Strategies to Enhance the Adenoma Detection Rate (ADR) and the Serrated Polyp Detection Rate (SPDR) in Colonoscopy: A Comprehensive Review

Abstract

Introduction: High-quality colonoscopy is influenced by several factors, with the adenoma detection rate (ADR) being one of the most studied indicators. A strong inverse relationship exists between ADR and the risk of developing post-colonoscopy colorectal cancer (PCCRC), prompting the European Society of Gastrointestinal Endoscopy guidelines to recommend a minimum ADR of 25%. In contrast, there is limited evidence supporting the clinical significance of the serrated polyp detection rate (SPDR), and no specific benchmark was established until a very recent update from the American societies. Main paper: This review examines the factors that influence ADR and SPDR, offering tips to improve these metrics. Effective interventions for enhancing ADR include training, colonoscopy feedback, adequate bowel preparation, longer withdrawal time, water-aided colonoscopy, right colon second look, and chromoendoscopy. The use of cap, devices, and specialized scopes also show promise, though these are often at higher costs. Artificial intelligence has generated great optimism, especially following positive results from early randomized controlled trials; however, its effectiveness has been less pronounced in real-world settings. Conclusions: Many of these approaches require further trials and meta-analyses to establish their ultimate efficacy. Moreover, future clinical head-to-head studies will help to identify the most effective interventions for reducing colorectal cancer incidence and the risk of PCCRC.

1. Introduction

Colonoscopy is widely considered the gold standard for colon investigation and one of the possible screening tools for colorectal cancer (CRC), either as an initial approach or following a positive fecal occult blood test (FOBT). To perform a high-quality colonoscopy, it is necessary to achieve several established benchmarks and targets, and one of the most studied is the adenoma detection rate (ADR), strongly linked to important clinical outcomes [1].

In a well-known study based on a colorectal cancer screening program using colonoscopy, endoscopist’s ADR was an independent predictor of the risk of interval CRC after screening colonoscopy. Specifically, an ADR < 20% correlates with a higher risk of CRC compared to an ADR > 20% [2]. Furthermore, it was demonstrated that the ADR was inversely correlated with CRC-associated mortality. Improving the ADR from below 19% to 34–53% has been shown to reduce the risk of post-colonoscopy colorectal cancer (PCCRC) by up to 62% [3]. For these reasons, European Society of Gastrointestinal Endoscopy (ESGE) guidelines recommend a minimum ADR of 25% [1]. A recent meta-analysis based on real-world observational studies found an overall pooled ADR of 26.5%, supporting ESGE’s benchmark [4].

In the last few years, interest in sessile serrated lesions (SSL) has grown, given that approximately 20–30% of CRC cases are thought to arise from the serrated polyp pathway, which is a unique carcinogenesis pathway that bypasses the traditional adenoma–carcinoma sequence [5,6]. Additionally, patients with a presence of SSLs are at increased risk of synchronous and metachronous advanced neoplasia compared to patients with conventional adenoma alone [6,7].

Recently, two population-based studies have highlighted how the serrated polyp detection rate (SPDR) is related to PCCRC- and CRC-related mortality [8,9].

Sessile serrated polyps are less frequent than adenomas, and in these two studies, the median SPDRs were 11.9% and 8.5%, respectively, with a moderate correlation with the ADR [8,9,10], which was likely due to variations in histological evaluations and differences in study methodologies [11].

Given the current evidence, no formal guidelines have been established regarding the SPDR as a benchmark. Considering the available evidence, there is a clear need to improve both the ADR and SPDR in order to reduce the risk of PCCRC and CRC mortality. While various factors and techniques that influence the ADR have been identified, there is a lack of evidence on factors related to the SPDR. This comprehensive review aims to provide an updated overview of the factors associated with both the ADR and SPDR, offering insights into strategies for improving these key performance metrics.

2. Materials and Methods

For the purpose of this narrative review, we searched in the PubMed and Embase databases for relevant papers using the following keywords: “adenoma detection rate”, “serrated polyp detection rate”, “colonoscopy”, “morning colonoscopy”, “afternoon colonoscopy”, “colonoscopy sedation”, “training colonoscopy”, “volume colonoscopy”, “endoscopist specialty”, “colonoscopy training”, “bowel preparation”, “colonoscopy devices”, “distal attachment”, “colonoscopy scopes”, “colonoscopy withdrawal time”, “colonoscopy dual observation”, “water-aided Colonoscopy”, “dynamic colonoscopy”, “right colon evaluation”, “chromoendoscopy”, and “artificial intelligence”, individually and in combination. The search included full-text articles published in the English language up to 31 September 2024. The bibliography of papers and reviews on relevant topics were also hand-searched to identify other potentially relevant papers.

3. Periprocedural Factors

3.1. Morning vs. Afternoon

Digestive endoscopy is a well-recognized physically and mentally demanding procedure, and the timing of the procedure may affect colonoscopy outcomes [12,13]. The ADR was found to be significantly higher in morning colonoscopies than in afternoon colonoscopies, as reported in several reports. This finding may be mainly explained by operator fatigue during the day [14,15,16,17]. For instance, a study by Gurudu et al. demonstrated that dividing endoscopy sessions into morning and afternoon shifts, with a change in the operator, helped to maintain a higher ADR by mitigating fatigue. Specifically, the ADR was significantly higher in the morning than in the afternoon in operators who performed endoscopies throughout the entire day [18]. Conversely, in endoscopy teams working half-day shifts, no significant difference in the ADR was observed between morning and afternoon sessions, further supporting the hypothesis that fatigue negatively impacts performance. Supportive devices, such as artificial intelligence and half-day endoscopy, may help maintain a high ADR throughout the day [17].

3.2. Sedation

Among the parameters analyzed during a colonoscopy, the type of sedation may affect the quality of the examination. However, several studies have shown that the use of deep sedation (propofol) compared to conscious sedation does not influence the polyp detection rate (PDR) and ADR [19,20]. Therefore, the sedation type does not appear to play a role in improving the ADR but rather influences other aspects of the colonoscopy, such as the procedural duration, technical difficulty and the cecal intubation rate.

4. Endoscopist Factors

4.1. Education and Training

Adequate endoscopic training is essential for performing high-quality endoscopic examinations, particularly in achieving an optimal ADR. One of the reasons for missing adenomas is a poor ability to recognize an adenomatous polyp due to inexperience/inadequate operator training [21]. Several endoscopic quality improvement programs have shown that structured training can dramatically increase the ADR and SPDR [22,23]. Additionally, the presence of colonoscopy performance feedback has demonstrated how it allows for significant improvements in learning and, among other things, an increase in the ADR, especially among endoscopists in training. In particular, they observed the change in the ADR in operators before and after subjecting them to ’feedback’ (including monitoring endoscopists’ ADRs and providing report cards or active interventions with endoscopy quality improvement training) [23].

4.2. Annual Volume and Endoscopist Specialty

Two recent meta-analyses explored whether the annual volume of colonoscopies and endoscopists’ specialty influence colonoscopy outcomes [24,25]. The first meta-analysis found that a higher annual volume of colonoscopies is associated with a higher cecal intubation rate but does not correlate with an increased ADR. The second analysis showed that colonoscopies performed by surgeons and non-gastroenterologists have a lower ADR compared to those performed by gastroenterologists. A multicenter American study confirmed the last finding, noting that gastroenterologists have a higher serrated polyp detection rate than other specialists; furthermore, it highlighted how a higher serrated polyp detection rate was associated with endoscopists with higher procedures volume [26].

4.3. Trainee Involvement

The impact of trainee involvement on colonoscopy quality has been the subject of several studies, reaching controversial results. A retrospective study by Buchner et al. [27] compared colonoscopies performed by trainees under the supervision of a specialist with those performed by specialists only, demonstrating how the presence of the trainee increased the rate of finding adenomas smaller than 5 mm. Further studies [28] confirm the finding that the presence of a trainee in endoscopy, supported by an expert physician, increases the quality of the examination, particularly ADR, likely due to double observations. Recently, a multicenter population-based cohort study conducted over more than 35,000 colonoscopies found no significant increase in the ADR when trainees were involved. However, it reported a lower serrated polyp detection rate in cases where trainees participated [29].

5. Bowel Preparation

An adequate bowel preparation (BP) is necessary to perform a high-quality colonoscopy, and the ESGE set a benchmark rate for a minimum adequate BP of 90%, with a target of ≥95% [1].

However, real-world studies have shown that this benchmark is not achieved outside of RCTs [30,31,32]; consequently, improving bowel preparation quality is essential to enhance the ADR and reduce the risk of PCCRC in routine daily practice.

Several BP regimens have been proven to enhance BP. Historically, 4L-PEG was the most commonly used one, offering good efficacy but lower patient compliance. In the last few years, several studies have suggested how lower-volume regimens, such as 1L-PEG and 2L-PEG, seem not to be inferior or even better than 4L-PEG; thus, their higher potential willingness to repeat colonoscopy may make them the first choices compared to 4L-PEG [32,33,34]. Independently of PEG-volume regimens, it is well-known that a BBPS < 6 or inadequate bowel preparation (lower than good for the Aronchick scale) is associated with a lower ADR [1,35,36]. Additionally, a recently published paper highlighted that a fair/inadequate BP was also associated with an increased risk of PCCRC-related death [37].

However, there are conflicting data regarding the correlation between the BP quality and ADR. As a matter of fact, it appears that there is no clear linear correlation between these two factors. In particular, when the BP is adequate, improving the quality of the BP (e.g., shifting from a partial BBPS score of 2 to 3 or from a good to excellent BP according to the Aronchick scale) does not always lead to significant differences in the ADR. Furthermore, statistical differences in the rate of adequate BP between two different BP regimens (e.g., 2L-PEG vs. 4L-PEG) or BP kits do not always result in significant differences in the ADR [32,36,38,39,40].

Similarly, while split-dose regimens have been shown to improve BP, this does not always translate into a significant increase in the ADR [41,42].

Fewer studies have explored the relationship between BP quality and the SPDR, but it appears that the challenging nature of detecting sessile serrated polyps makes more relevant the quality of BP. In a prospective study, serrated polyps were detected in a smaller percentage of patients with BBPS scores of two than scores of three [43]. Additionally, in the population-screening-based study by Zessner-Spitzenberg, the odds for detecting a serrated polyp were lower in patients with good and fair BP compared to excellent BP (excellent as reference, OR = 0.84 for good BP and OR = 0.83 for fair BP). Even combining excellent and good bowel preparation in a single category [37]. However, further studies are needed to confirm this preliminary data.

Simethicone (SIM) is an anti-foaming agent, occasionally added to laxatives to reduce bubble formation and potentially improve the quality of bowel cleansing. Two meta-analyses investigated the impact of SIM and colonoscopy outcomes [44,45]. Oral SIM appears to improve colon cleansing and Boston Bowel Preparation Scale scores, though its effectiveness seems limited to regimens other than 2L-PEG and non-split-dose regimens. There was no difference in the ADR between groups with or without SIM; however, the detection rate of polyps and adenomas in the right colon was significantly higher in the SIM group (respectively RR = 1.57 and RR = 1.15).

Patient education plays a crucial role in BP, with several studies indicating that giving enhanced instructions to patients before colonoscopy, such as verbal instructions, multimedia, written materials, or phone calls, leads to better BP, an increased ADR (RR = 1.37), and an increased sessile serrated ADR (OR = 1.76) [46].

Among the bowel preparation domain, the only interventions that improved ADRs were a split-dose regimen and enhanced instructions before the colonoscopy. These interventions are relatively simple to implement in endoscopy units, requiring minimal technological resources and financial investment.

6. Devices and Scopes

Several devices and scopes have been produced, starting from the fact that the more mucosa you can see, the more polyps can be found. With distal attachments on the tip of the colonoscope, it is possible to flatten the colonic folds, and new specialized scopes allow for a larger view and make it possible to see behind the fold, aiming to increase the area of the mucosa evaluated.

6.1. Cap

CAP acts as a small attachment on the tip of the colonoscope, which allows for better positioning and visibility around folds or bends. It also stabilizes the colonoscope tip, making it easier to navigate through challenging areas of the colon. Although a large RCT showed a better ADR and a faster cecal intubation time, two other meta-analyses of 7 and 16 RCTs, respectively, showed only a marginal advantage [47,48,49]. Furthermore, in another recent meta-analysis, a cap colonoscopy did not reach statistical significance in the detection of proximal colon adenomas compared to a standard colonoscopy [50].

Despite its low cost and wide availability, the conflicting results prevent the routine recommendation of a cap-assisted colonoscopy for screening or polyp surveillance.

6.2. Endocuff/Endocuff Vision

Endocuff (EC) and Endocuff Vision (ECV) (Olympus Corp., Tokyo, Japan) consist of a double and single row of fingers or flexible arms, respectively, designed to invert and flatten the colonic folds in order to improve the vision of the blind spots (Figure 1).

A large RCT demonstrates a significantly higher ADR (40.9% vs. 36.2%) and sessile serrated polyp detection rate (2.3% vs. 1.1%) for ECV, with no decrease in caecal intubation compared to a standard colonoscopy [51].

A subsequent meta-analysis of 23 RCTs involving 17,999 patients confirmed that EC/ECV improved ADRs, PDRs, and sessile serrated detection rates, particularly when a colonoscopy was performed by less experienced endoscopists with a baseline ADR below 50% [52]. Head-to-head studies are needed to clarify which Endocuff is better [53].

6.3. G-Eye

G EYE (ELUXEO EC-760 R endoscope, Fujifilm) is a colonoscope equipped with a balloon at the distal end of the tube. The balloon is kept deflated during insertion and then inflated during the withdrawal phase in order to flatten the folds and facilitate the stabilization of the endoscope. An RCT showed that a G-EYE colonoscopy offered advantages over a standard colonoscopy in ADRs and sessile serrated polyp detection rates [54]. A recent meta-analysis of three studies that pooled ADRs, PDRs, and SPDRs showed a statistically significant improvement with G-EYE (ADR, OR: 1.744 and serrated polyp detection rate, OR: 1.603) [55]. However, additional data are required to better understand its potential benefits in routine clinical practice.

6.4. Fuse

Full-spectrum endoscopy (Boston Scientific Corp., Marlborough, MA, USA) is a new technology featuring an endoscope equipped with three forward-facing and two side-facing lenses, providing a 330-degree field of view compared to the 170 degrees of a standard colonoscope. A large RCT and few retrospective studies failed to show an advantage of using FUSE compared to a standard frontal-view colonoscopy in ADR [56].

A meta-analysis by Facciorusso et al. of eight RCTs found no difference in terms of the PDR and ADR [57]. Due to the negative results, FUSE has fallen into disuse.

6.5. Third Eye Retroscope

A third eye retroscope (TER) (Avantis Medical Systems, Sunnyvale, CA, USA) is a through-the-scope device designed to provide a retrograde view of the colon, improving the detection of polyps located between the folds of the colon. In a prospective study of 298 subjects, the TER detected 14.8% more polyps compared to a standard colonoscopy [58]. In a RCT, the TER improves the ADR and PDR with a relative risk of missed diagnosis of 2.56 and 1.92, respectively. However, a post hoc analysis revealed that these results were consistent for a surveillance colonoscopy but not for a screening colonoscopy, likely due to the differing nature of the procedures [59]. Given the elevated costs, technical challenges, and limited application, TER usage has declined.

6.6. AmplifEYE

AmplifEYE (Cantel/Medivators, Minneapolis, MN, USA) is a device mounted on the end of the colonoscope and has a single row of arms, similar to Endocuff Vision. Only one RCT involving 344 patients compared amplifEYE to a standard colonoscopy. Although amplifEYE showed a higher PDR and serrated detection rate, the ADR did not show a statistically significant advantage. In a non-inferior RCT, AmplifEYE was not inferior to an ECV in terms of the adenoma-per-colonoscopy, ADR, and sessile serrated polyp detection rate. However, 73% of the procedures were performed by a single operator, which limits the generalizability of the results [60]. More research is needed to determine whether AmplifEYE can be routinely recommended for clinical use.

Among the various devices discussed, only Endocuff and Endocuff Vision have consistently demonstrated a positive impact on the ADR, making them the most promising options in this category. However, further studies are required for most of the other devices to establish their clinical utility and support their routine use in colonoscopies for screening and surveillance.

7. Intra-Procedural Strategies and Techniques

In the “intra-procedural strategies and techniques” domain, there are several potential interventions that can be easily applied immediately by most endoscopists without acquiring additional equipment with low financial impact.

7.1. Withdrawal Time

The withdrawal time (WT) is a well-known key quality indicator in colonoscopies, defined as the duration from cecal intubation to the removal of the colonoscope, excluding the time spent on polypectomy, terminal ileal intubation, or other therapeutic procedures. A randomized controlled trial showed that a 9 min WT significantly improved the ADR compared to a 6 min one (36.6% vs. 27.1%), with marked improvements in the proximal colon ADR, too (21.4% vs. 11.9%) [61]. Additionally, a meta-analysis conducted by Bhurwal et al. provides further evidence that at least a 9 min WT leads to a higher ADR, improving the serrated lesion detection rate [62]. Both the American Society for Gastrointestinal Endoscopy (ASGE) and the ESGE currently advise a higher withdrawal time than 6 min, with the ASGE recommending a minimum time of WT ≥ 8 min and the ESGE recommending an aspirational goal of at least 10 min [1,63].

7.2. Water-Aided Colonoscopy

Water-assisted colonoscopy offers several advantages over air insufflation, with proposed mechanisms including the straightening of the sigmoid colon, overall shortening of the colon, increased lubrication of the colonoscope, and reduced colonic spasms. Water-assisted colonoscopy typically involves two techniques: water immersion (WI) and water exchange (WE). WI entails infusing water during the insertion of the colonoscope, followed by aspiration during withdrawal, whereas WE involves the infusion of water while simultaneously aspirating the solution containing residual fecal contents.

A Cochrane Review of 16 RCTs involving over 2900 patients found that colonoscopies utilizing water-assisted techniques, including WI and WE, significantly improved the ADR compared to gas insufflation (RR 1.16) [64]. In a mixed-gender European population, WE is confirmed to be a superior insertion technique, showing a significant increase in <10 mm right colon adenoma detection compared to insufflation (19% vs. 12.1%), achieving the cleanest colon and lowest proportions of poor bowel preparation, which requires repeat procedures [65]. A subsequent multicenter, three-armed RCT comparing WE, WI, and air insufflation reaffirmed the effectiveness of WE techniques in enhancing the ADR compared to air insufflation (49.3% vs. 40.4%). Notably, in the same study, WE also achieved better outcomes in the proximal colon ADR, advanced ADR, adenoma per colonoscopy (APC), and cleanliness scores compared to air insufflation [66]. Lastly, a systematic review with a network meta-analysis of randomized controlled studies concluded by Fuccio et al. showed that WE significantly increased the overall ADR and also the right colon ADR in a screening colonoscopy [67].

Current evidence demonstrates the clear superiority of water-aided colonoscopy techniques, highlighting the importance of incorporating them into our clinical practice.

7.3. Dual Observation

To overcome the operator’s inattentional blindness, an easy solution is to include a second observer during the colonoscopy, especially during the withdrawal phase. Dual observation (DO) can be particularly beneficial in detecting flat or small adenomas that might be overlooked during the mucosal examination. Aziz et al. conducted a meta-analysis of five RCTs, showing a higher ADR in those colonoscopies performed with an experienced nurse as an additional observer (OR = 1.24) [68]. Additionally, the presence of a fellow has been associated with an increased mean number of adenomas per colonoscopy without increasing the risks of adverse events [69].

7.4. Dynamic Position Changes

Changing the patient’s position during a colonoscopy, such as rotating them from the left lateral to supine positions or to the right lateral decubitus, can improve the visualization of different segments of the colon. Positional changes during the withdrawal phase help redistribute gas and fluid within the colon, potentially improving mucosal visibility and facilitating the detection of lesions in difficult-to-see areas, such as the transverse colon and splenic flexure [70,71]. In a recent meta-analysis of five RCTs, the ADR was higher in the dynamic position change group compared to the static position during a colonoscopy (OR = 1.34), with no significant increase in WT [72].

Although dynamic positioning is easily performed during a colonoscopy, there are some possible limitations, determined by the difficulty or impossibility of turning the patient, such as in cases of patients under deep sedation or with certain physical and/or clinical limitations.

7.5. Multiple Examination of the Proximal Colon

While rectal retroflexion is an established procedure for evaluating the distal rectum and the ano-rectal junction, the effectiveness of other retroflexion maneuvers during colonoscopy for improving lesion detection remains a subject of ongoing debate. The proximal colon, which includes the cecum, the ascending colon, and the hepatic flexure, is a common site for missed lesions. Re-examining this segment, especially after completing the initial withdrawal, can be crucial. Lesions in the proximal colon are often flat or sessile and can be easily missed during a standard examination. The issue of the polyp miss rate (PMR) is particularly pronounced in the ascending colon, where it is significantly higher compared to the distal colon. This discrepancy is associated with an increased incidence of interval colorectal cancer (CRC), particularly in the proximal colon, with such cancers being diagnosed 3–5 years after an initial colonoscopy [73,74,75,76]. A pooled analysis by Desai et al. revealed that a second forward view (SFV) of the right colon increased the right-sided ADR by 10%, while retroflexion increased the right-sided ADR by 6% [77]. However, when comparing the SFV with right-colon retroflexion (RCR) in terms of the adenoma miss rate (AMR), an analysis of three eligible studies found no statistically significant difference. In addition, a recent meta-analysis of RCT trials [78] has demonstrated that an SFV of the right colon was associated with an 8% increase in the right-sided ADR, with no significant difference in the additional ADR between an SFV and RCR. Furthermore, when comparing an SFV to a standard colonoscopy (SC), there were no notable differences in WTs (Figure 2).

These findings support the use of an SFV as the preferred method for examining the right colon. However, emerging evidence suggests that an RCR following an SFV examination may further enhance the detection rate. In a prospective study by Lee et al. involving 1020 patients undergoing a screening or surveillance colonoscopy, all patients first underwent an SFV examination, followed by an RCR. Retroflexion was successfully and safely performed in the majority (82.4%) of cases [79]. Notably, the additional retroflexed examination from the cecum to the hepatic flexure significantly increased the adenoma detection rate from 25.5% after two standard forward-view examinations to 27.5%.

Miyamoto et al. also conducted a prospective multicenter study to evaluate the impact of an RCR after an SFV. Of the enrolling 777 participants, retroflexion was successful in 730 (94.0%) participants, with only a small number of participants showing minor bleeding (3.0%) or mucosal tears (0.8%). The repeated forward-view withdrawal technique detected 291 adenomas, while the third withdrawal in the right colon detected 53. The adenoma miss rate for the repeated forward-view withdrawal was 15.4% [80].

In both these studies, during the retroflexion, they found not only diminutive lesions but also serrated sessile adenomas/polyps (SSA/Ps) over 10 mm and mucosal cancer over 20 mm. These findings are particularly significant as they highlight the effectiveness of an RCR in detecting not only diminutive lesions but also more clinically relevant SSA/Ps, which are critical for preventing advanced colorectal cancer. These results underscore the potential role of RCRs performed after a double forward-view assessment. This approach may enhance our ability to identify not only diminutive lesions but also those with subtle and challenging morphologies, which are often more difficult to detect and are strictly connected with interval colorectal cancer.

The interventions outlined in the “intra-procedural strategies and techniques” appear to be effective in improving the ADR and serrated polyp detection rate so much that recent ASGE guidelines recommend incorporating all of these strategies as part of a screening colonoscopy [63].

8. Chromoendoscopy

Chromoendoscopy is an advanced imaging technique traditionally employed to enhance the characterization of mucosal lesions. This result can be obtained either by the application of dyes (vital or contrast staining administered through the working channel of the endoscope or taken orally) or advanced image processing technologies (virtual chromoendoscopy).

8.1. Dye-Based Chromoendoscopy

Indigo Carmine is a deep-blue contrast dye that penetrates into mucosal grooves, thereby enhancing the visualization of surface patterns (Figure 3). A multicenter RCT by Pohl et al. [81] demonstrated that a chromoendoscopy with Indigo-Carmine-spraying significantly increased the overall ADR adenomas (0.95 vs. 0.66 per patient), flat adenomas (0.56 vs. 0.28), and serrated lesions (1.19 vs. 0.49) compared to a standard colonoscopy. Conversely, Lesne et al. reported no significant difference in the ADR when comparing a blue-water infusion colonoscopy to a standard colonoscopy, although the adenoma-per-colonoscopy was significantly higher in the blue-water infusion group [82].

Methylene Blue is a blue dye originally described for detecting intestinal metaplasia in the gastric mucosa [83]. A meta-analysis of 10 RCTs by Antonelli et al. showed that in the dye-based group, the ADR was higher compared to a conventional colonoscopy (48.1% vs. 39.3%; RR, 1.20) [84]. In particular, this result was consistent for serrated sessile adenomas and broadly in right-sided lesions. However, a significant limitation of this technique, as with other dye-based methods, is the increased procedural time required. To address this issue, an oral formulation involving prolonged-release tablets of methylene blue (Methylene Blue MMX, Lumeblue, COSMO Pharma, Italy) has been proposed. A multicenter RCT demonstrated that the ADR was significantly higher in the 200 mg MB-MMX group compared to the placebo group (56.3% vs. 47.8%, OR = 1.41) [85]. These findings suggest that a dye-based chromoendoscopy has substantial potential to enhance the ADR, particularly for flat or non-pedunculated adenomas, despite its time-intensive nature. New formulations may mitigate this limitation by improving the manageability of the technique.

8.2. Virtual Chromoendoscopy

Autofluorescence imaging (AFI) is a technique that captures fluorescence emitted by tissues and translates these signals into distinct colors based on their characteristics: violet-magenta for colonic lesions and green for normal tissue. Several studies demonstrated that AFI improves the ADR (30% vs. 49%) and reduces the AMR. However, these benefits were primarily observed among less-experienced endoscopists, as experienced endoscopists already demonstrated a high ADR with high-resolution endoscopy [86,87]. While AFI shows potential to increase the ADR, its routine use in clinical practice remains limited.

Fuji Intelligent Color Enhancement (FICE, Fujifilm, Tokyo, Japan) is a digital chromoendoscopy technique that processes images by applying specific wavelengths to enhance surface details, pattern architecture, and vascular structures. In a single-center Korean RCT on 359 patients, FICE and WLI were compared, but no significant improvement in the adenoma miss rate (or detection rate) was found [88].

The iSCAN (Pentax, Tokyo, Japan) is a post-processing digital imaging technology that integrates three different algorithm modes (surface, contrast, and tone enhancement) to better delineate the mucosal surface. A systematic review and meta-analysis of five RCTs by Aziz et al. [89] demonstrated that the ADR was significantly higher using any iSCAN with surface and contrast enhancement compared to a high-definition colonoscopy (RR = 1.25), suggesting that this method holds considerable promise.

Narrow Band Imaging (NBI, Olympus, Tokyo, Japan) is an advanced optical imaging technology that filters white light into narrow bands using two laser light sources, enhancing the visualization of capillaries in the superficial mucosa and deeper vessels in mucosa and submucosa (Figure 4). A meta-analysis of 11 trials demonstrated that given optimal bowel preparation conditions, second-generation NBI had a higher ADR compared to WLI (OR = 1.28), while first-generation NBI showed no such advantage. NBI was particularly superior in detecting flat polyps. Notably, many of the additional polyps detected by NBI were non-adenomatous [90]. The role of NBI in enhancing the ADR remains debated, as other meta-analyses, such as the one by Dinesen et al., did not observe a significant improvement [91].

Blue Light Imaging (BLI, Fujifilm, Tokyo, Japan), similarly to NBI, exploits two monochromatic laser light sources to improve the delineation of vessels and structures in the superficial mucosa, providing bright images. However, evidence regarding the efficacy of BLI in increasing the ADR conflicts with RCTs [92,93].

Linked Color Imaging (LCI, Fujifilm, Tokyo, Japan) utilizes a pre-processing technology that enhances chromatic contrast through specific light wavelengths, improving the detection of subtle mucosal changes, such as early-stage lesions or inflammation. An RCT involving 995 patients found no significant difference in the ADR between LCI and WLI [94]. However, two subsequent RCTs conducted reported a higher ADR in the LCI group compared to WLI [95,96]. These results suggest that LCI may be a valuable tool in improving the ADR.

Several virtual chromoendoscopy technologies have shown promising results; however, the requirement for brand-new colonoscopy equipment and processors, along with the need for specialized training in lesion recognition using these systems, makes the widespread adoption of these technologies both challenging and financially burdensome. Moreover, its use in routine clinical practice may prolong the duration of the examination, so its specific role during colonoscopy needs to be fully evaluated. Currently, its application appears to be limited to assessing the pit and vascular patterns of colonic lesions rather than providing a complete evaluation of the colonic mucosa.

9. Artificial Intelligence

The use of deep-learning artificial intelligence (AI) computer-aided detection (CADe) systems to improve adenoma detection during a colonoscopy has been an active area of research in recent years, with many promising results. A meta-analysis by Barua et al. in 2020 synthesized data from five RCTs conducted between 2019 and 2020 and found that AI-assisted colonoscopy consistently improved the ADR, with a pooled relative risk (RR) of 1.52 in favor of AI-assisted colonoscopies [97].

Many other recent RCTs have confirmed the impact of AI-assisted colonoscopies on improving adenoma detection. Combined data from two randomized trials by Repici and colleagues evaluating the performance of AI-assisted colonoscopy by expert and non-expert endoscopists showed that AI significantly improved the ADR (53.3% vs. 44.5%), with an RR of 1.29 in favor of AI (GI Genius, Medtronic). Interestingly, this result was independent of the endoscopist’s experience level, which appeared to play no role [98].

In line with these results, a 2022 large US multicenter RCT including both academic and community medical centers found that the use of a novel AI system led to a significantly higher number of adenomas detected during a screening and surveillance colonoscopy. The study reported an increase in adenomas per colonoscopy (APC) from 0.83 to 1.05 with the use of AI [99]. More recent data also found similar increases in the APC with AI (0.71 vs. 0.51) as well as a similar reduction in the adenoma miss rate compared to a standard colonoscopy (19% vs. 36%) [100]. Moreover, a 2023 meta-analysis of RCTs highlighted how the use of AI can decrease the adenoma miss rate (AMR), which is another important performance metric of colonoscopy quality [101].

Additionally, AI appears to markedly improve adenoma detection among trainee endoscopists with <500 procedures and <3 years of experience (mean adenomas per colonoscopy 1.48 vs. 0.86) [102].

While much of the research has focused on adenoma detection, the impact of AI on sessile serrated lesion detection is also of significant interest, as they are known to be more challenging to detect, particularly those in the proximal colon. In this regard, a 2024 meta-analysis of seven RCTs found that, in addition to significantly reducing the adenoma miss rate (RR 0.46, p < 0.001), an AI-assisted colonoscopy also similarly reduced the miss rate for sessile serrated lesions, with a pooled RR of 0.43 [103].

Despite the promising results, several challenges and limitations still need to be addressed. A 2023 meta-analysis of studies evaluating an AI-assisted colonoscopy in non-randomized real-world settings found that although AI was statistically superior in terms of the ADR, the actual improvement in the ADR was very small (36.3% vs. 35.8%), suggesting the findings from RCTs may not translate directly into real-world settings [104]. False-positive detections with AI systems are another important issue, which could lead to unnecessary interventions or prolonged procedure times, and managing them remains an important issue [105]. The upfront costs of implementing these AI systems are also substantial, particularly in resource-limited settings, although estimates suggest that through the prevention of colorectal cancer cases, AI-assisted colonoscopies may ultimately be cost-effective and even cost-saving due to improved cancer prevention [106].

Ongoing research is focused on developing more sophisticated AI algorithms that can not only detect polyps but also classify them and predict malignant potential in real time. A 2024 meta-analysis pooling data from 10 studies found that although current AI systems have close to 90% sensitivity and specificity for predicting diminutive (≤5 mm) rectosigmoid polyp histology, there was no significant improvement with AI assistance compared to endoscopist predictions without AI in four studies that evaluated this [107].

Finally, in addition to enhancing the detection and optical diagnosis of lesions, AI systems are also being developed to improve colonoscopy quality in other ways. A recent study found that combining CADe with AI-assisted real-time monitoring of the withdrawal speed further improved the ADR compared to CADe alone [108]. Overall, the field of AI-assisted colonoscopies is undergoing rapid development and promises to significantly improve patient outcomes in the future, although at present, more data from real-world settings are required to ultimately confirm the benefits of this approach.

Table 1. Randomized controlled trials and meta-analysis on strategies to enhance adenoma detection rate and serrated polyp detection rate.

Domain	Intervention	Study	Type of Study	Indicator	Results	Interpretation
Bowel Preparation	Split-dose regimen (vs. day-before regimen)	Radaelli F. Gut, 2013 [41]	RCT	ADR	53.0% vs. 40.9%, p = 0.002	Split dose regimen improves ADR
Bowel Preparation	Simethicone (vs. without)	Liu X. J Clin Gastro, 2021 [45]	Meta-analysis	ADR	RR = 1.02; p = 0.68	Simethicone does not improve ADR
Devices	Endocuff Vision (vs. without)	Ngu WS. Gut, 2019 [51]	RCT	ADR	40.9% vs. 36.2%, p = 0.02	ECV improves ADR
Devices	G-EYE (vs. without)	Shirin H. GIE, 2019 [54]	RCT	ADR	48% vs. 37%, p = 0.0027	G-EYE improves ADR
Chromoendoscopy	Methylene blue (vs. without)	Antonelli G. GIE, 2022 [84]	Meta-analysis	ADR; SPDR	48.1% vs. 39.3%, RR, 1.20; 6.1% vs. 3.5%; RR, 1.68	Methylene blue improves right sided ADR and SPDR
Chromoendoscopy	iSCAN (vs. HD colonoscopy)	Aziz M. Endosc Int Open, 2022 [53]	Meta-analysis	ADR	43.4% vs. 39.7%, RR 1.20, p = 0.003	iSCAN improves ADR
Intra-procedural strategies	Withdrawal Time (≥9 min vs. 6–9 min)	Bhurwal A. JGH, 2021 [62]	Meta-analysis	ADR; SPDR	OR 1.54; OR 1.68	≥9 min of WT increases ADR and SPDR compared to 6–9 min WT
Intra-procedural strategies	Water-aided Colonoscopy (vs. air-insufflation)	Hafner S. Cochrane Database of Systematic Reviews, 2015 [64]	Meta-analysis	ADR	RR 1.16, 95% CI 1.04 to 1.30, p = 0.007	Water infusion improves ADR
Intra-proceduural strategies	Dual-Observation (vs. one observer)	Aziz M. CGH 2020 [68]	Meta-analysis	ADR	33.9% vs. 29.5%; RR 1.24; p = 0.004	DO improves ADR
Intra-proceduural strategies	Dynamic Position Changes (vs. without)	Li P. Surg Endosc, 2021 [72]	Meta-analysis	ADR	OR 1.34; 95% CI 1.13–1.59; p < 0.001	Dynamic position changes increases ADR
Intra-procedural strategies	Second examination of the right colon (vs. SC)	Desai M. GIE, 2019 [77]	Meta-analysis	rADR	rSFVand RCR increased ADR by 10% and 6% compared to SC	Both SFV and RCR improves ADR
Artificial Intelligence	CADe (GI Genius, Medtronic) (vs. SC)	Repici A. Gut, 2022 [98]	RCT	ADR	ADR the CADe than SC group (53.3% vs. 44.5%; RR = 1.22; p = 0.02 for superiority analysis	CADe improves ADR

Abbreviations: RCT, randomized controlled trial; ADR, adenoma detection rate; ECV, Endocuff Vision; SPDR, serrated polyp detection rate; WT, withdrawal time; SC, standard colonoscopy; DO, dual observation; rADR, right adenoma detection rate; rSFV right second forward view; RCR, right colon retroflexion; CADe, computer-aided detection.

synthesis RCTs and meta-analysis on strategies to enhance the ADR and the SPDR.

10. Discussion & Conclusions

Recent evidence regarding the relationship between the ADR, SPDR, and the risk of PCCRC highlights the necessity of developing strategies to improve these colonoscopy quality indicators. While several guidelines assess the optimal range of the ADR (

Table 2. Established thresholds for ADR and proposed thresholds for SPDR.

Study	Indicator	Population	Threshold	Guidelines
Kaminski M, Endoscopy, 2017 [1]	ADR	Screening	≥25%	Endoscopy, ESGE guidelines
Rex DK, GIE, 2015 [110]	ADR	Screening	Men, ≥ 30% Women, ≥ 20%	ASGE/ACG review
Rex DK, GIE/AJG 2024 [63]	ADR	Screening, surveillance, or diagnostic indications (except FOBT+ or CT-colonography)	Men, ≥40% Women, ≥30%	ASGE/ACG review
Rex DK, GIE/AJG, 2024 [63]	ADR	FOBT +	Men, ≥55% Women, ≥45%	ASGE/ACG review
Rex DK, GIE/AJG, 2024 [63]	SPDR	Screening, surveillance, or diagnostic colonoscopy	SSLDR ≥ 6%	ASGE/ACG review

Abbreviations: ADR, adenoma detection rate; ESGE, European Society of Gastrointestinal Endoscopy; ASGE, American Society for Gastrointestinal Endoscopy; ACG, American College of Gastroenterology; FOBT, fecal occult blood test; AJG, American Journal of Gastroenterology, SPDR, serrated polyp detection rate; SSLDR, sessile serrated lesion detection rate.

), no formal threshold has been proposed for serrated polyps until the very recent update of “Quality Indicators for Colonoscopy” by the ASGE and American College of Gastroenterology (ACG), which proposed a minimum for the sessile serrated lesion detection rate [63]. The necessity to include a threshold for serrated polyps is derived from the fact that several population studies have highlighted how their detection predicts PCCRC independent of the ADR [8,9,109,110]. This difficulty in proposing a universal threshold for serrated polyps is derived from several issues: firstly, there is a recognized interobserver variation among pathologists in the differential diagnosis between SSLs and hyperplastic polyps. Furthermore, the different definitions of serrated polyps, the inclusion of small hyperplastic polyps in the calculation, and also their size and location may be barriers to the SPDR threshold [63].

In the future, novel quality indicators may be proposed or integrated into guidelines, such as the adenoma miss rate (AMR) or adenoma per colonoscopy (APC). The AMR has been shown to be crucial, as higher rates are associated with an increased incidence of PCCRC. However, determining the AMR necessitates consecutive colonoscopies, which is challenging in routine clinical settings. Furthermore, there is a paucity of data from tandem colonoscopy studies to establish a reliable threshold for this metric [111].

This comprehensive review provides an overview of the most recent and effective approaches. Among the effective strategies, some older methods, such as a prolonged withdrawal time, the use of water-assisted colonoscopy, and a second evaluation of the right colon, offer significant improvements in the ADR without a considerable economic impact, making them accessible to most healthcare settings. On the contrary, certain newer technologies, such as artificial intelligence, G-EYE, and a third-eye endoscopy, appear promising but require substantial financial investment.

Future clinical studies comparing these technologies may help to find the most effective interventions, and, lastly, the combined use of endoscopic procedures and devices may further enhance the quality of colonoscopies, reducing colorectal cancer incidence and the risk of PCCRC.