1. Introduction
Colorectal cancer (CRC) remains the second leading cause of cancer-related mortality in the US[1] and the third leading cause in the world.[2] Over the last 3 decades, the health systems were able to decrease cancer-related death for many cancers including CRC by 30%.[1] Factors that contribute to the reduction in CRC mortality, include increasing awareness of CRC screening, achieving high penetrance of CRC screening close to 70% in the US,[3] and detecting advanced polyps and tumors at early stages.[4,5] Adenomatous polyp resections during colonoscopies have been shown to decrease the incidence of CRC.[6,7] Thus, adenoma detection rates (ADR) have been part of major quality indicators of screening colonoscopy, which includes quality of bowel preparation, cecal intubation rate, and withdrawal time.[8,9] The American Society of Gastrointestinal Endoscopy published guidelines recommended ADR of above 25% in both female and male populations (for men 30%, for women 20%) for healthy, asymptomatic patients during screening exams.[8] Recent studies have reported that colonoscopies are less effective in preventing right-sided CRC (from cecum to proximal two-thirds of transverse colon) compared with left-sided CRC (distal one-third of transverse colon to rectum).[10,11] Qualitative factors such as missed or incompletely resected lesions, suboptimal bowel preparation, and differences in tumorigenesis contribute to its ineffectiveness in detecting right-sided CRC.
Recent publications have established that gastroenterologists are more effective than surgeons, primary care physicians, and nurse practitioners at preventing CRC by colonoscopy.[10,12–14] During the last 2 decades, interest and opportunities for further training in specialized clinical practice after completing general gastroenterology fellowships have become widespread in the areas of interventional endoscopists, inflammatory bowel disease (IBD) specialists, motility specialists, and hepatologists. These subspecialists all continue to perform CRC screening colonoscopies, but data comparing their performances have not been evaluated. Due to their varying degree of clinical focus and exposure as well as training, we, therefore, aim to compare and identify any differences between the quality of the different specialist colonoscopy procedures and whether different milestones should be set.
2. Methods
2.1. Study design and participants
This is a retrospective study of all first-time screening colonoscopies performed from 2010 to 2020 at the academic outpatient ambulatory surgical center in Hershey, Pennsylvania. No trainees are involved with these procedures. The inclusion criteria were patients >45 years of age that have undergone their first screening colonoscopy. The exclusion criteria were patients with a history of a prior screening colonoscopy, aborted and incomplete colonoscopy procedures with documented reasons for incompletion, history of bowel resection in the past for any reason. The study was approved by the Institution Review Board Penn State University (16914).
2.2. Data collection
Patient-specific data included gender, age at the time of procedure, date of procedure, and personal and family history. Provider-specific data included age, gender, graduation year from general gastroenterology fellowship, additional training, graduation year from the additional training fellowship, and their focus on clinical practice. The focus of clinical practice is based on current focus of clinical practice and/or the completion of an extra year of training in the respected focus of practice.
To determine the quality of the screening colonoscopies, we collected patient-related data on the gender, the start and end time of the procedure, the time to reach the cecum, the quality of the bowel preparation, the completeness of the procedure, the polyp detection rate and the histologic characteristics of the polyp (adenoma, sessile serrated polyps [SSP], hyperplastic and inflammatory polyps). We recorded the number of adenomas and SSP total per procedure. We also gathered information on CRC risks and classified them as average versus high risk (family history of polyps or cancer). The cecal intubation time was defined as the time from the insertion of the colonoscope into the rectum until the identification of the cecum. It was calculated by taking the difference between the time to reach the cecum from the time the procedure began. The withdrawal time is defined as the time from the identification of the cecum to when the colonoscope was withdrawn from the anus. Similarly, it was calculated by taking the difference between the time the procedure ended from the time of cecal intubation. This time frame also included any polypectomies or biopsies that the endoscopist performed during the procedure. All times were rounded to the nearest minute. The quality of the bowel preparation was graded based on the overall procedure as described by the provider performing the colonoscopy and characterized as excellent, good, adequate, fair, poor, and inadequate.
2.3. Data management and statistical analysis
After initial quality analysis for the completeness of the data, cases with incomplete information were excluded (Fig. 1). A few new variables were generated using the available data. Quality of bowel preparation was condensed from 5 possible grades into 2 for the purpose of our analysis. We defined acceptable bowel prep as excellent, good, and adequate whereas inadequate bowel prep includes fair and poor. ADR alone and adenoma with SSP detection rate (ADR + SSP) were calculated. The year of performing the procedure was subtracted from the general gastroenterology fellowship graduation year and was calculated and defined as a separate variable named, years out of training. Endoscopists were divided into their respective subspecialty clinical focus (general/motility, hepatology, IBD, interventional endoscopy). Numerical variables are reported with mean +/− standard deviation. Univariate analysis was performed using chi-squared test for categorical variables and Student t test and ANOVA where appropriate. The primary outcome was ADR. The secondary outcome was adenoma and/or SSP detection rate. Binary logistic regression analysis using backward stepwise elimination based on the probability of the Wald statistic was performed including all clinical and statistically plausible variables. A 2-sided P value <0.05 was considered statistically significant. Data analysis was conducted by utilizing IBM Statistical Package for Social Sciences (IBM, Armonk, NY), version 27.
;)
Figure 1.: Patient procedural demographic breakdown.
3. Results
3.1. Baseline characteristics
A total of 5290 screening colonoscopies from January 1, 2010, to December 31, 2020, met the inclusion criteria. After excluding 19 patients (0.3%) due to incomplete or aborted procedures, data from 5271 procedures performed by 16 gastroenterologists were analyzed, Figure 1. Characteristics of eligible patients are shown in Table 1 and Figure 1. Forty-nine percent of all patients (n = 2590) who underwent screening colonoscopy were men. For patient risk factors, 1014 patients (19.2%) had an unknown risk factor history and could not be designated to the other remaining categories. A small portion of patients had either a family history of polyps (n = 565, 10.7%) or a family history of cancer (n = 404, 7.7%). The quality of the bowel prep in Table 1 was recategorized into 2 groups (acceptable or inadequate), which showed that 87.6% of patients had acceptable bowel prep (n = 4619). Characteristics of all the gastroenterologists performing screening colonoscopies at ambulatory surgical centers are shown in Table 2.
Table 1 -
Provider-related information.
|
Provider breakdown |
Procedures breakdown |
| Provider characteristics |
N = 16 |
N = 5,271 |
| Gender (Provider), N (%) |
|
|
| Male |
10 (62.5) |
3593 (68.2) |
| Female |
6 (37.5) |
1678 (31.8) |
| Out of training, median (range) yr |
11 (1–40) |
– |
| Subspecialty focus, N (%) |
|
|
| General/Motility |
3 (18.8%) |
964 (18.3) |
| Hepatology |
3 (18.8%) |
1324 (25.1) |
| IBD |
4 (25%) |
909 (17.2) |
| Interventional endoscopy |
6 (37.5%) |
2074 (39.3) |
| Total detected adenoma, N (%) |
|
|
| Detected |
– |
1784 (33.8) |
| Not detected |
– |
3419 (64.9) |
| Unknown |
– |
68 (1.3) |
| Total detected adenoma and SSP, N (%) |
|
|
| Detected |
– |
2050 (38.9) |
| Not detected |
– |
3192 (60.6) |
| Unknown |
– |
29 (0.6) |
IBD = inflammatory bowel disease, SSP = sessile serrated polyps
Table 2 -
Procedural breakdown by specialty focus.
|
General/Motility |
Hepatology |
IBD |
Interventional endoscopy |
|
N (%) |
N (%) |
N (%) |
N (%) |
| Patient gender |
|
|
|
|
| Male |
445 (46.2) |
611 (46.1) |
465 (51.2) |
1069 (51.5) |
| Risk |
|
|
|
|
| Average |
661 (79.4) |
815 (80.0) |
544 (77.8) |
1268 (74.3) |
| Family history of polyps |
93 (11.2) |
127 (12.5) |
97 (13.9) |
248 (14.5) |
| Family history of cancer |
79 (9.5) |
77 (7.5) |
58 (8.3) |
190 (11.1) |
| Prep quality |
|
|
|
|
| Excellent |
447 (46.6) |
212 (16.0) |
218 (24.0) |
447 (21.6) |
| Good |
389 (40.4) |
842 (63.6) |
420 (46.2) |
1234 (59.5) |
| Adequate |
49 (5.1) |
60 (4.5) |
189 (20.8) |
112 (5.4) |
| Fair |
71 (7.4) |
195 (14.7) |
11.1 (7.0) |
42.7 (11.9) |
| Poor |
8 (0.8) |
15 (1.1) |
18 (2.0) |
35 (1.7) |
| ADR |
|
|
|
|
| Detected |
264 (27.5) |
409 (31.4) |
346 (38.4) |
765 (37.5) |
| ADR + SSP |
|
|
|
|
| Detected |
298 (31.0) |
467 (35.5) |
396 (43.6) |
889 (43.2) |
ADR = adenoma detection rates, ADR + SSP = adenoma and sessile serrate polyp detection rate, IBD = inflammatory bowel disease.
3.2. Primary outcomes
The details of the procedures performed by the endoscopists are shown in Table 1. Out of the 5271 procedures, 68.2% of them (n = 3593) were performed by a male endoscopist while 31.8% of them (n = 1678) were performed by a female endoscopist. Adenomatous polyps were reported in 1274 (36%) and 510 (30.6%) patients by male and female endoscopists (P < .001), respectively. Approximately 44% of the providers received formal additional training after general fellowship. The median years out of training between the 16 providers is 11.0 years. When comparing the adenoma-detected group versus the undetected group, the mean years of being out of training (SD) were 12.6 (8.7) and 13.7 (9.6) years, respectively (P < .001). When comparing specialty focus, interventional endoscopists performed the most procedures (39.3%, n = 2074) and IBD specialists performed the least procedures (17.2%, n = 909). For 68 procedures, documentation of adenoma detection was not identified. Overall, the endoscopists had an ADR of 33.8% (n = 1784). For 29 procedures, documentation of adenoma and SSP were not identified. With this consideration, the total detection rate of adenoma and SSP is 38.9% (n = 2050).
The percentage of excellent bowel prep was reported the most by general/motility specialists (46.4%), and the percentage of good bowel prep was reported most frequently by hepatologists (63.6%) (Table 2). When comparing bowel prep quality in our condensed categories, acceptable bowel prep was reported with a higher percentage in general/motility (91.8%, n = 885) and IBD (91.0%, n = 827) specialists compared to hepatology (84.1%, n = 1114) and interventional endoscopy (86.5%, n = 1793) specialists. None of the general/motility specialists received any additional training. The ADR and ADR + SSP were most detected in IBD specialists (38.4%, 43.6% respectively) and they were least detected in general/motility specialists (27.5%, 31.0% respectively). Hepatologists and interventional endoscopists both had more procedures performed by endoscopists with additional training (61.4% and 58.8% respectively) compared to general/motility and IBD specialists (0% and 16.7% respectively).
The procedural times are detailed in Figure 2. For general/motility specialists, they had mean years out of training of 19.9 years (95% CI: 19.0–20.8), mean adenoma per procedure of 0.45 (95% CI: 0.39–0.51), and mean polyp per procedure of 0.88 (95% CI: 0.77–0.99). For hepatologists, they had a mean year out of training of 13.5 years (95% CI: 13.2–13.7), mean adenoma per procedure of 0.47 (95% CI: 0.42–0.52), mean polyp per procedure of 0.95 (95% CI: 0.89–1.02). For IBD specialists, they had a mean year out of training of 5.4 years (95% CI: 5.3–5.6), mean adenoma per procedure of 0.65 (95% CI: 0.58–0.73), and mean polyp per procedure of 1.20 (95% CI: 1.10–1.30). For interventional endoscopists, they had a mean year out of training of 13.6 years (95% CI: 13.2–13.9), mean adenoma per procedure of 0.64 (95% CI: 0.59–0.69), and mean polyp per procedure of 1.05 (95% CI: 1.07–1.20).
Figure 2.: Cecum intubation, withdrawal, and total procedural time divided based on specialty focus.
3.3. Multivariate analysis
We performed logistic regression analysis to determine if any of the variables affected the adenoma and/or polyp detection rates. By utilizing the backward stepwise elimination with the Wald statistic, we identified statistically significant odds ratios (OR) and reported them in Table 3. When comparing the analysis between ADR only and ADR + SSP, adequacy of bowel preparation with OR of 1.29 (95% CI: 1.06–1.56, P = .010) remained in the final model as an important independent factor in the detection of polyp in addition to other factors found in the model for ADR.
Table 3 -
Final model of regression analysis finding independent factors for detecting polyps.
|
Model A adenoma |
Model B adenoma &/or sessile serrated polyp |
| Variables |
OR |
CI (95%) |
P value |
OR |
CI (95%) |
P value |
| Male gender |
1.81 |
1.60–2.05 |
<.001 |
1.64 |
1.45–1.86 |
<.001 |
| Acceptable bowel preparation |
1.17 |
0.97–1.43 |
.106 |
1.29 |
1.06–1.56 |
.010 |
| Focus of clinical practice |
|
|
|
|
|
|
| General/Motility |
1.00 |
– |
– |
1.00 |
– |
– |
| Hepatology |
1.27 |
1.03–1.55 |
.022 |
1.30 |
1.07–1.59 |
.008 |
| IBD |
1.61 |
1.30–1.99 |
<.001 |
1.72 |
1.39–2.12 |
<.001 |
| Interventional endoscopy |
1.36 |
1.13–1.64 |
.001 |
1.44 |
1.20–1.72 |
<.001 |
| Withdrawal time |
1.16 |
1.14–1.18 |
<.001 |
1.20 |
1.18–1.22 |
<.001 |
Model A – looking for adenoma. Model B – looking for adenoma and/or sessile serrated polyps. Model A is controlled for provider gender, years being out of training, and cecum intubation time. Model B is controlled for provider gender, years being out of training, and cecum intubation time.
CI = confidence interval, IBD = inflammatory bowel disease, OR = odds ratio.
4. Discussions
The result of our study further shows high performance of gastroenterologists in regard to the quality metrics including procedure completeness, withdrawal time, bowel preparation, and ADR. However, further analysis revealed that the focus of clinical practice regardless of other provider-related factors was an independent factor in higher performance in both adenomatous polyp detection as well as adenoma &/or SSP detection rate.
Current study showed gastroenterology providers were able to perform higher than expected quality metrics in detecting polyps,[8] which is in accordance with prior reports.[15,16] Moreover, our study showed that the provider’s clinical focus can further enhance ADR as well as adenoma and SSP detection rates. Although female providers and a higher number of years being out of training were associated with lower polyp detection rates in the univariate analysis, their effects did not hold in the multivariate analysis. We found that the hepatologist, IBD specialists, and interventional endoscopists were 25%, 60%, and 35% respectively higher likely to detect adenomatous polyps compared to gastroenterologists with general or motility focus of practice. When the goal was set for detecting both adenoma and/or sessile polyp, the trend persisted (hepatologists: 30%, IBD specialists: 72%, and interventional endoscopists: 44%). The fact that the IBD specialist and interventional endoscopists were more likely to find precancerous polyps could be due to their focus on luminal diseases and their subsequent different approach to assessing the mucosa. Additionally, the general/motility group was all female, had more female patients to screen, and was out of training for longer periods of time which has been reported in prior studies inconsistently as a possible factor in detecting fewer adenoma polyps.[15–17] However, the effect of these factors was controlled in our regression analysis and did not seem to be a relevant contributing factor. It is interesting to mention that the weight of the clinical focus of practice in detecting adenomatous polyps or adenoma/sessile serrate polyps is second to patients’ gender, compared to bowel preparation and withdrawal time (Table 3).
Bowel preparation was adequate in >87% of our patients which is an acceptable quality metric. This factor was important for detecting adenomatous polyps in the univariate analysis in accordance with prior studies.[18–22] Interestingly, it did not stay as a significant factor in the multivariate analysis when adenomatous polyps were the outcome of interest, but when the goal was to detect sessile polyps as well, this became an important factor. This is in accordance with prior studies where bowel preparation was an important factor in detecting sessile polyps.[15,23]
From our study, male patients have higher OR for ADR (OR: 1.81) and ADR + SSP (OR: 1.64) than female patients. This finding is consistent with the general consensus that polyps, adenomas, and CRC are associated with men more than women.[24] We also noticed that higher withdrawal times were associated with both adenomatous polyps and sessile polyps which is in accordance with prior studies.[25–28]
The current study harbors limitations common to large retrospective studies where we did not have other patient-related factors such as smoking history, history of IBD, ethnicity, and dietary and lifestyle changes. Thus, these results may not be generalizable. Our inclusion criteria included all first-time screening colonoscopies, but this did not include all of our patients getting screened. The age at which they were screened can affect the adenoma and polyp detection during the colonoscopy. Additionally, we did not have information on the location or size of the polyps. Within the physicians included, we also have no general/motility specialists who received any additional training. Furthermore, we did not have data if any of the physicians took extra courses after their official training to enhance their procedural skills. The strength of this study is collecting a decade's worth of records on practicing providers in a single outpatient center. It also highlights the role of years being out of general gastroenterology training as a more accurate predictor of the most important quality metrics in colonoscopy – adenoma or sessile polyp detection rates. Furthermore, this study is one of the first of its kind to compare the specialty focus of gastroenterologists with the adenoma and polyp detection rates. By further identifying categorical and continuous variables that affect the adenoma and polyp detection in between the different gastroenterology specialties, we can explore other potential gaps in colorectal screening that can further enhance CRC prevention. Our study also covers a large population of patients which may eliminate any significant skewing of the data due to patient demographics.
In conclusion, this study emphasizes that gastroenterologists meet the expected quality metrics in screening colonoscopy. The clinical focus of the gastroenterologists may affect the quality of colorectal screening based on ADR and/or SSPDR which could be a focus of subsequent changes in quality metrics. Furthermore, bowel preparation continues to be an important factor in enhancing the detection rates of SSP. This result needs further validation with a larger number of providers from multiple centers with a special focus on training.
Author contributions
Conceptualization: Hadie Razjouyan, Myunghoon Kim, John Levenick, Kofi Clarke, Thomas McGarrity.
Data curation: Hadie Razjouyan, Myunghoon Kim.
Formal analysis: Hadie Razjouyan, Myunghoon Kim.
Methodology: Hadie Razjouyan, Myunghoon Kim, Kofi Clarke, Thomas McGarrity.
Project administration: Myunghoon Kim.
Software: Hadie Razjouyan.
Supervision: Hadie Razjouyan, Kofi Clarke.
Validation: Hadie Razjouyan, Myunghoon Kim, John Levenick, Thomas McGarrity.
Visualization: Myunghoon Kim.
Writing – original draft: Hadie Razjouyan, Myunghoon Kim.
Writing – review & editing: Hadie Razjouyan, Myunghoon Kim, John Levenick, Kofi Clarke, Thomas McGarrity.
References
[1]. Siegel RL, Miller KD, Fuchs HE, et al. Cancer statistics, 2021. CA Cancer J Clin. 2021;71:7–33.
[2]. Rawla P, Sunkara T, Barsouk A. Epidemiology of colorectal cancer: incidence, mortality, survival, and risk factors. Prz Gastroenterol. 2019;14:89–103.
[3]. Berkowitz Z, Zhang X, Richards TB, et al. Multilevel small-area estimation of colorectal cancer screening in the United States. Cancer Epidemiol Biomarkers Prev. 2018;27:245–53.
[4]. Allemani C, Weir HK, Carreira H, et al. Global surveillance of cancer survival 1995-2009: analysis of individual data for 25,676,887 patients from 279 population-based registries in 67 countries (CONCORD-2). Lancet. 2015;385:977–1010.
[5]. Schreuders EH, Ruco A, Rabeneck L, et al. Colorectal cancer screening: a global overview of existing programmes. Gut. 2015;64:1637–49.
[6]. Citarda F, Tomaselli G, Capocaccia R, et al. Efficacy in standard clinical practice of colonoscopic polypectomy in reducing colorectal cancer incidence. Gut. 2001;48:812–5.
[7]. Zauber AG, Winawer SJ, O’Brien MJ, et al. Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths. N Engl J Med. 2012;366:687–96.
[8]. Rex DK, Schoenfeld PS, Cohen J, et al.
Quality indicators for colonoscopy. Am J Gastroenterol. 2015;110:72–90.
[9]. Rex DK, Boland CR, Dominitz JA, et al. Colorectal cancer screening: recommendations for physicians and patients from the U.S. multi-society task force on colorectal cancer. Gastrointest Endosc. 2017;86:18–33.
[10]. Baxter NN, Goldwasser MA, Paszat LF, et al. Association of colonoscopy and death from colorectal cancer. Ann Intern Med. 2009;150:1–8.
[11]. Brenner H, Chang-Claude J, Seiler CM, et al. Protection from colorectal cancer after colonoscopy: a population-based, case-control study. Ann Intern Med. 2011;154:22–30.
[12]. Fejleh MP, Shen CC, Chen J, et al.
Quality metrics of screening colonoscopies performed by PAs. JAAPA. 2020;33:43–8.
[13]. Xirasagar S, Hurley TG, Sros L, et al.
Quality and safety of screening colonoscopies performed by primary care physicians with standby specialist support. Med Care. 2010;48:703–9.
[14]. Baxter NN, Sutradhar R, Forbes SS, et al. Analysis of administrative data finds endoscopist
quality measures associated with postcolonoscopy colorectal cancer. Gastroenterology. 2011;140:65–72.
[15]. Crockett SD, Gourevitch RA, Morris M, et al. Endoscopist factors that influence serrated polyp detection: a multicenter study. Endoscopy. 2018;50:984–92.
[16]. Mehrotra A, Morris M, Gourevitch RA, et al. Physician characteristics associated with higher adenoma detection rate. Gastrointest Endosc. 2018;87:778–786.e5.
[17]. Butterly L, Robinson CM, Anderson JC, et al. Serrated and adenomatous polyp detection increases with longer withdrawal time: results from the New Hampshire Colonoscopy Registry. Am J Gastroenterol. 2014;109:417–26.
[18]. Sherer EA, Imler TD, Imperiale TF. The effect of colonoscopy preparation
quality on adenoma detection rates. Gastrointest Endosc. 2012;75:545–53.
[19]. Wong MCS, Ching JYL, Chan VCW, et al. Determinants of bowel preparation
quality and its association with adenoma detection: a prospective colonoscopy study. Medicine (Baltim). 2016;95:e2251.
[20]. 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;73:1207–14.
[21]. Johnson DA, Barkun AN, Cohen LB, et al. Optimizing adequacy of bowel cleansing for colonoscopy: recommendations from the US multi-society task force on colorectal cancer. Gastroenterology. 2014;147:903–24.
[22]. Harewood GC, Sharma VK, de Garmo P. Impact of colonoscopy preparation
quality on detection of suspected colonic neoplasia. Gastrointest Endosc. 2003;58:76–9.
[23]. Clark BT, Laine L. High-
quality bowel preparation is required for detection of sessile serrated polyps. Clin Gastroenterol Hepatol. 2016;14:1155–62.
[24]. Lieberman DA, Williams JL, Holub JL, et al. Race, ethnicity, and sex affect risk for polyps >9 mm in average-risk individuals. Gastroenterology. 2014;147:351–8; quiz e14–5.
[25]. Yun GY, Eun HS, Kim JS, et al. Colonoscopic withdrawal time and adenoma detection in the right colon. Medicine (Baltim). 2018;97:e12113.
[26]. Lee TJ, Blanks RG, Rees CJ, et al. Longer mean colonoscopy withdrawal time is associated with increased adenoma detection: evidence from the bowel cancer screening programme in England. Endoscopy. 2013;45:20–6.
[27]. Jung Y, Joo YE, Kim HG, et al. Relationship between the endoscopic withdrawal time and adenoma/polyp detection rate in individual colonic segments: a KASID multicenter study. Gastrointest Endosc. 2019;89:523–30.
[28]. Barclay RL, Vicari JJ, Doughty AS, et al. Colonoscopic withdrawal times and adenoma detection during
screening colonoscopy. N Engl J Med. 2006;355:2533–41.
