This study was performed on a consecutive series of patients who had one to nine colorectal polyps and at least one adenoma sized more than 5 mm at the high-quality screening colonoscopy from May 2009 to September 2010 at Konkuk University Medical Center in Seoul, Republic of Korea. The enrolled patients underwent tandem colonoscopy with polypectomy within the next 3 months. The study protocol was approved by the institutional review board of the Konkuk University Medical Center.

Subjects were excluded if they met one of the following criteria: 1) the colonoscopy did not reach the cecum, 2) withdrawal time of index colonoscopy was less than 6 minutes, 3) patients with 10 or more polyps detected at the index colonoscopy and suspected of having polyposis syndrome, 4) tandem colonoscopy was performed 3 months after the index colonoscopy, 5) the bowel preparation at the tandem colonoscopy was fair, poor or inadequate based on the Aronchick scale, 6) patients with a history of bowel resection, and 7) patients with inflammatory bowel disease.

The bowel preparation status was assessed using previously published and validated bowel preparation scales: the Aronchick scale22 and the Ottawa bowel preparation quality scale (Ottawa scale).23 The Aronchick scale assesses the preparation quality of the entire colon as excellent (a small volume of clear liquid or greater than 95% of the surface seen), good (a large volume of clear liquid covering 5% to 25% of the surface but greater than 90% of the surface was seen), fair (some semisolid stool that could be suctioned or washed away, but greater than 90% of the surface was seen), poor (semisolid stool that could not be suctioned or washed away and less than 90% of the surface was seen), or inadequate (repeat preparation and colonoscopy was needed). The Ottawa scale assesses the cleanliness of the right (from the cecum to the ascending colon), mid (from the transverse colon to the descending colon), and recto-sigmoid colon individually by rating each colon segment on a scale of 0 to 4. The fluid quantity is a global value for the entire colon and this is rated from 0 to 2. The score of the Ottawa scale is calculated by adding the cleanliness scores and the fluid quantity score. Thus, the scale has a range from 0 (perfect) to 14 (solid stool in each colon segment and lots of fluid, i.e., a completely unprepared colon). Bowel preparation quality was scored after sufficient washing and suctioning of fecal debris.

Before applying the bowel preparation scales in this study, the participating endoscopists undertook a calibration exercise for achieving excellent inter-observer agreement (intraclass correlation coefficient, ICC >0.8). The calibration exercise was carried out using 10 testing colonoscopy images. If the ICC of the inter-observer agreement failed to reach 0.8, then a calibration exercise with discussions among the endoscopists was repeated. After 2 weeks of calibration exercises, the inter-observer agreement was re-measured using 10 different testing colonoscopy images. This calibration exercise was repeated until excellent inter-observer agreement was achieved among the endoscopists.

Four experienced endoscopists accredited by the Korean Society of Gastrointestinal Endoscopy participated in this study. The adenoma detection rates of the endoscopists ranged from 34% to 28%. Bowel cleansing was performed using polyethylene glycol (Colyte 4 L; Taejun Pharm. Co., Ltd., Seoul, Korea) or NaP (Fleet; Unimed Pharm. Inc., Seoul, Korea) as previously described.24 For all the study procedures, high-definition CF-H260AI colonoscope (Olympus, Tokyo, Japan) was used. During the index colonoscopy, the participating endoscopists recorded the adenoma characteristics including size, number, shape, and location, withdrawal time, and bowel preparation status assessed by the Aronchick and Ottawa scales. Tandem colonoscopy was performed within 3 months after the index colonoscopy. The median interval between the colonoscopies was 38 days (range, 5 to 89). Suboptimal bowel preparation at screening colonoscopy was associated with several causes, including the failure to follow preparation instructions, later start time of colonoscopy, and history of constipation. To improve the bowel preparation status at the tandem examination, we tried to identify whether or not the patients had consumed the preparation as prescribed. If a participant did not follow the preparation instructions, we strongly recommended him/her to follow the instructions. For participants who followed the preparation instruction, we recommended a longer period of dietary restriction to clear liquids or addition of bisacodyl for those with constipation in order to improve the bowel preparation status at the tandem colonoscopy.

All detected lesions were removed during the tandem colonoscopy using snare polypectomy or endoscopic mucosal resection. In order to find and remove the colorectal adenomas during the tandem colonoscopy, the size, location, and shape of the detected lesions were recorded in the data sheet in detail during the index colonoscopy. Since polyps of ≤5 mm is often impossible to find at tandem colonoscopy, they were removed during the index colonoscopy using cold biopsy polypectomy or cold snare polypectomy.

The adenoma size was estimated during the index colonoscopy using open-biopsy forceps. Adenoma were categorized as diminutive (≤5 mm), small (5 to 9 mm) or large (≥10 mm) according to their size. The location of the adenoma was classified as right (from cecum to ascending colon), mid (from transverse colon to descending colon), or recto-sigmoid colon as described in the Ottawa scale.23 The shape of a colorectal adenoma was classified as pedunculated, sessile, or flat/depressed. A flat/depressed lesion was defined as an endoscopically visible flat and/or depressed mucosal lesion with a height less than half the diameter of the lesion.25,26

Continuous variables are expressed as the mean±standard deviation, while categorical variables are presented as absolute values and percentages. Differences between the continuous variables were analyzed using the unpaired Student's t-test, and differences between the categorical variables were analyzed using the χ² test and Fisher's exact test as appropriate.

To assess the inter-observer agreement, the ICC and 95% predictive interval (PI) for the Aronchick and Ottawa scales were calculated. An ICC greater than 0.80 is defined as excellent agreement.27 The statistical correlation between the Aronchick and Ottawa scales was calculated using Spearman's rank correlation test.

The relationship between the bowel preparation status assessed by the Aronchick scale and the missed polyps, adenomas, or advanced adenomas was analyzed using the χ² test for trends. The per-patient miss rate was calculated by dividing the number of patients with missing lesions at the index colonoscopy by the total number of patients.28 The per-polyp miss rate was calculated by dividing the number of missing lesions at the index colonoscopy by the total number of lesions found either on the index or tandem colonoscopies.28

To investigate the risk of missing polyps, adenomas or advanced adenomas in a patient according to the bowel preparation status assessed by Aronchick scale, multivariate analysis was performed using logistic regression analysis adjusted with age, gender, withdrawal time, and number of polyps detected at index colonoscopy. For each variable, the odds ratio (OR) and 95% confidence interval (CI) were reported. A p-value of less than 0.05 was considered to indicate statistical significance. The analyses were performed with SPSS software version 12.0K (SPSS Inc., Chicago, IL, USA).

At 2 weeks after the initial calibration exercise, the inter-observer agreement for assessing the bowel preparation status was first measured using the ICCs for the Aronchick and Ottawa scales, which were 0.749 (95% PI, 0.495 to 0.919) and 0.862 (95% PI, 0.690 to 0.958), respectively. After the second calibration exercise, the ICCs for the Aronchick and Ottawa scales reached 0.822 (95% PI, 0.615 to 0.945) and 0.880 (95% PI, 0.724 to 0.964), respectively.

A total of 277 patients with 942 polyps and 714 adenomas completed the 1st and tandem colonoscopies with polyp removal within 3 months. Table 1 shows the clinical characteristics of the study patients. The mean age was 56.2±11.3 years, and 191 patients were male. There were no significant differences in gender or number of polyps and adenomas detected at the index colonoscopy between the patients with and without missed polyps, adenomas or advanced adenomas. However, the patients with missed adenomas were older than those without a missed advanced adenoma. There was a strong positive correlation between the Aronchick and Ottawa scales (r=0.917; p<0.001).

The bowel preparation of the index colonoscopy, according to the Aronchick scale, was described as excellent in 88 patients (32%), good in 114 patients (41%), fair in 56 patients (20%), poor in 17 patients (6%), and inadequate in two patients (1%). Table 2 shows the per-patient miss rate analysis of polyp, adenoma, and advanced adenoma according to bowel preparation status. When the bowel preparation status was assessed by the Aronchick scale, the per-patient miss rate of polyps, adenomas and advanced adenomas increased significantly as bowel preparation declined from excellent, to good, to fair, and to poor/inadequate (per-patient polyp miss rate, 27%, 35%, 36%, and 58%, respectively, p=0.024; per-patient adenoma miss rate, 21%, 27%, 27%, and 47%, respectively, p=0.040; per-patient advanced adenoma miss rate, 9%, 17%, 18%, and 37%, respectively, p=0.006). In addition, when the bowel preparation was assessed by the Ottawa scale, the score was higher in patients with missed lesions, compared to those without a missed lesion (polyp, 3.4±2.4 vs. 4.2±3.2, p=0.036; adenoma, 3.5±2.5 vs. 4.3±3.2, p=0.046; advanced adenoma, 3.5±2.6 vs. 4.6±3.2, p=0.015) (Table 3).

To identify whether the bowel preparation status is an independent variable associated with missed polyps, adenomas or advanced adenomas, multivariate analyses were performed (Table 4). The patients with poor/inadequate bowel preparation were independently associated with an increased risk of having missed polyps (OR, 3.21; 95% CI, 1.13 to 9.15), missed adenomas (OR, 3.04; 95% CI, 1.04 to 8.88), or missed advanced adenomas (OR, 5.28; 95% CI, 1.58 to 17.68) compared to those with excellent bowel preparation.

Among the total 942 polyps, 187 polyps were missed at the screening colonoscopy, with 19.9% of polyp miss rate. Among the total 714 adenomas, 127 adenomas were missed at the screening colonoscopy, producing 17.8% of adenoma miss rate. When the bowel preparation was excellent, good, fair, or poor/inadequate, the miss rate of polyps, adenomas, and advanced adenomas significantly increased (polyp miss rate, 14%, 17%, 26%, and 48%, respectively, p<0.001; adenoma miss rate, 12%, 17%, 22%, and 40%, respectively, p<0.001; advanced adenoma miss rate, 9%, 6%, 19%, and 58%, respectively, p<0.001) (Table 5).