Fellow Involvement During Colonoscopy Does Not Reduce Adenoma Detection Rate
1
difference in sex, difficulty of cannulation, pancreatic opacification and SOD between the severe and non-severe pancreatitis groups. Medication before and after ERCP was not related to the severity of the pancreatitis. Conclusion: Post-ERCP pancreatitis follows benign course and malignancy, incomplete drainage of biliary obstruction and repeated ERCP are the risk factors for severe ERCP-related pancreatitis. M1407 A Prospective Evaluation of Fluoroscopy Time At ERCP Mark Mcloughlin, Edward Y. Kim, Eric C. Lam, Michael F. Byrne, Jennifer J. Telford, Jack Amar, Robert A. Enns Background: Fluoroscopy is essential during ERCP but carries a risk of radiation exposure to patients and staff. In a previous retrospective study in our institution factors such as age, diagnosis, endoscopist and the nature and number of interventions performed were shown to significantly affect fluoroscopy time. Elderly patients with stones or a malignant stricture were likely to require the longest duration of fluoroscopy. Objectives: To prospectively determine factors which influence fluoroscopy time. Methods: We performed a prospective analysis of 432 consecutive ERCPs carried out by 5 physicians at 2 tertiary referral hospitals over a 9 month period. Information including fluoroscopy time, endoscopist, patient demographics, indication, types of interventions, instruments used, presence/absence of a technician or GI Fellow and diagnosis was recorded prospectively. Results: 388 procedures were analyzed (44 excluded for incomplete data) with a mean fluoroscopy time of 6.77 (95% CI 6.15 to 7.39) minutes. Only 1 endoscopist used a radiology technician to control fluoroscopy during ERCP. The mean fluoroscopy time was significantly lower for 2 of the other endoscopists when compared with the endoscopist/radiology technician combination (mean of 4.16 minutes less [95% CI -5.48 to -2.48] for each). For the remaining 2 endoscopists there was no significant difference. In a multivariable analysis the other variables which were found to significantly increase fluoroscopy time were stent insertion (increase of 3.11 minutes [95% CI 1.91 to 4.30]), lithotripsy (þ5.74 mins [95% CI 0.931 to 10.5]), needle knife sphincterotomy (þ4.44 mins [95% CI 2.20 to 6.67]), biopsies (þ2.11 mins [95% CI 0.025 to 4.18]), guidewire use (þ1.55 mins [95% CI 0.025 to 3.07]), additional wires other than the standard (þ5.61 mins [95% CI 2.69 to 8.51]) and use of a balloon catheter (þ4.27 mins [95% CI 3.00 to 5.53]). When a Fellow was involved in the procedure (nZ318) the mean fluoroscopy time was 6.62 minutes, compared to 4.44 minutes when not present (nZ70) pZ0.005. Conclusions: Several factors affect the duration of fluoroscopy required during an ERCP procedure. The use of instrumentation such as a needle knife, guidewires, and balloon increase duration interval and corresponding fluoroscopy time and in many cases this probably relates to case complexity. The involvement of a radiology technician and a GI Fellow may also influence fluoroscopy time. Appropriate caution (ie. shielding) is critical especially during procedures with predicted high radiation exposure. M1408 Feasibility of Measuring Quality Indicators in Real-Time for Gastrointestinal Endoscopic Procedures Using Visual Tools and Computer Automation Anita B. Saini, Pradeep K. Bekal, Alan V. Safdi, Dan Walker, Bharat Saini, Manish Chokshi, Michael Safdi Introduction: Quality indicators are objective measures which are being used to establish competence in performing gastrointestinal endoscopic procedures and to help identify areas for continuous quality improvement. A task force consisting of expert endoscopists created a list of quality indicators. In order to implement these indicators, endoscopists need access to continuous quality data. It is possible to collect, analyze and present data in an automated fashion with the use of technology. We present how one such technology product, eMergeEndo is being used in endoscopy centers to measure quality indicators in real-time. We measured the intra-procedure indicator for colonoscopy-withdrawal time. Mean withdrawal time should be R6 minutes in colonoscopies with normal results performed in patients with intact anatomy. eMergeEndo is an automated workflow system that captures data at every step of endoscopic procedures. Procedure start time, cecal intubation time and procedure end time are documented by a nurse interacting with eMergeEndo via patented voice command navigation. Additionally, the withdrawal time is displayed live in bold numbers giving endoscopists visual feedback during the procedure. Summary of this data is available for endoscopist to review for any time period. Methods: We retrospectively analyzed data collected by eMergeEndo at one center, for cecal withdrawal time, for 15 endoscopists over a 10 month period. 2348 screening colonoscopies were studied for the period 1/30/08 to 11/26/08. Endoscopists started reviewing their summary data 11/3/08 onwards. Results: The percent of colonoscopies with cecal withdrawal time !6 minutes decreased after endoscopists had access to quality data. With just 3 weeks of data, mean cecal withdrawal times have improved although not statistically significant. P- value is progressively improving(pZ0.58, pZ0.44, pZ0.21) and we will continue to monitor to see if this improvement trend becomes statistically significant. Conclusion: Our analysis shows that by giving physicians access to real-time and visual quality indicators data, there is a trend towards improvement in cecal withdrawal time. Thus, using automated systems like eMergeEndo make measuring, analyzing and presenting quality indicators data feasible in real-time. Period Total normal colonoscopies Colonoscopies with cecal withdrawal time !6minutes % of colonoscopies with cecal withdrawal time ! 6minutes Mean cecal withdrawal time 1/30/08 to 11/26/08 2348 244 10.39 8 min 37 sec 11/03/08 to 11/26/08 293 26 8.87 8 min 45 sec 11/10/08 to 11/26/08 222 19 8.55 8 min 51 sec 11/17/08 to 11/26/08 140 8 5.71 9 min 9 sec Continuous improvement in cecal withdrawal time after giving endoscopists access to quality data. M1409 Fellow Involvement During Colonoscopy Does Not Reduce Adenoma Detection Rate Mark Friedman, Eddie L. Irions, Gloria G. Guptill, Geeta Arora, Amy Kakkanatt, Jesse A. Green Background: Colorectal cancer (CRC) is the second leading cause of cancer deaths in the United States. Colonoscopy has been effective in reducing the incidence of CRC. However, colonoscopy is a complex technical procedure that requires training and experience to maximize accuracy, and an adequate procedure volume is clearly necessary to achieve competence. Few studies have examined if gastroenterology (GI) fellows, who have less endoscopy experience than their attending physicians, are less likely to identify adenomas. Objective: To assess whether performance of screening colonoscopy by a gastroenterology fellow affects adenoma detection rate (ADR). Methods: A retrospective chart review of 1190 patients who underwent age- appropriate, average-risk screening colonoscopy from July 2007 to July 2008 was conducted. The patients’ gender and race were obtained. Colonoscopy insertion and withdrawal times, prep quality, fellow involvement, and findings during the procedure were also recorded. A database was created using Microsoft Excel. Statistical analysis was performed using chi-square with significance set at p!0.05. The study was approved by the university IRB. Results: In total, 971 colonoscopies were performed by one of 5 GI and 3 colorectal surgery attendings alone. 219 were performed by a GI fellow under the supervision of one of the same 5 GI attendings. There was no significant difference in patient age, gender, or colonoscopy prep quality between the patients in the attending group and the trainee group. The polyp detection rate (PDR) in the attending group was 43.3%, as opposed to 48.8% in the fellow group (pO0.05). The adenoma detection rate (ADR) was 25.6% in the attending group, as opposed to 27.9% in the fellow group (pO0.05). The detection of advanced adenomas was 6.5% in the attending group, as opposed to 5.9% in the fellow group (pO0.05). Conclusions: ADR is increasingly used as a quality measure to assess colonoscopic competency. In our retrospective study, the ADR during procedures performed by a GI fellow, under the supervision of an attending, was comparable to those performed by an attending alone. Although GI fellows have less endoscopic experience than GI attendings, our results demonstrated similar ADRs between the two groups. M1410 EchoBrush vs. Standard Endoscopic Ultrasound-Fine Needle Aspiration (EUS-FNA) Techniques for Cytologic Evaluation of Cystic Pancreatic Lesions (CPLs): Final Results of Blinded Prospective Comparison Study Kanwar R. Gill, Mohammad A. Al-Haddad, Murli Krishna, Laith H. Jamil, Muhammad K. Hasan, Timothy A. Woodward, Massimo Raimondo, Michael B. Wallace Background: Cystic pancreatic lesions (CPLs) are associated with a potential for malignant transformation. Endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) is a widely used method for evaluating these lesions, but has limited sensitivity. Previously we have shown our preliminary analysis that the cytology samples obtained by through-the-needle cytologic brush at the time of EUS (EchoBrush, Cook Endoscopy, Winston-Salem, NC) has superior diagnostic yield compared to conventional FNA. (Al-Haddad et al, 2007). Objectives: To determine if Echobrush is superior to FNA for EUS sampling of CPLs in blinded prospective study. Methods: Consecutive patients with cystic pancreatic lesions were included. All cysts were sampled by standard EUS-FNA (50% of cyst volume) followed by EchoBrush cytology, then by aspiration of the remaining fluid. Fluid samples were separately submitted (EUS-FNA and EchoBrush) and were read by a study pathologist who was unaware of the sampling technique and findings in the corresponding paired sample. Cytologic specimens were evaluated for the presence of mucinous epithelium/intracellular mucin and grade of dysplasia. The two methods were compared for presence of mucinous epithelium and grade of dysplasia. Statistical analyses of the paired samples were done by McNemar’s test. Abstracts AB234 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 5 : 2009 www.giejournal.org
-
Upload
mark-friedman -
Category
Documents
-
view
215 -
download
1
Transcript of Fellow Involvement During Colonoscopy Does Not Reduce Adenoma Detection Rate
Abstracts
difference in sex, difficulty of cannulation, pancreatic opacification and SODbetween the severe and non-severe pancreatitis groups. Medication before andafter ERCP was not related to the severity of the pancreatitis. Conclusion: Post-ERCPpancreatitis follows benign course and malignancy, incomplete drainage of biliaryobstruction and repeated ERCP are the risk factors for severe ERCP-relatedpancreatitis.
M1407
A Prospective Evaluation of Fluoroscopy Time At ERCPMark Mcloughlin, Edward Y. Kim, Eric C. Lam, Michael F. Byrne, JenniferJ. Telford, Jack Amar, Robert A. EnnsBackground: Fluoroscopy is essential during ERCP but carries a risk of radiationexposure to patients and staff. In a previous retrospective study in our institutionfactors such as age, diagnosis, endoscopist and the nature and number ofinterventions performed were shown to significantly affect fluoroscopy time.Elderly patients with stones or a malignant stricture were likely to require thelongest duration of fluoroscopy. Objectives: To prospectively determine factorswhich influence fluoroscopy time. Methods: We performed a prospective analysisof 432 consecutive ERCPs carried out by 5 physicians at 2 tertiary referral hospitalsover a 9 month period. Information including fluoroscopy time, endoscopist,patient demographics, indication, types of interventions, instruments used,presence/absence of a technician or GI Fellow and diagnosis was recordedprospectively. Results: 388 procedures were analyzed (44 excluded for incompletedata) with a mean fluoroscopy time of 6.77 (95% CI 6.15 to 7.39) minutes. Only 1endoscopist used a radiology technician to control fluoroscopy during ERCP. Themean fluoroscopy time was significantly lower for 2 of the other endoscopists whencompared with the endoscopist/radiology technician combination (mean of 4.16minutes less [95% CI -5.48 to -2.48] for each). For the remaining 2 endoscopiststhere was no significant difference. In a multivariable analysis the other variableswhich were found to significantly increase fluoroscopy time were stent insertion(increase of 3.11 minutes [95% CI 1.91 to 4.30]), lithotripsy (þ5.74 mins [95% CI0.931 to 10.5]), needle knife sphincterotomy (þ4.44 mins [95% CI 2.20 to 6.67]),biopsies (þ2.11 mins [95% CI 0.025 to 4.18]), guidewire use (þ1.55 mins [95% CI0.025 to 3.07]), additional wires other than the standard (þ5.61 mins [95% CI 2.69to 8.51]) and use of a balloon catheter (þ4.27 mins [95% CI 3.00 to 5.53]). Whena Fellow was involved in the procedure (nZ318) the mean fluoroscopy time was6.62 minutes, compared to 4.44 minutes when not present (nZ70) pZ0.005.Conclusions: Several factors affect the duration of fluoroscopy required during anERCP procedure. The use of instrumentation such as a needle knife, guidewires,and balloon increase duration interval and corresponding fluoroscopy time and inmany cases this probably relates to case complexity. The involvement of a radiologytechnician and a GI Fellow may also influence fluoroscopy time. Appropriatecaution (ie. shielding) is critical especially during procedures with predicted highradiation exposure.
M1408
Feasibility of Measuring Quality Indicators in Real-Time for
Gastrointestinal Endoscopic Procedures Using Visual Tools and
Computer AutomationAnita B. Saini, Pradeep K. Bekal, Alan V. Safdi, Dan Walker, Bharat Saini,Manish Chokshi, Michael SafdiIntroduction: Quality indicators are objective measures which are being used toestablish competence in performing gastrointestinal endoscopic procedures and tohelp identify areas for continuous quality improvement. A task force consisting ofexpert endoscopists created a list of quality indicators. In order to implement theseindicators, endoscopists need access to continuous quality data. It is possible tocollect, analyze and present data in an automated fashion with the use oftechnology. We present how one such technology product, eMergeEndo is beingused in endoscopy centers to measure quality indicators in real-time. We measuredthe intra-procedure indicator for colonoscopy-withdrawal time. Mean withdrawaltime should be R6 minutes in colonoscopies with normal results performed inpatients with intact anatomy. eMergeEndo is an automated workflow system thatcaptures data at every step of endoscopic procedures. Procedure start time, cecalintubation time and procedure end time are documented by a nurse interactingwith eMergeEndo via patented voice command navigation. Additionally, thewithdrawal time is displayed live in bold numbers giving endoscopists visualfeedback during the procedure. Summary of this data is available for endoscopist toreview for any time period. Methods: We retrospectively analyzed data collected byeMergeEndo at one center, for cecal withdrawal time, for 15 endoscopists over a 10month period. 2348 screening colonoscopies were studied for the period 1/30/08 to11/26/08. Endoscopists started reviewing their summary data 11/3/08 onwards.Results: The percent of colonoscopies with cecal withdrawal time !6 minutesdecreased after endoscopists had access to quality data. With just 3 weeks of data,mean cecal withdrawal times have improved although not statistically significant. P-value is progressively improving(pZ0.58, pZ0.44, pZ0.21) and we will continue to
AB234 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 5 : 2009
monitor to see if this improvement trend becomes statistically significant.Conclusion: Our analysis shows that by giving physicians access to real-time andvisual quality indicators data, there is a trend towards improvement in cecalwithdrawal time. Thus, using automated systems like eMergeEndo makemeasuring, analyzing and presenting quality indicators data feasible in real-time.
Colonoscopies% of colonoscopieswith cecal
Period
Totalnormalcolonoscopies
with cecalwithdrawaltime !6minutes
withdrawaltime! 6minutes
www.giejo
Mean cecalwithdrawaltime
1/30/08 to 11/26/08
2348 244 10.39 8 min 37 sec 11/03/08 to 11/26/08 293 26 8.87 8 min 45 sec 11/10/08 to 11/26/08 222 19 8.55 8 min 51 sec 11/17/08 to 11/26/08 140 8 5.71 9 min 9 secContinuous improvement in cecal withdrawal time after giving endoscopists accessto quality data.
M1409
Fellow Involvement During Colonoscopy Does Not Reduce
Adenoma Detection RateMark Friedman, Eddie L. Irions, Gloria G. Guptill, Geeta Arora,Amy Kakkanatt, Jesse A. GreenBackground: Colorectal cancer (CRC) is the second leading cause of cancer deathsin the United States. Colonoscopy has been effective in reducing the incidence ofCRC. However, colonoscopy is a complex technical procedure that requires trainingand experience to maximize accuracy, and an adequate procedure volume is clearlynecessary to achieve competence. Few studies have examined if gastroenterology(GI) fellows, who have less endoscopy experience than their attending physicians,are less likely to identify adenomas. Objective: To assess whether performance ofscreening colonoscopy by a gastroenterology fellow affects adenoma detection rate(ADR). Methods: A retrospective chart review of 1190 patients who underwent age-appropriate, average-risk screening colonoscopy from July 2007 to July 2008 wasconducted. The patients’ gender and race were obtained. Colonoscopy insertionand withdrawal times, prep quality, fellow involvement, and findings during theprocedure were also recorded. A database was created using Microsoft Excel.Statistical analysis was performed using chi-square with significance set at p!0.05.The study was approved by the university IRB. Results: In total, 971 colonoscopieswere performed by one of 5 GI and 3 colorectal surgery attendings alone. 219 wereperformed by a GI fellow under the supervision of one of the same 5 GI attendings.There was no significant difference in patient age, gender, or colonoscopy prepquality between the patients in the attending group and the trainee group. Thepolyp detection rate (PDR) in the attending group was 43.3%, as opposed to 48.8%in the fellow group (pO0.05). The adenoma detection rate (ADR) was 25.6% in theattending group, as opposed to 27.9% in the fellow group (pO0.05). The detectionof advanced adenomas was 6.5% in the attending group, as opposed to 5.9% in thefellow group (pO0.05). Conclusions: ADR is increasingly used as a quality measureto assess colonoscopic competency. In our retrospective study, the ADR duringprocedures performed by a GI fellow, under the supervision of an attending, wascomparable to those performed by an attending alone. Although GI fellows haveless endoscopic experience than GI attendings, our results demonstrated similarADRs between the two groups.
M1410
EchoBrush vs. Standard Endoscopic Ultrasound-Fine Needle
Aspiration (EUS-FNA) Techniques for Cytologic Evaluation of
Cystic Pancreatic Lesions (CPLs): Final Results of Blinded
Prospective Comparison StudyKanwar R. Gill, Mohammad A. Al-Haddad, Murli Krishna, Laith H. Jamil,Muhammad K. Hasan, Timothy A. Woodward, Massimo Raimondo,Michael B. WallaceBackground: Cystic pancreatic lesions (CPLs) are associated with a potential formalignant transformation. Endoscopic ultrasound-guided fine needle aspiration(EUS-FNA) is a widely used method for evaluating these lesions, but has limitedsensitivity. Previously we have shown our preliminary analysis that the cytologysamples obtained by through-the-needle cytologic brush at the time of EUS(EchoBrush, Cook Endoscopy, Winston-Salem, NC) has superior diagnostic yieldcompared to conventional FNA. (Al-Haddad et al, 2007). Objectives: To determine ifEchobrush is superior to FNA for EUS sampling of CPLs in blinded prospectivestudy. Methods: Consecutive patients with cystic pancreatic lesions were included.All cysts were sampled by standard EUS-FNA (50% of cyst volume) followed byEchoBrush cytology, then by aspiration of the remaining fluid. Fluid samples wereseparately submitted (EUS-FNA and EchoBrush) and were read by a studypathologist who was unaware of the sampling technique and findings in thecorresponding paired sample. Cytologic specimens were evaluated for thepresence of mucinous epithelium/intracellular mucin and grade of dysplasia. Thetwo methods were compared for presence of mucinous epithelium and grade ofdysplasia. Statistical analyses of the paired samples were done by McNemar’s test.
urnal.org
