Dig Dis Sci DOI 10.1007/s10620-015-3636-6
ORIGINAL ARTICLE
Deployment of a Short, Single-Opening Endoscopic Clip Versus a Long, Reopening Endoscopic Clip in Clinical Practice Eric Wee1 • Mathew Philip Sachin1 • Uthamanand Chinnappa1 • Su Chang2 Cherng Hann Benjamin Yip1
•
Received: 16 January 2015 / Accepted: 17 March 2015 Ó Springer Science+Business Media New York 2015
Abstract Background Endoscopic clips vary in their designs and costs. Clip wastage is a common problem, and this is dependent on the success of its deployment. Aims The aim of this study is to compare the rates of successful deployment between two different commonly used endoscopic clips. Methods A single-center, retrospective study was conducted. Endoscopy reports of patients with clips deployed over 24 months were reviewed. We compared a long-pronged, reopening endoscopic clip (type A: Resolution clip; Boston Scientific, Natick, MA, USA) versus a short-pronged, single-opening clip (type B: QuickClip2; Olympus Medical Systems Corp, Japan). The main outcome was clip deployment success rate. Secondary outcomes were predictors of successful deployment, cost, and wastage.
Results Of 14,690 endoscopic cases, 472 clips (171 type A and 301 type B) were deployed in 262 procedures. Type A clips had a significantly higher successful deployment rate (147/171, 86.0 %) than type B clips (221/301, 73.4 %) (p = 0.002). On multivariate analysis, variables independently associated with successful deployment included using type A clips (OR 2.07, 95 % CI 1.20–3.55; p = 0.009) and clips placed in the lower gastrointestinal tract (OR 3.48, 95 % CI 1.64–7.40; p = 0.001). The cost of using type A clips was higher than type B clips (p \ 0.001). Type B clips were associated with more wastage (p = 0.049). Conclusions Long-pronged, reopening clips (type A) have a better deployment rate than short-pronged, singleopening clips (type B). Although type A clips had less wastage, the cost per procedure was higher. Keywords Endoscopic clip Resolution clip QuickClip2 Deployment Cost
& Eric Wee [email protected] Mathew Philip Sachin [email protected] Uthamanand Chinnappa [email protected] Su Chang [email protected] Cherng Hann Benjamin Yip [email protected] 1
Division of Gastroenterology, Department of General Medicine, Khoo Teck Puat Hospital, 90 Yishun Central, Singapore 768828, Singapore
2
Clinical Research Unit, Khoo Teck Puat Hospital, 90 Yishun Central, Singapore 768828, Singapore
Introduction Endoscopic clips were initially designed with the aim of achieving hemostasis in patients with bleeding lesions [1– 4]. Other modalities used for hemostasis include epinephrine injection, coaptive coagulation probes, coagulation grasper forceps, argon plasma coagulation, hemostatic powder spray, and over-the-scope clips [5–7]. The role of endoscopic clips has since expanded to include prophylactic hemostasis for large polyp resections, endoscopic closure (e.g., perforations, fistulas, post-mucosal resection defects), anchoring devices at risk of migration (e.g., jejunal tubes, fully covered self-expandable enteral stents), and marking lesions for subsequent localization [8–12].
123
Dig Dis Sci
Although clips are frequently used during therapeutic endoscopy procedures, they are known to fail or malfunction during deployment, resulting in wastage and additional costs. Factors which influence the deployment of clips are unknown. There is also a lack of clinical data to suggest whether the different designs truly influence their ability to deploy successfully. An endoscopy clipping device consists of a clip, delivery catheter, and handle with a deployment assembly. Endoscopy clips are manufactured in different configurations. Some devices are supplied preloaded and may be used only once, while others have a reusable catheter which requires the clip to be loaded before it can be fired. Preloaded devices can be used rapidly, because they do not require assembly. Reusable devices require time to assemble, but have clips which cost less, although the reusable catheter may cost more [8]. Clips may be double- or triple-pronged, single-opening or reopening, and rotatable or non-rotatable. In addition, clips come in different lengths and open to varying diameters. Different combinations of these features define each clip as a unique device. Clips with prongs that open widely have better tissue retention rates and stay onto the mucosa for a longer period of time [13]. A clip with the reopening function may allow closure in the optimal position before it is finally released. This is believed to reduce the incidence of suboptimal deployment although clinical data are lacking. The aim of this study is to compare the successful deployment of 2 different endoscopic clips available in clinical practice. We compared a long-pronged, reopening endoscopic clip (Resolution clip; Boston Scientific, Natick, MA, USA) versus a shorter-pronged, single-opening clip (QuickClip2; Olympus Medical Systems Corp, Japan). The secondary aims of this study are to identify predictors of successful clip deployment, the costs of using each type of clip, and the rates of clip wastage.
standard 2.8-mm instrument channel. The clip has a reopening function which allows the two prongs to be opened and closed up to five times before its final deployment. This allows repositioning of the clip if tissue acquisition is not ideal. The clip has elongated prongs and opens to a maximum diameter of 11 mm, theoretically allowing more tissue to be captured [8] (Fig. 1). The QuickClip2 is a single-use, preloaded, dual-pronged, rotatable clip. In contrast to the Resolution clip, the prongs of the QuickClip2 are unidirectional and cannot be reopened once closed. The clip has a maximum diameter of 9.5 mm between the prongs and is shorter facilitating deployment in tight spaces [8] (Fig. 2). This study was conducted in a real-world clinical setting. Therefore, clips were used by endoscopists with expertise ranging from advanced endoscopy trainees to accredited endoscopists. All trainees were supervised during endoscopy. Trained endoscopy nurses also participated in the deployment by handling the clip rotation, opening, closing, and firing upon the orders of the doctor. The type of clip and number of clips used were decided by the endoscopist. Endoscopy clips were considered used once they were extended out of the delivery sheath and opened. We determined the deployment success rates based on images in the database as this was objective and accurate. Highquality images of the clipped lesion at the end of the procedure were routinely taken and available. Details of the number of successfully deployed clips were not always available from the endoscopy reports as it was not a standard practice to capture this information. All images taken during and at the end of the procedure were retrieved and reviewed. The gastroenterologist reviewing the images was
Methods Data for the study were retrieved from the endoscopy reporting module of Khoo Teck Puat Hospital, Singapore. The recruitment period was from January 2012 to December 2013. We included all endoscopy cases in which clips were used. Procedures which were performed out of the endoscopy suites (e.g., intensive care units and operating theaters) were excluded, as data on clip utilization were unavailable. This study was approved by our domainspecific research board (DSRB) and local ethics committee. Permission for a consent waiver was granted. The Resolution clip and QuickClip2 were used in these procedures. The Resolution clip is a rotatable, stainless steel, preloaded, single-use clip that can be inserted via a
123
Fig. 1 Resolution clip placed prophylactically on the broad stalk of a resected colonic polyp. Inset Endoscopic snare resection of the large pedunculated polyp
Dig Dis Sci
Fig. 2 Endoscopic hemostasis using the QuickClip2 for a postpolypectomy bleed in the colon
Fig. 3 Endoscopic clip placed on a Forrest Ib pyloric channel ulcer. One clip has been deployed successfully, while the other has been dislodged
blinded to the report. When there was ambiguity, another gastroenterologist was consulted to review the images independently and the findings were corroborated. The number of clips present on the targeted lesion at the end of the procedure represented the number of clips successfully deployed.
Data Collection We collected data on baseline characteristics such as the age, gender, location of the clip in the gastrointestinal (GI) tract, indication of endoscopy, diagnosis, presence of active bleeding (oozing or arterial spurting lesions), type of clip used, and indication of clip application. We also collected data on the number of clips used and wasted (unsuccessfully deployed). Successful deployment was defined as a clip which was placed in the area of interest and retained its position at the end of the procedure. Therapeutic success, such as hemostasis, was not measured since this was not the aim of the study and may be influenced by multiple factors independent of the clip. These factors include accuracy when targeting the vessel, visibility (e.g., bloody field), and cleanliness of the lumen. In addition, multiple modalities may be used for hemostasis (e.g., epinephrine injection/coagulation/proton pump inhibitors) [14]. Deployment of a clip was regarded as unsuccessful if it malfunctioned (e.g., failed closure, jammed, premature closure, dislodged from the delivery mechanism) or fell off the lesion during endoscopy (e.g., by peristalsis, flushing, scope contact) (Figs. 3, 4). The cost of clip application included the costs of both successfully and unsuccessfully deployed clips. The
Fig. 4 Malfunctioned clip in the stomach with the prongs failing to close. Deployment was unsuccessful in a retroflexed position for a gastric ulcer at the cardia
current cost of each QuickClip2 in Singapore is USD 47, whereas the cost of each Resolution clip is USD 94. Statistical Analysis Sample size estimation was based on following criteria: QuickClip2 has a 70 % successful deployment rate based on clinical observations. A 15 % difference would be required to detect an alpha error of 0.05 with a power of 80 %. Recruitment was set as two QuickClip2 clips for each Resolution clip based on the utilization. We calculated that at least 186 QuickClip2 and 93 Resolution clips were required to reject the null hypothesis.
123
Dig Dis Sci
Statistical analysis was carried out using SPSS Statistics version 21.0 (SPSS Inc, Chicago, IL, USA) and STATA 13 (StataCorp LP, TX, USA). The Shapiro–Wilk test was used to determine the normality of continuous variables. Variables were summarized as mean (standard deviation) or median (quartiles) for normally distributed and skewed data, respectively. Categorical variables were presented as frequency (percentage). Group comparisons were evaluated using the Student’s t test or Mann–Whitney U test as appropriate for continuous variables. The Pearson’s Chisquare test or Fisher’s exact test was used to compare categorical variables. When analyzing the outcomes, adjustment of covariates and logistic regression analysis were performed to overcome any bias from confounding variables in the baseline demographics such as the indication. Univariate statistical analysis was performed followed by multivariate analysis to identify predictors of successful clip deployment. Variables which were significant or close to significance in univariate analysis were included in the multivariate analysis. The degree of association was reflected as an odds ratio (OR) and its 95 % confidence interval (CI). A p value of less than 0.05 was considered significant.
Results A total of 14,690 endoscopic procedures were performed from January 2012 to December 2013. The Resolution clip was used in 99 procedures and the QuickClip2 in 163
procedures. Images from five cases were reviewed by a second gastroenterologist. A total of 171 Resolution clips and 301 QuickClip2 were deployed. Baseline demographics for the Resolution clip and QuickClip2 groups are given in Tables 1 and 2. Per procedure analysis showed that the median age of patients receiving a Resolution clip was 69 (59, 73) years and QuickClip2 66 (56, 74) years (p = 0.764). Sixty-four (64.7 %) males received the Resolution clip and 99 (60.7 %) males received the QuickClip2 (p = 0.527). The proportion of clips deployed in the upper GI and lower GI tracts was comparable between both clips (p = 0.409). There were no differences in the indications of endoscopy between both clips (p = 0.261) (Table 1). The indications of clip deployment were significantly different between the Resolution clip and QuickClip2 (p = 0.001). Resolution clips were used more frequently in ulcer therapy (29.8 %), post-polypectomy/biopsy hemorrhage (33.3 %), and large polyp resection (12.9 %), while the QuickClip2 was used frequently in ulcer therapy (46.2 %), post-polypectomy/biopsy hemorrhage (35.2 %), and therapy of vascular lesions (9 %) (Table 2). Analysis of the primary outcome showed that the Resolution clip had a significantly higher successful deployment rate (147/ 171, 86.0 %) than the QuickClip2 (221/301, 73.4 %) (p = 0.002). This significance remained after adjustment for covariates such as location of the clip and indication (p = 0.009) Univariate logistic regression was performed to identify predictors of a successful deployment. The type of clip used, location of clip, post-polypectomy/biopsy
Table 1 Baseline characteristics per endoscopic procedure Variable
Procedures using Resolution clip (n = 99)
Procedures using QuickClip2 (n = 163)
p value
Age (years)
69 (59, 73)
66 (56, 74)
0.764
64 (64.6) 35 (35.4)
99 (60.7) 64 (39.3)
0.527
Upper GI tract
55 (55.6)
99 (60.7)
0.409
Lower GI tract
44 (44.4)
64 (39.3)
Anemia
26 (26.3)
24 (14.7)
Overt GI hemorrhage
43 (43.4)
85 (52.1)
Lower GI symptoms
6 (6.1)
16 (9.8)
Upper GI symptoms
6 (6.1)
8 (4.9)
16 (16.2)
26 (16.0)
2 (2.0)
4 (2.5)
Gender [n (%)] Male Female Location of clip application [n (%)]
Indication of endoscopy [n (%)]
Polyp surveillance/screening Others
Median (quartiles)
* p \ 0.05
123
0.261
Dig Dis Sci Table 2 Indication and success of clip deployment
Variable
Resolution clip (n = 171)
QuickClip2 (n = 301)
p value
0.001*
Indication [n (%)] Ulcer therapy
51 (29.8)
139 (46.2)
Post-polypectomy/biopsy hemorrhage
57 (33.3)
106 (35.2)
Vascular lesion
4 (2.3)
27 (9.0)
Defect closure
20 (11.7)
11 (3.7)
Prophylactic for large polyp resection
22 (12.9)
7 (2.3)
Others
17 (10.0)
11 (3.7)
147 (86.0)
221 (73.4)
Successfully deployed [n (%)]
0.002*
* p \ 0.05
Table 3 Logistic regression analysis of potential variables associated with successful clip deployment Variable
Univariate
Multivariate
Odds ratio (95 % CI)
p value
Odds ratio (95 % CI)
p value
2.07 (1.20–3.55)
0.009*
3.48 (1.64–7.40)
0.001*
Type of clip (reference: QuickClip2)
2.22 (1.34–3.66)
0.002*
Age
1.00 (0.98–1.02)
0.80
Gender (reference: male) Location of clip (reference: upper GI tract)
1.28 (0.81–2.03) 4.46 (2.49–8.01)
0.30 \0.001*
Post-polypectomy/biopsy hemorrhage
3.14 (1.79–5.50)
\0.001*
1.45 (0.73–2.89)
0.29
Vascular lesion
0.92 (0.41–2.08)
0.85
0.80 (0.35–1.86)
0.61
Indication (reference: ulcer)
Prophylactic clipping for polyp resection
0.94 (0.36–2.43)
0.90
0.68 (0.25–1.82)
0.44
Defect closure
3.81 (1.11–13.08)
0.03*
1.08 (0.27–4.32)
0.92
Others
2.27 (0.90–5.74)
0.08
0.97 (0.35–2.69)
0.95
* p \ 0.05
hemorrhage, and defect closure were found to be significant (p \ 0.05). Subsequently, after covariate adjustment in the multivariate logistic regression, only the type of clip used (reference: QuickClip2; OR 2.07, 95 % CI 1.20–3.55; p = 0.009) and location of clip application (reference: upper GI tract; OR 3.48, 95 % CI 1.64–7.40; p = 0.001) were independently associated with successful deployment. The odds ratio of successful deployment using the Resolution clip was 1.07 times higher than the QuickClip2 and 2.48 times higher when clips were deployed in the lower GI tract as compared to the upper GI tract (Table 3). The number of clips used per procedure was similar to both types of clips (p = 0.534). The cost of using the Resolution clip per procedure was significantly higher than QuickClip2 [Resolution clip USD 94 (94, 188) vs. USD 47 (47, 94); p \ 0.001]. When QuickClip2 was used, there was a higher incidence of wasted clips as compared to the Resolution clip (QuickClip2 32.5 % vs. Resolution clip 21.2 %; p = 0.049) (Table 4).
Based on the deployment success rates, we compared the costs of treating 100 subjects with endoscopic clips using four different strategies. These were using the Resolution clip only (strategy 1), QuickClip2 only (strategy 2), QuickClip2 once, failing which Resolution clips were used (strategy 3), and QuickClip2 twice, failing which Resolution clips were used (strategy 4). Strategy 1 would result in a maximum of 3 Resolution clips deployed before success was achieved, and 116 Resolution clips would be used. This would cost USD 10,904 for 100 subjects (average of USD109.04 per person). Strategy 2 would result in a maximum of 5 QuickClip2 deployed, a total of 137 QuickClip2 costing USD 6439 for 100 subjects (USD 64.39 per person). Strategy 3 would result in a maximum of 1 QuickClip2 and 3 Resolution clips. A total of 100 QuickClip2 and 32 Resolution clips would cost USD 7708 for 100 subjects (USD 77.08 per person). Finally, strategy 4 would result in a maximum of 2 QuickClip2 and 2 Resolution clips. 113 QuickClip2 and 8 Resolution clips costing
123
Dig Dis Sci Table 4 Utilization and costs of endoscopic clipping Variable
Procedures using resolution clip (n = 99)
Procedures using QuickClip2 (n = 163)
1
52 (52.5)
91 (55.8)
2
31 (31.3)
41 (25.2)
p value
Number of clips per procedure [n (%)]
16 (16.2)
31 (19.0)
Total cost of clips per procedure (USD)
C3
94 (94, 188)
47 (47.94)
Procedures with clips wasted during deployment [n (%)]
21 (21.2)
53 (32.5)
0.534
\0.001* 0.049*
Median (quartiles)
* p \ 0.05
USD 6298 would be used for 100 subjects (USD 62.98 per person).
Discussion Endoscopic clips are available in a variety of designs; however, studies comparing these clips in humans are limited. The few studies comparing the Resolution clip and the QuickClip2 were performed on animal models and on the benchtop [15, 16]. Jensen et al. conducted two canine studies using the Resolution clip and QuickClip2. The Resolution clip was found to have a higher retention rate at 3 and 4 weeks as compared to the QuickClip2 [15, 16]. When clips are deployed, not all of them will successfully anchor onto the targeted lesion. Sometimes, clips are wasted. This study has illustrated that different clip designs can give rise to different deployment rates. The longpronged, reopening clip had an 86.0 % successful deployment rate , which was significantly higher than 73.4% of the short-pronged, single-opening clip. Clips that were longer and reopened had significantly less wastage as well. On multivariate analysis, both the long-pronged, reopening clip and deployment in the lower GI tract were independently associated with successfully deployment. We have observed several mechanisms in which deployment may fail. The clip may be dislodged before it is fired, the prongs may shut prematurely, the clip may malfunction, and the clip may dislodge soon after it is deployed onto the tissue. The QuickClip2 being a singleopening clip was particularly prone to wastage from premature prong closure during clip opening. Both the QuickClip2 and Resolution clips had incidences of clip malfunctioning during deployment when the distal delivery catheter was in a tightly angulated or retroflexed position. The superiority of the Resolution clip may be attributed to several features. Its reopening function allows misalignment to be corrected before release. This facilitates optimal positioning and reduces wastage.
123
Another advantage of this clip is that its prongs are longer and opening angles wider allowing more tissue to be grasped. A benchtop study by Daram et al. [17] showed that the Resolution clip was superior to the QuickClip2 in closure strength and pulling strength. These qualities enhance its ability to anchor securely and facilitate a successful deployment. In addition, an observational study by Swellengrebel et al. [13] showed that the Resolution clip had a longer tissue retention duration as compared to the QuickClip2. This may be important if clips are needed to close defects or to localize lesions. Clips placed in the lower GI tract were shown to be associated with successful deployment. The thin colonic mucosa allows tissue to be tented and grasped more readily. In contrast, gastric mucosa is thicker and harder to grasp. Proper anchorage is also difficult when clips are deployed on fibrotic ulcers in the upper GI tract. Although the Resolution clip had a higher deployment rate, the median cost per procedure was higher. This is due to the higher cost of a Resolution clip (USD 94) as compared to QuickClip2 (USD 47). When comparing different strategies of clip utilization, using Resolution clips only led to the least number of clips, but cost the most (USD 109.04 per person). Starting with the QuickClip2 and switching to the Resolution clip after 2 unsuccessful QuickClip2 applications was the most cost-effective strategy (USD 62.98 per person). Other factors which may influence the type of clip used include familiarity with the device and whether the indication can only be performed successfully with a wideropening clip for technical reasons (e.g., large defect closure). Therefore, if the lesion is technically amenable to therapy using the QuickClip2, one should consider using it for a maximum of 2 attempts, failing which the Resolution clip should be considered. This study design was retrospective since it allowed us to collect a larger number of cases for a well-powered analysis. A retrospective study is ideal when it is not possible to know before a procedure whether the
Dig Dis Sci
intervention (clips) will be performed. In our study, only 262 procedures utilized clips out of 14,690 procedures (1.8 %). Of these, more than half were diagnostic procedures in which the use of a clip was not anticipated. In conclusion, this study showed that a long-pronged, reopening endoscopic clip had a significantly higher deployment success rate than the shorter-pronged, singleopening clip. The former clip was associated with a significantly lower incidence of wastage. Independent variables associated with a successful deployment were a longpronged, reopening clip and deployment in the lower GI tract. The cost of the Resolution clip was significantly greater than QuickClip2. Conflict of interest
None.
References 1. Hayashi T, Yonezawa M, Kuwabara T, Kudoh I. The study on staunch clip for the treatment by endoscopy. Gastroenterol Endosc. 1975;17:92–101. 2. Hachisu T. Evaluation of endoscopic hemostasis using an improved clipping apparatus. Surg Endosc. 1988;2:13–17. 3. Hachisu T, Miyazaki S, Hamaguchi K. Endoscopic clip-marking of lesions using the newly developed HX-3L clip. Surg Endosc. 1989;3:142–147. 4. Hachisu T, Yamada H, Satoh S, Kouzu T. Endoscopic clipping with a new rotatable clip-device and a long clip. Dig Endosc. 1996;8:127–133. 5. Leung Ki EL, Lau JY. New endoscopic hemostasis methods. Clin Endosc. 2012;45:224–229. 6. Kirschniak A, Kratt T, Stu¨ker D, Braun A, Schurr MO, Ko¨nigsrainer A. A new endoscopic over-the-scope clip system for treatment of lesions and bleeding in the GI tract: first clinical experiences. Gastrointest Endosc. 2007;66:162–167.
7. Chan SM, Chiu PW, Teoh AY, Lau JY. Use of the over-the-scope clip for treatment of refractory upper gastrointestinal bleeding: a case series. Endoscopy. 2014;46:428–431. 8. Technology Assessment Committee, Chuttani R, Barkun A, Carpenter S, Chotiprasidhi P, et al. Endoscopic clip application devices. Gastrointest Endosc. 2006;63:746–750. 9. Ginsberg GG, Lipman TO, Fleischer DE. Endoscopic clip-assisted placement of enteral feeding tubes. Gastrointest Endosc. 1994;40:220–222. 10. Sriram PV, Das G, Rao GV, Reddy DN. Another novel use of endoscopic clipping: to anchor an esophageal endoprosthesis. Endoscopy. 2001;33:724–726. 11. Baron TH, Gostout CJ, Herman L. Hemoclip repair of a sphincterotomy-induced duodenal perforation. Gastrointest Endosc. 2000;52:566–568. 12. Luigiano C, Ferrara F, Ghersi S, Fabbri C, et al. Endoclip-assisted resection of large pedunculated colorectal polyps: technical aspects and outcome. Dig Dis Sci. 2010;55:1726–1731. 13. Swellengrebel HA, Marijnen CA, Vincent A, Cats A. Evaluating long-term attachment of two different endoclips in the human gastrointestinal tract. World J Gastrointest Endosc. 2010;2: 344–348. 14. Hwang JH, Fisher DA, Ben-Menachem T, Chandrasekhara V, et al. Standards of Practice Committee of the American Society for Gastrointestinal Endoscopy. The role of endoscopy in the management of acute non-variceal upper GI bleeding. Gastrointest Endosc. 2012;75:1132–1138. 15. Jensen DM, Machicado GA, Hirabayashi K. Randomized controlled study of 3 different types of hemoclips for hemostasis of bleeding canine acute gastric ulcers. Gastrointest Endosc. 2006;64:768–773. 16. Jensen DM, Machicado GA. Hemoclipping of chronic canine ulcers: a randomized, prospective study of initial deployment success, clip retention rates, and ulcer healing. Gastrointest Endosc. 2009;70:969–975. 17. Daram SR, Tang SJ, Wu R, To SD. Benchtop testing and comparisons among three types of through-the-scope endoscopic clipping devices. Surg Endosc. 2013;27:1521–1529.
123
ORIGINAL ARTICLE
Deployment of a Short, Single-Opening Endoscopic Clip Versus a Long, Reopening Endoscopic Clip in Clinical Practice Eric Wee1 • Mathew Philip Sachin1 • Uthamanand Chinnappa1 • Su Chang2 Cherng Hann Benjamin Yip1
•
Received: 16 January 2015 / Accepted: 17 March 2015 Ó Springer Science+Business Media New York 2015
Abstract Background Endoscopic clips vary in their designs and costs. Clip wastage is a common problem, and this is dependent on the success of its deployment. Aims The aim of this study is to compare the rates of successful deployment between two different commonly used endoscopic clips. Methods A single-center, retrospective study was conducted. Endoscopy reports of patients with clips deployed over 24 months were reviewed. We compared a long-pronged, reopening endoscopic clip (type A: Resolution clip; Boston Scientific, Natick, MA, USA) versus a short-pronged, single-opening clip (type B: QuickClip2; Olympus Medical Systems Corp, Japan). The main outcome was clip deployment success rate. Secondary outcomes were predictors of successful deployment, cost, and wastage.
Results Of 14,690 endoscopic cases, 472 clips (171 type A and 301 type B) were deployed in 262 procedures. Type A clips had a significantly higher successful deployment rate (147/171, 86.0 %) than type B clips (221/301, 73.4 %) (p = 0.002). On multivariate analysis, variables independently associated with successful deployment included using type A clips (OR 2.07, 95 % CI 1.20–3.55; p = 0.009) and clips placed in the lower gastrointestinal tract (OR 3.48, 95 % CI 1.64–7.40; p = 0.001). The cost of using type A clips was higher than type B clips (p \ 0.001). Type B clips were associated with more wastage (p = 0.049). Conclusions Long-pronged, reopening clips (type A) have a better deployment rate than short-pronged, singleopening clips (type B). Although type A clips had less wastage, the cost per procedure was higher. Keywords Endoscopic clip Resolution clip QuickClip2 Deployment Cost
& Eric Wee [email protected] Mathew Philip Sachin [email protected] Uthamanand Chinnappa [email protected] Su Chang [email protected] Cherng Hann Benjamin Yip [email protected] 1
Division of Gastroenterology, Department of General Medicine, Khoo Teck Puat Hospital, 90 Yishun Central, Singapore 768828, Singapore
2
Clinical Research Unit, Khoo Teck Puat Hospital, 90 Yishun Central, Singapore 768828, Singapore
Introduction Endoscopic clips were initially designed with the aim of achieving hemostasis in patients with bleeding lesions [1– 4]. Other modalities used for hemostasis include epinephrine injection, coaptive coagulation probes, coagulation grasper forceps, argon plasma coagulation, hemostatic powder spray, and over-the-scope clips [5–7]. The role of endoscopic clips has since expanded to include prophylactic hemostasis for large polyp resections, endoscopic closure (e.g., perforations, fistulas, post-mucosal resection defects), anchoring devices at risk of migration (e.g., jejunal tubes, fully covered self-expandable enteral stents), and marking lesions for subsequent localization [8–12].
123
Dig Dis Sci
Although clips are frequently used during therapeutic endoscopy procedures, they are known to fail or malfunction during deployment, resulting in wastage and additional costs. Factors which influence the deployment of clips are unknown. There is also a lack of clinical data to suggest whether the different designs truly influence their ability to deploy successfully. An endoscopy clipping device consists of a clip, delivery catheter, and handle with a deployment assembly. Endoscopy clips are manufactured in different configurations. Some devices are supplied preloaded and may be used only once, while others have a reusable catheter which requires the clip to be loaded before it can be fired. Preloaded devices can be used rapidly, because they do not require assembly. Reusable devices require time to assemble, but have clips which cost less, although the reusable catheter may cost more [8]. Clips may be double- or triple-pronged, single-opening or reopening, and rotatable or non-rotatable. In addition, clips come in different lengths and open to varying diameters. Different combinations of these features define each clip as a unique device. Clips with prongs that open widely have better tissue retention rates and stay onto the mucosa for a longer period of time [13]. A clip with the reopening function may allow closure in the optimal position before it is finally released. This is believed to reduce the incidence of suboptimal deployment although clinical data are lacking. The aim of this study is to compare the successful deployment of 2 different endoscopic clips available in clinical practice. We compared a long-pronged, reopening endoscopic clip (Resolution clip; Boston Scientific, Natick, MA, USA) versus a shorter-pronged, single-opening clip (QuickClip2; Olympus Medical Systems Corp, Japan). The secondary aims of this study are to identify predictors of successful clip deployment, the costs of using each type of clip, and the rates of clip wastage.
standard 2.8-mm instrument channel. The clip has a reopening function which allows the two prongs to be opened and closed up to five times before its final deployment. This allows repositioning of the clip if tissue acquisition is not ideal. The clip has elongated prongs and opens to a maximum diameter of 11 mm, theoretically allowing more tissue to be captured [8] (Fig. 1). The QuickClip2 is a single-use, preloaded, dual-pronged, rotatable clip. In contrast to the Resolution clip, the prongs of the QuickClip2 are unidirectional and cannot be reopened once closed. The clip has a maximum diameter of 9.5 mm between the prongs and is shorter facilitating deployment in tight spaces [8] (Fig. 2). This study was conducted in a real-world clinical setting. Therefore, clips were used by endoscopists with expertise ranging from advanced endoscopy trainees to accredited endoscopists. All trainees were supervised during endoscopy. Trained endoscopy nurses also participated in the deployment by handling the clip rotation, opening, closing, and firing upon the orders of the doctor. The type of clip and number of clips used were decided by the endoscopist. Endoscopy clips were considered used once they were extended out of the delivery sheath and opened. We determined the deployment success rates based on images in the database as this was objective and accurate. Highquality images of the clipped lesion at the end of the procedure were routinely taken and available. Details of the number of successfully deployed clips were not always available from the endoscopy reports as it was not a standard practice to capture this information. All images taken during and at the end of the procedure were retrieved and reviewed. The gastroenterologist reviewing the images was
Methods Data for the study were retrieved from the endoscopy reporting module of Khoo Teck Puat Hospital, Singapore. The recruitment period was from January 2012 to December 2013. We included all endoscopy cases in which clips were used. Procedures which were performed out of the endoscopy suites (e.g., intensive care units and operating theaters) were excluded, as data on clip utilization were unavailable. This study was approved by our domainspecific research board (DSRB) and local ethics committee. Permission for a consent waiver was granted. The Resolution clip and QuickClip2 were used in these procedures. The Resolution clip is a rotatable, stainless steel, preloaded, single-use clip that can be inserted via a
123
Fig. 1 Resolution clip placed prophylactically on the broad stalk of a resected colonic polyp. Inset Endoscopic snare resection of the large pedunculated polyp
Dig Dis Sci
Fig. 2 Endoscopic hemostasis using the QuickClip2 for a postpolypectomy bleed in the colon
Fig. 3 Endoscopic clip placed on a Forrest Ib pyloric channel ulcer. One clip has been deployed successfully, while the other has been dislodged
blinded to the report. When there was ambiguity, another gastroenterologist was consulted to review the images independently and the findings were corroborated. The number of clips present on the targeted lesion at the end of the procedure represented the number of clips successfully deployed.
Data Collection We collected data on baseline characteristics such as the age, gender, location of the clip in the gastrointestinal (GI) tract, indication of endoscopy, diagnosis, presence of active bleeding (oozing or arterial spurting lesions), type of clip used, and indication of clip application. We also collected data on the number of clips used and wasted (unsuccessfully deployed). Successful deployment was defined as a clip which was placed in the area of interest and retained its position at the end of the procedure. Therapeutic success, such as hemostasis, was not measured since this was not the aim of the study and may be influenced by multiple factors independent of the clip. These factors include accuracy when targeting the vessel, visibility (e.g., bloody field), and cleanliness of the lumen. In addition, multiple modalities may be used for hemostasis (e.g., epinephrine injection/coagulation/proton pump inhibitors) [14]. Deployment of a clip was regarded as unsuccessful if it malfunctioned (e.g., failed closure, jammed, premature closure, dislodged from the delivery mechanism) or fell off the lesion during endoscopy (e.g., by peristalsis, flushing, scope contact) (Figs. 3, 4). The cost of clip application included the costs of both successfully and unsuccessfully deployed clips. The
Fig. 4 Malfunctioned clip in the stomach with the prongs failing to close. Deployment was unsuccessful in a retroflexed position for a gastric ulcer at the cardia
current cost of each QuickClip2 in Singapore is USD 47, whereas the cost of each Resolution clip is USD 94. Statistical Analysis Sample size estimation was based on following criteria: QuickClip2 has a 70 % successful deployment rate based on clinical observations. A 15 % difference would be required to detect an alpha error of 0.05 with a power of 80 %. Recruitment was set as two QuickClip2 clips for each Resolution clip based on the utilization. We calculated that at least 186 QuickClip2 and 93 Resolution clips were required to reject the null hypothesis.
123
Dig Dis Sci
Statistical analysis was carried out using SPSS Statistics version 21.0 (SPSS Inc, Chicago, IL, USA) and STATA 13 (StataCorp LP, TX, USA). The Shapiro–Wilk test was used to determine the normality of continuous variables. Variables were summarized as mean (standard deviation) or median (quartiles) for normally distributed and skewed data, respectively. Categorical variables were presented as frequency (percentage). Group comparisons were evaluated using the Student’s t test or Mann–Whitney U test as appropriate for continuous variables. The Pearson’s Chisquare test or Fisher’s exact test was used to compare categorical variables. When analyzing the outcomes, adjustment of covariates and logistic regression analysis were performed to overcome any bias from confounding variables in the baseline demographics such as the indication. Univariate statistical analysis was performed followed by multivariate analysis to identify predictors of successful clip deployment. Variables which were significant or close to significance in univariate analysis were included in the multivariate analysis. The degree of association was reflected as an odds ratio (OR) and its 95 % confidence interval (CI). A p value of less than 0.05 was considered significant.
Results A total of 14,690 endoscopic procedures were performed from January 2012 to December 2013. The Resolution clip was used in 99 procedures and the QuickClip2 in 163
procedures. Images from five cases were reviewed by a second gastroenterologist. A total of 171 Resolution clips and 301 QuickClip2 were deployed. Baseline demographics for the Resolution clip and QuickClip2 groups are given in Tables 1 and 2. Per procedure analysis showed that the median age of patients receiving a Resolution clip was 69 (59, 73) years and QuickClip2 66 (56, 74) years (p = 0.764). Sixty-four (64.7 %) males received the Resolution clip and 99 (60.7 %) males received the QuickClip2 (p = 0.527). The proportion of clips deployed in the upper GI and lower GI tracts was comparable between both clips (p = 0.409). There were no differences in the indications of endoscopy between both clips (p = 0.261) (Table 1). The indications of clip deployment were significantly different between the Resolution clip and QuickClip2 (p = 0.001). Resolution clips were used more frequently in ulcer therapy (29.8 %), post-polypectomy/biopsy hemorrhage (33.3 %), and large polyp resection (12.9 %), while the QuickClip2 was used frequently in ulcer therapy (46.2 %), post-polypectomy/biopsy hemorrhage (35.2 %), and therapy of vascular lesions (9 %) (Table 2). Analysis of the primary outcome showed that the Resolution clip had a significantly higher successful deployment rate (147/ 171, 86.0 %) than the QuickClip2 (221/301, 73.4 %) (p = 0.002). This significance remained after adjustment for covariates such as location of the clip and indication (p = 0.009) Univariate logistic regression was performed to identify predictors of a successful deployment. The type of clip used, location of clip, post-polypectomy/biopsy
Table 1 Baseline characteristics per endoscopic procedure Variable
Procedures using Resolution clip (n = 99)
Procedures using QuickClip2 (n = 163)
p value
Age (years)
69 (59, 73)
66 (56, 74)
0.764
64 (64.6) 35 (35.4)
99 (60.7) 64 (39.3)
0.527
Upper GI tract
55 (55.6)
99 (60.7)
0.409
Lower GI tract
44 (44.4)
64 (39.3)
Anemia
26 (26.3)
24 (14.7)
Overt GI hemorrhage
43 (43.4)
85 (52.1)
Lower GI symptoms
6 (6.1)
16 (9.8)
Upper GI symptoms
6 (6.1)
8 (4.9)
16 (16.2)
26 (16.0)
2 (2.0)
4 (2.5)
Gender [n (%)] Male Female Location of clip application [n (%)]
Indication of endoscopy [n (%)]
Polyp surveillance/screening Others
Median (quartiles)
* p \ 0.05
123
0.261
Dig Dis Sci Table 2 Indication and success of clip deployment
Variable
Resolution clip (n = 171)
QuickClip2 (n = 301)
p value
0.001*
Indication [n (%)] Ulcer therapy
51 (29.8)
139 (46.2)
Post-polypectomy/biopsy hemorrhage
57 (33.3)
106 (35.2)
Vascular lesion
4 (2.3)
27 (9.0)
Defect closure
20 (11.7)
11 (3.7)
Prophylactic for large polyp resection
22 (12.9)
7 (2.3)
Others
17 (10.0)
11 (3.7)
147 (86.0)
221 (73.4)
Successfully deployed [n (%)]
0.002*
* p \ 0.05
Table 3 Logistic regression analysis of potential variables associated with successful clip deployment Variable
Univariate
Multivariate
Odds ratio (95 % CI)
p value
Odds ratio (95 % CI)
p value
2.07 (1.20–3.55)
0.009*
3.48 (1.64–7.40)
0.001*
Type of clip (reference: QuickClip2)
2.22 (1.34–3.66)
0.002*
Age
1.00 (0.98–1.02)
0.80
Gender (reference: male) Location of clip (reference: upper GI tract)
1.28 (0.81–2.03) 4.46 (2.49–8.01)
0.30 \0.001*
Post-polypectomy/biopsy hemorrhage
3.14 (1.79–5.50)
\0.001*
1.45 (0.73–2.89)
0.29
Vascular lesion
0.92 (0.41–2.08)
0.85
0.80 (0.35–1.86)
0.61
Indication (reference: ulcer)
Prophylactic clipping for polyp resection
0.94 (0.36–2.43)
0.90
0.68 (0.25–1.82)
0.44
Defect closure
3.81 (1.11–13.08)
0.03*
1.08 (0.27–4.32)
0.92
Others
2.27 (0.90–5.74)
0.08
0.97 (0.35–2.69)
0.95
* p \ 0.05
hemorrhage, and defect closure were found to be significant (p \ 0.05). Subsequently, after covariate adjustment in the multivariate logistic regression, only the type of clip used (reference: QuickClip2; OR 2.07, 95 % CI 1.20–3.55; p = 0.009) and location of clip application (reference: upper GI tract; OR 3.48, 95 % CI 1.64–7.40; p = 0.001) were independently associated with successful deployment. The odds ratio of successful deployment using the Resolution clip was 1.07 times higher than the QuickClip2 and 2.48 times higher when clips were deployed in the lower GI tract as compared to the upper GI tract (Table 3). The number of clips used per procedure was similar to both types of clips (p = 0.534). The cost of using the Resolution clip per procedure was significantly higher than QuickClip2 [Resolution clip USD 94 (94, 188) vs. USD 47 (47, 94); p \ 0.001]. When QuickClip2 was used, there was a higher incidence of wasted clips as compared to the Resolution clip (QuickClip2 32.5 % vs. Resolution clip 21.2 %; p = 0.049) (Table 4).
Based on the deployment success rates, we compared the costs of treating 100 subjects with endoscopic clips using four different strategies. These were using the Resolution clip only (strategy 1), QuickClip2 only (strategy 2), QuickClip2 once, failing which Resolution clips were used (strategy 3), and QuickClip2 twice, failing which Resolution clips were used (strategy 4). Strategy 1 would result in a maximum of 3 Resolution clips deployed before success was achieved, and 116 Resolution clips would be used. This would cost USD 10,904 for 100 subjects (average of USD109.04 per person). Strategy 2 would result in a maximum of 5 QuickClip2 deployed, a total of 137 QuickClip2 costing USD 6439 for 100 subjects (USD 64.39 per person). Strategy 3 would result in a maximum of 1 QuickClip2 and 3 Resolution clips. A total of 100 QuickClip2 and 32 Resolution clips would cost USD 7708 for 100 subjects (USD 77.08 per person). Finally, strategy 4 would result in a maximum of 2 QuickClip2 and 2 Resolution clips. 113 QuickClip2 and 8 Resolution clips costing
123
Dig Dis Sci Table 4 Utilization and costs of endoscopic clipping Variable
Procedures using resolution clip (n = 99)
Procedures using QuickClip2 (n = 163)
1
52 (52.5)
91 (55.8)
2
31 (31.3)
41 (25.2)
p value
Number of clips per procedure [n (%)]
16 (16.2)
31 (19.0)
Total cost of clips per procedure (USD)
C3
94 (94, 188)
47 (47.94)
Procedures with clips wasted during deployment [n (%)]
21 (21.2)
53 (32.5)
0.534
\0.001* 0.049*
Median (quartiles)
* p \ 0.05
USD 6298 would be used for 100 subjects (USD 62.98 per person).
Discussion Endoscopic clips are available in a variety of designs; however, studies comparing these clips in humans are limited. The few studies comparing the Resolution clip and the QuickClip2 were performed on animal models and on the benchtop [15, 16]. Jensen et al. conducted two canine studies using the Resolution clip and QuickClip2. The Resolution clip was found to have a higher retention rate at 3 and 4 weeks as compared to the QuickClip2 [15, 16]. When clips are deployed, not all of them will successfully anchor onto the targeted lesion. Sometimes, clips are wasted. This study has illustrated that different clip designs can give rise to different deployment rates. The longpronged, reopening clip had an 86.0 % successful deployment rate , which was significantly higher than 73.4% of the short-pronged, single-opening clip. Clips that were longer and reopened had significantly less wastage as well. On multivariate analysis, both the long-pronged, reopening clip and deployment in the lower GI tract were independently associated with successfully deployment. We have observed several mechanisms in which deployment may fail. The clip may be dislodged before it is fired, the prongs may shut prematurely, the clip may malfunction, and the clip may dislodge soon after it is deployed onto the tissue. The QuickClip2 being a singleopening clip was particularly prone to wastage from premature prong closure during clip opening. Both the QuickClip2 and Resolution clips had incidences of clip malfunctioning during deployment when the distal delivery catheter was in a tightly angulated or retroflexed position. The superiority of the Resolution clip may be attributed to several features. Its reopening function allows misalignment to be corrected before release. This facilitates optimal positioning and reduces wastage.
123
Another advantage of this clip is that its prongs are longer and opening angles wider allowing more tissue to be grasped. A benchtop study by Daram et al. [17] showed that the Resolution clip was superior to the QuickClip2 in closure strength and pulling strength. These qualities enhance its ability to anchor securely and facilitate a successful deployment. In addition, an observational study by Swellengrebel et al. [13] showed that the Resolution clip had a longer tissue retention duration as compared to the QuickClip2. This may be important if clips are needed to close defects or to localize lesions. Clips placed in the lower GI tract were shown to be associated with successful deployment. The thin colonic mucosa allows tissue to be tented and grasped more readily. In contrast, gastric mucosa is thicker and harder to grasp. Proper anchorage is also difficult when clips are deployed on fibrotic ulcers in the upper GI tract. Although the Resolution clip had a higher deployment rate, the median cost per procedure was higher. This is due to the higher cost of a Resolution clip (USD 94) as compared to QuickClip2 (USD 47). When comparing different strategies of clip utilization, using Resolution clips only led to the least number of clips, but cost the most (USD 109.04 per person). Starting with the QuickClip2 and switching to the Resolution clip after 2 unsuccessful QuickClip2 applications was the most cost-effective strategy (USD 62.98 per person). Other factors which may influence the type of clip used include familiarity with the device and whether the indication can only be performed successfully with a wideropening clip for technical reasons (e.g., large defect closure). Therefore, if the lesion is technically amenable to therapy using the QuickClip2, one should consider using it for a maximum of 2 attempts, failing which the Resolution clip should be considered. This study design was retrospective since it allowed us to collect a larger number of cases for a well-powered analysis. A retrospective study is ideal when it is not possible to know before a procedure whether the
Dig Dis Sci
intervention (clips) will be performed. In our study, only 262 procedures utilized clips out of 14,690 procedures (1.8 %). Of these, more than half were diagnostic procedures in which the use of a clip was not anticipated. In conclusion, this study showed that a long-pronged, reopening endoscopic clip had a significantly higher deployment success rate than the shorter-pronged, singleopening clip. The former clip was associated with a significantly lower incidence of wastage. Independent variables associated with a successful deployment were a longpronged, reopening clip and deployment in the lower GI tract. The cost of the Resolution clip was significantly greater than QuickClip2. Conflict of interest
None.
References 1. Hayashi T, Yonezawa M, Kuwabara T, Kudoh I. The study on staunch clip for the treatment by endoscopy. Gastroenterol Endosc. 1975;17:92–101. 2. Hachisu T. Evaluation of endoscopic hemostasis using an improved clipping apparatus. Surg Endosc. 1988;2:13–17. 3. Hachisu T, Miyazaki S, Hamaguchi K. Endoscopic clip-marking of lesions using the newly developed HX-3L clip. Surg Endosc. 1989;3:142–147. 4. Hachisu T, Yamada H, Satoh S, Kouzu T. Endoscopic clipping with a new rotatable clip-device and a long clip. Dig Endosc. 1996;8:127–133. 5. Leung Ki EL, Lau JY. New endoscopic hemostasis methods. Clin Endosc. 2012;45:224–229. 6. Kirschniak A, Kratt T, Stu¨ker D, Braun A, Schurr MO, Ko¨nigsrainer A. A new endoscopic over-the-scope clip system for treatment of lesions and bleeding in the GI tract: first clinical experiences. Gastrointest Endosc. 2007;66:162–167.
7. Chan SM, Chiu PW, Teoh AY, Lau JY. Use of the over-the-scope clip for treatment of refractory upper gastrointestinal bleeding: a case series. Endoscopy. 2014;46:428–431. 8. Technology Assessment Committee, Chuttani R, Barkun A, Carpenter S, Chotiprasidhi P, et al. Endoscopic clip application devices. Gastrointest Endosc. 2006;63:746–750. 9. Ginsberg GG, Lipman TO, Fleischer DE. Endoscopic clip-assisted placement of enteral feeding tubes. Gastrointest Endosc. 1994;40:220–222. 10. Sriram PV, Das G, Rao GV, Reddy DN. Another novel use of endoscopic clipping: to anchor an esophageal endoprosthesis. Endoscopy. 2001;33:724–726. 11. Baron TH, Gostout CJ, Herman L. Hemoclip repair of a sphincterotomy-induced duodenal perforation. Gastrointest Endosc. 2000;52:566–568. 12. Luigiano C, Ferrara F, Ghersi S, Fabbri C, et al. Endoclip-assisted resection of large pedunculated colorectal polyps: technical aspects and outcome. Dig Dis Sci. 2010;55:1726–1731. 13. Swellengrebel HA, Marijnen CA, Vincent A, Cats A. Evaluating long-term attachment of two different endoclips in the human gastrointestinal tract. World J Gastrointest Endosc. 2010;2: 344–348. 14. Hwang JH, Fisher DA, Ben-Menachem T, Chandrasekhara V, et al. Standards of Practice Committee of the American Society for Gastrointestinal Endoscopy. The role of endoscopy in the management of acute non-variceal upper GI bleeding. Gastrointest Endosc. 2012;75:1132–1138. 15. Jensen DM, Machicado GA, Hirabayashi K. Randomized controlled study of 3 different types of hemoclips for hemostasis of bleeding canine acute gastric ulcers. Gastrointest Endosc. 2006;64:768–773. 16. Jensen DM, Machicado GA. Hemoclipping of chronic canine ulcers: a randomized, prospective study of initial deployment success, clip retention rates, and ulcer healing. Gastrointest Endosc. 2009;70:969–975. 17. Daram SR, Tang SJ, Wu R, To SD. Benchtop testing and comparisons among three types of through-the-scope endoscopic clipping devices. Surg Endosc. 2013;27:1521–1529.
123
