Automated detection of cecal intubation with variable bowel preparation using a deep convolutional neural network

Background and study aims Colonoscopy completion reduces post-colonoscopy colorectal cancer. As a result, there have been attempts at implementing artificial intelligence to automate the detection of the appendiceal orifice (AO) for quality assurance. However, the utilization of these algorithms has not been demonstrated in suboptimal conditions, including variable bowel preparation. We present an automated computer-assisted method using a deep convolutional neural network to detect the AO irrespective of bowel preparation. Methods A total of 13,222 images (6,663 AO and 1,322 non-AO) were extracted from 35 colonoscopy videos recorded between 2015 and 2018. The images were labelled with Boston Bowel Preparation Scale scores. A total of 11,900 images were used for training/validation and 1,322 for testing. We developed a convolutional neural network (CNN) with a DenseNet architecture pre-trained on ImageNet as a feature extractor on our data and trained a classifier uniquely tailored for identification of AO and non-AO images using binary cross entropy loss. Results The deep convolutional neural network was able to correctly classify the AO and non-AO images with an accuracy of 94 %. The area under the receiver operating curve of this neural network was 0.98. The sensitivity, specificity, positive predictive value, and negative predictive value of the algorithm were 0.96, 0.92, 0.92 and 0.96, respectively. AO detection was > 95 % regardless of BBPS scores, while non-AO detection improved from BBPS 1 score (83.95 %) to BBPS 3 score (98.28 %). Conclusions A deep convolutional neural network was created demonstrating excellent discrimination between AO from non-AO images despite variable bowel preparation. This algorithm will require further testing to ascertain its effectiveness in real-time colonoscopy.

Optimal Timing of Simethicone Supplement for Bowel Preparation: A Prospective Randomized Controlled Trial

Background and Aims. Simethicone (SIM), as an antifoaming agent, has been shown to improve bowel preparation during colonoscopy. However, the optimal timing of SIM addition remained undetermined. We aimed to investigate the optimal timing of SIM addition to polyethylene glycol (PEG) to improve bowel preparation. Methods. Eligible patients were randomly assigned to two groups: the SIM evening group (SIM addition to PEG in the evening of the day prior to colonoscopy) and the SIM morning group (SIM addition to PEG in the morning of colonoscopy). The primary outcome was Bubble Scale (BS). The secondary outcomes were Boston Bowel Preparation Scale (BBPS) and adenoma detection rate (ADR). Results. A total of 419 patients were enrolled in this study. The baseline characteristics of the patients were similar in both groups. No significant differences were observed in terms of BS (8.76 ± 0.90 vs. 8.65 ± 1.16, P = 0.81), ADR (34.1% vs. 30.8%, P = 0.47), Boston Bowel Preparation Scale (BBPS) (8.59 ± 0.94 vs. 8.45 ± 1.00, P = 0.15), and withdrawal time (8.22 ± 2.04 vs. 8.01 ± 2.51, P = 0.094) between the two groups. Moreover, safety and compliance were similar in both groups. However, the SIM evening group was associated with shorter cecal intubation time (3.80 ± 1.81 vs. 4.42 ± 2.03, P < 0.001), higher BS (2.95 ± 0.26 vs. 2.88 ± 0.38, P = 0.04) in the right colon, and diminutive ADR (62.5% vs. 38.6%, P = 0.022) in the right colon, when compared to the SIM evening group. Conclusions. The SIM addition to PEG in the evening of the day prior to colonoscopy can shorten cecal intubation time and improve BS scores and diminutive ADR of the right colon compared with the SIM addition to PEG in the morning of colonoscopy in bowel preparation.

Colonoscopy quality across Europe: a report of the European Colonoscopy Quality Investigation (ECQI) Group

Background and study aims The European Colonoscopy Quality Investigation (ECQI) Group comprises expert colonoscopists and investigators with the aim of raising colonoscopy standards. We assessed the levels of monitoring and achievement of European Society of Gastrointestinal Endoscopy (ESGE) performance measures (PMs) across Europe using responses to the ECQI questionnaires. Methods The questionnaire comprises three forms: institution and practitioner questionnaires are completed once; a procedure questionnaire is completed on multiple occasions for individual total colonoscopies. ESGE PMs were approximated as closely as possible from the data collected via the procedure questionnaire. Procedure data could provide rate of adequate bowel preparation, cecal intubation rate (CIR), withdrawal time, polyp detection rate (PDR), and tattooing resection sites. Results We evaluated ECQI questionnaire data collected between June 2016 and April 2018, comprising 91 practitioner and 52 institution questionnaires. A total of 6445 completed procedure forms were received.Institution and practitioner responses indicate that routine recording of PMs is not widespread: adenoma detection rate (ADR) is routinely recorded in 29 % of institutions and by 34 % of practitioners; PDR by 42 % and 47 %, CIR by 62 % and 64 %, bowel preparation quality by 56 % and 76 %, respectively.Procedure data showed a rate of adequate bowel preparation of 84.2 %, CIR 73.4 %, PDR 40.5 %, mean withdrawal time 7.8 minutes and 12.2 % of procedures with possible removal of a non-pedunculated lesion ≥ 20 mm reporting tattooing. Conclusions Our findings clearly show areas in need of quality improvement and the importance of promoting quality monitoring throughout the colonoscopy procedure.

Effect of physician-provided education on the quality of bowel preparation

Background and study aims : Inadequate bowel preparation inpatients scheduled for colonoscopy is an important problem. Inour study, we aimed to investigate the effect of physician-providedbowel preparation education on the quality of bowel preparationand process. Patients and methods : A total of 150 outpatients who were referredto Kocaeli University Medical Faculty Hospital GastroenterologyUnit for colonoscopy between May 2019 and October 2019 wereenrolled in our prospective, endoscopist-blinded study. Patientswere divided into two groups. Group 1 (education group) included73 patients who received 10 minutes of verbal information froma physician in addition to a written information form. Group 2(control group) included 75 patients who received informationfrom a medical secretary in addition to a written information form.During colonoscopy, the quality of bowel preparation was assessedusing the Boston bowel preparation scale (BBPS). A BBPS score≥ 5 was considered adequate bowel preparation. The mean BBPSscore, polyp detection rate, cecal intubation rate and time, andprocedure time were also evaluated. Results : The rate of adequate bowel preparation (BBPS score≥ 5) was 90.4% and 74.7% in groups 1 and 2, respectively (p =0.021). The odds ratio for having a BBPS score ≥ 5 in the educationgroup was 3.199 compared with the control group (95% confidenceinterval = 1.254-8.164; p = 0.015). The cecal intubation rates were91.8% and 88% in groups 1 and 2, respectively (p > 0.05). The cecalintubation time, procedure time, and adenoma detection rates weresimilar between the groups. The relationships of age, educationlevel, sex, diabetes mellitus, medicine use, procedure time, andintraabdominal surgery with inadequate bowel preparationwere analysed using a logistic regression model. Univariate andmultivariate analyses revealed no significant factors associatedwith inadequate bowel preparation. Conclusions : Patient education on the bowel preparationprocess via a physician improved the quality of bowel preparation.

Effect of Using Personal Protective Equipment during the COVID-19 Pandemic on the Quality Indicators of Screening Colonoscopies

Background. Coronavirus Disease 2019 (COVID-19) has affected many facets of the practice of medicine including screening colonoscopies. Aims. Our study looks to observe if there has been an effect on the quality of colonoscopies, as indicated by quality measures such as the cecal intubation rate (CIR), cecal intubation time (CIT), scope withdrawal time (SWT), and adenoma detection rate (ADR) with the adoption of standard COVID-19 precautions. Methods. We conducted a retrospective chart review to analyze the effects of the COVID-19 pandemic on screening colonoscopies. The study utilized data on CIR, CIT, SWT, and ADR from outpatient, nonemergent procedures conducted at 3 endoscopy suites of St. Luke’s University Health Network. All inpatient and emergent procedures were excluded. Results. Our study demonstrated that the total number of screening colonoscopies was decreased between 2019 and 2020 (318 in 2019 vs. 157 in 2020, p = 0.005 ). CIT ( 320 ± 105 seconds in 2019 vs. 392 ± 107 seconds in 2020, p = 0.001 ) and SWT ( 706 ± 232 seconds in 2019 vs. 830 ± 241 seconds in 2020, p = 0.001 ) were increased while CIR (98.2% in 2019 vs. 96.6% in 2020, p = 0.04 ) was decreased between 2019 and 2020 likely due to PPE introduction. ADR was similar between the two groups (38.23 (12.50-66.66) in 2019 vs. 38.18 (16.66-66.00) in 2020, p = 0.8 ). Conclusion. Our study showed that quality indices for screening colonoscopies like CIR, CIT, and SWT were negatively impacted during the COVID-19 time period. ADR, however, was similar. Thus, the efficiency of the procedures was affected by the use of PPE but it did not affect the colonoscopy’s clinical benefit.

Efficacy of Macrogol 4000 Plus Electrolytes in Bowel Preparation for Colonoscopy in Patients with Chronic Constipation

Background: Chronic constipation is a significant factor in poor bowel preparation for colonoscopy. Macrogol 4000 plus electrolytes (Movicol, EA Pharma, Tokyo, Japan), containing polyethylene glycol (PEG) and electrolytes, have been used recently to treat patients with constipation. However, prospective studies on the use of macrogol 4000 for bowel cleansing for colonoscopy are lacking. This study aimed to investigate the efficacy and safety of macrogol 4000 in addition to PEG administered in patients with chronic constipation.Methods: This single-center, single-arm prospective study enrolled patients with chronic constipation who were scheduled to undergo colonoscopy. The primary endpoint was the proportion of good bowel preparation assessed using the Boston Bowel Preparation Scale (BBPS) (6 or more points). The secondary endpoints were the time from when pPEG (MoviPrep, EA Pharma, Tokyo, Japan) was taken until colonoscopy could be started, amount of PEG taken, number of defecations, whether additional PEG doses were taken, and adverse events. Endoscopy-related endpoints included cecal intubation rate, insertion time, observation time, adenoma detection rate (ADR), and polyp detection rate (PDR). The tolerability of PEG and macrogol 4000 was assessed using a questionnaire.Results: Forty patients were included in the analysis. The median BBPS was 7 (range, (3–9)) and ³6 points in 37 cases (92.5%). The median time until colonoscopy can be started was 210 min (90–360 min), the median volume of PEG taken was 1500 mL (1000–2000 mL), and the median number of defecations was 7 (3-20). No adverse events were observed. Fourteen patients required an additional dose of PEG. Cecal intubation was achieved in all cases, the median insertion time was 6.0 min (range, 2.3–22 min), and the median observation time was 8.8 min (range, 4.0–16.0 min). The ADR and PDR were 60.0% and 75.0%, respectively. A greater proportion of patients rated the tolerability of macrogol 4000 as good compared with that of PEG (95.0% vs. 50.0%, p < 0.01).Conclusions: Intake of macrogol 4000 in addition to PEG is effective and safe for colonoscopy in patients with chronic constipation.

Weekend and evening planned colonoscopy activity: a safe and effective way to meet demands

Background and study aims Colonoscopy is the “gold standard” investigation for assessment of the large bowel that detects and prevents colorectal cancer, as well as non-neoplastic conditions. The Joint Advisory Group (JAG) on Gastrointestinal Endoscopy recommends monitoring key performance indicators such as cecal intubation rate (CIR) and adenoma detection rate (ADR). We aimed to investigate the quality of colonoscopies carried out during evening and Saturday lists in our unit and compare them against JAG standards of quality for colonoscopies. Patients and methods We retrospectively collected and analyzed demographical and procedure-related data for non-screening colonoscopies performed between January 2016 and November 2018. Evenings and Saturdays were defined as the out-of-hour (OOH) period. We compared the outcomes of the procedures done in these against the working hours of the weekdays. We also wanted to explore whether the outcomes were different among certain endoscopists. Other factors that could affect the KPIs, such as endoscopist experience and bowel preparation, were also analyzed. Results There were a total of 17634 colonoscopies carried out; 56.9 % of the patients (n = 10041) < 70 years old. Key Performance Indicators (KPIs) of weekday, evening, and Saturday colonoscopies regarding the CIR and ADR met the JAG standards as they were above 93 % and 24 %, respectively. Advanced colonoscopists had better KPIs when compared to the non-advanced colonoscopists, with CIR at 97.6 % vs. 93.2 % and ADR at 40.8 % vs. 26 %, respectively. Conclusions JAG standards were maintained during colonoscopies done on weekdays, evenings, and Saturdays. Advanced colonoscopists had higher CIR and ADRs.