- Email: [email protected]
Sa1508 The Role of Ultrathin TransNasal Endoscopy: Meta Analysis
Abstracts
significance of positive FOBT and predictive factors of CRC with positive FOBT for the elderly aged R85 years. Methods: The records of consecutive 127 patients aged R85 years who tested FOBT positive and then underwent colonoscopy from January 2010 to September 2014 in Nagoya University Hospital and Tokai Central Hospital were reviewd. Uni- and multivariate logistic regression analyses were performed to describe the association. Results: 127 patients (M/F: 66/61) aged R85 years who tested FOBT(+) and underwent colonoscopy were enrolled. Median age at colonoscopy was 86.6 years (85-97years). Baseline characteristics of enrolled patients were 43 cases (33.9%) of hypertension or cardiovascular disease, 26 cases (20.5%) of malignancy, 21 cases (16.5%) of dementia, 19 cases (15.0%) of DM, 17 cases (13.4%) of CKD, 12 cases (9.4%) of COPD and 35 cases (27.6%) of antiplatelets or anticoagulants use (including overlapping cases). The cecal intubation rate was 94.5%. In terms of post colonoscopy complications, there was 1 case (0.8%) of perforation. The adenoma detection rate and prevalence of CRC were 37 cases (29.1%) and 12 cases (9.4%), respectively. The outcomes of enrolled patients were 30 cases of endoscopic resection, and 11 cases of surgery (including perforation case). Uni- and multivariate analysis using a logistic regression model identified presence of two or more sign or symptom among body weight loss, anemia, abd pain, appetite loss, and rectal bleeding at colonoscopy as a significant independent predictive factor for CRC in patients aged R85 years with positive FOBT (OR 13.8, 95%CI 2.8-66.8). There were no significant differences in sex ratio, age at colonoscopy, baseline characteristics, antiplatelets or aniticoagulants use, laboratory data at colposcopy, presence of one sign or symptom, past history of CRC or adenoma, past history of positive FOBT, and family history of CRC. Conclusion: The predictive factor that correlated with CRC in the elderly aged R85 years with positive FOBT was presence of two or more sign or symptom among body weight loss, anemia, abd pain, appetite loss, and rectal bleeding at colonoscopy.
Sa1506 A Cost-Utilization Analysis of Non-Hematemesis GIB Diagnostic Management Strategies Salmaan Jawaid*, Louise Maranda, Neil B. Marya, David R. Cave Univ of Massachusetts, Worcester, MA Introduction: Non-hematemesis gastrointestinal bleeding [NH GIB] can originate anywhere within the gastrointestinal tract. Current practice usually dictates an upper and lower endoscopy, followed by other procedures as needed. As a consequence, the diagnostic evaluation may incur unnecessary health care costs and diagnostic times. The use of a cost decision-analytical model of the current diagnostic management strategy applied to patients presenting with NH GIB may reveal alternative strategies for the evaluation of NH GIB. Methods: The model was designed to retrospectively follow the diagnostic course of 231 consecutive patients presenting with NHGIB to a tertiary care center emergency department [ED]. Calculations of total cost were based on cost of each esophagogastroduodenoscopy [EGD], colonoscopy [COLO], and video capsule endoscopy [VCE] plus the cost of pursuing secondary procedures after non-diagnostic primary procedures. Variables incorporated in the model were the diagnostic/non-diagnostic rate with each respective procedure and the average time taken to reach each procedure. Our analysis was based primarily on the effect of selecting a specific procedure, as the primary diagnostic modality, on the cost and time to diagnosis. Results: When performed as a primary diagnostic procedure in NHGIB patients admitted to the floor, VCE had a diagnostic rate of 68% vs. 45% and 48% for a primary EGD and primary COLO, respectively. The non-diagnostic EGDs [55%] and COLs [52%] underwent a second procedure; the non diagnostic VCEs [32%] did not undergo further procedures. Combining the diagnostic rates for each primary procedure with the cost of performing subsequent procedures [after nondiagnostic primary procedures], demonstrates the use VCE as a primary diagnostic procedure [nZ9] results in a total cost of $12,146 vs. $12,746 and $13,162 for a primary EGD [nZ47] and COLO [nZ33], respectively. In a one way sensitivity analysis, even if the diagnostic rate of VCE reached 100%, requiring a guaranteed second procedure, the cost of performing a primary VCE [$12,445] would still be less than a primary diagnostic EGD [$12,746] or COLO [$13,161]. Combining the diagnostic and non-diagnostic rate of each primary procedure with the time taken to each primary procedure [29 hours to EGD, 38 hours to COLO, 37 hours to VCE], demonstrates the use of VCE as a primary diagnostic procedure in NHGIB patients admitted to the floor would take 74 unit hours to reach a diagnosis compared to 104 and 131 for EGD and COLO, respectively. In a one way sensitivity analysis, the diagnostic rate of EGD and COLO would have to reach 73% and 84%, respectively, to diagnose lesions faster than a primary VCE. Conclusion: Initial use of VCE for the diagnosis of acute NHGIB, may reduce time to diagnosis and management costs.
AB242 GASTROINTESTINAL ENDOSCOPY Volume 81, No. 5S : 2015
Sa1507 Use of the Glasgow-Blatchford Bleeding Score in Bed Assignment for Patients With Suspected UGI Bleeding Drew B. Schembre*1, Janice M. Connolly2, Robson E. Ely1 1 Gastroenterology, Swedish Medical Center, Seattle, WA; 2Hospital Medicine, Swedish Medical Center, Seattle, WA Background: UGI bleeding is a frequent cause for hospitalization. Admission to ICU versus floor bed is usually based on subjective criteria. A more objective tool may help prevent errors in placement that delay needed care or waste resources. The Glasgow-Blatchford bleeding score (GBBS) has been shown to predict hospital mortality and the need for interventions in patients admitted with non-variceal UGI bleeding. Using the GBBS at admission may help stratify patients appropriate for floor care versus intermediate care (IMCU) or ICU. Methods: We reviewed patients admitted with possible UGI bleeding over 32 months to a single tertiary hospital. During the first 16 months, bleeding scores were infrequently calculated. For the last 16 months, scores were calculated by almost all ER physicians. For score !8, floor bed was recommended, 8-12 IMCU and ICU O12. We compared outcomes for patients admitted with and without a GBBS. We also calculated and reviewed GBBS among patients who either died or had admissions O10 days. Results: From October 2011 to June 2014, 456 patients were admitted for presumed UGI bleeding. 281 patients were assigned to beds based on bleeding score and 175 based on physicians’ assessment of risk. Patients stratified by GBBS were more likely to be assigned to IMCU and less likely to go to a floor bed or ICU than those without bleeding scores. Among 68 scored patients placed in floor beds, 7 patients were transferred to a higher level of care and 11 of 140 patients admitted to IMCU were eventually transferred to ICU, compared to 5 of 72 unscored patients admitted to the floor who were later transferred to a higher level of care (pZ0.56) and 13 of 64 IMCU patients transferred to ICU (pZ0.04). Of 31 patients who died or had an admission O10 days, 12 GBBS patients had hospitalizations longer than 10 days and 5 died compared to 8 unscored patients with LOS O10 days and 6 deaths (pZNS). After calculating GBBS for the unscored patients, 27 of 31 (87%) scored 9 or higher and 29 (94%) 7 or higher. No GBBS patient appeared to have died or had a prolonged admission due to a low level of care at admission. One unscored patient admitted to a floor bed, who would have scored 16, died after transfer to ICU and may have benefitted from earlier ICU placement. Conclusion: Use of the GBBS for patients with UGI bleeding may provide more uniform bed assignment and reduce ICU transfers but does not appear to impact LOS, cost or mortality. The great majority of individuals at risk for death or prolonged admission scored O6. The GBBS, along with physician judgment, appears to be a safe and useful tool to direct patient placement. The ideal criteria for bed placement should be determined by larger studies and may vary by facility.
Total Admitted to Floor Bed Admitted to IMCU Admitted to ICU LOS Median Direct Costs Transferred up from Floor Bed Transferred IMCU to ICU Admission O10 Days Death
Patients with GBBS
Patients without GBBS
P value
281 68 (24%) 140 (50%) 73 (26%) 2.75 Days \$5507 7 (11%) 11 (8%) 12 (4%) 5 (2%)
175 72 (41%) 64 (37%) 39 (22%) 2.87 Days \$5481 5 (7%) 13 (20%) 8 (5%) 6 (3%)
!0.01 0.01 0.05 0.82 0.35 0.56 0.04 1.0 0.35
Sa1508 The Role of Ultrathin TransNasal Endoscopy: Meta Analysis Tarek Sawas*, Kia Saeian, Reza Shaker Gastroenterology & Hepatology, Medical College of Wisconsin, Milwaukee, WI Background: Esophagogastroduodenoscopy (EGD) is a common procedure that is usually performed under sedation. The use of sedation improves patients’ tolerance and acceptance. However, It also increases the risk of complications, requires closer monitoring and subsequently increases the cost of the EGD. Ultrathin transnasal endoscopy (T-EGD) was first proposed by Shaker et al. (1993). Since then many randomized controlled trials have evaluated the role of T-EGD in general population and high-risk patients. We report a meta-analysis of prospective controlled trials comparing T-EGD to the conventional regular size sedated and unsedated endoscopy (C-EGD). Method: We searched Pubmed, Embase and Cochrane database for prospective controlled trials that compared T-EGD to regular size C-EGD. Two reviewers selected studies. Quality for each included study was assessed by consort system. No study was eliminated based on quality. Heterogeneity was analyzed by Cochran’s Q statistics. Mantel Haenszel odds ratio (OR) and mean differences (MD) were calculated with fix effect or random effect mode depending on heterogeneity. RevMan 5.2 from Cochrane was used for analysis. Results: 22 prospective controlled trials involving 2921 patients (1626 C-EGD, 1573 T-EGD) were included. Success rate was higher in the C-EGD 98.4% compared to 93.4% in
www.giejournal.org
Abstracts
T-EGD (OR: 0.24, CI: 0.16, 0.36, P!0.001). Failures in T-EGD were due to nasal stenosis. Comfort visual analog scale (VAS) where patients scored their comfort level from 1-10 was not significantly different between the two groups (MD: 0.54, CI: -0.20, 1.28, P: 0.15). Subgroup analysis of patients who underwent unsedated C-EGD showed that T-EGD was significantly more comfortable (MD: 0.94, CI: 0.41, 1.48, P: 0.005). Patients who had T-EGD experienced less gag reflex with VAS MD: -1.78 (CI: -2.83, -0.72, P: 0.001). They also had less nausea and vomiting compared to unsedated C-EGD (MD: -1.73, CI: -2.25, -1.20, P!0.001). Procedure duration was longer in the T-EGD with MD: 74.8 seconds (CI: 50.07, 99.50, P: 0.004). However the recovery duration was shorter in the T-EGD compared to sedated C-EGD (MD -10 min, CI: -10.7, -9.5, P !0.001). Developing hypoxia was not different between T-EGD and unsedated C-EGD (OR: 0.17, CI, 0.02, 1.45, P: 0.11). Patient who underwent T-EGD had lower Heart rate during the procedure (MD -7.16, CI: -11.50, -2.81, P: 0.01) and lower SBP (MD: -7.16, CI: -11.50, -2.81. P: 0.001). 78% in T-EGD group were willing to repeat procedure through same rout VS 58.6% in C-EGD group (OR: 2.25, CI: 1.23, 4.15, P: 0.01). Conclusion: Use of T-EGD is superior to un-sedated C-EGD in term of comfort and should be used as first modality in patients who are at higher risk under sedation. T-EGD should also be offered as an alternative to sedated C-EGD since it decreases recovery duration and subsequently resuming their life quicker.
Sa1509 Financial Analysis of Single Balloon Enteroscopy At a Tertiary Care Center: Where Does the Revenue Come From? Nikhil Banerjee*1, Michael V. Presta2, Matthew T. Charous2, Neil Gupta1 1 Gastroenterology, Loyola University Medical Center, Maywood, IL; 2 Anesthesiology, Loyola University Medical Center, Maywood, IL Background: Resource-intensive endoscopic procedures such as ERCP and EUS have shown to generate more costs than revenue. Lack of financial viability has been mostly in part due to the poor rates of reimbursement under the current payment system in the United States. Double balloon enteroscopy, another resource intensive technology used to evaluate the small bowel has been shown to generate smaller profit margins than predicted upon its implementation in a prior financial analysis. Single balloon enteroscopy (SBE) has never been evaluated for its financial impact at tertiary care hospitals. Aim: 1) Determine the sources of revenue that SBE procedures generate. 2) Determine whether SBE can generate downstream revenues from referrals for the procedure. Methods: All procedures using the SBE system during the first year the equipment was available at a single tertiary care center were identified using the electronic medical record system. Procedure related and financial data in the form of revenue collected were recorded for all SBE procedures. Results were expressed as both total and average standard deviation. Paired and unpaired t-tests were used to determine statistical significance with a p value of ! 0.05 deemed significant. Revenues were analyzed for both physicians and facility as a function of payor type and whether an intervention was performed during the SBE. A subanalysis comparing revenue collected from gastroenterology physician fees and anesthesiology physician fees was also done. Results: 52 single balloon enteroscopy cases were performed between November 1, 2012 and October 31, 2013. Total revenue generated for all SBE procedures was $123,714 including $64,475 dollars from physician fees and $59,239 dollars from the facility fees. Average revenues for the physician and facility fees were $1239506 and $21151923 per case respectively. Revenue generated by anesthesia fees were higher from medicare cases compared to private insurance cases, however revenues from facility fees were higher from private insurance cases compared to medicare cases (Table1 ). Overall, revenues from anesthesiology physician fees were more than revenues from gastroenterology physician fees. This was driven predominately from higher revenues generated from medicare cases and it remained true for both SBE cases that included and did not include endoscopic interventions (Table 2). Only one referred case from outside hospital generated downstream revenue totaling $4727. Conclusion: A larger than expected proportion of revenues generated from SBE cases come in the form of ancillary services provided by anesthesia. Medicare seems to contribute more revenues for these services that private insurance. Projected revenue generation (and it sources) should be considered when establishing a device assisted enteroscopy program.
Sa1510 Clinical Impact of Confocal LASER Endomicroscopy Probe (P - CLE) in the Management of Gastrointestinal Neoplasic and Non - Neoplasic Lesion Carlos Robles-Medranda*, Jesenia Ospina, Miguel Puga-Tejada, Miguel Soria Alcívar, Gladys Bravo Velez, Carlos A. Robles-Jara, Hannah P. Lukashok Endoscopy, Ecuadorian Institute of Digestive Diseases, Guayaquil, Ecuador Background: It has been demonstrated that endoscopic and histopathology findings don’t have good correlation, leading sometimes to diagnostic and therapeutic doubts. Confocal laser endomicroscopy probe (p-CLE) is a new technology which permits in-vivo cellular view of gastrointestinal mucosa. It includes esophagus, stomach, small intestine, colon, biliary tract and pancreas. Studies showed a diagnostic accuracy above 90% for neoplastic (N) and non - neoplastic (NN) lesions, with specific characteristics and classifications. However, few studies have been oriented to determine the clinical impact of p-CLE for diagnosis and management in patients with diagnostic doubts. The aim of this study is to put in evidence the clinical impact of p-CLE in this group of patients. Patients and method: Prospective study, performed from November 2013 to November 2014, in consecutive patients in whom pCLE (CellvizioÒ, Mauna Kea Technology) was indicated due to diagnostic doubts (absence of endoscopic-histological relationship). Baseline characteristics, indications, previous diagnosis studies, findings at p-CLE, clinical management and histopathological outcomes were evaluated. Indications for p-CLE were N and NN lesions. Lesions include: adenomas, dysplasia or cancer located in any gastrointestinal tract level, Barrett’s esophagus, inflammatory bowel disease or pancreatic cysts. Previous diagnostic studies included: high definition magnification with digital chromoendoscopy, ERCP with brushing, EUS. Other studies were: CT-scan, cholangiography by MRI, tumor markers. Interventions based on the findings of p-CLE were analyzed according to the records, and included: drug treatment, other diagnostic studies, endoscopic or surgical treatments. The diagnostic yield was determined thought sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and match measurement. Diagnostic and therapeutic management, redirection of biopsies and need of other diagnostic methods were evaluated. Results: 43 patients were included. 51% were male (22/43) with mean age of 49 19 years. N lesions were 37% (15/43) of cases, located at: stomach (6), esophagus (4), bile duct (3) and colon (2). The sensitivity was 87 % and specificity of 93%,(AUCZ0,9582) with a PPV of 87% and NPV 93%. p value !0.05. The observed concordance was 90.6 % with a Kappa value of 0.80, corresponding to a force of substantial agreement according to Landis & Koch criteria. Changes in diagnostic and therapeutic approach were evident in 39.5% of cases (17/43), directing the sampling in 100% of cases (17/17) and avoiding diagnostic or therapeutic methods in all cases. Conclusion: p-CLE is an essential diagnostic tool for patients with diagnostic doubts, as it allows in-vivo display cell and helps to direct biopsies. It has a significant clinical impact on the diagnosis and treatment of patients
Table 1 Medicare
Private
p value
\$276.40 \$1,175.85 \$1249.63
\$346.34 \$667.07 \$5271.37
0.38 !0.01 !0.01
Gastroenterologist
Anesthesiologist
p value
\$278.23 \$276.40 \$346.34 \$166.06 \$311.88
\$961.67 \$1,175.85 \$667.07 \$1,096.65 \$921.18
!0.01 !0.01 0.06 !0.01 !0.01
Gastroenterologist Revenue Anesthesiologist Revenue Facility Revenue (outpatient only) Revenues are per case (mean)
Table 2 Total Medicare Private No Intervention Intervention Revenues are per case (mean)
www.giejournal.org
Volume 81, No. 5S : 2015 GASTROINTESTINAL ENDOSCOPY AB243
significance of positive FOBT and predictive factors of CRC with positive FOBT for the elderly aged R85 years. Methods: The records of consecutive 127 patients aged R85 years who tested FOBT positive and then underwent colonoscopy from January 2010 to September 2014 in Nagoya University Hospital and Tokai Central Hospital were reviewd. Uni- and multivariate logistic regression analyses were performed to describe the association. Results: 127 patients (M/F: 66/61) aged R85 years who tested FOBT(+) and underwent colonoscopy were enrolled. Median age at colonoscopy was 86.6 years (85-97years). Baseline characteristics of enrolled patients were 43 cases (33.9%) of hypertension or cardiovascular disease, 26 cases (20.5%) of malignancy, 21 cases (16.5%) of dementia, 19 cases (15.0%) of DM, 17 cases (13.4%) of CKD, 12 cases (9.4%) of COPD and 35 cases (27.6%) of antiplatelets or anticoagulants use (including overlapping cases). The cecal intubation rate was 94.5%. In terms of post colonoscopy complications, there was 1 case (0.8%) of perforation. The adenoma detection rate and prevalence of CRC were 37 cases (29.1%) and 12 cases (9.4%), respectively. The outcomes of enrolled patients were 30 cases of endoscopic resection, and 11 cases of surgery (including perforation case). Uni- and multivariate analysis using a logistic regression model identified presence of two or more sign or symptom among body weight loss, anemia, abd pain, appetite loss, and rectal bleeding at colonoscopy as a significant independent predictive factor for CRC in patients aged R85 years with positive FOBT (OR 13.8, 95%CI 2.8-66.8). There were no significant differences in sex ratio, age at colonoscopy, baseline characteristics, antiplatelets or aniticoagulants use, laboratory data at colposcopy, presence of one sign or symptom, past history of CRC or adenoma, past history of positive FOBT, and family history of CRC. Conclusion: The predictive factor that correlated with CRC in the elderly aged R85 years with positive FOBT was presence of two or more sign or symptom among body weight loss, anemia, abd pain, appetite loss, and rectal bleeding at colonoscopy.
Sa1506 A Cost-Utilization Analysis of Non-Hematemesis GIB Diagnostic Management Strategies Salmaan Jawaid*, Louise Maranda, Neil B. Marya, David R. Cave Univ of Massachusetts, Worcester, MA Introduction: Non-hematemesis gastrointestinal bleeding [NH GIB] can originate anywhere within the gastrointestinal tract. Current practice usually dictates an upper and lower endoscopy, followed by other procedures as needed. As a consequence, the diagnostic evaluation may incur unnecessary health care costs and diagnostic times. The use of a cost decision-analytical model of the current diagnostic management strategy applied to patients presenting with NH GIB may reveal alternative strategies for the evaluation of NH GIB. Methods: The model was designed to retrospectively follow the diagnostic course of 231 consecutive patients presenting with NHGIB to a tertiary care center emergency department [ED]. Calculations of total cost were based on cost of each esophagogastroduodenoscopy [EGD], colonoscopy [COLO], and video capsule endoscopy [VCE] plus the cost of pursuing secondary procedures after non-diagnostic primary procedures. Variables incorporated in the model were the diagnostic/non-diagnostic rate with each respective procedure and the average time taken to reach each procedure. Our analysis was based primarily on the effect of selecting a specific procedure, as the primary diagnostic modality, on the cost and time to diagnosis. Results: When performed as a primary diagnostic procedure in NHGIB patients admitted to the floor, VCE had a diagnostic rate of 68% vs. 45% and 48% for a primary EGD and primary COLO, respectively. The non-diagnostic EGDs [55%] and COLs [52%] underwent a second procedure; the non diagnostic VCEs [32%] did not undergo further procedures. Combining the diagnostic rates for each primary procedure with the cost of performing subsequent procedures [after nondiagnostic primary procedures], demonstrates the use VCE as a primary diagnostic procedure [nZ9] results in a total cost of $12,146 vs. $12,746 and $13,162 for a primary EGD [nZ47] and COLO [nZ33], respectively. In a one way sensitivity analysis, even if the diagnostic rate of VCE reached 100%, requiring a guaranteed second procedure, the cost of performing a primary VCE [$12,445] would still be less than a primary diagnostic EGD [$12,746] or COLO [$13,161]. Combining the diagnostic and non-diagnostic rate of each primary procedure with the time taken to each primary procedure [29 hours to EGD, 38 hours to COLO, 37 hours to VCE], demonstrates the use of VCE as a primary diagnostic procedure in NHGIB patients admitted to the floor would take 74 unit hours to reach a diagnosis compared to 104 and 131 for EGD and COLO, respectively. In a one way sensitivity analysis, the diagnostic rate of EGD and COLO would have to reach 73% and 84%, respectively, to diagnose lesions faster than a primary VCE. Conclusion: Initial use of VCE for the diagnosis of acute NHGIB, may reduce time to diagnosis and management costs.
AB242 GASTROINTESTINAL ENDOSCOPY Volume 81, No. 5S : 2015
Sa1507 Use of the Glasgow-Blatchford Bleeding Score in Bed Assignment for Patients With Suspected UGI Bleeding Drew B. Schembre*1, Janice M. Connolly2, Robson E. Ely1 1 Gastroenterology, Swedish Medical Center, Seattle, WA; 2Hospital Medicine, Swedish Medical Center, Seattle, WA Background: UGI bleeding is a frequent cause for hospitalization. Admission to ICU versus floor bed is usually based on subjective criteria. A more objective tool may help prevent errors in placement that delay needed care or waste resources. The Glasgow-Blatchford bleeding score (GBBS) has been shown to predict hospital mortality and the need for interventions in patients admitted with non-variceal UGI bleeding. Using the GBBS at admission may help stratify patients appropriate for floor care versus intermediate care (IMCU) or ICU. Methods: We reviewed patients admitted with possible UGI bleeding over 32 months to a single tertiary hospital. During the first 16 months, bleeding scores were infrequently calculated. For the last 16 months, scores were calculated by almost all ER physicians. For score !8, floor bed was recommended, 8-12 IMCU and ICU O12. We compared outcomes for patients admitted with and without a GBBS. We also calculated and reviewed GBBS among patients who either died or had admissions O10 days. Results: From October 2011 to June 2014, 456 patients were admitted for presumed UGI bleeding. 281 patients were assigned to beds based on bleeding score and 175 based on physicians’ assessment of risk. Patients stratified by GBBS were more likely to be assigned to IMCU and less likely to go to a floor bed or ICU than those without bleeding scores. Among 68 scored patients placed in floor beds, 7 patients were transferred to a higher level of care and 11 of 140 patients admitted to IMCU were eventually transferred to ICU, compared to 5 of 72 unscored patients admitted to the floor who were later transferred to a higher level of care (pZ0.56) and 13 of 64 IMCU patients transferred to ICU (pZ0.04). Of 31 patients who died or had an admission O10 days, 12 GBBS patients had hospitalizations longer than 10 days and 5 died compared to 8 unscored patients with LOS O10 days and 6 deaths (pZNS). After calculating GBBS for the unscored patients, 27 of 31 (87%) scored 9 or higher and 29 (94%) 7 or higher. No GBBS patient appeared to have died or had a prolonged admission due to a low level of care at admission. One unscored patient admitted to a floor bed, who would have scored 16, died after transfer to ICU and may have benefitted from earlier ICU placement. Conclusion: Use of the GBBS for patients with UGI bleeding may provide more uniform bed assignment and reduce ICU transfers but does not appear to impact LOS, cost or mortality. The great majority of individuals at risk for death or prolonged admission scored O6. The GBBS, along with physician judgment, appears to be a safe and useful tool to direct patient placement. The ideal criteria for bed placement should be determined by larger studies and may vary by facility.
Total Admitted to Floor Bed Admitted to IMCU Admitted to ICU LOS Median Direct Costs Transferred up from Floor Bed Transferred IMCU to ICU Admission O10 Days Death
Patients with GBBS
Patients without GBBS
P value
281 68 (24%) 140 (50%) 73 (26%) 2.75 Days \$5507 7 (11%) 11 (8%) 12 (4%) 5 (2%)
175 72 (41%) 64 (37%) 39 (22%) 2.87 Days \$5481 5 (7%) 13 (20%) 8 (5%) 6 (3%)
!0.01 0.01 0.05 0.82 0.35 0.56 0.04 1.0 0.35
Sa1508 The Role of Ultrathin TransNasal Endoscopy: Meta Analysis Tarek Sawas*, Kia Saeian, Reza Shaker Gastroenterology & Hepatology, Medical College of Wisconsin, Milwaukee, WI Background: Esophagogastroduodenoscopy (EGD) is a common procedure that is usually performed under sedation. The use of sedation improves patients’ tolerance and acceptance. However, It also increases the risk of complications, requires closer monitoring and subsequently increases the cost of the EGD. Ultrathin transnasal endoscopy (T-EGD) was first proposed by Shaker et al. (1993). Since then many randomized controlled trials have evaluated the role of T-EGD in general population and high-risk patients. We report a meta-analysis of prospective controlled trials comparing T-EGD to the conventional regular size sedated and unsedated endoscopy (C-EGD). Method: We searched Pubmed, Embase and Cochrane database for prospective controlled trials that compared T-EGD to regular size C-EGD. Two reviewers selected studies. Quality for each included study was assessed by consort system. No study was eliminated based on quality. Heterogeneity was analyzed by Cochran’s Q statistics. Mantel Haenszel odds ratio (OR) and mean differences (MD) were calculated with fix effect or random effect mode depending on heterogeneity. RevMan 5.2 from Cochrane was used for analysis. Results: 22 prospective controlled trials involving 2921 patients (1626 C-EGD, 1573 T-EGD) were included. Success rate was higher in the C-EGD 98.4% compared to 93.4% in
www.giejournal.org
Abstracts
T-EGD (OR: 0.24, CI: 0.16, 0.36, P!0.001). Failures in T-EGD were due to nasal stenosis. Comfort visual analog scale (VAS) where patients scored their comfort level from 1-10 was not significantly different between the two groups (MD: 0.54, CI: -0.20, 1.28, P: 0.15). Subgroup analysis of patients who underwent unsedated C-EGD showed that T-EGD was significantly more comfortable (MD: 0.94, CI: 0.41, 1.48, P: 0.005). Patients who had T-EGD experienced less gag reflex with VAS MD: -1.78 (CI: -2.83, -0.72, P: 0.001). They also had less nausea and vomiting compared to unsedated C-EGD (MD: -1.73, CI: -2.25, -1.20, P!0.001). Procedure duration was longer in the T-EGD with MD: 74.8 seconds (CI: 50.07, 99.50, P: 0.004). However the recovery duration was shorter in the T-EGD compared to sedated C-EGD (MD -10 min, CI: -10.7, -9.5, P !0.001). Developing hypoxia was not different between T-EGD and unsedated C-EGD (OR: 0.17, CI, 0.02, 1.45, P: 0.11). Patient who underwent T-EGD had lower Heart rate during the procedure (MD -7.16, CI: -11.50, -2.81, P: 0.01) and lower SBP (MD: -7.16, CI: -11.50, -2.81. P: 0.001). 78% in T-EGD group were willing to repeat procedure through same rout VS 58.6% in C-EGD group (OR: 2.25, CI: 1.23, 4.15, P: 0.01). Conclusion: Use of T-EGD is superior to un-sedated C-EGD in term of comfort and should be used as first modality in patients who are at higher risk under sedation. T-EGD should also be offered as an alternative to sedated C-EGD since it decreases recovery duration and subsequently resuming their life quicker.
Sa1509 Financial Analysis of Single Balloon Enteroscopy At a Tertiary Care Center: Where Does the Revenue Come From? Nikhil Banerjee*1, Michael V. Presta2, Matthew T. Charous2, Neil Gupta1 1 Gastroenterology, Loyola University Medical Center, Maywood, IL; 2 Anesthesiology, Loyola University Medical Center, Maywood, IL Background: Resource-intensive endoscopic procedures such as ERCP and EUS have shown to generate more costs than revenue. Lack of financial viability has been mostly in part due to the poor rates of reimbursement under the current payment system in the United States. Double balloon enteroscopy, another resource intensive technology used to evaluate the small bowel has been shown to generate smaller profit margins than predicted upon its implementation in a prior financial analysis. Single balloon enteroscopy (SBE) has never been evaluated for its financial impact at tertiary care hospitals. Aim: 1) Determine the sources of revenue that SBE procedures generate. 2) Determine whether SBE can generate downstream revenues from referrals for the procedure. Methods: All procedures using the SBE system during the first year the equipment was available at a single tertiary care center were identified using the electronic medical record system. Procedure related and financial data in the form of revenue collected were recorded for all SBE procedures. Results were expressed as both total and average standard deviation. Paired and unpaired t-tests were used to determine statistical significance with a p value of ! 0.05 deemed significant. Revenues were analyzed for both physicians and facility as a function of payor type and whether an intervention was performed during the SBE. A subanalysis comparing revenue collected from gastroenterology physician fees and anesthesiology physician fees was also done. Results: 52 single balloon enteroscopy cases were performed between November 1, 2012 and October 31, 2013. Total revenue generated for all SBE procedures was $123,714 including $64,475 dollars from physician fees and $59,239 dollars from the facility fees. Average revenues for the physician and facility fees were $1239506 and $21151923 per case respectively. Revenue generated by anesthesia fees were higher from medicare cases compared to private insurance cases, however revenues from facility fees were higher from private insurance cases compared to medicare cases (Table1 ). Overall, revenues from anesthesiology physician fees were more than revenues from gastroenterology physician fees. This was driven predominately from higher revenues generated from medicare cases and it remained true for both SBE cases that included and did not include endoscopic interventions (Table 2). Only one referred case from outside hospital generated downstream revenue totaling $4727. Conclusion: A larger than expected proportion of revenues generated from SBE cases come in the form of ancillary services provided by anesthesia. Medicare seems to contribute more revenues for these services that private insurance. Projected revenue generation (and it sources) should be considered when establishing a device assisted enteroscopy program.
Sa1510 Clinical Impact of Confocal LASER Endomicroscopy Probe (P - CLE) in the Management of Gastrointestinal Neoplasic and Non - Neoplasic Lesion Carlos Robles-Medranda*, Jesenia Ospina, Miguel Puga-Tejada, Miguel Soria Alcívar, Gladys Bravo Velez, Carlos A. Robles-Jara, Hannah P. Lukashok Endoscopy, Ecuadorian Institute of Digestive Diseases, Guayaquil, Ecuador Background: It has been demonstrated that endoscopic and histopathology findings don’t have good correlation, leading sometimes to diagnostic and therapeutic doubts. Confocal laser endomicroscopy probe (p-CLE) is a new technology which permits in-vivo cellular view of gastrointestinal mucosa. It includes esophagus, stomach, small intestine, colon, biliary tract and pancreas. Studies showed a diagnostic accuracy above 90% for neoplastic (N) and non - neoplastic (NN) lesions, with specific characteristics and classifications. However, few studies have been oriented to determine the clinical impact of p-CLE for diagnosis and management in patients with diagnostic doubts. The aim of this study is to put in evidence the clinical impact of p-CLE in this group of patients. Patients and method: Prospective study, performed from November 2013 to November 2014, in consecutive patients in whom pCLE (CellvizioÒ, Mauna Kea Technology) was indicated due to diagnostic doubts (absence of endoscopic-histological relationship). Baseline characteristics, indications, previous diagnosis studies, findings at p-CLE, clinical management and histopathological outcomes were evaluated. Indications for p-CLE were N and NN lesions. Lesions include: adenomas, dysplasia or cancer located in any gastrointestinal tract level, Barrett’s esophagus, inflammatory bowel disease or pancreatic cysts. Previous diagnostic studies included: high definition magnification with digital chromoendoscopy, ERCP with brushing, EUS. Other studies were: CT-scan, cholangiography by MRI, tumor markers. Interventions based on the findings of p-CLE were analyzed according to the records, and included: drug treatment, other diagnostic studies, endoscopic or surgical treatments. The diagnostic yield was determined thought sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and match measurement. Diagnostic and therapeutic management, redirection of biopsies and need of other diagnostic methods were evaluated. Results: 43 patients were included. 51% were male (22/43) with mean age of 49 19 years. N lesions were 37% (15/43) of cases, located at: stomach (6), esophagus (4), bile duct (3) and colon (2). The sensitivity was 87 % and specificity of 93%,(AUCZ0,9582) with a PPV of 87% and NPV 93%. p value !0.05. The observed concordance was 90.6 % with a Kappa value of 0.80, corresponding to a force of substantial agreement according to Landis & Koch criteria. Changes in diagnostic and therapeutic approach were evident in 39.5% of cases (17/43), directing the sampling in 100% of cases (17/17) and avoiding diagnostic or therapeutic methods in all cases. Conclusion: p-CLE is an essential diagnostic tool for patients with diagnostic doubts, as it allows in-vivo display cell and helps to direct biopsies. It has a significant clinical impact on the diagnosis and treatment of patients
Table 1 Medicare
Private
p value
\$276.40 \$1,175.85 \$1249.63
\$346.34 \$667.07 \$5271.37
0.38 !0.01 !0.01
Gastroenterologist
Anesthesiologist
p value
\$278.23 \$276.40 \$346.34 \$166.06 \$311.88
\$961.67 \$1,175.85 \$667.07 \$1,096.65 \$921.18
!0.01 !0.01 0.06 !0.01 !0.01
Gastroenterologist Revenue Anesthesiologist Revenue Facility Revenue (outpatient only) Revenues are per case (mean)
Table 2 Total Medicare Private No Intervention Intervention Revenues are per case (mean)
www.giejournal.org
Volume 81, No. 5S : 2015 GASTROINTESTINAL ENDOSCOPY AB243