Abstract
Background
Inadequate bowel preparation (IBP) is associated with reduced adenoma detection. However, limited research has examined the impact of different commercial bowel preparations (CBPs) on IBP and adenoma detection. We aim to determine whether type of CBP used is associated with IBP or adenoma detection.
Methods
We retrospectively evaluated outpatient, screening or surveillance colonoscopies performed in the Cleveland Clinic health system between January 2011 and June 2017. IBP was defined by the Aronchick scale. Multilevel mixed-effects logistic regression was performed to assess the association between CBP type and IBP and adenoma detection. Fixed effects were defined as demographics, comorbidities, medication use, and colonoscopy factors. Random effect of individual endoscopist was considered.
Results
Of 153,639 colonoscopies, 75,874 records met inclusion criteria. Median age was 54; 50% were female; 17.7% had IBP, and adenoma detection rate was 32.6%. In adjusted analyses, compared to GoLYTELY, only NuLYTELY [OR 0.66 (95% CI 0.60, 0.72)] and SuPREP [OR 0.53 (95% CI 0.40, 0.69)] were associated with reduced IBP. Adenoma detection did not vary based on the type of bowel preparation used.
Conclusions
Among patients referred for screening or surveillance colonoscopy, choice of CBP was not associated with adenoma detection. Decisions about CBP should be based on other factors, such as tolerability, cost, or safety.
Keywords: Commercial bowel preparation, Inadequate bowel preparation, Bowel preparation quality, Adenoma detection rate
Introduction
The National Colorectal Cancer Roundtable (NCCRT), established by the Center for Disease Control and American Cancer Society, promulgated the “80% by 2018” colorectal screening initiative in 2014 as a means to maximize colorectal cancer (CRC) screening for all eligible adults ages 50 years or older [1]. If successful, it is estimated that 277,000 cases of CRC and 203,000 CRC related deaths will be prevented by 2030 [2]. However, the effectiveness of any screening colonoscopy is predicated on optimal bowel preparation [3]. Anywhere from 14 to 32% of all outpatient colonoscopies are estimated to have inadequate bowel preparation (IBP) [4-6]. Thus, even if physicians meet the NCCRT goals, unless all procedures are performed with adequate bowel preparation, a substantial number of people will not receive optimal screening.
IBP has been associated with multiple patient factors, including age, male sex, medical comorbidities, and post-surgical anatomy, most of which are not modifiable prior to colonoscopy [7-11]. One modifiable factor is the choice of commercial bowel preparation (CBP) ordered prior to colonoscopy. CBPs have been compared in small randomized trials and in large observational studies. Compared to generic polyethylene glycol, CBPs have been promoted as having greater tolerability and resulting in better cleaning, perhaps justifying their higher price. However, no studies have examined the impact of CBP on adenoma detection rate (ADR). Because the purpose of screening colonoscopy is to identify precancerous adenomas, the ADR represents a more clinically relevant outcome measure than cleaning quality. Because small differences in ADR among the different CBPs could result in large numbers of adenomas missed, it is important to identify whether one is better than the others. In this observational study, we compared CBPs in terms of ADR, IBP, and total procedure time.
Methods
With permission from the Cleveland Clinic’s Institutional Review board (IRB), records of all patients who underwent outpatient screening or surveillance colonoscopies between January 2011 and June 2017 were identified. Given that this study involved retrospective chart review only with no practical means of obtaining informed consent, the IRB exempted the study from needing informed consent from patients. Colonoscopies lacking description of the quality of bowel preparation, employing infrequently used bowel preparations (used < 0.1% of the time), or performed on patients less than 18 years of age were excluded from analysis. For patients who had more than one colonoscopy during this period, only the first colonoscopy was included.
Baseline Characteristics and Possible Predictors of IBP
Demographic information including patient age at the time of colonoscopy, sex, ethnicity, as well as substance use (any history of tobacco use and diagnosis of alcohol abuse), insurance/payor status, and language spoken was obtained from the EHR. International Classification of Disease 9th and 10th revision, Clinical Modifications (ICD-9/10-CM) codes were used to identify patient comorbidities known to be associated with IBP, adenoma detection, or CRC risk. These included diabetes, cirrhosis, dementia, stroke, constipation, coronary artery disease, heart failure, obesity, dysphagia, gastroparesis, hypothyroidism, inflammatory bowel disease (IBD), and irritable bowel syndrome (IBS) [7-11]. We noted all prescriptions for medications associated with these same outcomes, including angiotensin converting enzyme inhibitors (ACE-inhibitors), angiotensin receptor blocker (ARB), calcium channel blockers (CCBs), calcium supplements, vitamin D, estrogen replacement therapy, bisphosphonates, and opiate medications, if they were active at the time of colonoscopy [8-14].
Patient’s risk of CRC was categorized as high or average based on the referring provider’s assessment in the colonoscopy order. Other secondary predictors included in the analysis were colonoscopy indication, timing (before or afternoon, month, and year), location where the colonoscopy was performed, and distance from the patient’s home to the colonoscopy suite.
The brand of CBP ordered, volume of CBP, recommendations to use the CBP as split (taking half of the dose the night prior with the remainder in the morning), and the number of days that elapsed between picking up the CBP and the date of the colonoscopy were extracted. The concordance rate between the type of CBP ordered by the provider and the type of CBP given to the patient at the pharmacy was examined for patients who had Medical Mutual of Ohio’s health insurance.
Outcomes
The primary endpoint for our study was detection of at least one adenoma. Secondary endpoints were IBP and total procedure time (TPT). Adenoma detection was assessed using pathology reports that corresponded to the colonoscopy. Colonoscopy and pathology reports were reviewed using an internally validated natural language processing algorithm to identify presence of adenomas. Determination of IBP was based on the endoscopist’s report. Physician selections of excellent and good in the Provation® system were regarded as adequate bowel preparation (ABP). Descriptions of fair, poor, or inadequate were considered inadequate bowel preparation (IBP). These descriptors were based on Aronchick scale [15]. In addition, our endoscopy software allowed the endoscopist to categorize the bowel preparation as adequate and this too was placed in the adequate category. To understand the relationship between IBP and adenoma detection, we also examined total procedure time (TPT) as another endpoint, hypothesizing that longer procedures could mitigate the effects of IBP on adenoma detection. The TPT was defined as time (in minutes) elapsed from insertion to removal of the colonoscope. Given that intra-procedural cleansing in the setting of poor quality bowel preparation occurs during both insertion and withdrawal (depending on the provider), we have decided to use TPT over the more common process metric of withdrawal time.
Analysis
Baseline characteristics of all included patients were enumerated, and rates of IBP and adenoma detection based on these characteristics were computed. This was reported as frequencies with percentages. Two-sided Chi-square analysis was used to assess significance of differences between categorical variables. Multilevel mixed-effects logistic regression analysis was used to adjust for random effects of endoscopist and fixed effects of patient demographic characteristics, comorbidities, and medication use (“Appendix”).
When encountering inadequately prepared colons, endoscopists frequently spend additional time cleaning the colon, which in turn may lead to improved adenoma detection. Therefore, differences in TPT and adenoma detection were computed by individual descriptors in the Aronchick scale. The significance of difference in TPTs was assessed using Kruskal–Wallis H test, and the significance of difference in adenoma detection was assessed using Chi-square test.
Results
Overall, 75,874 patients were included. Median age was 54 years (25th percentile-44 years; 75th percentile-65 years), and 50.3% were female. The majority of patients were White (69.3%). Overall IBP rate was 17.7%, and adenoma detection rate was 32.6%. Baseline characteristics are described in Table 1. In a subgroup analysis of 876 patients who had Medical Mutual of Ohio health insurance, concordance rate of 96% was noted between the type of CBP prescribed and the type of CBP dispensed at the pharmacy. Differences in demographic characteristics, comorbidities, medication use, and colonoscopic factors based on the defined endpoints of IBP and adenoma detection were calculated (Table 1).
Table 1.
Baseline characteristics along with IBP rate and adenoma detection rate based on patient and colonoscopy-associated characteristics
| Variable | Overall | IBP |
Adenoma detection |
||
|---|---|---|---|---|---|
| N (%) | P value | N (%) | P value | ||
| All colonoscopies | 75,874 (100.0%) | 13,408 (17.7%) | – | 24,719 (32.6%) | – |
| Age | |||||
| < 65 | 23,813 (31.4%) | 4585 (19.3%) | < 0.001 | 9475 (39.8%) | < 0.001 |
| 65+ | 52,061 (68.6%) | 8823 (16.9%) | 15,244 (29.3%) | ||
| Gender | |||||
| Female | 38,189 (50.3%) | 6319 (16.5%) | < 0.001 | 10,260 (26.9%) | < 0.001 |
| Male | 37,685 (49.7%) | 7089 (18.8%) | 14,459 (38.4%) | ||
| Race | |||||
| White | 52,604 (69.3%) | 8486 (16.1%) | < 0.001 | 17,127 (32.6%) | 0.327 |
| Black | 11,871 (15.6%) | 2517 (21.2%) | 3936 (33.2%) | ||
| Hispanic | 6242 (8.2%) | 1307 (20.9%) | 2029 (32.5%) | ||
| Asian | 1759 (2.3%) | 294 (16.7%) | 546 (31%) | ||
| Other | 3398 (4.5%) | 804 (23.7%) | 1081 (31.8%) | ||
| Tobacco use | |||||
| No | 47,699 (62.9%) | 7554 (15.8%) | < 0.001 | 14,305 (30.0%) | < 0.001 |
| Yes | 28,175 (37.1%) | 5854 (20.8%) | 10,414 (37.0%) | ||
| Alcohol use | |||||
| No | 71,115 (93.7%) | 12,466 (17.5%) | < 0.001 | 23,108 (32.5%) | 0.053 |
| Yes | 4759 (6.3%) | 942 (19.8%) | 1611 (33.9%) | ||
| Insurance | |||||
| Private insurance | 46,835 (61.7%) | 7472 (16%) | < 0.001 | 13,986 (29.9%) | < 0.001 |
| Medicare | 25,012 (33.0%) | 5073 (20.3%) | 9432 (37.7%) | ||
| Medicaid | 3372 (4.4%) | 725 (21.5%) | 1102 (32.7%) | ||
| Other | 655 (0.9%) | 138 (21.1%) | 199 (30.4%) | ||
| Primary language | |||||
| English | 73,105 (96.4%) | 12,808 (17.5%) | < 0.001 | 23,741 (32.5%) | 0.002 |
| Other | 2769 (3.6%) | 600 (21.7%) | 978 (35.3%) | ||
| Comorbidities | |||||
| Diabetes | 15,266 (20.1%) | 3742 (24.5%) | < 0.001 | 5836 (38.2%) | < 0.001 |
| Cirrhosis | 4130 (5.4%) | 1050 (25.4%) | < 0.001 | 1481 (35.9%) | < 0.001 |
| Dementia | 1449 (1.9%) | 358 (24.7%) | < 0.001 | 590 (40.7%) | < 0.001 |
| Stroke | 431 (0.6%) | 99 (23.0%) | 0.004 | 169 (39.2%) | 0.003 |
| Constipation | 8395 (11.1%) | 1726 (20.6%) | < 0.001 | 2574 (30.7%) | < 0.001 |
| CAD | 9046 (11.9%) | 2053 (22.7%) | < 0.001 | 3614 (40.0%) | < 0.001 |
| CHF | 3473 (4.6%) | 972 (28%) | < 0.001 | 1468 (42.3%) | < 0.001 |
| Obesity | 15,189 (20%) | 3158 (20.8%) | < 0.001 | 5429 (35.7%) | < 0.001 |
| Dysphagia | 5122 (6.8%) | 1052 (20.5%) | < 0.001 | 1649 (32.2%) | 0.543 |
| Gastroparesis | 299 (0.4%) | 99 (33.1%) | < 0.001 | 95 (31.8%) | 0.766 |
| Hypothyroid | 9108 (12%) | 1658 (18.2%) | 0.156 | 2857 (31.4%) | 0.009 |
| Inflammatory bowel disease | 1554 (2.0%) | 332 (21.4%) | < 0.001 | 205 (13.2%) | < 0.001 |
| Irritable bowel syndrome | 3420 (4.5%) | 567 (16.6%) | 0.087 | 947 (27.7%) | < 0.001 |
| Hospitalization within the previous year | |||||
| No | 65,787 (86.7%) | 11,122 (16.9%) | < 0.001 | 21,274 (32.3%) | < 0.001 |
| Yes | 10,087 (13.3%) | 2286 (22.7%) | 3445 (34.2%) | ||
| Medication use | |||||
| Angiotensin converting enzyme inhibitors | 23,582 (31.1%) | 4881 (20.7%) | < 0.001 | 8550 (36.3%) | < 0.001 |
| Angiotensin receptor blocker | 9933 (13.1%) | 2083 (21%) | < 0.001 | 3804 (38.3%) | < 0.001 |
| Calcium channel blocker | 14,607 (19.3%) | 3201 (21.9%) | < 0.001 | 5559 (38.1%) | < 0.001 |
| Statin | 24,761 (32.6%) | 4648 (18.8%) | < 0.001 | 9212 (37.2%) | < 0.001 |
| Aspirin | 24,303 (32%) | 4543 (18.7%) | < 0.001 | 8729 (35.9%) | < 0.001 |
| Calcium supplements | 29,724 (39.2%) | 5370 (18.1%) | 0.022 | 8929 (30%) | < 0.001 |
| Vitamin D | 41,228 (54.3%) | 7400 (17.9%) | 0.029 | 13,013 (31.6%) | < 0.001 |
| Estrogen | 6049 (8.0%) | 1035 (17.1%) | 0.233 | 1637 (27.1%) | < 0.001 |
| Bisphosphonate | 7006 (9.2%) | 1297 (18.5%) | < 0.001 | 2099 (30%) | < 0.001 |
| Opiates | 5672 (7.5%) | 1416 (25%) | < 0.001 | 1997 (35.2%) | < 0.001 |
| Indication | |||||
| Screening | 52,511 (69.2%) | 9078 (17.3%) | < 0.001 | 15,493 (29.5%) | < 0.001 |
| Surveillance | 23,363 (30.8%) | 4330 (18.5%) | 9226 (39.5%) | ||
| Risk of colorectal cancer | |||||
| High risk | 26,063 (34.4%) | 4528 (17.4%) | 0.119 | 9863 (37.8%) | < 0.001 |
| Average risk | 49,811 (65.6%) | 8880 (17.8%) | 14,856 (29.8%) | ||
| Preparation type | |||||
| GoLYTELY | 50,621 (66.7%) | 9166 (18.1%) | < 0.001 | 16,455 (32.5%) | < 0.001 |
| MoviPrep | 12,619 (16.6%) | 2124 (16.8%) | 4037 (32%) | ||
| NuLYTELY | 5911 (7.8%) | 1198 (20.3%) | 2097 (35.5%) | ||
| CoLyte | 3374 (4.4%) | 454 (13.5%) | 957 (28.4%) | ||
| PREPOPIK | 1354 (1.8%) | 166 (12.3%) | 407 (30.1%) | ||
| SuPREP | 1332 (1.8%) | 113 (8.5%) | 554 (41.6%) | ||
| HalfLytely | 298 (0.4%) | 115 (38.6%) | 97 (32.6%) | ||
| GaviLyte | 224 (0.3%) | 42 (18.8%) | 70 (31.3%) | ||
| OsmoPrep | 141 (0.2%) | 30 (21.3%) | 45 (31.9%) | ||
| Volume of preparation | |||||
| 4000 mL | 60,130 (79.2%) | 10,860 (18.1%) | < 0.001 | 19,579 (32.6%) | 0.837 |
| < 4000 mL | 15,744 (20.8%) | 2548 (16.2%) | 5140 (32.6%) | ||
| Split preparation | |||||
| No | 72,219 (95.2%) | 12,907 (17.9%) | < 0.001 | 23,390 (32.4%) | < 0.001 |
| Yes | 3655 (4.8%) | 501 (13.7%) | 1329 (36.4%) | ||
| Time of colonoscopy | |||||
| AM | 60,374 (79.6%) | 10,577 (17.5%) | 0.030 | 19,632 (32.5%) | 0.474 |
| PM | 15,500 (20.4%) | 2831 (18.3%) | 5087 (32.8%) | ||
| Sedation | |||||
| Conscious sedation | 63,941 (84.3%) | 10,072 (15.8%) | < 0.001 | 20,675 (32.3%) | 0.001 |
| Anesthesia | 11,933 (15.7%) | 3336 (28%) | 4044 (33.9%) | ||
| Year | |||||
| 2011 | 6558 (8.6%) | 1277 (19.5%) | < 0.001 | 2097 (32%) | < 0.001 |
| 2012 | 10,821 (14.3%) | 2549 (23.6%) | 3414 (31.5%) | ||
| 2013 | 11,765 (15.5%) | 2763 (23.5%) | 3700 (31.4%) | ||
| 2014 | 12,507 (16.5%) | 2333 (18.7%) | 4062 (32.5%) | ||
| 2015 | 13,989 (18.4%) | 1955 (14%) | 4617 (33%) | ||
| 2016 | 13,581 (17.9%) | 1725 (12.7%) | 4491 (33.1%) | ||
| 2017 | 6653 (8.8%) | 806 (12.1%) | 2338 (35.1%) | ||
| Location | |||||
| Main campus (Tertiary center) | 17,175 (22.6%) | 3667 (21.4%) | < 0.001 | 6010 (35%) | < 0.001 |
| Community hospitals | 15,634 (20.6%) | 3839 (24.6%) | 5142 (32.9%) | ||
| Ambulatory health centers | 43,065 (56.8%) | 5902 (13.7%) | 13,567 (31.5%) | ||
| Distance to center | |||||
| ≤ 10 mi | 54,876 (72.3%) | 9727 (17.7%) | 0.528 | 17,480 (31.9%) | < 0.001 |
| > 10 mi | 20,998 (27.7%) | 3681 (17.5%) | 7239 (34.5%) | ||
| Prep pickup date (prior to procedure) | |||||
| ≥ 3 days | 69,017 (91.0%) | 11,491 (16.6%) | < 0.001 | 22,504 (32.6%) | 0.609 |
| < 3 days | 6857 (9.0%) | 1917 (28.0%) | 2215 (32.3%) | ||
Comparison of CBPs
In unadjusted analyses, significant differences were noted among the CBPs for each of the three endpoints. The highest and the lowest IBP rates were noted with HalfLytely® (38.6%) and SuPREP® (8.5%), respectively. With regard to adenoma detection, SuPREP® had the highest adenoma detection rate (41.6%), while CoLyte® had the lowest adenoma detection rate (28.4%). Univariable analysis demonstrating these differences in terms of odds ratios is included in Table 2. After adjusting for variables listed in “Appendix,” compared to GoLYTELY®, NuLYTELY® [odds ratio (OR), 0.66; 95% confidence interval (95% CI), 0.60, 0.72] and SuPREP® (OR 0.53; 95% CI 0.40, 0.69) had reduced odds of IBP. The other CBPS were not significantly different from GoLYTELY® (Table 3). Adjusted rates of ADR do not differ significantly by the type of CBP used. Adjusted rates of IBP and ADR by bowel preparation are depicted in Fig. 1a, b, respectively. Multivariable analysis has also identified that split preparations were associated with reduced odds of IBP (OR 0.77; 95% CI 0.69, 0.86). Afternoon colonoscopies (OR 0.99; 95% CI 0.94, 1.04) or use of small volume bowel preparations—defined as less than four liters (OR 1.03; 95% CI 0.97, 1.09)—was not associated with bowel preparation quality after this adjusted analysis.
Table 2.
Unadjusted analysis comparing commercial bowel preparations with GoLYTELY
| Bowel preparation | IBP |
Adenoma detection |
||
|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | |
| MoviPrep | 0.92 (0.87, 0.96) | < 0.001 | 0.98 (0.94, 1.02) | 0.269 |
| NuLYTELY | 1.15 (1.07, 1.23) | < 0.001 | 1.14 (1.08, 1.21) | < 0.001 |
| CoLyte | 0.70 (0.64, 0.78) | < 0.001 | 0.82 (0.76, 0.89) | < 0.001 |
| PREPOPIK | 0.63 (0.54, 0.74) | < 0.001 | 0.89 (0.79, 1.00) | 0.058 |
| SuPREP | 0.42 (0.35, 0.51) | < 0.001 | 1.48 (1.32, 1.65) | < 0.001 |
| HalfLytely | 2.84 (2.25, 3.59) | < 0.001 | 1 (0.79, 1.28) | 0.987 |
| GaviLyte | 1.04 (0.75, 1.46) | 0.803 | 0.94 (0.71, 1.25) | 0.689 |
| OsmoPrep | 1.22 (0.82, 1.83) | 0.330 | 0.97 (0.68, 1.39) | 0.881 |
Table 3.
Adjusted analysis comparing commercial bowel preparations with GoLYTELY
| Bowel preparation | IBP |
Adenoma detection |
||
|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | |
| MoviPrep | 0.97 (0.91, 1.03) | 0.318 | 0.96 (0.92, 1.01) | 0.141 |
| NuLYTELY | 0.66 (0.60, 0.72) | < 0.001 | 1.04 (0.96, 1.12) | 0.340 |
| CoLyte | 0.96 (0.84, 1.09) | 0.537 | 1.04 (0.94, 1.15) | 0.488 |
| PREPOPIK | 1.06 (0.87, 1.27) | 0.575 | 0.89 (0.78, 1.02) | 0.087 |
| SuPREP | 0.53 (0.40, 0.69) | < 0.001 | 0.99 (0.83, 1.17) | 0.903 |
| HalfLytely | 1.16 (0.89, 1.53) | 0.270 | 1.02 (0.79, 1.33) | 0.851 |
| GaviLyte | 1.31 (0.9, 1.92) | 0.164 | 0.84 (0.62, 1.14) | 0.259 |
| OsmoPrep | 1.03 (0.66, 1.62) | 0.898 | 1.20 (0.83, 1.74) | 0.329 |
Please refer to “Appendix” for variables included in the adjusted analysis
Fig. 1.
a Adjusted IBP rate and 95% confidence interval for each commercial bowel preparation. NuLYTELY and SuPREP are lower than the others (P < 0.001). b Adjusted ADR and 95% confidence interval for each commercial bowel preparation. Variables included in the adjusted analysis are included in “Appendix”
TPT and Adenoma Detection by Individual Descriptors of Aronchick Scale
To understand why IBP did not translate into lower adenoma detection, we examined the adenoma detection rate by the description of bowel preparation quality. The adenoma detection rate for excellent, good, and fair was all similar, while poor and inadequate preps had lower adenoma detection rates (Fig. 2a). Thus, although some CBPs resulted in lower rates of IBP, these were primarily the result of fewer fair preps. We then examined the procedure times associated with each description (Fig. 2b) to see whether fair preps may have been salvaged by more assiduous cleaning on the part of the endoscopists. We found that patients with fair preps indeed had procedure times that were 4 min longer than either excellent or good preps (P < 0.01).
Fig. 2.
a Unadjusted ADR according to Aronchick scale descriptor (P < 0.001). b Unadjusted TPT according to Aronchick scale descriptor (P < 0.001)
In this analysis, it should be noted that 34.7% of the included patients did not have TPT. The proportion of missing TPTs were unequally distributed by the individual descriptors on the Aronchick scale. Whereas 67% missing TPTs were noted in patients described as having inadequate on Aronchick scale, only 17% of those with fair bowel preparation had missing TPTs.
Discussion
In this large study of patients undergoing screening and surveillance colonoscopies within a single large health system, we found that many patient and procedural factors were associated with both IBP and adenoma detection. After adjusting for these factors, compared to GoLYTELY, patients receiving SuPREP and NuLYTELY had 48% and 35% reduced odds of having IBP. However, this advantage did not translate into a difference in adenoma detection. The reason for this appeared to depend on the definition of IBP. Compared to patients who received GoLYTELY, those who received SuPREP and NuLYTELY were more likely to have preps described as good, which was considered adequate, and less likely to have ones described as fair, which was considered inadequate. However, we found that there was no difference in adenoma detection rate between good and fair preps, whereas patients with preps described as poor or inadequate did have lower adenoma detection rate. This was likely due to the ability of endoscopists to salvage fair preps through additional cleaning during the procedure. This idea is supported by our finding that patients with fair preps had longer procedure times.
Thus far, only two large-scale studies have evaluated the impact of CBP on IBP. Martin et al. analyzed 28,782 preps from a database of patients who underwent screening or surveillance colonoscopies in Central Illinois. After adjusting for a limited set of variables, they too found that sodium sulfate-based bowel preparation (such as SuPREP®) was superior to polyethylene glycol (PEG)-based preparations [16]. Gu et al. evaluated 4339 colonoscopies and noted that SuPREP® (OR = 7.28) had the greatest odds of cleansing followed by MoviPrep® (OR = 7.11) and MiraLAX with Gatorade (OR = 7.09) [17]. Further research is needed to understand the pathophysiological or practical reasons for this difference. In addition, they concluded that patients using sodium phosphate-based bowel preparations (such as OsmoPrep®) had improved bowel preparation quality and that magnesium sulfate-based bowel preparations were associated with worse bowel preparation quality—two findings we did not observe. Although not as large as the previous studies, Gu et al. demonstrated MoviPrep®, SuPREP®, and MiraLAX® mixed with Gatorade® had superior bowel cleansing [17]. None of these studies examined how use of commercial bowel preparations affected adenoma detection, which is arguably the most important quality indicator for a screening or surveillance colonoscopy.
In a randomized-control study, sulfate-free PEG (such as NuLYTELY®) bowel preparations were shown to be non-inferior to PEG–electrolyte lavage solution (such as GoLYTELY®) [18]; we found that NuLYTELY® bowel preparations were associated with reduced rates of IBP compared to GoLYTELY®. Better taste associated with sulfate-free preparations compared to the traditional PEG–electrolyte lavage solution may have contributed to tolerability of the former [18]. This in turn may have led to better adherence, contributing to improved bowel preparation quality. Further studies are needed to clarify specific mechanisms by which sulfate-free PEG might contribute to improved quality of bowel preparation.
Our finding that differences in IBP rate did not translate into higher adjusted adenoma detection rate appears novel and has important implications for clinical practice. Adenoma detection, not adequate prep, is the primary goal of colonoscopy. Although multiple studies have noted the importance of bowel preparation quality to adenoma detection [3, 19], it appears that the differences in IBP rates produced by the type of CBP used occur primarily in the fair to good range, which has minimal impact on adenoma detection, because even fair preps can be salvaged by a diligent endoscopist. Therefore, in determining the type of CBP to use for a patient undergoing an outpatient screening or surveillance colonoscopy, patient tolerability, safety profile, costs, etc., should be accorded a higher weight. Indeed, the cost of a CBP can range anywhere from over $220 dollars for OsmoPrep® to $9 dollars for GoLYTELY® [20], and since cost may pose a barrier to screening, using the cheapest bowel preparation that is tolerable makes sense, especially when considering the costs at a population level. One caveat is that CBPs that produce fair results may require more time in the endoscopy suite.
Regardless of CBP chosen, split dose preparations were associated with a 33% reduction in IBP in our study. Other studies have also demonstrated benefits of using a split preparation [21-23]. Because the improvement in IBP rates in our study did not translate into improved adenoma detection, the primary benefit of split dosing may be reduced procedure time, although a recent prospective study also noted that patients receiving split dosing had reduced inconvenience, and decreased intensity and duration of bowel movements [24]. Consistent with previous research, IBP and adenoma detection did not vary significantly by bowel preparation volume [25].
Previous studies have investigated timing of colonoscopy (morning versus afternoon) as another modifiable factor. Some have noted improved bowel preparation quality associated with morning colonoscopies [26-28], while others have not [29]. We found slightly lower IBP rates for morning colonoscopies, but adjusting for other variables largely mitigated the association. Additionally, in keeping with previous research, no difference in adenoma detection rate was appreciated based on colonoscopy timing [30]. Given these findings, there is no need for colonoscopy centers to accommodate more patients during morning hours. Instead, factors such as patient convenience, endoscopist availability, and costs ought to play a role in these decisions.
Our conclusions should be considered within the limitations of our study design. First, due to the retrospective nature of our analysis, our data may be subject to misclassification. Diagnostic codes used to identify comorbidities can vary widely in accuracy depending on the disease condition [31]. Next, although modifiable factors such as written, nursing initiated, or video-administered patient education have demonstrably reduced IBP [32, 33], we did not have access to this information. However, it is unlikely that these factors would significantly affect adenoma detection. Further, medications prescribed by providers outside of our health system may not have been captured. We were also unable to ascertain whether a patient who was prescribed certain medications had adhered to the physician recommendation. However, with regard to the type of CBP, our subgroup analysis noted that over 95% of the bowel preparations ordered by the physicians are the ones actually being given to patients.
Although standard printed instructions are given to all patients undergoing colonoscopy, pertinent variables such as the level of patient’s comprehension of instructions, the nature of verbal instructions provided to the patients, patient’s incomplete consumption of bowel preparation volume, and assessment of effluent clarity after bowel preparation were not able to be evaluated. Our study’s finding that less than five percent of patients were instructed to split the bowel preparation dose may also be related our inability to obtain instructions given through other means, including verbal. Alternatively, the low rate could also be explained by the fact that split dosing has only become progressively popular within our institution over the study period.
Finally, it must also be noted that assessment of the quality of bowel preparation is to some degree subjective. However, in addition to some data showing fair to substantial inter-observer reliability with the Aronchick scale [34], our use of multilevel logistic regression controls for random effects of endoscopist. Given greater number of studies validating Boston Bowel Preparation [35], use of this metric to assess IBP would have been ideal. Further, with Aronchick scale, it is not clear whether endoscopists are rating the quality of bowel preparation during insertion or after cleansing. However, given that Aronchick scale was the only required measure elicited by our endoscopic electronic medical record, very few providers used Boston Bowel Preparation score. Therefore, we had to rely on Aronchick scale alone in determining bowel preparation quality.
Conclusions
In this retrospective study, using data from a large health system, SuPREP and NuLYTELY were associated with reduced IBP compared to GOLYTELY, as was use of split preparation. Although IBP itself was associated with reduced adenoma detection, the type of CBP and the use of split preparation were not. In choosing a commercial bowel preparation, factors such as patient tolerability, cost, and safety profile should take precedence.
Acknowledgments
Shashank Sarvepalli is supported by a National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Institutional National Service Award (T32) from the National Institutes of Health under Award Number T32 DK083266.
Appendix
Multilevel mixed-effects logistic regression was performed to evaluate the effect of CBP type on IBP and ADR. Analysis was adjusted for random effects of individual providers and fixed effects such as patient’s age, gender, ethnicity, any history of tobacco use, alcohol abuse (as determined by clinician), insurance/payor type, language spoken, comorbid conditions (diabetes, cirrhosis, dementia, stroke, constipation, coronary artery disease, congestive heart failure, obesity, dysphagia, gastroparesis, hypothyroidism, inflammatory bowel disease, irritable bowel syndrome), hospitalization within the last year, medication use (angiotensin converting enzyme receptor inhibitor, angiotensin receptor blocker, calcium channel blocker, statin, aspirin, calcium, vitamin D, estrogen, bisphosphonate, opiates use), indication, CRC risk, timing of colonoscopy (before noon vs. afternoon), sedation type, total procedure time, miles from colonoscopy center, location of the colonoscopy center, when bowel preparation was picked up.
Footnotes
Conflict of interest Authors have no conflict of interest to disclose pertaining to this study.
References
- 1.Becker EA, Buscemi J, Fitzgibbon ML, et al. Society of Behavioral Medicine (SBM) position statement: SBM supports the National Colorectal Cancer Roundtable’s (NCCRT) call to action to reach 80% colorectal cancer screening rates by 2018. Transl Behav Med. 2016;6:306–308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Meester RG, Doubeni CA, Zauber AG, et al. Public health impact of achieving 80% colorectal cancer screening rates in the United States by 2018. Cancer. 2015;121:2281–2285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Chokshi RV, Hovis CE, Hollander T, Early DS, Wang JS. Prevalence of missed adenomas in patients with inadequate bowel preparation on screening colonoscopy. Gastrointest Endosc. 2012;75:1197–1203. [DOI] [PubMed] [Google Scholar]
- 4.Cheng RW, Chiu YC, Wu KL, et al. Predictive factors for inadequate colon preparation before colonoscopy. Tech Coloproctol. 2015;19:111–115. [DOI] [PubMed] [Google Scholar]
- 5.Hassan C, Fuccio L, Bruno M, et al. A predictive model identifies patients most likely to have inadequate bowel preparation for colonoscopy. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc. 2012;10:501–506. [DOI] [PubMed] [Google Scholar]
- 6.Van Dongen M. Enhancing bowel preparation for colonoscopy: an integrative review. Gastroenterol Nurs. 2012;35:36–44. [DOI] [PubMed] [Google Scholar]
- 7.Jawa H, Mosli M, Alsamadani W, et al. Predictors of inadequate bowel preparation for inpatient colonoscopy. Turk J Gastroenterol. 2017;28:460–464. [DOI] [PubMed] [Google Scholar]
- 8.Gandhi K, Tofani C, Sokach C, Patel D, Kastenberg D, Daskalakis C. Patient characteristics associated with quality of colonoscopy preparation: a systematic review and meta-analysis. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc. 2018;16:357–369. [DOI] [PubMed] [Google Scholar]
- 9.Rex DK. Optimal bowel preparation–a practical guide for clinicians. Nat Rev Gastroenterol Hepatol. 2014;11:419–425. [DOI] [PubMed] [Google Scholar]
- 10.Chaptini L, Laine L. Can I improve my adenoma detection rate? J Clin Gastroenterol. 2015;49:270–281. [DOI] [PubMed] [Google Scholar]
- 11.Brenner H, Kloor M, Pox CP. Colorectal cancer. Lancet (London, England). 2014;383:1490–1502. [DOI] [PubMed] [Google Scholar]
- 12.Kumar K, Mohindra S, Raj M, Choudhuri G. Procalcitonin as a marker of sepsis in alcoholic hepatitis. Hepatol Int. 2014;8:436–442. [DOI] [PubMed] [Google Scholar]
- 13.Bardou M, Barkun A, Martel M. Effect of statin therapy on colorectal cancer. Gut. 2010;59:1572–1585. [DOI] [PubMed] [Google Scholar]
- 14.Sehdev A, O’Neil BH. The role of aspirin, vitamin d, exercise, diet, statins, and metformin in the prevention and treatment of colorectal cancer. Curr Treat Options Oncol. 2015;16:43. [DOI] [PubMed] [Google Scholar]
- 15.Aronchick CA, Lipshutz WH, Wright SH, Dufrayne F, Bergman G. A novel tableted purgative for colonoscopic preparation: efficacy and safety comparisons with Colyte and Fleet Phospho-Soda. Gastrointest Endosc. 2000;52:346–352. [DOI] [PubMed] [Google Scholar]
- 16.Martin D, Walayat S, Ahmed Z, et al. Impact of bowel preparation type on the quality of colonoscopy: a multicenter community-based study. J Commun Hosp Intern Med Perspect. 2016;6:31074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Gu P, Lew D, Oh SJ, et al. Comparing the real-world effectiveness of competing colonoscopy preparations: results of a prospective trial. Am J Gastroenterol. 2019;114:305–314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.DiPalma JA, Marshall JB. Comparison of a new sulfate-free polyethylene glycol electrolyte lavage solution versus a standard solution for colonoscopy cleansing. Gastrointest Endosc. 1990;36:285–289. [DOI] [PubMed] [Google Scholar]
- 19.Ko CW, Dominitz JA, Green P, Kreuter W, Baldwin LM. Specialty differences in polyp detection, removal, and biopsy during colonoscopy. Am J Med. 2010;123:528–535. [DOI] [PubMed] [Google Scholar]
- 20.Good Rx. Stop paying too much for your prescriptions. 2019
- 21.Woo DH, Kim KO, Jeong DE, et al. Prospective analysis of factors associated with inadequate bowel preparation for colonoscopy in actual clinical practice. Intest Res. 2018;16:293–298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Siddiki H, Ravi S, Al-Qaisi MT, et al. Comparison of small versus large volume split dose preparation for colonoscopy: a study of colonoscopy outcomes. Dig Dis Sci. 2018;63:2413–2418. 10.1007/s10620-018-5102-8. [DOI] [PubMed] [Google Scholar]
- 23.Perreault G, Goodman A, Larion S, et al. Split-versus single-dose preparation tolerability in a multiethnic population: decreased side effects but greater social barriers. Ann Gastroenterol. 2018;31:356–364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Horton N, Garber A, Hasson H, Lopez R, Burke CA. Impact of single- vs. split-dose low-volume bowel preparations on bowel movement kinetics, patient inconvenience, and polyp detection: a prospective trial. Am J Gastroenterol.. 2016;111:1330–1337. [DOI] [PubMed] [Google Scholar]
- 25.Kang X, Zhao L, Zhu Z, et al. Same-day single dose of 2 liter polyethylene glycol is not inferior to the standard bowel preparation regimen in low-risk patients: a randomized, controlled trial. Am J Gastroenterol. 2018;113:601–610. [DOI] [PubMed] [Google Scholar]
- 26.Sanaka MR, Shah N, Mullen KD, Ferguson DR, Thomas C, McCullough AJ. Afternoon colonoscopies have higher failure rates than morning colonoscopies. Am J Gastroenterol. 2006;101:2726–2730. [DOI] [PubMed] [Google Scholar]
- 27.Wells CD, Heigh RI, Sharma VK, et al. Comparison of morning versus afternoon cecal intubation rates. BMC Gastroenterol. 2007;7:19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Gupta T, Mandot A, Desai D, Abraham P, Joshi A, Shah S. Comparison of two schedules (previous evening versus same morning) of bowel preparation for colonoscopy. Endoscopy. 2007;39:706–709. [DOI] [PubMed] [Google Scholar]
- 29.Eun CS, Han DS, Hyun YS, et al. The timing of bowel preparation is more important than the timing of colonoscopy in determining the quality of bowel cleansing. Dig Dis Sci. 2011;56:539–544. 10.1007/s10620-010-1457-1. [DOI] [PubMed] [Google Scholar]
- 30.Singh S, Dhawan M, Chowdhry M, Babich M, Aoun E. Differences between morning and afternoon colonoscopies for adenoma detection in female and male patients. Ann Gastroenterol. 2016;29:497–501. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Quan H, Li B, Saunders LD, et al. Assessing validity of ICD-9-CM and ICD-10 administrative data in recording clinical conditions in a unique dually coded database. Health Serv Res. 2008;43:1424–1441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Hsueh FC, Wang HC, Sun CA, et al. The effect of different patient education methods on quality of bowel cleanliness in outpatients receiving colonoscopy examination. Appl Nurs Res. 2014;27:e1–e5. [DOI] [PubMed] [Google Scholar]
- 33.Prakash SR, Verma S, McGowan J, et al. Improving the quality of colonoscopy bowel preparation using an educational video. Can J Gastroenterol. 2013;27:696–700. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Parmar R, Martel M, Rostom A, Barkun AN. Validated scales for colon cleansing: a systematic review. Am J Gastroenterol. 2016;111:197–204. (quiz 205). [DOI] [PubMed] [Google Scholar]
- 35.Kastenberg D, Bertiger G, Brogadir S. Bowel preparation quality scales for colonoscopy. World J Gastroenterol. 2018;24:2833–2843. [DOI] [PMC free article] [PubMed] [Google Scholar]