Gradual stiffness versus magnetic imaging-guided variable stiffness colonoscopes: A randomized noninferiority trial

Abstract

Background

Colonoscopes with gradual stiffness have recently been developed to enhance cecal intubation.

Objective

We aimed to determine if the performance of gradual stiffness colonoscopes is noninferior to that of magnetic endoscopic imaging (MEI)-guided variable stiffness colonoscopes.

Methods

Consecutive patients were randomized to screening colonoscopy with Fujifilm gradual stiffness or Olympus MEI-guided variable stiffness colonoscopes. The primary endpoint was cecal intubation rate (noninferiority limit 5%). Secondary endpoints included cecal intubation time. We estimated absolute risk differences with 95% confidence intervals (CIs).

Results

We enrolled 475 patients: 222 randomized to the gradual stiffness instrument, and 253 to the MEI-guided variable stiffness instrument. Cecal intubation rate was 91.7% in the gradual stiffness group versus 95.6% in the variable stiffness group. The adjusted absolute risk for cecal intubation failure was 4.3% higher in the gradual stiffness group than in the variable stiffness group (upper CI border 8.1%). Median cecal intubation time was 13 minutes in the gradual stiffness group and 10 minutes in the variable stiffness group (p < 0.001).

Conclusions

The study is inconclusive with regard to noninferiority because the 95% CI for the difference in cecal intubation rate between the groups crosses the noninferiority margin. (ClinicalTrials.gov identifier: NCT01895504).

Keywords

Endoscopy colonoscopy innovation

Introduction

The proportion of procedures completed to the cecum, the cecal intubation rate, is a key colonoscopy performance indicator.^1,2 Successful cecal intubation depends on the endoscopist’s skills but also on the technical features of the colonoscope. Technological innovations that have been shown to increase the cecal intubation rate include variable stiffness colonoscopes and the use of magnetic endoscopic imaging (MEI).^3,4 Variable stiffness allows the endoscopist to adjust the stiffness of the insertion section to prevent bending and thereby improve insertion force to pass colonic bends and flexures. A typical example is to apply the stiffener to avoid looping in the sigmoid colon when the tip is pushed through the splenic flexure. In addition, variable stiffness colonoscopes are compatible with the MEI, which is currently the only external imaging device that does not employ ionizing radiation. MEI-compatible variable stiffness colonoscopes are equipped with magnetic field generator coils within the shaft. The magnetic fields are detected by sensor coils in an external portable unit positioned alongside the patient that in turn generates a real-time computer graphic image of the colonoscope position within the abdomen.⁵ Although the variable stiffness mechanism and the external MEI unit are independent features not necessarily used together, the MEI image provides useful information about colonoscope bending and guides the appropriate use of the variable stiffness. The MEI is, however, an expensive auxiliary tool for colonoscopy, and although MEI-guided colonoscopy may be regarded as the gold standard for colonoscope insertion, the evidence of benefit is limited.⁴

Colonoscopes with gradual stiffness in the insertion section intended to ease cecal intubation have recently been developed. Gradual stiffness means that the stiffness is changing over the length of the scope being softest at the distal end and hardest at the proximal end. The gradual stiffness insertion section is designed to optimize force and torque transmission during colonoscope insertion and thereby enhance insertion capability without the need for an external imaging device. The comparative performance of MEI-guided variable stiffness colonoscopes versus gradual stiffness colonoscopes is currently unknown.

In this investigator-initiated trial we aimed to compare the performance of a new colonoscope (Fujifilm EC-590WI4 with gradual stiffness (ColoAssist), Fujifilm, Tokyo, Japan) with an MEI-guided variable stiffness colonoscope (Olympus CF-H180DI colonoscopes and ScopeGuide, Olympus, Tokyo, Japan).

Materials and methods

Patients consecutively enrolled in a screening colonoscopy trial at two Norwegian centers (Kristiansand and Arendal) were eligible to participate in the present trial. The screening colonoscopy trial (Nordic-European Initiative on Colorectal Cancer; NordICC) has been described elsewhere.⁶

Patients aged 55 to 64 years were randomized to undergo screening colonoscopy with the test instrument (gradual stiffness group) or the MEI-guided variable stiffness colonoscope (variable stiffness group). A study investigator not involved with patient recruitment or colonoscopies at the study centers performed a 1:1 cluster randomization in blocks of four. Sealed, opaque envelopes were prepared containing information on group allocation for each cluster. Up to eight screening patients were scheduled per center per day, four in the morning and four in the afternoon. One cluster was all screening colonoscopies performed during the morning session and another cluster was all procedures in the afternoon session at each center. An endoscopy assistant opened one envelope containing group allocation for all morning patients before the first patient entered the endoscopy suite in the morning; the next envelope was opened at noon before the afternoon patients entered the suite. Cluster randomization was chosen for logistic purposes to reduce the time needed for changing endoscopy racks between each patient. All patients provided informed consent before inclusion in the trial. The Regional Committee for Medical and Health Research Ethics approved the trial. The trial was registered at www.clinicaltrials.gov (NCT01895504).

Patients were blinded to group allocation; endoscopists and assistants were not. Seven experienced endoscopists participated in the trial. All endoscopists had several years of experience with MEI-guided variable stiffness colonoscopes before the start of the trial. The gradual stiffness colonoscope was available for training for all endoscopists for three months before the trial began.

The primary trial endpoint was cecal intubation rate. Secondary endpoints were cecal intubation time, total procedure time, polyp detection rate, adenoma detection rate, proportion of patients requesting sedation or analgesia, and procedural pain scored by the patients on a four-point Likert scale (no, slight, moderate or severe pain) using a validated questionnaire on the day after the procedure.⁷ The application of external abdominal pressure was recorded (yes/no).

Instruments and equipment

The test colonoscope (Fujifilm EC-590WI4) is equipped with a newly developed gradual but non-adjustable stiffness insertion technology (ColoAssist) intended to ease insertion. Olympus CF-H180DI adult colonoscopes with variable stiffness were used in the control group. The MEI (ScopeGuide, Olympus, Tokyo, Japan) was used with the Olympus colonoscopes. No external imaging device was used during procedures with the gradual stiffness colonoscopes. Carbon dioxide was used as insufflation gas in all procedures. Water to aid insertion was used at the discretion of the endoscopist. Technical specifications of the two colonoscope types are summarized in Table 1.

Table 1.

Technical specifications of the gradual stiffness and variable stiffness colonoscopes

	Gradual stiffness	Variable stiffness
Colonoscope model	Fujifilm EC-590WI4	Olympus CF-H180DI
Insertion tube diameter (mm)	12.8	13.2
Working length (cm)	152	133
Working channel diameter (mm)	3.8	3.7
Angle of view (degrees)	140	170
Image quality	HD	HD

HD: high definition.

Power estimates and statistical analyses

A consistent cecal intubation rate of ≥95% with the standard MEI-guided variable stiffness instrument was documented at the two study centers before the start of the current trial. We hypothesized that the cecal intubation rate with the gradual stiffness colonoscope would be noninferior to the MEI-guided variable stiffness colonoscope using a predefined noninferiority margin of five percentage points. Choosing a one-sided 95% confidence interval (CI), 472 participants (236 per group) were needed to have 80% power to rule out the possibility that the cecal intubation rate was five percentage points lower or worse with the gradual stiffness colonoscope than with the MEI-guided variable stiffness colonoscope.

Categorical data were compared using the chi squared or Fisher’s exact test, as appropriate. Continuous data were visually checked for normality using histograms; normally distributed data were compared with the Student’s t-test and non-normally distributed data were compared with the Mann-Whitney U test. Because of a skewed allocation of sex between the two treatment groups, we performed univariate and multivariate logistic regression analyses with adjustment for patient age and sex; use of water-aided colonoscopy for insertion; and individual endoscopist to analyze predictors of cecal intubation failure other than treatment group allocation. The final multivariate logistic regression model was fitted using backward elimination of variables not significantly associated with the outcome of interest at p < 0.25. Variables included in the model and subsequently removed were: patient age, endoscopist, insertion technique (carbon dioxide insufflation with or without water aid), and an interaction term between endoscopists and insertion technique. The derived adjusted risk difference with one-sided 95% CI for cecal intubation failure was calculated with the final multivariate logistic regression model. Other applied tests were two-sided. A p value of < 0.05 was considered statistically significant. Statistical analyses were performed with Stata 13.1 (StataCorp, College Station, TX, USA), and the adjusted risk difference was derived from the logistic regression analyses using the user-written Stata program adjrr.⁸ The trial was planned, conducted and reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) statement for noninferiority trials.⁹

Results

Between April 2013 and May 2014 (with different starting dates at the trial centers), a total of 475 patients were enrolled; 222 were allocated to the gradual stiffness group and 253 were allocated to the variable stiffness group. Cluster randomization resulted in a slight, nonsignificant imbalance in the number of participants in the two groups (p = 0.17). The study flowchart is presented in Figure 1. Baseline characteristics were similar in the two groups (Table 2).

Figure 1.

Study flowchart.

Table 2.

Baseline characteristics and procedural outcomes in the gradual stiffness group and the variable stiffness group

	Gradual stiffness group N = 222	Variable stiffness group N = 253	p
Women, n (%)	96 (43)	130 (51)	0.08
Age (years), mean (SD)	61 (2.9)	61 (2.7)	0.26
Cecal intubation rate (intention to treat), n (%)	199 (90)	238 (94)	0.08
Adjusted cecal intubation rate^a, n (%)	199 (91.7)	238 (95.6)	0.12
Cecal intubation time (min), median (IQR)	13 (10–20)	10 (7–14)	<0.001
Total procedure time (min), median (IQR)	26 (20–34)	21 (16–27)	<0.001
External abdominal pressure, n (%)	137 (63)	122 (49)	0.002
Sedation/analgesia, n (%)	26 (12)	31 (12)	0.85
Water-aided colonoscope insertion n (%)	132 (61)	178 (71)	0.02
Polyp detection rate, n (%)	99 (45)	104 (42)	0.43
Adenoma detection rate, n (%)	74 (33)	75 (30)	0.21
Sessile-serrated adenoma detection rate, n (%)	12 (5)	12 (5)	0.31

SD: standard deviation; IQR: interquartile range; MEI: magnetic endoscopic imaging.

Cecal intubation failures due to poor bowel preparation excluded from analysis (ColoAssist group n = 5; MEI group n = 4).

Cecal intubation failed because of poor bowel preparation in five patients in the gradual stiffness group and four patients in the variable stiffness group, and these were excluded from the analyses of cecal intubation rate (Figure 1). Cecal intubation was achieved in 199 out of 217 patients (91.7%) in the gradual stiffness group and 238 out of 249 patients (95.6%) in the variable stiffness group (p = 0.12). In the multivariate regression analysis, the gradual stiffness group had a 4.3% higher absolute risk for cecal intubation failure than the variable stiffness group (one-sided 95% CI 4.3–8.1%) (Figure 2).

Figure 2.

Observed absolute risk difference for cecal intubation failure with the gradual stiffness colonoscope (test instrument) compared with the MEI-guided variable stiffness colonoscope (standard instrument) after adjustment for patient gender (patients with cecal intubation failure due to poor bowel preparation are excluded from the analysis). Error bar: one-sided 95% confidence interval; Δ: noninferiority margin.

The vast majority of the colonoscopies were performed without sedation, and sedation rates were similar in the two groups (Table 2). Cecal intubation time and total procedure time was significantly longer in the gradual stiffness group than in the variable stiffness group (Table 2). The detection rates of polyps (including all histologic types), adenomas and sessile-serrated adenomas were similar in the two groups (Table 2). A total of 150 out of 222 patients (68%) in the gradual stiffness group and 172 out of 253 patients (68%) in the variable stiffness group responded to the pain questionnaire (p = 0.92). Pain scores were similar in the two groups (Figure 3). No adverse events occurred in either group.

Figure 3.

Patient pain during colonoscopy in the two treatment groups (p for trend = 0.26).

Discussion

In the current noninferiority trial we compared the performance of a new gradual stiffness colonoscope with the variable stiffness instrument currently in use at the study centers. In addition to differences in colonoscope features, the variable stiffness colonoscope was used in combination with the MEI (ScopeGuide) to aid insertion whereas the gradual stiffness instrument had no magnetic or other imaging device.

After adjustment for differences at baseline, the absolute risk for cecal intubation failure was 4.3% (one-sided 95% CI 4.3–8.1%) higher with the gradual stiffness instrument than with the MEI-guided variable stiffness instrument. Thus, we were not able to conclude that the performance of the gradual stiffness instrument was noninferior to that of the MEI-guided variable stiffness colonoscope, although the difference may still be less than the pre-defined five percentage point noninferiority margin. A cecal intubation rate of 95% or higher is recommended for colorectal cancer screening.² The current results indicate that the gradual stiffness instrument performed slightly under this threshold, while the MEI-guided variable stiffness instrument performed above the threshold.

There were important differences between the two instruments with regard to secondary endpoints. The significantly shorter cecal intubation time and total procedure time indicate a performance advantage with the MEI-guided variable stiffness colonoscope. However, all endoscopists had used the MEI during colonoscope insertion for most colonoscopies during the last years, and four out of the seven endoscopists had used the MEI device during their early-career colonoscopy training. In a previous trial we compared MEI-guided colonoscopy with standard colonoscopy with fluoroscopy on demand, and we found that fluoroscopy was deemed necessary in 38.9% of procedures when the MEI tool was not available.¹⁰ In the current trial, however, a cecal intubation rate of 92% was achieved in the gradual stiffness group without the availability of additional external imaging. With regard to patient pain and discomfort, there was no significant difference between the two groups.

Using the MEI is not standard of care in most centers, and it can be argued that a study comparing cecal intubation rate with the new gradual stiffness instrument versus the variable stiffness instrument without the MEI would be more appropriate. Although there is no established gold-standard colonoscopy equipment, there is evidence to support the use of variable stiffness colonoscopes over conventional colonoscopes and MEI guidance over no external imaging for cecal intubation.^3,4 In addition, there are no apparent drawbacks with using the MEI during colonoscopy except for the prime cost. We therefore argue that MEI-guided variable stiffness colonoscopes can appropriately serve as the gold-standard equipment to which new equipment should be compared.

The strengths of this study are its randomized, population-based design and the inclusion of two endoscopy centers.

Limited statistical power is the main limitation that hampers more definite conclusions according to confirmatory studies. Further, the cluster randomization resulted in a somewhat unbalanced number of patients in the two groups that should not, however, influence the internal validity of our results. The endoscopists’ relative inexperience with the gradual stiffness instrument compared with the MEI-guided variable stiffness instrument is another limitation, and this may have introduced bias in favor of the latter instrument. Although sedation rates were similar in the two groups, the fact that most trial procedures were unsedated may affect generalizability. Finally, only patients, but not endoscopists, could be blinded to group allocation.

In summary the study is inconclusive. We were not able to determine if the cecal intubation rate with the novel gradual stiffness colonoscope is noninferior to the MEI-guided variable stiffness colonoscope because the 95% CI for the difference crosses the noninferiority margin. Nevertheless, secondary endpoints suggest a performance advantage with the variable stiffness instrument.

Footnotes

Funding

The work was supported by a research grant from the Regional Health Board of South-East Norway, grant no. 2013099.

Acknowledgment

FUJIFILM Europe provided the test instruments used in this trial.

Conflicts of interest

Kjetil Garborg has received travel support and research funding from Olympus medical systems, Norway. All other authors have nothing to declare.

References

Rembacken

Hassan

Riemann

. Quality in screening colonoscopy: Position statement of the European Society of Gastrointestinal Endoscopy (ESGE). Endoscopy 2012; 44: 957–968.

Rex

Schoenfeld

Cohen

. Quality indicators for colonoscopy. Am J Gastroenterol 2015; 110: 72–90.

Othman

Bradley

Choudhary

. Variable stiffness colonoscope versus regular adult colonoscope: Meta-analysis of randomized controlled trials. Endoscopy 2009; 41: 17–24.

Mark-Christensen

Brandsborg

Iversen

. Magnetic endoscopic imaging as an adjuvant to elective colonoscopy: A systematic review and meta-analysis of randomized controlled trials. Endoscopy 2015; 47: 251–261.

Shah

Saunders

. Aids to insertion: Magnetic imaging, variable stiffness, and overtubes. Gastrointest Endosc Clin N Am 2005; 15: 673–686.

Kaminski

Bretthauer

Zauber

. The NordICC Study: Rationale and design of a randomized trial on colonoscopy screening for colorectal cancer. Endoscopy 2012; 44: 695–702.

Hoff

Bretthauer

Huppertz-Hauss

. The Norwegian Gastronet project: Continuous quality improvement of colonoscopy in 14 Norwegian centres. Scand J Gastroenterol 2006; 41: 481–487.

Norton

Miller

Kleinman

. Computing adjusted risk ratios and risk differences in Stata. Stata J 2013; 13: 492–509.

Piaggio

Elbourne

Altman

. Reporting of noninferiority and equivalence randomized trials: An extension of the CONSORT statement. JAMA 2006; 295: 1152–1160.

10.

Holme

Höie

Matre

. Magnetic endoscopic imaging versus standard colonoscopy in a routine colonoscopy setting: A randomized, controlled trial. Gastrointest Endosc 2011; 73: 1215–1222.