3 Liter Polyethylene Glycol is Not Inferior to The Standard Bowel Preparation in Chinese Population with Equal Ecacy: A Randomized, Controlled Trial

Ideal bowel cleansing regimens should be both effective and well tolerated. The European Society of Gastrointestinal Endoscopy recommends 4L Polyethylene Glycol (PEG) as the standard regimen for the bowel preparation (BP). This large volume of PEG ensures a better-quality bowel cleansing but might be poorly tolerated amongst the Chinese population given their different characteristics. The current study compared the 3L and 4L PEG with regard to their effectiveness, tolerability, and safety amongst Chinese patients to identify the better suitable bowel cleansing method for this population. glycol; BP, bowel preparation; IBD, inammatory bowel disease.


Background
High-quality bowel preparation is essential for successful colonoscopy. Ideal bowel cleansing regimens should be both effective and well tolerated. For patients, bowel preparation requires tolerance rst, which can become an effective strategy. The European Society of Gastrointestinal Endoscopy (ESGE) [1] recommends 4L of split-dose polyethylene glycol (PEG) as the standard regimen for the bowel preparation. Because of the different characteristics of Chinese patients (e.g., smaller body size, lower body weight, and different diet habits), however, large volumes (i.e., 4L) of PEG might be poorly tolerated by the Chinese population, despite ensuring a better-quality bowel cleansing. As such, a balance should be maintained between volume and effectiveness. The current study compared the effectiveness, tolerability, and safety of 3L and 4L split-doses with regard to Chinese patients to identify the most suitable method of bowel cleansing for this population.

Study design
This prospective, randomized, reviewer-blinded, and controlled trial enrolled patients and collected relevant data at the Digestive Endoscopic Centre, Changhai Hospital of Shanghai between December 2017 and February 2018. Our study follows CONSORT guidelines. Two designated, experienced endoscopists (each having completed more than 1,000 endoscopies) performed the colonoscopies. CF-H260AI and CF-H290I endoscopes (Olympus, Japan) were used in this study. All patients were awake during the examination and without analgesia or sedation intervention. Biopsies were performed for suspected polyps and tumors. The nal diagnosis was con rmed based on a histopathological examination performed by pathologists who were unaware of the experiment at Changhai Hospital.
Eligible patients were randomly and blindly assigned to the 3L-PEG group or the 4L-PEG group through concealed allocation by a technician. Random numbers were generated using SPSS software V20.0.
Allocation concealment was achieved using sequentially numbered sealed opaque envelopes. The technician who generated the randomization table was not involved in the colonoscopy procedure. The endoscopist or nurse were not informed of the patients' preparation methods.
Each patient signed the informed consent document and then received relevant training on bowel preparation by the doctors. The protocol complied with the Declaration of Helsinki and was approved by the Ethics Committee of Changhai Hospital (ClinicalTrials.gov registration number: NCT03356015).

Patient selection
Patients scheduled for colonoscopy were eligible for this study. Additional inclusion criteria were aged from 18 to 75 years, undergoing diagnostic colonoscopy, and providing informed consent.
Patients were excluded from this study if they underwent therapeutic colonoscopy (i.e., colonic polypectomy, endoscopic mucosal resection, or endoscopic submucosal dissection); had acute myocardial infarction over the past 6 months; had severe heart, brain, lung, or kidney comorbidities; had intestinal obstruction; had limited mobility; had in ammatory bowel disease; had previous colon surgery; were pregnant or lactating; had participated in other clinical observational studies; or had participated in other clinical studies in the past 60 days.
Patients who did not come as scheduled for their colonoscopy or did not complete a colonoscopy for any reason after study entry were considered as drop-outs.
2) Bowel preparation regimens: The patients in the 4L-PEG group were instructed to take two split doses. The patients were advised to take the rst 2L dose (all of the contents of two bags of PEG, dissolved and mixed in 2,000 mL of clear warm water) at 19:00 the night before the examination by drinking 250 mL every 15 minutes and completing the dose within 2 hours. They were advised to take the second 2L dose 4-6 hours before the examination following the same procedure as the rst dose.
The patients in the 3L-PEG group were instructed to take two split doses. The patients were advised to take the rst 1L dose (all of the contents of one bag of PEG, dissolved and mixed in 1,000 mL of clear warm water) at 19:00 the night before the examination by drinking 250 mL every 15 minutes and completing the dose within 1 hour. They were advised to take the second 2L dose (all of the contents of two bags of PEG, dissolved and mixed into 2,000 mL of clear warm water) 4-6 hours before the examination by drinking 250 mL every 15 minutes and completing the dose within 2 hours.
Moreover, the patients were instructed to have little to no residue and maintain a semi-liquid diet (e.g., porridge and noodles; no vegetables or fruits) during the day before the examination. In addition, the patients were instructed to fast beginning at 19:00 the night before the examination. The patients were able to engage in light exercise during dosing to reduce discomfort such as nausea, vomiting, and bloating.
Body mass index (BMI) was calculated using height and weight. The participants were asked to complete a study-speci c questionnaire that included information such as age, sex, constipation, history of abdominal surgery, history of diabetes, history of hypertension, and other endoscopy-related indicators such as endoscopic insertion success rate, insertion time, and withdrawal time. Insertion time was de ned as the time from endoscopic insertion into the anus to advancement to the caecum. Withdrawal time was de ned as the time from endoscopic withdrawal from the caecum to the anus, excluding the time of the biopsy. The same nurse or doctor recorded the data. In addition, gastrointestinal symptoms during bowel preparation such as nausea, vomiting, abdominal pain, bloating, and discomfort during endoscopy were recorded. Patient satisfaction regarding bowel preparation and willingness to undergo the same bowel preparation in the future were evaluated.

Evaluation measures
The primary evaluation measures was the Boston Bowel Preparation Scale (BBPS) score [2,3]. The entire colon was evaluated in three segments [4]: right colon (caecum and ascending colon), transverse colon (liver and splenic exures), and left colon (descending colon to rectum). The BBPS is a 9-point scale used to evaluate the quality of a bowel preparation and was used several times in our previous studies [5,6], where 0 denotes a colon lled with faeces and faecal residue, resulting in the termination of the colonoscopy; 1 represents the presence of a large amount of faeces and faecal residue, obscuring some of the intestines; 2 denotes a small amount of faeces and faecal residue, with no major effect on intestinal visibility; and 3 represents a clean intestine with no faeces or faecal residue. All ratings were performed after bowel irrigation and aspiration. High-quality bowel preparation was de ned as a total BBPS score of ≥ 6. Inadequate bowel preparation was de ned as a total BBPS score of < 6 [7]. The secondary measure was the adenoma detection rate (ADR), which was de ned as the percentage of participants found to have colon adenoma, which was histologically con rmed by at least one colon polyp [5]. Additional measures to be recorded were endoscopic insertion time and withdrawal time.
Insertion time was de ned as the time from endoscopic insertion into the anus to advancement to the caecum. Withdrawal time was de ned as the time from endoscopic withdrawal from the caecum to the anus, excluding the time of biopsy. The same doctor or nurse recorded all of the data. Patient satisfaction regarding bowel preparation was rated using a Likert scale, where 1: unsatisfactory, 2: fair, 3: satisfactory. Patient willingness to undergo the same bowel preparation in the future was rated as "yes" or "no".

Sample size estimate and statistical analysis
Our previous experience showed that the BBPS score is ≥ 6 in approximately 85% of patients undergoing regular bowel preparation. For a statistical power of ≥ 80% and a level of 0.05 (one-sided), 158 participants were expected to enrol in each group to improve the percentage of participants with a BBPS score of ≥ 6 by 10% in the 4L-PEG group over the 3L-PEG group. Given a drop-out rate of 5%, 165 participants were planned for enrolment in each group, for a total of 330 participants. Analyses of intention-to-treat (ITT) and per-protocol (PP) were performed to compared the e cacy of 3L-PEG vs. standard regimen. Categorical variables were analysed using the Chi-square test or Fisher exact test according to appropriate situation. Continuous variables were expressed as means ± standard deviations and analysed utilizing Mann-Whitney U test or Student's t-test. Logistic regression was applied to identify the potential impact factors of inadequate BP. Subgroup analysis were conducted with the following variables: age, BMI, gender, smoking, drinking, comorbidities, constipation, abdominal or pelvic surgery history and colonoscopy history. SPSS v20.0 (SPSS, Chicago, IL, USA) was used for data processing. P < 0.05 was considered as signi cant.

Study population
The study ow chart is shown in Figure 1. Between December 2017 and February 2018, a total of 330 participants were enrolled in this study. In these patients, six of whom did not complete the examination, including seven participants in the 4L-PEG group (poor bowel preparation: n=5, personal reasons: n=2) and ve in the 3L-PEG group (poor bowel preparation: n=3, personal reasons: n=2). Therefore, 318 participants completed the examination (4L-PEG group: n=158, 3L-PEG group: n=160).

Baseline characteristics
The clinical and demographic characteristics are shown in Table 1

Parameters associated with procedures and adverse events
The procedure-related parameters and adverse events are shown in Table 3. In terms of incorrect diet restriction, no difference was shown in the two groups (4L vs. 3L, 10.9% vs. 9.7%, P=0.717). In the 4L-PEG group, 9.1% of participants' interval of preparation-to-colonoscopy was ≥ 8h and this ratio was 6.7% in the 3L-PEG group (P=0.414). There was also no difference in preparation-to-colonoscopy interval between the two groups (P=0.698). The patient satisfaction rates with "fair" and "satisfactory" were 89.1% for the 4L-PEG group and 97.5% for the 3L-PEG group, and this difference was signi cant (P=0.009). Regarding patient willingness to undergo the same bowel preparation as needed in the future, 86.7% in the 4L-PEG group and 96.4% in the 3L-PEG group chose "yes"; this difference was signi cant (P=0.002).

Subgroup analysis
In the Figure 2, the effects of the two PEG methods on the adequate BP were estimated respectively.
There was no signi cant difference in these subgroups (P all >0.05).

Discussion
Colorectal cancer (CRC) is one of the most common malignant tumours. The prognosis of advanced CRC is poor; therefore, early detection and treatment play key roles in improving patient prognosis. Colonoscopies are essential for the diagnosis and treatment of CRC. When used as a CRC screening tool, colonoscopies contribute to the early detection of cancer and the resection of precancerous lesions such as adenoma [8][9][10][11][12][13], thereby reducing CRC-related morbidity and mortality rates. However, colonoscopies are invasive as well as associated with signi cant discomfort and certain risks; furthermore, they cause fear and anxiety among most patients. Successful colonoscopies require extensive colonoscopic experience, high-quality colon cleansing, and patient cooperation [14]. Inadequate bowel preparation is considered as the most important adverse factor for colonoscopies because it prolongs insertion time, causes more discomfort, and is associated with a low ADR and potentially earlier repeat colonoscopy [15,16]. Therefore, researchers have focusing on optimal bowel preparation strategies. Current clinical guidelines recommend several intestinal cleansing regimens, including 3L-PEG (see the guidelines for intestinal preparation related to digestive endoscopy in China), 4L-PEG (the ESGE), and high-permeability low-volume solution (e.g., magnesium sulphate, mannitol, and sodium phosphate oral solution). In 2010, the US Food and Drug Administration disclosed that oral phosphate-containing intestinal cleansing solution can cause hypovolemia or severe electrolyte imbalance in patients scheduled for colonoscopy [17]; it might even cause renal failure in patients with phosphatase nephropathy. The 2013 ESGE Guidelines for Bowel Preparation for Colonoscopy [1] recommend a 4-L PEG split-dose regimen for routine bowel preparation. PEG binds to water molecules in the colon via its hydrogen bonding site, thereby increasing faecal water content. Moreover, as an isotonic whole-intestinal lavage solution, PEG maintains the approximate isotonic condition of the intestine. PEG is not rapidly absorbed or metabolised; thus, it does not affect uid or electrolyte balance. In addition, it does not produce acid or gas and is mild and non-irritating. Furthermore, it is not rapidly absorbed by the intestinal mucosa, nor can it cause an excessive exudation of body uid, dehydration, or weight loss [18]. PEG is recommended as a rst-line bowel preparation solution given the physical cleansing and satisfactory cleansing effects associated with its high volume. A recent meta-analysis [19] showed that a 4-L PEG split-dose regimen was superior to other bowel cleansing regimens for bowel preparation. However, no studies have investigated whether this high-volume bowel preparation regimen, typically used for European and American populations, is suitable for Asians who are often smaller.
A high-volume PEG solution helps ensure high-quality bowel cleansing; however, but it inevitably affects tolerability. In fact, bowel preparation regimens with poor tolerability negatively affect the quality of bowel preparation. From the perspective of patients, any bowel preparation regimen must rst have acceptable tolerance to become an effective product. Fayad et al [20] showed that BMI was an independent factor regarding the quality of bowel preparation. Intense bowel preparation regimens are recommended for patients with high BMIs. Thus, research is needed to investigate whether the volume of the cleansing solution can be reduced in Asian populations with lower BMIs, while maintaining the quality of the bowel preparation. If low-volume bowel cleansing achieves the same bowel preparation quality, then it might improve patient experience, reduce drug costs, and provide a better option for patients. To this end, we conducted the current randomized controlled clinical study to evaluate the effectiveness, tolerability, and safety of the ESGE-recommended 4-L PEG split-dose regimen versus a 3-L PEG split-dose regimen that is used for clinical trials in China for bowel preparation.
In this study, the 3L-PEG and 4L-PEG split-dose regimens demonstrated similar bowel preparation quality with regard to the Chinese population. The 3L-PEG split-dose regimen was superior to the 4L-PEG splitdose regimen in terms of patient satisfaction, tolerability, and safety. Five of the 330 patients enrolled, including ve patients in the 4L-PEG group and three patients in the 3L-PEG group did not undergo colonoscopy because they were unable to take the prescribed high-volume PEG within the speci ed timeframe (i.e., they took less than 75% or vomited more than 25% of the prescribed dose). The 3L-PEG group demonstrated better tolerability than the 4L-PEG group, although this difference was not signi cant. Of the 330 patients included in the ITT analysis, the clinical characteristics (sex, age, BMI, indications for examination) of the patients in 4L-PEG and 3L-PEG groups were consistent; no signi cant between-group differences were observed with regard to the endoscopic insertion success rate, insertion time, withdrawal time, or ADR. Moreover, no signi cant between-group differences were observed with regard to the quality of bowel preparation, including the overall and segmental BBPS, bubble rate scores, or the percentage of patients with adequate bowel preparation (BBPS score ≥ 6). Importantly, the patient satisfaction rate was 97.5% in the 3L-PEG group, which was signi cantly higher than that in the 4L-PEG group (vs. 89.1%; P = 0.009). Furthermore, the rates of nausea, vomiting and bloating were signi cantly lower in the 3L-PEG group than in the 4L-PEG group.
Zhao et al [21] evaluated the effect of three oral PEG regimens on bowel preparation before colonoscopy. The results showed that the 4L-PEG split-dose regimen was superior to the 3L-PEG split-dose regimen and the 4L-PEG single-dose regimen, indicating that the 4L-PEG split-dose regimen achieved better cleansing results; these ndings differ from those of the present study. This discrepancy might be related to the following factors. Zhao et al enrolled patients aged 18 to 60, whereas this study enrolled patients aged 18 to 75. Thus, the patient population in Zhao et al was younger and showed better tolerability for highvolume regimens, which contributed to better cleansing results. Moreover, Zhao et al measured weight but not BMI; BMI is an independent factor for bowel preparation quality. Patients with higher BMIs might require a higher volume of cleansing solution. The sample size of Zhao et al was relatively small (67, 54, and 59 patients in each of the three groups), which might have contributed to different conclusions compared with those of the current study if the mean BMI was relatively high. In addition, Zhao et al did not evaluate tolerability or adverse reactions. Future research should evaluate the effect of the 3L-PEG single-dose regimen versus the 4L-PEG single-dose regimen. In the current study, we did not monitor chemistry (electrolytes) or renal function after bowel preparation; however, no patient experienced any serious adverse reactions.
In summary, the 3L-PEG regimen was equally effective for bowel preparation and showed better tolerability and safety than the 4L-PEG regimen in a Chinese population. Thus, the former treatment might be an acceptable alternative to the latter regimen. We recommend using a 3-L PEG split-dose regimen for routine bowel preparation before colonoscopy amongst Chinese patients.

Limitations
This study was conducted in single center, which might cause bias in results. The safety and e cacy of 3L-PEG demands further muti-center randomized trial to examine. Conclusions 3L-PEG bowel cleansing represents an optimal alternative to a 4L-PEG preparation for Chinese people, showing a similar e cacy and superior levels of satisfaction, acceptability, and safety amongst users. We recommend 3L PEG be the routine regimen in the clinical setting for Chinese patients.

Consent for publication
All authors have read the article and approved the publication.

Availability of data and material
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
The work was funded by Hainan Health Industry Scienti c research Project (20A200397 20A200444 and 20A200073). The funding source of the study had no role in study design, data collection, analysis, interpretation or writing the report. The rst author and corresponding authors had full access to all the study data and nal responsibility for submission for publication.    Data are presented as mean ± standard deviation or number (percentage) as appropriate. PEG, polyethylene glycol; BP, bowel preparation; IBD, in ammatory bowel disease. Figure 1 Study ow diagram. PEG, polyethylene glycol; ITT, intention-to-treat; PP, per-protocol.

Figure 2
Subgroup analyses of adequate bowel cleansing. The effects of 4L-PEG and 3L-PEG regime on adequacy of bowel preparation were analyzed in subgroups. The strati cation factors included age, BMI, gender, smoking, drinking, comorbidities, constipation, previous abdominal or pelvic surgery and prior history of colonoscopy. A P value <0.05 was considered to be statistically signi cant. BMI, body mass index; OR, odds ratio; CI, con dence interval.