CC BY-NC-ND 4.0 · Endosc Int Open 2018; 06(10): E1214-E1223
DOI: 10.1055/a-0650-4258
Original article
Owner and Copyright © Georg Thieme Verlag KG 2018

Cap-assisted colonoscopy: a meta-analysis of high-quality randomized controlled trials

Venkat Nutalapati
1  Department of Internal Medicine, The University of Kansas Medical Center, Kansas City, Kansas, United States
,
Vijay Kanakadandi
2  Department of Gastroenterology, The University of Kansas Medical Center, Kansas City, Kansas, United States
,
Madhav Desai
2  Department of Gastroenterology, The University of Kansas Medical Center, Kansas City, Kansas, United States
,
Mojtaba Olyaee
2  Department of Gastroenterology, The University of Kansas Medical Center, Kansas City, Kansas, United States
,
Amit Rastogi
2  Department of Gastroenterology, The University of Kansas Medical Center, Kansas City, Kansas, United States
› Author Affiliations
Further Information

Corresponding author

Amit Rastogi, MD
Department of Gastroenterology & Hepatology
University of Kansas Medical Center
3901 Rainbow Blvd
Kansas City, KS 66160
USA   

Publication History

submitted 25 January 2018

accepted after revision 09 May 2018

Publication Date:
08 October 2018 (online)

 

Abstract

Background and study aims Standard colonoscopy (SC) is the preferred modality for screening for colon cancer; however, it carries a significant polyp/adenoma miss rate. Cap-assisted colonoscopy (CC) has been shown to improve polyp/adenoma detection rate, decrease cecal intubation time and increase cecal intubation rate when compared to standard colonoscopy (SC). However, data on adenoma detection rate (ADR) are conflicting. The aim of this meta-analysis was to compare the performance of CC with SC for ADR among high-quality randomized controlled trials.

Patients and methods We performed an extensive literature search using MEDLINE, EMBASE, Scopus, Cochrane and Web of Science databases and abstracts published at national meetings. Only comparative studies between CC and SC were included if they reported ADR, adenoma per person (APP), cecal intubation rate, and cecal intubation time. The exclusion criterion for comparing ADR was studies with Jadad score ≤ 2. The odds ratio (OR) was calculated using Mantel-Haenszel method. I2 test was used to measure heterogeneity among studies.

Results Analysis of high-quality studies (Jadad score ≥ 3, total of 7 studies) showed that use of cap improved the ADR with the results being statistically significant (OR 1.18, 95 % CI 1.03 – 1.33) and detection of 0.16 (0.02 – 0.30) additional APP. The cecal intubation rate in the CC group was 96.3 % compared to 94.5 % with SC (total of 17 studies). Use of cap improved cecal intubation (OR 1.61, 95 % CI 1.33 – 1.95) when compared to SC (P value < 0.001). Use of cap decreased cecal intubation time by an average of 0.88 minutes (95 % CI 0.37 – 1.39) or 53 seconds.

Conclusions Meta-analysis of high-quality studies showed that CC improved the ADR compared to SC.


#

Introduction

Population-based colorectal cancer (CRC) screening has been shown to reduce incidence of colon cancer and related mortality [1] [2]. Among patients at average risk, the most favored cancer prevention test is colonoscopy every 10 years, beginning at age 50 (45 for African-Americans) [3]. Screening per 1000 patients using colonoscopy, a gain of 270 life-years and a decrease in 24 deaths from CRC has been estimated [4].

However, despite being the reference standard, colonoscopy is far from a perfect test. Studies using compute tomography colonography have estimated the sensitivity of colonoscopy for detecting advanced adenomas to be 88 % [5]. Tandem colonoscopy studies have shown that up to one-quarter of polyps are missed during colonoscopy [6]. Adenoma detection rate (ADR) has been shown to be associated with interval colon cancer and related mortality [7] [8]. ADR ≥ 30 % for men and ≥ 20 % for women has been recommended as a quality indicator for colonoscopy [9]. Wide variations in ADRs for endoscopists have been reported [10] [11]. Therefore, various methods have been employed in attempts to improve ADR, including brief educational interventions [12], use of distal attachments such as caps [13], third-eye retroscopes, newer-generation wide-angle colonoscopes, cuffs and EndoRings.

Cap-assisted colonoscopy (CC) has been extensively studied as a modality to improve ADR. The cap is a straightforward attachment on the distal end of the endoscope that extends outward beyond the tip of tje colonoscope to varying lengths. The cap helps in deflecting and flattening the mucosal folds, and by keeping the mucosa away from the lens prevents a red-out. These maneuvers expose the proximal aspects of colonic folds and thereby help in detecting polyps in these otherwise blind mucosal areas. Use of cap has been shown to decrease cecal intubation time, increase cecal intubation rate and improve polyp detection rate. However, data on ADR are rather conflicting. The aim of this meta-analysis was to compare the performance of CC with standard colonoscopy (SC) for ADR among high-quality randomized controlled trials (RCT).


#

Patients and methods

Search strategy

An electronic search was performed in MEDLINE, EMBASE, Google scholar, Cochrane database and Web of science. The search for studies of relevance was performed using the following key words and corresponding Medical Subject Heading/Entree terms when possible: “CAP assisted colonoscopy,” “colonoscopy with distal attachment,” “adenoma detection rate,” “adenoma per person,” “cecal intubation rate,” “cecal intubation time” with varying combinations with and/or. We retrieved 2558 abstracts ([Fig. 1]). Abstracts published in major international conferences, including Digestive Disease Week, United European Gastroenterology Week and Asia Pacific Digestive Week over the past 10 years were manually searched. References from major trials and review articles were manually searched.

Zoom Image
Fig. 1 Study flow diagram depicting search strategy, screening and studies of cap-assisted colonoscopy identified for inclusion in the meta-analysis of adenoma detection rate.

From the 2400 records, 2358 records were removed (1473 studies, 927 abstracts) because they were not relevant to the comparison between CC and SC. Of the remaining 42 records, 23 were excluded for the following reasons: duplicity, case report, review article, editorial, abstract only. Of the 19 full-text articles that were accepted, only 7 met the criteria of prospective RCTs, Jadad score  ≥3 (see [Table 1]), reported ADR, and these studies were used for ADR and APP (adenomas detected per person) [14] [15] [16] [17] [18] [19] [20]. Of the 42 records, 17 studies were included that compared cecal intubation rate between CC and SC [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]. Thirteen studies were included that compared cecal intubation time between CC and SC [14] [15] [16] [17] [19] [21] [22] [23] [25] [26] [30] [31] [32]. For analysis of cecal intubation and cecal intubation time, even studies with Jadad score < 3 were included. ADR alone was the primary aim of the study. We removed the constraints for cecal intubation time or rate as we wanted to be less stringent and more inclusive for these endpoints. While ADR is a cornerstone quality indicator for colonoscopy, the other two are not.

Table 1

Studies and their respective Jadad scores.

Study

Final score

Tada 1997

Paper

0

Matsushita 1998

Paper

1

Kondo 2007

Paper

3 (No ADR/APP reported)

Horiuchi 2008

Paper

3

Shida 2008

Paper

0

Takano 2008

Abstract

0

Lee 2009

Paper

1

Choi 2009

Paper

0

Harada 2009

Paper

1

Sato 2009

Prelim Report

3 (No ADR/APP reported)

Takeuchi 2010

Paper

3

Tee 2010

Paper

3 (No ADR/APP reported)

Dai 2010

Paper

0

Hewett 2010

Paper

3

Park 2012

Paper

3

Rastogi 2012

Paper

3

De Wijkerslooth 2012

Paper

4

Frieling 2013

Paper

3 (No ADR/APP reported)

Pohl 2015

Paper

3


#

Data extraction

Two investigators (VN and MD) independently reviewed the studies and imported the data into a standardized form. In case of lack of consensus, the senior investigator (AR) reviewed the study independently and then made a final decision regarding the data point.

Data extracted were patient demographics, year of publication, study location, number of subjects, size of adenomas, number of adenomas detected, cecal intubation rate, cecal intubation time and study quality. Individual study and patient characteristics are shown in [Table 2].

Table 2

Study characteristics.

Author

Country

Sample

CC

SC

Age

Male (%)

Tada et al. [32]

Japan

 140

  70

  70

60

73

Matsushita et al. [26]

Japan

  24

  12

  12

59

63

Kondo et al. [24]

Japan

 456

 221

 235

61

60

Horiuchi et al. [16]

Japan

 835

 424

 411

64

65

Shida et al. [28]

Japan

 178

  82

  96

64

51

Takano et al. [29]

Japan

2502

1287

1215

NA

NA

Harada et al. [23]

Japan

 592

 289

 303

63

66

Lee et al. [25]

Hong Kong

1000

 499

 501

53

46

Sato et al. [27]

Japan

 221

 110

 111

NA

NA

Dai et al. [31]

China

 250

 121

 129

51

54

Hewett et al. [15]

United States

 100

  52

  48

62

57

Takeuchi et al. [20]

Japan

 274

 141

 133

64

70

Tee et al. [30]

Australia

 400

 200

 200

54

48

De Wijkerslooth et al. [14]

Netherlands

1339

 656

 683

60

51

Choi et al. [21]

Korea

 228

 114

 114

NA

NA

Rastogi et al. [19]

United States

 420

 210

 210

61

95

Park et al. [17]

Korea

 600

 300

 300

62

52

Frieling et al. [22]

Germany

 504

 252

 252

60 ± 15.5

182

Pohl et al. [18]

United States

1113

 562

 551

62

64


#

Statistical analysis

Meta-analyses were performed using Mantel-Haenszel method combining the results from different trials comparing CC and SC. Meta-Analysis was performed according to the PRISMA statement. A complete checklist is provided in [Table 3] [33]. A random effects model was used for statistical heterogeneity across trials and a fixed effect model was used if no significant heterogeneity was present. Relative risks (RR) with corresponding 95 % CI were calculated. Heterogeneity was calculated using I2 test. Publication bias was assessed using a funnel plot. Statistical analyses were performed using RevMan software (Review Manager version 5.3; The Nordic Cochrane Centre, Copenhagen, Demark, The Cochrane Collaboration 2015).

Table 3

PRISMA checklist.

TITLE

Title

 1

Identify the report as a systematic review, meta-analysis, or both.

Mentioned as meta-analysis

ABSTRACT

Structured summary

 2

Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number:

A detailed abstract with the necessary information has been provided

INTRODUCTION

Rationale/

 3

Describe the rationale for the review in the context of what is already known.

Provided

Objectives

 4

Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).

Provided

METHODS

Protocol and registration/

 5

Indicate if a review protocol exists, if and where it can be accessed (e. g., Web address), and, if available, provide registration information including registration number.

Not applicable with Meta-analysis

Eligibility criteria

 6

Specify study characteristics (e. g., PICOS, length of follow-up) and report characteristics (e. g., years considered, language, publication status) used as criteria for eligibility, giving rationale.

Provided

Information sources

 7

Describe all information sources (e. g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.

Provided

Search

 8

Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.

Provided

Study selection

 9

State the process for selecting studies (i. e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

Provided

Data collection process

10

Describe method of data extraction from reports (e. g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.

Provided

Data items

11

List and define all variables for which data were sought (e. g., PICOS, funding sources) and any assumptions and simplifications made.

Provided

Risk of bias in individual studies

12

Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.

Provided

Summary measures

13

State the principal summary measures (e. g., risk ratio, difference in means).

Provided

Synthesis of results

14

Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e. g., I2) for each meta-analysis.

Provided

Risk of bias across studies

15

Specify any assessment of risk of bias that may affect the cumulative evidence (e. g., publication bias, selective reporting within studies).

Provided

Additional analyses

16

Describe methods of additional analyses (e. g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-specified.

Provided

RESULTS

Study selection

17

Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.

Provided

Study characteristics

18

For each study, present characteristics for which data were extracted (e. g., study size, PICOS, follow-up period) and provide the citations.

Provided

Risk of bias within studies

19

Present data on risk of bias of each study and, if available, any outcome-level assessment (see Item 12).

Provided

Results of individual studies

20

For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group and (b) effect estimates and confidence intervals, ideally with a forest plot.

Provided

Synthesis of results

21

Present results of each meta-analysis done, including confidence intervals and measures of consistency.

Provided

Risk of bias across studies

22

Present results of any assessment of risk of bias across studies (see Item 15).

Provided

Additional analysis

23

Give results of additional analyses, if done (e. g., sensitivity or subgroup analyses, meta-regression [see Item 16]).

Provided

DISCUSSION

Summary of evidence

24

Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e. g., health care providers, users, and policy makers).

Provided

Limitations

25

Discuss limitations at study and outcome level (e. g., risk of bias), and at review level (e. g., incomplete retrieval of identified research, reporting bias).

Provided

Conclusions

26

Provide a general interpretation of the results in the context of other evidence, and implications for future research.

Provided

FUNDING

Funding

27

Describe sources of funding for the systematic review and other support (e. g., supply of data); role of funders for the systematic review

Provided


#
#

Results

Adenoma detection rate

An initial pooled analysis of eight RCTs (5681 patients) was performed, which showed a numerically higher ADR in the CC group compared to the SC group, but results were not statistically significant (OR 1.08, 95 % CI 0.97 – 1.21; I2 56 %) ([Fig. 2a]). However, when only high-quality RCTs were included (Jadad score  ≥ 3) as per the primary aim of this study, there were seven RCTs with a total of 4,681 patients (2,344 patients in the CC group, 2,337 patients in the SC group). We were unbale to include some studies with a score of 3 or more, as they lacked information regarding ADR/APP [22] [24] [30]. ADR was significantly higher in the CC group (OR 1.18, 95 % CI 1.03 – 1.33) ([Fig. 2b]). There was no significant heterogeneity in the ADR analysis (I2 = 0 %). Publication bias for studies included for ADR was assessed using a funnel plot ([Fig. 3]).

Zoom Image
Fig. 2 Forest plot of pooled estimates of adenoma detection rate using cap-assisted colonoscopy compared to standard colonoscopy. a Results with all eligible studies. b Results with only high-quality studies (Jadad score ≥ 3). c Results with only high-quality studies using random effects.
Zoom Image
Fig. 3 Funnel plot showing publication bias.

Analysis was also performed using a random effects model. Analysis of the seven high-quality RCTs using the random effects model showed significantly higher ADR in the CC group (OR 1.104, 95 % CI 1.02 – 1.18) ([Fig. 2c]).

Sensitivity analysis was not performed based on our stringent criteria to include only high-quality studies with Jadad score ≥ 3 which carry a very low risk for bias [34] [35] [36].


#

Mean adenomas detected per person

Analysis for APP included six RCTs with 4,368 patients. There were 2184 patients in each group. Use of cap led to a mean difference of 0.16 (95 % CI 0.02 – 0.30) additional APP ([Fig. 4]). Significant heterogeneity was found in the studies reporting mean APP (I2 = 68 %).

Zoom Image
Fig. 4 Forest plot of pooled estimate of adenoma per person (APP) showing higher detection of average adenoma per person using cap compared to standard colonoscopy.

#

Large adenoma detection rate

Analysis for large adenomas ( ≥ 10 mm) included four RCTs with 2468 patients. There were 1247 patients in the CC group compared to 1221 patients in the SC group. Use of cap led to a statistically significantly higher rate of detection of large adenomas (OR 1.49, 95 % CI 1.03 – 2.15, P < 0.005) with heterogeneity of (I2 = 44 %) ([Fig. 5]).

Zoom Image
Fig. 5 Figure plot of pooled estimate of adenomas > 10 mm, showing significant improved detection with CAP assisted colonoscopy compared to standard colonoscopy

#

Sessile serrated adenoma detection rate

Analysis for sessile serrated adenoma (SSA) included only three RCTs with 2872 patients. There were 1427 patients in the CC group compared to 1445 patients in the SC group. Use of cap did not lead to any significant difference in detection of SSA with (OR 1.12, 95 % CI 0.66 – 1.88) and a significant heterogeneity of (I2 = 76 %) ([Fig. 6]).

Zoom Image
Fig. 6 Figure plot of pooled estimate of sessile serrated adenoma (SSA) showing no significant improvement in the detection of proximal adenomas.

#

Cecal intubation rate and time

Pooled analysis of 17 studies that included 5416 patients in the CC and 5401 patients in the SC groups were utilized to evaluate the cecal intubation rate ([Fig. 7a]). The cecal intubation rate in the CC group was 96.3 % compared to 94.5 % with SC. Use of cap improved cecal intubation (OR 1.61, 95 % CI 1.33 – 1.95) when compared to SC (P < 0.001). Low heterogeneity was identified among studies (I2  = 2 %).

Zoom Image
Fig. 7 Forest plot of pooled estimates of cecal intubation rate (a) and cecal intubation time (b) showing improved rates and lesser time with cap compared to standard colonoscopy

Thirteen studies were used to analyze the impact of cap on cecal intubation time ([Fig. 7b]). The CC group included 3014 patients and the SC group included 3037 patients. Use of cap decreased the cecal intubation time by an average of 0.88 minutes (95 % CI 0.37 – 1.39) or 53 seconds. However, significant heterogeneity was detected among these studies (I2  = 87 %).


#
#

Discussion

Results of our meta-analysis indicate that use of cap improves detection of adenomas. An improvement in ADR, mean number of adenomas detected per patient and large adenomas was seen with CC. For ADR we included only trials with a Jadad score ≥ 3 to ensure only high-quality trials. The Jadad score is the most widely used scale to measure the quality of RCTs. Overall, we found seven RCTs with a Jadad score ≥ 3. This study differs from a previous meta-analysis [13] in that we excluded the study by Lee [25] as it employed suboptimal techniques for randomization. Proper technique includes a statistician and computer-generated randomization, where as in the study by Lee et al, only sealed envelopes were used without mention of statistician or a computer-generated sequence [25]. Furthermore, in that study, the quality of bowel preparation was significantly less satisfactory. They classified the quality of their bowel preparation into three categories: “excellent,” “fair,” and “poor.” In the results, they noted that a higher proportion of patients in the CC group had less satisfactory bowel preparation (excellent/fair/poor bowel preparation in CC group were 52.7:33.5:13.8 % vs. SC group's 62.3:28.1:9.6 %, respectively, P  = 0.006). They also reported an ADR that was lower with use of CAP. The inferior bowel preparation in the CC group could have negatively impacted the ADR. As a matter of fact, this is the only trial where use of CAP has been associated with lower ADR compared to standard colonoscopy. All other trials have shown either no difference or higher ADR with CAP.

ADR is a quality indicator for colonoscopy and has been shown to be associated with improved outcomes related to interval cancer and colorectal cancer-related mortality. While this meta-analysis shows an overall improvement in ADR with CC, individual studies have shown variable results. The study by Pohl et al. [18] which was the largest study evaluating CC in the United States showed that the impact on the individual endoscopist ADR is variable. The range of impact was from 20 % improvement to 15 % decrease in the individual ADR with CC. They also showed that those who preferred

CAP showed an improvement in ADR. We have also shown an improvement in the average number of adenomas detected per patient.

CC also improved detection of large adenomas, however, a statistically significant improvement in mean number of diminutive adenomas was not found. We suspect this may be due, in part, to the differing sizes of small adenomas reported (5 mm vs. 6 mm). There was no significant improvement in detection of proximal adenomas or SSAs as the RCTs that were performed were not adequately powered to detect any difference in the above outcomes.

Our meta-analysis has some limitations. The study populations in the studies were very diverse with studies being performed in Asia, North America, and Europe. That, however, improves generalizability of the results. Given the obvious lack of blinding of the endoscopists and the nature of such studies evaluating devices to improve ADR, investigator bias is unavoidable. Endoscopist experience in the different studies also varies widely and could not be accounted for with respect to the impact of CC on ADR. Use of a cap with colonoscopy requires some training, adjustment, and experience. This factor was not adjusted for or studied in the trials, making it difficult to gauge the impact of that on the results.

A cap is a simple, inexpensive and easy-to-use tool to improve the quality of colonoscopy. The cost of the cap, albeit low, appears to be the only negative factor weighing against its use in daily clinical practice. To derive maximum benefit from cap, endoscopists need to gain experience with the device. As the cap projects outside the tip of the colonoscope, it may appear to limit the angle of view. This must be compensated for withi adequate deflection of the tip and use of the edge of the cap to flatten the haustral folds to expose their proximal aspects and derive the maximum benefit. Furthermore, the benefit of CC has been shown to significantly extend visualization of the right colon in a colonoscopic training model [37]. Use of cap offers other secondary benefits such as improved cecal intubation rates and stabilization of the tip of the scope during polypectomy.


#

Conclusion

In conclusion, this meta-analysis showed that there is a marginal and statistically significant benefit to use of a cap during colonoscopy to improve ADR and cecal intubation rate and reduce cecal intubation time. Further research needs to be conducted to determine if there are specific patient subgroups that may benefit more from use of a cap, whether to train endoscopists in use of the device, and identify appropriate training methods.


#
#

Competing interest

None


Corresponding author

Amit Rastogi, MD
Department of Gastroenterology & Hepatology
University of Kansas Medical Center
3901 Rainbow Blvd
Kansas City, KS 66160
USA   


Zoom Image
Fig. 1 Study flow diagram depicting search strategy, screening and studies of cap-assisted colonoscopy identified for inclusion in the meta-analysis of adenoma detection rate.
Zoom Image
Fig. 2 Forest plot of pooled estimates of adenoma detection rate using cap-assisted colonoscopy compared to standard colonoscopy. a Results with all eligible studies. b Results with only high-quality studies (Jadad score ≥ 3). c Results with only high-quality studies using random effects.
Zoom Image
Fig. 3 Funnel plot showing publication bias.
Zoom Image
Fig. 4 Forest plot of pooled estimate of adenoma per person (APP) showing higher detection of average adenoma per person using cap compared to standard colonoscopy.
Zoom Image
Fig. 5 Figure plot of pooled estimate of adenomas > 10 mm, showing significant improved detection with CAP assisted colonoscopy compared to standard colonoscopy
Zoom Image
Fig. 6 Figure plot of pooled estimate of sessile serrated adenoma (SSA) showing no significant improvement in the detection of proximal adenomas.
Zoom Image
Fig. 7 Forest plot of pooled estimates of cecal intubation rate (a) and cecal intubation time (b) showing improved rates and lesser time with cap compared to standard colonoscopy