied bowel preparation regimens. The efficacy of PEG solutions is well established. However, the generally poor palatability and patient tolerance (owing in part to its large volume of 4 L) of PEG-based solutions influences compliance with preparation. Therefore, adjunct therapies (eg, bisacodyl), split-dose regimens, and lower-volume regimens have been used in an attempt to improve efficacy and patient compliance, both of which impact the quality of screening colonoscopy. We hypothesize that 4 L of split-dose PEG-based bowel preparation solution is superior in bowel cleansing to other bowel preparations. Therefore, we conducted a systematic review and meta-analysis to qualitatively and quantitatively summarize prior high-quality randomized controlled trials (RCTs) that compared 4-L split-dose PEG with any other bowel preparation in terms of bowel preparation quality. We aimed to also statistically summarize secondary outcomes such as favorable overall patient experience, adherence to prescribed bowel preparation regimen, patient willingness to repeat the same bowel preparation, and side effects including nausea, bloating, cramping, and sleep disturbance. Methods The Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) statement and guidelines were Abbreviations used in this paper: CI, confidence interval; OR, odds ratio; PEG, polyethylene glycol; RCT, randomized controlled trial. 2012 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2012.08.029

1226 ENESTVEDT ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 11 consulted during all stages of design, analysis, and reporting of this meta-analysis. 7 On an a priori basis, the primary and secondary outcomes and study inclusion and exclusion criteria were identified and are described later. Study Selection Criteria All RCTs of adult patients ( 18 y) comparing the bowel preparation efficacy of 4-L split-dose PEG solution (1 3 L the day before colonoscopy and 1 3 L the day of colonoscopy) with any other bowel preparation were included in the analysis. Search Strategy MEDLINE, Cochrane Central Register of Controlled Trials & Database, recent abstracts from major conference proceedings (Digestive Diseases Week 2009 2011, American College of Gastroenterology 2009 2011), Google Scholar, and http://clinicaltrials.gov were searched from January 1980 through September 2011 by 2 independent investigators (B.K.E., C.T.). The search terms used were as follows: bowel preparation, colonoscopy, polyethylene glycol, PEG, split dose, Golytely (Braintree Lab, Inc, Braintree, MA), CoLyte (UCB Inc, Rochester, NY), TriLytely (UCB Inc), and NuLytely (Braintree Lab, Inc). The reference lists of selected reviews and RCTs were searched manually to ensure that no pertinent articles were missed. The search was limited to studies published in English. To ensure completeness, the 2 investigators (B.K.E., C.T.) compared their lists of potentially eligible titles and abstracts and came to a consensus on which reports warranted full review. This agreed-upon list of titles underwent full review by both investigators independently to determine if the studies met inclusion criteria. Disagreement was resolved by mutual discussion and consultation with the principal investigator (M.B.F.), if necessary. Assessment of Methodologic Quality Every RCT was assessed for quality using a modified Jadad scoring system. 8 This 5-point assigns a single point for each of the following: (1) the study is described as randomized; (2) the study is described as double-blinded; (3) there is a description of withdrawals and drop-outs; (4) the randomization method is described and appropriate; and (5) the blinding method is described and appropriate. A score of 5 suggests excellent quality and a score of 0 implies a poor-quality RCT. Because double blinding cannot be logistically executed with bowel preparation studies, we adopted a modified Jadad scoring criteria in which we assigned a point for single (endoscopist) blinding and an additional point for description of an appropriate single-blinding method. To achieve points, the study had to specifically state that all the endoscopists were blinded to the bowel preparation and that staff, nurses, and patients were instructed not to discuss the bowel preparation with the endoscopist. Only higher-quality RCTs, defined as a score of 4 to 5, were included for analysis. Data Collection and Extraction Two investigators (B.K.E., C.T.) performed independent data extraction using a standardized collection form. Disagreement was resolved by mutual discussion and, if required, by consultation with the principal investigator (M.B.F.). If data were missing from a study, the investigator was contacted to provide the missing data if possible. Outcomes All outcomes were determined on an a priori basis. The primary outcome was bowel preparation efficacy. This was prespecified as a Boston Bowel Preparation Scale score of 7 or greater, or an Ottawa score less than 5, or an excellent or good bowel preparation designation on the Aronchik or other unvalidated 4- or 5-point s (ie, excellent, good, fair, poor, very poor, and so forth). As an assessment of bowel preparation tolerability and side effects, a patient s subjective evaluation of their level of satisfaction and acceptability of the bowel preparation was recorded by several studies that administered a periprocedural nonstandardized questionnaire to the patient. The data for the secondary outcomes were extracted from the results of these questionnaires. The outcome of favorable overall experience was defined as the proportion of patients who described their overall bowel preparation experiences as easy or acceptable, mild or no symptoms/side effects, or good tolerability. Compliance with bowel preparation was defined as adherence to the bowel preparation as prescribed or consuming at least 75% of the prescribed bowel preparation. Additional secondary outcomes of willingness to repeat the same bowel preparation, bloating, nausea, abdominal cramping, or sleep disturbance represented affirmative responses to the relevant question from the questionnaires. Statistical Analysis The treatment effect was computed using a pooled estimate of odds ratio (OR) and 95% confidence limits for the proportions of subjects attaining an excellent or good bowel preparation or appropriate secondary outcome. In studies with multiple comparator arms, the outcome events for all comparator arms were aggregated for comparison. A random-effects model was used to calculate the pooled OR when heterogeneity existed (P.10). Otherwise, a fixed-effect model was used. Sensitivity analysis was performed to assess for the influence of studies that contained multiple bowel preparation comparators on the primary outcome of the meta-analysis. Separate metaanalyses were performed on the following subgroups: (1) studies that had the same bowel preparations (including MiraLAX [Merck & Co Inc, Whitehouse Station, NJ]/Gatorade [Pepsi Co, Purchase, NY], or 2 L PEG with ascorbic acid) as the comparator; and (2) studies with single-dose PEG bowel preparations. Heterogeneity was calculated using the chi-square test withn-1degrees of freedom, where n represents the number of studies contributing to the meta-analysis. Heterogeneity was considered significant if the P value was less than.10. An additional measure of heterogeneity was calculated with the I 2 test of inconsistency (I 2 40% suggests heterogeneity). Publication bias was assessed by conducting a funnel plot. RevMan 5 (The Cochrane Collaboration; Oxford, UK) and STATA (Stata- Corp, College Station, TX) were used for statistical analysis. Results The initial search yielded 1123 potentially relevant references (Figure 1). After limiting the search to English studies and RCTs, 237 abstracts were reviewed. Twenty-three RCTs were identified whose full text was reviewed. We excluded 14 RCTs that did not include a full 4-L split-dose PEG arm or whose 4-L split-dose PEG arm included the use of adjunct therapy. Therefore, a total of 9 higher-quality RCTs, including

November 2012 META-ANALYSIS: BOWEL PREPARATION 1227 Figure 1. Study selection for meta-analysis. *ClinicalTrials.gov and review of references from selected RCTs. a total of 2477 patients, met the inclusion criteria and were included in this meta-analysis. 9 17 Table 1 summarizes the characteristics of these 9 studies. The mean age of the patients ranged from 52 to 59 years. The study population in 8 of the 9 studies was subjects older than the age of 18 years who underwent colonoscopy for any indication including screening for colorectal cancer. One study s population included only average-risk patients undergoing colorectal cancer screening. 14 One arm of every study, as defined in the initial study design, used a 4-L split-dose PEG regimen. All studies used a split regimen of4lin2equally divided doses (2 L each) except for Corporaal et al, 11 whose split-dose regimen depended on the timing of colonoscopy (morning procedure: 3 L the night before and 1 L the morning of the procedure; afternoon procedure: 2 L the night before and 2 L the morning of the procedure), and Park et al 17 who split their preparation into a 3-L dose the night before the procedure and a 1-L dose the morning of the procedure. An excellent or good bowel preparation was defined as a Boston Bowel Preparation Scale score of 7 or greater, an Ottawa score greater than 5, an excellent or good rating according to a historical or Aronchick (6 studies), and a very good or good historical rating (2 studies). Four studies used a validated bowel preparation (Ottawa, Aronchik, or Boston Bowel Preparation Scale), 14 17 and the remaining 5 studies used an unvalidated. Primary End Point: Excellent/Good Bowel Preparation Figure 2 contains the forest plot showing results of the primary end point for the individual studies and for the aggregated studies. A 4-L split-dose of PEG had a significantly higher proportion of patients with excellent or good bowel preparations compared with other bowel preparation comparators in 7 of the 9 studies. The summary statistic for comparison of split-dose 4-L PEG vs other comparators in the 9 studies was an OR of 3.46 (95% confidence interval [CI], 2.45 4.89), although significant heterogeneity was present (P.008; I 2 62%). Split-dose 4-L PEG was not significantly better than the comparator in only 2 of the 9 studies, both of which used a comparator of 2-L split-dose PEG plus ascorbic acid. 11,12 The results of both these studies showed a trend toward superiority of 4-L split-dose PEG, and when aggregated in a post hoc subgroup meta-analysis (fixed-effect model), the OR for an excellent or good preparation with 4-L split preparation vs 2-L split preparation was 2.27 (95% CI, 1.24 4.30). Additional subgroup analyses that aggregated all RCTs that used the same comparator revealed that 4-L split dose PEG was superior to MiraLAX/Gatorade and to single-dose PEG regimens (Figure 3). To explore potential explanations for the heterogeneity in the primary end point, 2 separate sensitivity analyses were performed. The first analysis was conducted by sequentially eliminating studies with multiple comparator arms. We performed a second entirely separate sensitivity analysis of those studies with similar preparation comparators. Neither sensitivity analysis appeared to significantly change the primary outcome OR or eliminate statistical heterogeneity (I 2 40%). Secondary Outcomes Table 2 contains the summation OR for each of the secondary outcomes. There was no significant difference in the proportion of patients with favorable overall experience, bowel preparation compliance, willingness to repeat bowel preparation, bloating, nausea, abdominal cramping, or sleep disturbance between the 4-L split-dose arm and the other comparators. Statistically significant heterogeneity was present for all secondary outcomes except abdominal cramping and sleep disturbance. Publication Bias There was no significant publication bias detected for the primary outcome of excellent or good bowel preparation efficacy by funnel plot (Figure 4), Egger (P.69; 95% CI, 6.05 to 4.24), or Begg Mazumdar (Kendall s tau, 2; P.84) bias indicators. Discussion In this meta-analysis of 9 higher-quality bowel preparation RCTs, 4-L split-dose PEG bowel preparation showed superiority over other bowel preparation comparators. The 4-L split-dose PEG regimen had a significantly higher proportion of patients with excellent-good bowel preparation than its comparators in 7 of 9 studies, and a trend to superiority in the other 2 studies. The summary results for the aggregated studies also indicate superiority of the 4-L split-dose PEG preparation (OR, 3.46; 95% CI, 2.45 4.89). Significant heterogeneity was present and most likely has a multifactorial source such as varying bowel preparation protocols including timing of preparation ingestion and diet the day before the examination and multiple bowel preparation comparator arms within single studies. There was no significant difference in the subjects favorable overall experience with the bowel preparation, compliance with bowel preparation, willingness to repeat, or adverse effects between 4-L split-dose PEG and its comparators. Split-dosed PEG bowel preparation regimens have been shown in a previous meta-analysis to be superior in bowel cleansing efficacy to single-dose PEG preparations with an OR of 3.70 (95% CI, 1.06 2.91; P.03). 18 In addition, Kilgore et al 18 reported that split-dose PEG was associated with a higher willingness to repeat the same preparation, fewer preparation discontinuations, and less nausea than single-dose PEG. In the

Table 1. Summary of Studies Included in Meta-analysis Study n Location Abdul-Baki et al, 9 2008 Aoun et al, 10 2005 Corporaal et al, 11 2010 Ell et al, 12 2008 Ell et al, 13 2003 Enestvedt et al, 14 2011 Hjelkrem et al, 15 2011 Marmo et al, 16 2010 Park et al, 17 2010 Bowel preparation 290 Beirut, Lebanon Unvalidated 4-point 141 Beirut, Lebanon Unvalidated 4-point 307 Groningen, The Unvalidated 5-point Netherlands 308 Germany Unvalidated 5-point 173 Germany Unvalidated 5-point 190 Portland, OR Boston Bowel Preparation Scale 403 Fort Sam Houston, TX Population Elective outpatient, 18 y Elective outpatient, 18 y Diagnostic colonoscopy, 18 y Inpatients, age 18 85 y, any indication 18 75 y, elective colonoscopy Average-risk screening Mean age, y Bowel preparation comparator 55 14 4-L PEG single-dose plus placebo the night before procedure 4-L PEG single dose plus tegaserod the night before procedure 57 14 4-L PEG single dose the night before procedure 54 15 2-L split-dose PEG ascorbic acid: 1 L the night before and 1 L the morning of procedure 59 15 2-L split-dose PEG ascorbic acid: 1 L the night before and 1 L the morning of procedure 56 14 3-L split-dose PEG: 2 L the night prior to and 1 L the morning of procedure 90-mL split-dose NaPhos: 45 ml the morning and night before the procedure 57 6 MiraLAX/Gatorade 64-oz split dose: 32 oz the night before and 32 oz the morning of the procedure Ottawa Screening, 18 y 54 6 MiraLAX/Gatorade 64-oz split dose: 32 oz the night before and 32 oz the morning of the procedure MiraLAX/Gatorade 64-oz split dose 24 g lubiprostone the day before procedure MiraLAX/Gatorade 64-oz split dose 10 mg bisacodyl the day before procedure 58 15 4-L PEG single dose the night indication before procedure 433 Rome, Italy Ottawa 18, any 232 Seoul, Korea Aronchik 18, any colonoscopy 52 11 4-L PEG single dose the night before procedure 250-mL Mg citrate night before and then 2-L PEG at 5 AM day of procedure Randomization b, Absent in RCT;, Performed in RCT. a Jadad score was modified to account for single blinding because double blinding is not logistically possible in bowel preparation studies. b The study was described as randomized. c The randomization method was described and appropriate. d The study was described as blinded. e The blinding method was described and appropriate. Appropriate randomization method c Modified Jadad score (0 5) a Blinding d Appropriate blinding method e Description of withdrawals 4 4 4 4 4 Total 1228 ENESTVEDT ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 11

November 2012 META-ANALYSIS: BOWEL PREPARATION 1229 Figure 2. Forest plot showing a significantly higher number of excellent or good bowel preparations with 4-L split-dose PEG. present study, we confirmed the results of the meta-analysis by Kilgore et al 18 in a post hoc subgroup meta-analysis favoring 4-L split-dose PEG (OR, 3.47; 95% CI, 1.96 6.14) over singledose PEG in bowel preparation efficacy. However, our study did not show a difference in willingness to repeat the same bowel preparation or any adverse effects in comparison with the Kilgore et al 18 study. In addition, in a post hoc analysis, we aggregated the results of 2 recent randomized clinical trials that showed the superiority of split-dose PEG to other split regimens such as PEG 3350 (MiraLAX) combined with a sports drink. 14,15 Split-dose 4-L PEG was favored significantly (OR, 3.40; 95% CI, 2.28 5.06). What this meta-analysis adds to our understanding of optimal bowel preparation is that not only is split dosing clearly superior to single-dose regimens, but also that 4-L PEG-based solutions in a split-dose regimen provide improved bowel preparation compared with any other solution against which it has been studied. In our analysis, 2 individual studies comparing the 4-L split-dose PEG preparation with a lower-volume 2-L split-dose PEG plus ascorbic acid preparation each failed to show a significant difference in the primary end point of excellent-good bowel preparation. However, when a post hoc analysis aggre- Figure 3. Post hoc subgroup meta-analyses showing a higher number of excellent or good bowel preparations with 4-L split-dose PEG than for (A) MiraLAX/Gatorade or (B) single-dose 4-L PEG preparations. (A) A 4-L split-dose PEG vs MiraLAX/Gatorade. (B) A 4-L split-dose PEG vs single-dose PEG.

1230 ENESTVEDT ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 11 Table 2. Secondary Outcomes of 4-L Split-Dose vs Other Comparators Outcome Studies, n N OR 95% CI Favorable overall experience a 4 439 1.58 0.82 3.03 Bowel preparation 6 718 1.41 0.65 3.05 compliance a Willingness to repeat a 4 870 0.79 0.43 1.45 Bloating a 3 851 0.69 0.13 3.64 Nausea a 6 1587 0.92 0.61 1.39 Abdominal cramping 5 1391 0.98 0.75 1.27 Sleep disturbance 2 373 0.91 0.58 1.45 a Indicates statistically significant heterogeneity was present. Figure 4. Funnel plot showing no significant publication bias for primary outcome. gating the 2 studies (given that they had the same preparation comparator) was performed, the 4-L split-dose PEG solution did indeed show significantly better results (OR, 2.27; 95% CI, 1.24 4.30). We eliminated 2 RCTs by Park et al 19 and Seo et al 20 from analysis because of lower Jadad scores. Seo et al 20 performed an RCT comparing bowel preparation efficacy using an Ottawa of 4-L split-dose PEG vs split-dose sodium phosphate solution (90 ml total) in adult outpatients. Patients consumed a solid diet for the entire day before colonoscopy. They found no significant difference in the mean Ottawa score between the 2 arms (5.85 1.85 vs 5.50 1.94; P.181). The study by Park et al 19 compared single- vs split-dose 4-L PEG, with no dietary restriction the day before colonoscopy in either arm. In those patients who consumed more than 75% of their bowel preparation, the mean Ottawa score was significantly lower in the split-dose PEG group (5.9 2.6 vs 8.5 2.5; P.01). In both studies, the proportion of patients in each arm that achieved an Ottawa score of less than 5 (the defined outcome of this meta-analysis) was not reported and therefore post hoc metaanalysis could not be performed to assess the influence of these studies on the primary outcome of bowel preparation efficacy. Identifying bowel preparation regimens that produce excellent bowel visualization and patient compliance will continue to be of great importance given the current emphasis on ensuring high-quality colonoscopy. Bowel preparation quality is a predictor and surrogate of adenoma detection. 5,6 Therefore, the quality of colonoscopy may be influenced strongly by the bowel regimen prescribed. Based on the results of our systematic review and meta-analysis, we suggest that a 4-L split-dose PEG regimen should be the gold standard with which new preparations are compared. At present, a split-dose regimen has received support from the American College of Gastroenterology in its 2009 guidelines for colorectal cancer screening as a key measure for improving quality and cost effectiveness on screening colonoscopy. 21 Adenoma detection rates may have been a more clinically relevant end point of this study; however, this outcome was not reported in any of the studies in this metaanalysis and has yet to be the primary end point of any bowel preparation study. Therefore, it was not possible for adenoma detection to be the primary end point of this meta-analysis, but should be considered in future bowel preparation studies. Bowel visualization therefore remains the current standard for assessment of bowel preparation quality. Given the detected heterogeneity, a sensitivity analysis was performed by sequentially eliminating studies that had multiple comparator arms and studies with similar bowel preparations. Heterogeneity was impacted minimally. There is likely some element of clinical heterogeneity derived from varying patient populations, differing bowel preparation protocols including timing of preparation ingestion as well as variable preprocedure diets, and use of different bowel preparation s. We therefore opted for a more conservative estimate of treatment effect with a random effects model. In addition to the finding of heterogeneity, this study had several other limitations. There was no standardization of bowel preparation, and bowel preparation assessments are inherently subjective. Unfortunately, there were too few studies that used the same validated bowel preparation to perform subgroup meta-analyses. Future studies of bowel preparations should use only validated s such as the Boston Bowel Preparation Scale 22 or the Ottawa. Although the majority of studies in this analysis used an evenly split bowel preparation regimen (2 L the day before and 2 L the day of the colonoscopy), the results of this study cannot be extrapolated to define the ideal amount of PEG solution to consume on each day or to indicate that a split dose is superior to a full preparation given solely on the morning of the procedure. We included only English language studies in the literature search; this is likely to impact the results of the analysis only minimally, if at all, because prior literature suggests that language restrictions do not appear to bias the estimates of a conventional intervention s effectiveness. 23 In addition, as mentioned earlier, although bowel preparation has been reported to be a surrogate for adenoma detection, the end points used in these studies were not clinical outcomes such as adenoma detection rate or reduction of colorectal cancer. In summary, 4-L split-dose PEG solution was superior to all other comparators studied for bowel cleansing. Significant differences were not apparent in other outcomes such as patient experience, willingness to repeat the same bowel preparation, compliance, or adverse events. We therefore suggest that the 4-L split-dose PEG bowel preparation regimen be used as the standard with which new bowel preparations are compared in future RCTs.

November 2012 META-ANALYSIS: BOWEL PREPARATION 1231 References 1. Zauber AG, Winawer SJ, O Brien MJ, et al. Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths. N Engl J Med 2012;366:687 696. 2. Quintero E, Castells A, Bujanda L, et al. Colonoscopy versus fecal immunochemical testing in colorectal-cancer screening. N Engl J Med 2012;366:697 706. 3. Chokshi RV, Hovis CE, Hollander T, et al. Prevalence of missed adenomas in patients with inadequate bowel preparation on screening colonoscopy. Gastrointest Endosc 2012;75:1197 1203. 4. Lebwohl B, Kastrinos F, Glick M, et al. The impact of suboptimal bowel preparation on adenoma miss rates and the factors associated with early repeat colonoscopy. Gastrointest Endosc 2011; 73:1207 1214. 5. Harewood GC, Sharma VK, de Garmo P. Impact of colonoscopy preparation quality on detection of suspected colonic neoplasia. Gastrointest Endosc 2003;58:76 79. 6. Froehlich F, Wietlisbach V, Gonvers JJ, et al. Impact of colonic cleansing on quality and diagnostic yield of colonoscopy: the European panel of appropriateness of gastrointestinal endoscopy European multicenter study. Gastrointest Endosc 2005;61:378 384. 7. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009;151:W65 W94. 8. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1 12. 9. Abdul-Baki H, Hashash JG, Elhajj II, et al. A randomized, controlled, double-blind trial of the adjunct use of Tegaserod in whole-dose or split-dose polyethylene glycol electrolyte solution for colonoscopy preparation. Gastrointest Endosc 2008;68:294 300; quiz, 334, 336. 10. Aoun E, Abdul-Baki H, Azar C, et al. A randomized single-blind trial of split-dose PEG-electrolyte solution without dietary restriction compared with whole dose PEG-electrolyte solution with dietary restriction for colonoscopy preparation. Gastrointest Endosc 2005;62:213 218. 11. Corporaal S, Kleibeuker JH, Koornstra JJ. Low-volume PEG plus ascorbic acid versus high-volume PEG as bowel preparation for colonoscopy. Scand J Gastroenterol 2010;45:1380 1386. 12. Ell C, Fischbach W, Bronisch HJ, et al. Randomized trial of low-volume PEG solution versus standard PEG electrolytes for bowel cleansing before colonoscopy. Am J Gastroenterol 2008; 103:883 893. 13. Ell C, Fischbach W, Keller R, et al. A randomized, blinded, prospective trial to compare the safety and efficacy of three bowelcleansing solutions for colonoscopy (HSG-01*). Endoscopy 2003;35:300 304. 14. 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;33:33 40. 15. Hjelkrem M, Stengel J, Liu M, et al. MiraLAX is not as effective as GoLytely in bowel cleansing before screening colonoscopies. Clin Gastroenterol Hepatol 2011;9:326 332.e1. 16. Marmo R, Rotondano G, Riccio G, et al. Effective 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;72:313 320. 17. Park SS, Sinn DH, Kim YH, et al. Efficacy 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: 1319 1326. 18. Kilgore TW, Abdinoor AA, Szary NM, et al. Bowel preparation with split-dose polyethylene glycol before colonoscopy: a meta-analysis of randomized controlled trials. Gastrointest Endosc 2011; 73:1240 1245. 19. Park JS, Sohn CI, Hwang SJ, et al. Quality and effect of single dose versus split dose of polyethylene glycol bowel preparation for early-morning colonoscopy. Endoscopy 2007;39:616 619. 20. Seo EH, Kim TO, Kim TG, et al. Efficacy and tolerability of splitdose PEG compared with split-dose aqueous sodium phosphate for outpatient colonoscopy: a randomized, controlled trial. Dig Dis Sci 2011;56:2963 2971. 21. Rex DK, Johnson DA, Anderson JC, et al. American College of Gastroenterology guidelines for colorectal cancer screening 2009 [corrected]. Am J Gastroenterol 2009;104:739 750. 22. Lai EJ, Calderwood AH, Doros G, et al. The Boston bowel preparation : a valid and reliable instrument for colonoscopyoriented research. Gastrointest Endosc 2009;69:620 625. 23. Moher D, Pham B, Lawson ML, et al. The inclusion of reports of randomised trials published in languages other than English in systematic reviews. Health Technol Assess 2003;7:1 90. Reprint requests Address requests for reprints to: Brintha K. Enestvedt, MD, MBA, Division of Gastroenterology, University of Pennsylvania, 1 Convention Boulevard, Penn Tower 9th Floor, Philadelphia, Pennsylvania 19104. e-mail: brintha.e@gmail.com; fax: (215) 707-9629. Conflicts of interest The authors disclose no conflicts.