Authors: Chun‐Wei Chen (Department of Gastroenterology and Hepatology, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan), Cheng‐Tang Chiu (Department of Gastroenterology and Hepatology, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan; College of Medicine, Chang Gung University, Taoyuan, Taiwan), Ming‐Yao Su (Department of Gastroenterology and Hepatology, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan; College of Medicine, Chang Gung University, Taoyuan, Taiwan), Chun‐Jung Lin (Department of Gastroenterology and Hepatology, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan), Chen‐Ming Hsu (Department of Gastroenterology and Hepatology, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan), Siew‐Na Lim (Department of Neurology, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan), Chau‐Ting Yeh (Department of Gastroenterology and Hepatology, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan; College of Medicine, Chang Gung University, Taoyuan, Taiwan; Liver Research Center, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan), Wey‐Ran Lin (Department of Gastroenterology and Hepatology, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan; College of Medicine, Chang Gung University, Taoyuan, Taiwan; Liver Research Center, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan)
Categories: Original Articles, colonoscopy, polyp detection rate, sedation
Source: The Kaohsiung Journal of Medical Sciences
Doi: 10.1002/kjm2.12090
Authors: Chun‐Wei Chen, Cheng‐Tang Chiu, Ming‐Yao Su, Chun‐Jung Lin, Chen‐Ming Hsu, Siew‐Na Lim, Chau‐Ting Yeh, Wey‐Ran Lin
The polyp detection rate (PDR) is an important quality indicator for colonoscopy. Several factors have been shown to be associated with PDR. However, whether the moderate sedation is a factor for polyp detection remains controversial. This study aims to assess factors associated with polyp detection including the moderate sedation factor during colonoscopy. Patients who underwent colonoscopy from June 1, 2014 to May 31, 2015 were enrolled into this retrospective study. Patients with poor colon preparation and failure to reach the cecum were excluded. The clinical factors, including patient's sex, age, midazolam/fentanyl sedation, indications, endoscopist colonoscopy volume, and use of antispasmodic agent were evaluated by multivariate analysis. A total of 3373 patients were included in this study. The mean age was 55.8 years, and 1980 patients (58.7%) were male. Among the 3373 patients, 2513 (74.5%) underwent midazolam/fentanyl‐based sedated colonoscopy. The multivariate analysis showed that male sex, age over 50 years old, midazolam/fentanyl sedation and indications of screening and surveillance were significantly associated with polyp detection. Moreover, when stratified by sex and age, the midazolam/fentanyl sedation was associated with polyp detection in male patients and patients over 50 years old. This study has highlighted the role of midazolam/fentanyl sedation administered by colonoscopists as a modifiable factor that may increase polyp detection during colonoscopy.
Colonoscopy is currently the best modality for the detection of colorectal polyps and colorectal cancer (CRC), which is the third most common malignancy in the United States and worldwide.1, 2 In Taiwan, CRC is also the most common malignancy and is the third leading cause of cancer‐related death.3 The development of CRC can be prevented when the detection of colon adenoma is followed by polypectomy.4 However, colonoscopy is usually regarded as a highly invasive procedure and the uncomfortable sensation experienced during colonoscopy has been a major obstacle that deters patients from undergoing the procedure.
In everyday colonoscopy practice, an increasing number of patients have been requesting sedation. Previous studies have demonstrated that colonoscopy with sedation not only reduced patient discomfort, but also increased colonoscopist satisfaction.5, 6 In general, there are four levels of sedation and analgesia.7
In patients who require sedation during colonoscopy, propofol is used for deep sedation and midazolam/fentanyl is used for moderate sedation.8 Although propofol has some advantages, such as rapid induction and quick recovery, it requires administration by trained general anesthetists who can monitor the patient's clinical condition throughout the procedure.7 In contrast to propofol‐based deep sedation, midazolam/fentanyl‐based moderate sedation usually allows patients to maintain cardiovascular and ventilator function as well as responses to verbal or light tactile stimuli. This permits the administration of midazolam/fentanyl to patients under the colonoscopists' supervision, which eliminates the need for anesthesiologists, and thus renders this approach more cost effective than propofol sedation.
The adenoma detection rate (ADR) has been validated as an important quality indicator for colonoscopy, and is an independent risk factor for interval CRC.9 Several studies suggest that the polyp detection rate (PDR) can serve as a good surrogate for ADR, which the average adenoma to PDR quotient was 0.64 to 0.68.10, 11, 12, 13, 14 The following factors have been reported to influence the ADR or PDR: the cecal intubation rate (CIR), withdrawal time of at least 6 minutes, bowel preparation quality, image enhancement equipment (narrow band image and high definition image), and the colonoscopist's performance.15, 16, 17 However, it remains controversial whether moderate sedation could improve quality indicators for colonoscopy, such as the PDR. Radaelli et al demonstrated that sedation increased polyp detection in routine colonoscopy.18 In contrast, Bannert et al showed that adenoma detection or polyp detection were not increased by sedation.19 Moreover, the effects of moderate sedation on polyp detection during colonoscopy have not been well examined in Asian populations. The aim of this study was to clarify whether midazolam/fentanyl‐based moderate sedation was a factor for polyp detection during colonoscopy.
From June 1, 2014 to May 31, 2015, all patients who underwent colonoscopy for screening, surveillance and diagnosis indication at a therapeutic endoscopic center of a tertiary medical center were retrospectively included in this study. Patients with poor colon preparation, failure to reach the cecum, and withdrawal time less than 6 minutes were excluded from this study. This study was approved by the ethics committee of Chang Gung Memorial Hospital (reference No: 201601349B0).
All colonoscopies were performed with the Olympus CF Q260AL (Tokyo, Japan) by 35 board‐certified gastroenterologists. Carbon dioxide was supplied to all patients throughout the procedure. Patients were instructed to use two liters of polyethylene glycol 4000 solution or 90 mL of fleet phospho‐soda as split‐dose for bowel preparation. The bowel preparation quality was recorded on the Aronchick scale as excellent (more than 95% of the mucosa was observed), good (clear liquid covering up to 25% of the mucosa and more than 90% of the mucosa was observed) and fair (semisolid stool could not be suctioned and more than 90% of the mucosa was observed).20, 21
Prior to the procedure, the clinical benefits and side effects of sedation were explained to the patients. The decision was then made by the patients to either undergo colonoscopy with midazolam/fentanyl moderate sedation administered by the colonoscopist, or an unsedated colonoscopy. The colonoscopist intravenously administered adequate midazolam (1‐5 mg) and fentanyl (20‐100 μg) before and during the procedure. The dosage was then adjusted taking into account the patient's age, body weight, and underlying disease to achieve a moderate sedation level. The antispasmodic drug hyoscine 20 mg was given intravenously to each patient with the exception of patients with heart disease, benign prostate hyperplasia or glaucoma.
Demographic data (age, sex), colonoscopy volume of the colonoscopist, indications for colonoscopy, use of antispasmodic medication, and midazolam/fentanyl sedation were ascertained before colonoscopy. On average, each colonoscopist performs at least 100 colonoscopies every year. Thus, to be considered experienced, a colonoscopist has to perform a minimum of 500 colonoscopies. Colon preparation qualities, cecal intubation time, withdrawal time, size of polyp and patient numbers of at least one polyp detected were recorded. The cecal intubation time was recorded from the beginning of insertion to reach the ileocecal valve or appendiceal orifice. The withdrawal time was defined as inspection time from cecum to anus. The time taken for irrigation and aspiration to clear the mucosa and time taken for the procedure to remove polyps, such as polypectomy or endoscopic mucosa resection was included in the withdrawal time.
All continuous data are shown as mean values and SD, and the categorical variables are presented as absolute and relative frequencies. Comparisons of continuous data were performed using the Mann‐Whitney U test, and comparisons of discrete variables were conducted with the Chi‐square test. Univariate and multivariate logistic regression analyses were used to examine the associations between independent and dependent variables of the PDR. Odds ratios (ORs) and 95% confidence intervals (CIs) were also calculated. Statistical significance was defined as a P value <.05. The statistical software SPSS 22 (IBM Corp., Armonk, New York) was used for the analysis.
A total of 4159 patients were performed colonoscopy in our endoscopic therapeutic center. One hundred and forty one patients were excluded due to failure to reach the cecum and 695 patients were excluded due to poor colon preparation. Of these patients, 3373 patients were included in this retrospective study. The mean age was 55.8 ± 13.09 years and 1980 patients (58.7%) were male. The demographic characteristics of the patients are shown in Table 1. Among the 3373 patients, 2513 (74.5%) patients underwent colonoscopy with midazolam/fentanyl‐based sedation and 860 (25.5%) patients underwent colonoscopy without sedation. The percentage of patients with sedation was significantly higher in the female population than in the male population (79.9% vs 70.7%, P < .001). Two thousand and twenty (59.9%) patients had at least one polyp detected. The number of polyps of size ≥10 mm was 273 (13.5%) and number of polyps of size <10 mm was 1747 (86.5%). The mean cecal intubation time was 7.7 ± 6.3 minutes. The number of patients with withdrawal time >8 minutes was 1824 (54.1%). In colonoscopy indications, 474 (14.1%) patients underwent colonoscopy for screening and a positive fecal occult blood test, 1488 (44.1%) patients for surveillance and 1411 (41.8%) patients for symptom diagnosis. With regard to the colonoscopist's experience, 2052 (60.8%) colonoscopy exams were performed by colonoscopists who had performed >500 colonoscopies. There were 3047 (90.3%) patients who underwent colonoscopy with an anti‐spasmodic agent. Regarding colon preparation, the quality was rated excellent in 126 (3.7%) patients, good in 1850 (54.8%) patients, and fair in 1397 (41.5%) patients. The characteristics of patients divided by sedation and unsedation was listed in Table S1.
The PDR is an important indicator of the colonoscopy quality. Therefore, factors associated with polyp detection shown in Table 2 were further investigated. The multivariate regression analysis showed that several independent factors predicated polyp male sex, age ≥50 years, midazolam/fentanyl sedation, and indications of screening and surveillance.
To evaluate the factors of the PDR within specific groups, the patients were stratified by sex and age. The subgroup multivariate regression analysis for independent factors of polyp detection in male and female patients, is shown in Table 3. In male patients, age ≥50 years, midazolam/fentanyl sedation and indications of screening and surveillance were independent factors of polyp detection (OR: 2.07; 95% CI: 1.68‐2.55; P < .001; OR: 1.31; 95% CI: 1.06‐1.62; P = .013; OR: 2.66; 95% CT: 2.18‐3.24; P < .001, respectively). In female patients, age ≥50 years and indications of screening and surveillance were independent factors of polyp detection (OR: 1.77; 95% CI: 1.39‐2.26; P < .001; OR: 2.36; 95% CI: 1.89‐2.94; P < .001 respectively).
The subgroup multivariate regression analysis for independent factors of polyp detection in patients aged ≥50 years and <50 years is shown in Table 4. In patients aged ≥50 years, the independent factors of polyp detection were male sex, midazolam/fentanyl sedation and indications of screening and surveillance (OR: 1.73; 95% CI: 1.45‐2.08; P < .001; OR: 1.22; 95% CI: 1.00‐1.49; P = .048; OR: 2.56; 95% CI: 2.15‐3.07; P < .001, respectively).
Our results showed that male sex, age ≥50 years, the combination of midazolam/fentanyl during colonoscopy, and indications of screening and surveillance were predictive factors of colon polyp detection. Our findings were also compatible with data from recently published prospective studies on increasing adenoma detection.22, 23 Moreover, we found that midazolam/fentanyl sedation was another independent factor for polyp detection. This suggested that moderate sedation administered by colonoscopists may increase the quality of colonoscopy.
Although routine colonoscopy is generally performed in patients under sedation, the effects of this factor on polyp detection have remained uncertain and inconclusive. A study conducted by Radaelli et al reported that polyp detection was improved by various sedative regimens,18 while another study conducted by Bannert et al reported that polyp detection was not improved by all levels of sedation in screening colonoscopy.19 In this study, we found that a single sedative regimen with midazolam/fentanyl administration by colonoscopists may enhance colon polyp detection. Although the possible mechanisms were beyond the scope of this study, we speculated that sedation of the patients allowed the colonoscopists to pay more attention during the procedures. Furthermore, when patients are sedated, their increased tolerance to procedure‐related pain allows for more air to be insufflated, which may allow the examiner to inspect the mucosa in more detail and facilitate polyp detection.19 In addition, our data demonstrated that midazolam/fentanyl sedation was one factor of polyp detection in males and patients older than 50 years of age. Interestingly, the sedation was not a factor of colon polyp detection in female patients. In our patients, the mean age was 55.8 years. The prevalence of adenoma in male with 55 to 59 years old is 23.0% and 13.2% in female with 55 to 59 years old.24 The prevalence of adenoma in female is almost 10% less compared to the prevalence of adenoma in male patients. Therefore, we speculated that sedation could not increase the colon polyp detection in the female patients, who are a low adenoma prevalence group.
Two previous studies demonstrated the feasibility of unsedated colonoscopy in selective screening patients.25, 26 However, these studies did not compare quality indicators, such as the PDR between sedated and unsedated groups. It is agreed that unsedated colonoscopy has some advantages such as no sedation‐related cost and risk, and no physical activity restriction after the procedure. However, according to patient experiences and physician assessments, the patients had significantly higher average pain scores and lower average satisfaction scores in unsedated colonoscopy procedures compared to sedated procedures.25 Once repeated colonoscopy is indicated, patients who experienced pain in a previous unsedated colonoscopy exam may be less willing to undergo an unsedated colonoscopy or may even be hesitant to undergo another colonoscopy exam.25
Propofol‐based sedation is another sedative regimen used during colonoscopy exams and it provides a deeper level of sedation and a higher percentage of patient satisfaction than the midazolam/fentanyl combination. In general, when propofol is administered, the team should include one trained individual dedicated to the patient's clinical condition with the ability to rescue patients who become unresponsive or fail to maintain spontaneous respiratory or cardiovascular functions.7 However, one meta‐analysis study revealed no difference in the incidence of bradycardia, hypoxemia, and hypotension in midazolam‐based and propofol‐based sedation.6 Additionally, one randomized controlled trial demonstrated that the effect of propofol‐based sedation on polyp detection during colonoscopy was similar to that of midazolam‐based sedation.27 In other words, propofol‐based sedation is more expensive and does not seem to improve patient safety or polyp detection during colonoscopy when compared with midazolam/fentanyl sedation, which can be administered by the colonoscopist alone.
Our multivariate regression analysis showed that the colonoscopists' experience was not a factor for polyp detection. This finding is different from a previous study conducted by Jover et al,28 where the colonoscopists' age and life‐long number of colonoscopies performed are factors associated with adenoma detection. However, there was no sedation‐related factor analyzed in their study. In our study, weighting both the colonoscopists' experience and sedation regimen together, the rate of polyp could be similar in colonoscopists' with different level of experience. A possible explanation is that moderate sedation might help less‐experienced colonoscopists to achieve similar performance during colonoscopy compared to more experienced colonoscopists.
We acknowledge that this study has several limitations. First, our study was retrospective in nature and was conducted in only one medical center. Second, selection bias may have occurred based on the patients' willingness to be sedated. Third, the withdrawal time could affect the PDR17, 29 but was not analyzed in our study. The majority of detected polyps were immediately removed by therapeutic procedures such as biopsy, snare polypectomy, and endoscopic mucosa resection. However, the time taken for these procedures was inevitably included in the withdrawal times, which could result in longer withdrawal times in patients with polyps. Therefore, to avoid this type of error, we did not include the withdrawal time as a factor of multivariate regression analysis in our study. Fourth, the percentages of colon preparation in excellent and good condition were 58.5%, which was lower than 90%. The possible reason was that the patient sources in our study were from outpatient, inpatient, and emergent department. In addition, the patients may have higher rate of comorbidities such as diabetes mellitus, hypertension, heart failure coronary artery disease, stroke, chronic kidney disease, liver cirrhosis and malignancy disease, which may affect the colon preparation results.30, 31 However, our endoscopic databases were not linked to patients' underlying condition. Finally, pathological data were not presented in our study due to the fact that this study was not linked to pathological databases, and we were unable to evaluate the ADR. However, a previous study noted that the PDR correlates well with the ADR.12
In summary, this study highlighted factors associated with polyp detection in patients with acceptable colon preparation and complete cecal intubation. These factors were male sex, age ≥50 years, the combination of midazolam/fentanyl during colonoscopy, and indications of screening and surveillance. Our findings suggest midazolam/fentanyl sedation may increase polyp detection in males and patients older than 50 years. Further randomized studies are required to evaluate the impact of midazolam/fentanyl on indicators of colonoscopy quality.
All authors declare no conflict of interests.