- Email: [email protected]
Development and validation of a web-based electronic application in managing antithrombotic agents in patients undergoing GI endoscopy
ORIGINAL ARTICLE
Development and validation of a web-based electronic application in managing antithrombotic agents in patients undergoing GI endoscopy Venkat Nutalapati, MD,1 Krishna Teja Tokala, BTech,2 Madhav Desai, MD, MPH,1 Vijay Kanakadandi, MD,1 Mojtaba Olyaee, MD,1 Sravanthi Parasa, MD,3 Amit Rastogi, MD1 Kansas City, Kansas; Phoenix, Arizona; Seattle, Washington, USA
Background and Aims: Antithrombotic therapy among patients undergoing GI procedures is frequently encountered and can impact the procedure and patient outcomes. The American Society for Gastrointestinal Endoscopy (ASGE) guidelines help to manage these medications before endoscopy depending on the patient’s clinical status and the type of GI procedure. However, currently there is no readily available electronic tool that can assist in decision-making regarding preprocedural management of these agents. Our aim was to develop an electronic application, endoscopy þ aid (ENDOAID), to help manage antithrombotic agents before endoscopy and to perform a validation study to test its accuracy. Methods: ENDOAID, a web-based application, was developed using JavaScript software (Oracle Corporation, Redwood Shores, Calif, USA) based on an algorithm to categorize patients and procedures into low and high risk as outlined in the updated ASGE guidelines published in 2016. Once pertinent information regarding a patient’s clinical status and the procedure are entered, the application generates recommendations for the management of antithrombotic agents based on their cardiovascular risk and published ASGE guidelines. We performed a validation study with 52 patients who were referred to endoscopy and were taking antithrombotic agents. The patients were divided into groups of 5, and in the simulation each patient had 4 procedures. Different GI procedures were used in the simulation for each group of patients to ensure the entire spectrum of procedures were covered for analysis. Every simulation was then run through ENDOAID. The results from ENDOAID were compared with recommendations based on ASGE guidelines. The latter was derived by consensus between 2 endoscopists (the criterion standard). The personnel using the ENDOAID and those using the ASGE guidelines were different to avoid bias. Any clinical scenario that was unclear or not clearly outlined in ASGE guidelines was discussed with expert endoscopists for a final decision. We evaluated ASGE recommendations and calculated concordance rates between guidelines and ENDOAID results. The Pearson correlation coefficient (r) was calculated to assess the correlation between ENDOAID results to guidelines. Results: There was a total of 208 simulated encounters, including 26 procedures. Initial concordance between ENDOAID recommendations and the criterion standard was seen in 206 encounters (99.03%). The 2 encounters that needed further review occurred among patients with Factor V Leiden mutation and deep vein thrombosis from antiphospholipid antibody syndrome and who were undergoing high-risk procedures that had ambiguous guidelines. ENDOAID suggested consultations with an expert before the elective procedure. This suggestion was agreed on by expert endoscopist consensus. Thus, ENDOAID showed a 100% concordance with the ASGE guideline for managing antithrombotics. There was a high degree of correlation (r Z .996, P < .01) between ENDOAID results with ASGE. Conclusions: We have developed and validated an easy-to-use web-based application that can help in periprocedural management of antithrombotics. Such an application has the potential to simplify the management of these agents and potentially prevent procedural delays, cancellations, or unnecessary consults. (Gastrointest Endosc 2019;-:1-7.)
(footnotes appear on last page of article)
www.giejournal.org
Volume
-,
No.
-
: 2019 GASTROINTESTINAL ENDOSCOPY 1
Web-based electronic application to manage antithrombotic agents before endoscopy
Anticoagulants and antiplatelet agents (APAs), also referred to collectively as antithrombotic agents, are widely prescribed to reduce the risk of thromboembolic events in patients with conditions such as atrial fibrillation, acute coronary syndrome and recent percutaneous coronary intervention, deep vein thrombosis (DVT), hypercoagulable states, and endoprostheses and have established benefits for these patients.1 The rising prevalence and incidence of atrial fibrillation currently requires about 2% of the population in developed countries to be on long-term anticoagulation, with levels reaching 8% to 10% in elderly patients.2 Periprocedural management of antithrombotic agents is a commonly encountered clinical dilemma for gastroenterologists. Annually, 10% of patients on anticoagulants require temporary discontinuation of therapy for surgical or other invasive elective procedures, which in many cases are represented by the GI endoscopy.3 Balancing the risk of periprocedural/postprocedural bleeding with that of a thromboembolic event by the continuation or discontinuation of antithrombotics, respectively, can be a daunting challenge for the endoscopist. Conventional GI procedures are divided into high risk (HR) and low risk (LR) based on their potential to induce bleeding, as per American Society for Gastrointestinal Endoscopy (ASGE) guidelines. HR GI procedures are associated with a higher possibility for bleeding that requires an intervention, such as hospitalization, transfusion, endoscopic treatment, or surgery.4 Similarly, the risk of a cardiovascular event on temporary discontinuation of antithrombotics is also divided into HR and LR. The possibility of a thromboembolic event related to the temporary interruption of an antithrombotic therapy for a GI procedure depends on the indication for the therapy and individual patient characteristics. For example, patients with nonvalvular atrial fibrillation and a CHADSVASC (C, Congestive heart failure history, H, Hypertension history, A, Age, D, Diabetes history, S, Stroke history, Vasc, Vascular disease history) score 2 have a high risk for thromboembolism.5 Furthermore, identifying scenarios at HR for thrombosis, such as patients on APAs for drugeluting coronary stents 12 months, can be challenging.6 Patients with venous thromboembolism and the presence of underlying thrombophilia are at higher risk of future thromboembolic risk.7 Comprehensive knowledge of these different clinical scenarios and the recommendations for managing antithrombotics for this wide array of possibilities is often overwhelming and time-consuming for gastroenterologists. Often, patients are referred back to the prescribing physician to obtain recommendations, thereby delaying procedures and adding to the workload of the prescriber. Procedure cancellations and delays can also result from unnecessary continuation of these agents, which negatively impacts the efficiency of the endoscopy block. ASGE has published evidence-based guidelines that help to manage these medications before endoscopy 2 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2019
Nutalapati et al
depending on the patient’s clinical status and the type of GI procedure.7 This is a current standard of care. However, currently there are no readily available tools or electronic applications that can assist in decision-making regarding periprocedural management of these antithrombotic agents. Resorting to ASGE guidelines for every patient using either a paper or electronic format may cause delays in decision-making. It also complicates the process for nonphysician providers who are triaging patients before endoscopy becasue those providers have to either consult an endoscopist frequently or schedule procedures without careful, efficient planning. We aimed to develop an electronic application, endoscopy þ aid (ENDOAID), to help manage antithrombotic agents before endoscopy and performed a validation study to test its performance and accuracy.
METHODS ENDOAID is an electronic application, currently webbased URL (Uniform Resource Locator) that we have developed using JavaScript software for use on personal computers, laptops, smartphones, iPads, or any electronic devices with Internet capability. ENDOAID was built using data and recommendations provided in the updated ASGE 2016 guidelines for the management of antithrombotic agents in patients undergoing GI endoscopy.7 Because ENDOAID does not require input of patient identifiers, there is no risk of HIPAA violation. ENDOAID uses an algorithm to categorize patients and procedures into LR and HR based on the risk of cardiovascular or thromboembolic events related to the interruption of antithrombotic agents and the risk of bleeding associated with the procedures. Although the procedures are characterized as LR and HR based on the risk of bleeding associated with them, this categorization requires pertinent information regarding a patient’s clinical status and the procedure to be entered into the application. Based on ASGE guidelines, the application includes an array of common and some rare clinical conditions that require patients to be on antithrombotic agents. Along with this, all the different GI procedures listed in the ASGE guidelines were incorporated. The application also includes a list of commonly used antithrombotic agents. After all pertinent information is entered by selecting the different options, the application generates recommendations for the management of antithrombotics (Figs. 1 and 2) Furthermore, while categorizing patients into LR versus HR, the application specifically divides them into risk for venous thromboembolism and risk for cardiovascular events based on their respective clinical conditions. The application generates the following recommendations: continue antithrombotics without interruption during the GI procedure, discontinue all antithrombotics and restart after completion of GI procedure, discontinue www.giejournal.org
Nutalapati et al
Web-based electronic application to manage antithrombotic agents before endoscopy
Figure 1. Pictorial demonstration of home page, clinical conditions, medications page, procedure selection page, and any additional information page. Available at: http://www.endoaid.net/. ENDOAID, endoscopy þ aid.
Figure 2. Recommendations page. Available at: http://www.endoaid.net/patient_report.
all antithrombotics with bridge therapy, and defer the procedure until completion of antithrombotics (except aspirin [ASA]) or after a consultation with an expert based on the procedure and the clinical condition. www.giejournal.org
Finally, the half-life of the respective antithrombotic agents is provided to ensure discontinuation at the appropriate time. All details pertinent to clinical scenario and antithrombotic agents are added to the required Volume
-,
No.
-
: 2019 GASTROINTESTINAL ENDOSCOPY 3
Web-based electronic application to manage antithrombotic agents before endoscopy
TABLE 1. Demographics and comorbidities Characteristic
Value 66.3 12.35
Age, y Gender Male
30/52 (57.7)
Female
22/52 (42.3)
Comorbidities Atrial fibrillation
39/52 (75)
Diabetes mellitus
17/52 (32.7)
Congestive heart failure
6/52 (11.5)
Hypertension
28/52 (53.8)
Deep vein thrombosis
9/52 (17.3)
Pulmonary embolism
5/52 (9.6)
Transient ischemic attack
3/52 (5.7)
Peripheral vascular disease/coronary artery disease
25/52 (48)
Antiphospholipid antibody syndrome
1/52 (1.9)
Factor V Leiden
1/52 (1.9)
Aortic valve replacement
4/52 (7.7)
Drug-eluting stent
6/52 (11.5)
Bare metal stent
1/52 (1.9)
Values are mean standard deviation or n/N (%).
information in the application. The application then generates a specific set of directions for each antithrombotic agent that will lead to recommendations (results) for the clinical question. To demonstrate the accuracy of the application’s recommendations for managing antithrombotics, we performed a pilot validation study. We selected consecutive patients who were on antithrombotic agents and who were referred for various elective GI procedures to the Department of Gastroenterology at the University of Kansas Medical Center between January 2017 and August 2017. After selection, patients were divided in groups of 5. Each patient in a group was simulated to undergo 4 different GI procedures, thereby covering most GI endoscopic procedures. Different GI procedures were used in the simulation for each group of patients to ensure the entire spectrum of procedures was covered to avoid ascertainment bias. Selection of procedures was made initially to cover 2 LR and 2 HR procedures among the 4 different procedures each patient was simulated to undergo. After all GI procedures were included in at least 1 simulation, they were repeated with more emphasis on including HR procedures given their higher risk of bleeding and therefore increased the complexity in the management of antithrombotics. Data input in ENDOAID was made by 2 different investigators (V.N. and M.D.). The results from ENDOAID were compared with recommendations for managing antithrombotic agents based on ASGE guidelines. The latter were manually derived by consensus between 2 endoscopists 4 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2019
Nutalapati et al
(S.P. and V.K.) and were considered the criterion standard against which the output from ENDOAID was compared. The personnel entering the data in the ENDOAID and those manually using ASGE guidelines to generate the recommendations were different to avoid any measurement bias. Any clinical scenario that was unclear, rare, or not readily available in ASGE guidelines were referred for discussion with expert endoscopists for a final decision (M.O. and A.R.). This study was approved with a waiver of consent by the Institutional Review Board at the University of Kansas Medical Center. We evaluated ASGE recommendations and calculated concordance rates between guidelines and ENDOAID results. Pearson’s correlation coefficient (r) was calculated to assess correlation between ASGE recommendations and results from the ENDOAID application. All statistical analysis was performed using Microsoft Excel 2016 (Microsoft, Redmond, Wash, USA).
RESULTS Fifty-two consecutive patients who were referred and scheduled for elective outpatient GI endoscopy procedures while being on anticoagulants, APAs, or both between January 2017 and August 2017 were included in the analysis. The mean age was 66.3 years, with 30 men and 22 women. The population was representative of 13 different comorbidities. Table 1 delineates further demographics and comorbidity details. After dividing patients into groups of 5, there were 208 simulated procedures including 4 for each patient. Twentysix GI procedures were included covering most endoscopic procedures. Details and distribution are presented in Table 2. Patients were on a variety of different antithrombotics, with 5 different anticoagulant agents and 3 different APAs. Their distribution is detailed in Table 3. Of the 208 simulated procedures, 119 (57%) were HR procedures and 89 (43%) were LR procedures. The recommendation regarding management of antithrombotics that was manually derived by consensus between 2 endoscopists (criterion standard) who provided their recommendations based on the ASGE 2016 guidelines are presented in Table 4. Of the 208 simulations, a high concordance rate between the ENDOAID recommendations and the criterion standard was observed in 206 encounters (99.03%). The 2 encounters that needed further review occurred among patients with Factor V Leiden mutation and DVT from APLA and were simulated to undergo HR procedures, which have more ambiguous guidelines given the rarity of the clinical scenarios. ENDOAID suggested a consultation with a hematologist before an elective procedure as an additional recommendation after considering them to be HR for venous thromboembolism. The manual review (by V.N. and M.D.) recommended further discussion with expert endoscopists (A.R. and M.O.). The latter gave www.giejournal.org
Nutalapati et al
Web-based electronic application to manage antithrombotic agents before endoscopy
TABLE 2. Distribution of procedures
Procedures
TABLE 3. Classifications and specific medications No. of simulations
Low risk (n Z 89)
n/N (%) Antiplatelets only
5/52 (9.6)
Antiplatelets and anticoagulants
18/52 (34.6) 29/52 (55.7)
EGD
10
Anticoagulants only
Barrett’s ablation
10
Rivaroxaban (Janssen Pharmaceuticals, Titusville, NJ)
13/52 (25)
EUS without FNA
5
3/52 (5.8)
Argon plasma coagulation
10
Dabigatran (Boehringer Ingelheim Pharmaceuticals, Ingelheim am Rhein, Germany)
Colonoscopy with biopsy sampling
10
Coumadin (Bristol-Myers Squibb, New York, NY)
15/52 (29)
Flexible sigmoidoscopy
10
Apixaban (Bristol-Myers Squibb, New York, NY)
11/52 (21)
Diagnostic balloon-assisted enteroscopy
5
Enoxaparin (Sanofi-Aventis, Bridgewater, NJ)
3/52 (5.8)
Push enteroscopy
5
Clopidogrel (Bristol-Myers Squibb, New York, NY)
7/52 (13.4)
Capsule endoscopy
6
Ticagrelor (AstraZeneca, Cambridge, UK)
2/52 (3.8)
ERCP/papillary balloon dilation without sphincterotomy
5
Aspirin
ERCP/stent (biliary) without sphincterotomy
5
ERCP/stent (pancreatic) without sphincterotomy
5
17/52 (32.7)
TABLE 4. Procedure and recommendations based on ASGE guidelines
High risk (n Z 119) Bougie dilation
7
Procedure and recommendations
ESD
9
High-risk procedures
119/208 (57)
EMR
11
Low-risk procedures
89/208 (43)
Colonoscopy with polypectomy
11
Continue antithrombotic
88/208 (42)
ERCP w/sphincterotomy
11 10
Discontinue anticoagulation without bridge
27/208 (13)
Variceal bleeding management PEG
10
Discontinue anticoagulation with bridge
77/208 (37)
Consult specialist/defer until completion of antithrombotic therapy
16/208 (7.7)
PEJ
5
Pneumatic dilation
10
Therapeutic balloon-assisted enteroscopy
10
EUS/FNA
10
Tumor ablation
5
Cystogastrostomy
6
Ampullary resection
5
ESD, Endoscopic submucosal dissection; PEJ, percutaneous endoscopic jejunostomy.
a similar recommendation of consultation with a hematology expert also. Finally, there was 100% concordance between the management of antithrombotics recommended by ENDOAID and the manual recommendations by ASGE guidelines. Therefore, the accuracy of ENDOAID in this set of patients undergoing various endoscopic procedures was 100%. Correlation coefficient (r) was .996 with P < .01 suggestive of a high degree of correlation between the ASGE recommendations and results derived from the application.
DISCUSSION The accurate recommendations for managing antithrombotics before endoscopic procedures relates to balancing the risk of peri- and postprocedural bleeding www.giejournal.org
n/N (%)
ASGE recommendations were derived manually. ASGE, American Society for Gastrointestinal Endoscopy.
versus the risk of thromboembolic events with the continuation and interruption of therapy, respectively. For this purpose the endoscopic procedures are divided into LR and HR for bleeding, whereas patients are divided into HR and LR of a thromboembolic event when antithrombotic therapy is interrupted. Over the years, major societies like the ASGE, the British Society of Gastroenterology, and the European Society of Gastrointestinal Endoscopy have published guidelines for managing patients on antithrombotics undergoing endoscopic procedures. For the endoscopist, remembering the details and intricacies of these guidelines can be cumbersome and difficult, thereby making accurate recommendations a labor-intensive process. Given recent advances and widespread availability of smartphones and the Internet, innovative ways to implement guidelines need to be developed. In the field of endoscopy, data show that bowel preparation for colonoscopy can be improved with a smartphone application.8 Similarly, if a computer-based application can make Volume
-,
No.
-
: 2019 GASTROINTESTINAL ENDOSCOPY 5
Web-based electronic application to manage antithrombotic agents before endoscopy
Nutalapati et al
TABLE 5. Concordance between ASGE recommendations and ENDOAID ASGE recommendations ENDOAID
Continue antithrombotic Continue antithrombotic
Discontinue anticoagulation without bridge
Discontinue anticoagulation Consult specialist/defer until with bridge completion of antithrombotic therapy
88/88 (100)
Discontinue anticoagulation without bridge
27/27 (100)
Discontinue anticoagulation with bridge
75/77 (97.4) ENDOAID added consultation recommendation in patients with antiphospholipid antibody syndrome and Factor V Leiden
Consult specialist/defer until completion of antithrombotic therapy
16/16 (100)
Values are n/N (%). ASGE, American Society for Gastrointestinal Endoscopy; ENDOAID, endoscopy þ aid.
accurate recommendations for managing antithrombotics in an automated fashion, then this task can become less taxing. We used the most recent ASGE recommendations for the development of the ENDOAID application. ENDOAID as an application was developed to ensure easy applicability and availability. The recommendations made by ENDOAID showed an initial concordance of 99% with those made by 2 endoscopists following the ASGE guidelines (Table 5). The scenarios where the recommendations needed further discussion occurred among patients with Factor V Leiden and patients with DVT and APLA simulated undergoing HR procedures. These procedures were EMR for the patient with Factor V Leiden and ampullary resection for the patient with APLA. ASGE recommended considering these conditions as medium risk/HR for venous thromboembolism. ENDOAID provided recommendations to consult with hematologists in addition to ASGE guidelines given the rarity of the scenarios and lack of specific knowledge regarding these conditions and recommendations.7 This recommendation was agreed on after consultation with 2 expert endoscopists, thereby yielding a 100% concordance between ENDOAID recommendations and those derived manually following ASGE guidelines. In addition, the high correlation between guideline recommendations and ENDOAID results further validates the findings. A special mention must be made regarding patients with conditions such as DVT/pulmonary embolism 3 months or patients with drug-eluting stents 12 months and those with bare metal stents <1 month. In these situations ENDOAID recommended to defer the HR procedure until completion of the antithrombotic regimen and consult with a specific specialist, such as a cardiologist, regarding periprocedural management of these agents if there is clinical necessity to perform the procedure. For patients with 6 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2019
rare clinical conditions (eg, APLA, protein C/S deficiency, or Factor V Leiden) with unclear procedure recommendations, ENDOAID generates the recommendation for consultation with the necessary specialty like hematology to prevent delays and adverse events.7 These specialists are better versed with the nuances of these clinical conditions than a gastroenterologist, and their input is invaluable. ENDOAID is limited in that it is a reflection of the guidelines on which the application was built. A limitation of building an application from such guidelines is the challenge of categorizing some procedures to a certain risk level. For example, PEG tube is listed under HR procedure in the ASGE guidelines with the caveat that it is LR if the patient is on single APA, either aspirin or thienopyridines. In this iteration of the application we considered PEG tube as a HR procedure based on further documentation in the ASGE guidelines that recommended discontinuation of thienopyridines at least 5 to 7 days before HR endoscopic procedure or switching to ASA monotherapy and continuing until the thienopyridine can be safely resumed. Furthermore, most patients on thienopyridines are usually also on aspirin therapy. ASGE does not categorize every procedure (eg, enteral stent placement, ERCP with ductoscopy), which led to some procedure exclusions. Finally, we were unable to input all possible clinical scenarios, such as patients with cancer, because giving accurate recommendations in such scenarios would need extensive clinical history input, making the application usage cumbersome. The application may also need updating whenever newer iterations of AGSE guidelines are published. The application cannot completely incorporate and adjudicate on all rare clinical scenarios and cannot be used in emergent situations. Further validation of the application needs to be performed with the help of prospective studies. www.giejournal.org
Nutalapati et al
Web-based electronic application to manage antithrombotic agents before endoscopy
Future updates can include incorporation of medication stoppage based on the creatinine clearance as well as any changes in future recommendations by ASGE. The use of the ENDOAID application is based on access to electronic devices and Internet service. Because most hospitals and providers based in the United States have access to such resources, this limitation is redundant but should still be considered. A similar application was developed by James et al9 for periprocedural management of antithrombotics. However, their study was a quality improvement initiative for fellow education. It was not a validation study like ours where we assessed the concordance and accuracy of the application recommendations with ASGE guidelines. In conclusion, we successfully developed and validated an electronic application that can help in periprocedural management of antithrombotics. ENDOAID can easily be used on multiple devices, such as personal computers and smartphones. It is a user-friendly application with the potential to simplify the management of antithrombotic agents. If the accuracy observed in this study is further validated by others, the application has the potential to be more broadly used for the management of antithrombotics. It may even be used by ancillary staff at the time of scheduling endoscopic procedures to assist with decisions regarding management of antithrombotic agents.
4. Eisen GM, Baron TH, Dominitz JA, et al. Guideline on the management of anticoagulation and antiplatelet therapy for endoscopic procedures. Gastrointest Endosc 2002;55:775-9. 5. Lip GY, Nieuwlaat R, Pisters R, et al. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the Euro Heart Survey on atrial fibrillation. Chest 2010;137:263-72. 6. Jneid H, Anderson JL, Wright RS, et al. 2012 ACCF/AHA focused update of the guideline for the management of patients with unstable angina/non-ST-elevation myocardial infarction (updating the 2007 guideline and replacing the 2011 focused update): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2012;60:645-81. 7. ASGE Standards of Practice Committee; Acosta RD, Abraham NS, Chandrasekhara V, et al. The management of antithrombotic agents for patients undergoing GI endoscopy. Gastrointest Endosc 2016;83: 3-16. 8. Desai M, Nutalapati V, Bansal A, et al. Use of smart phone application to improve quality of bowel prep for colonoscopy: a systematic review and meta-analysis. Endosc Int Open 2019;7:E216-24. 9. James TW, Strauss AT, Mathews SC. Fellow education improved through mobile clinical decision support application: a multi-center approach involving peri-procedural antithrombotic use. Gastroenterology 2018;155:2014-5.
Abbreviations: APA, antiplatelet agent; APLA, antiphospholipid antibody syndrome; ASGE, American Society for Gastrointestinal Endoscopy; DVT, deep vein thrombosis; ENDOAID, endoscopy þ aid; HR, high risk; LR, low risk. DISCLOSURE: All authors disclosed no financial relationships relevant to this publication.
REFERENCES 1. Holbrook A, Schulman S, Witt DM, et al. Evidence-based management of anticoagulant therapy: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidencebased clinical practice guidelines. Chest 2012;141:e152S-84S. 2. Chugh SS, Havmoeller R, Narayanan K, et al. Worldwide epidemiology of atrial fibrillation: a Global Burden of Disease 2010 Study. Circulation 2014;129:837-47. 3. Douketis JD, Spyropoulos AC, Spencer FA, et al. Perioperative management of antithrombotic therapy: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 2012;141:e326S-50S.
www.giejournal.org
Copyright ª 2019 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 https://doi.org/10.1016/j.gie.2019.06.015 Received February 10, 2019. Accepted June 7, 2019. Current affiliations: Department of Gastroenterology, University of Kansas Medical Center, Kansas City, Kansas, USA (1), Department of Computer Sciences, Clairvoyant LLC, Phoenix, Arizona, USA (2), Department of Gastroenterology, Swedish Medical Center, Seattle, Washington, USA (3). Reprint requests: Amit Rastogi, MD, Department of Gastroenterology, University of Kansas Medical Center, Kansas City, KS 66160. Tel: 816-8614700.
Volume
-,
No.
-
: 2019 GASTROINTESTINAL ENDOSCOPY 7
Development and validation of a web-based electronic application in managing antithrombotic agents in patients undergoing GI endoscopy Venkat Nutalapati, MD,1 Krishna Teja Tokala, BTech,2 Madhav Desai, MD, MPH,1 Vijay Kanakadandi, MD,1 Mojtaba Olyaee, MD,1 Sravanthi Parasa, MD,3 Amit Rastogi, MD1 Kansas City, Kansas; Phoenix, Arizona; Seattle, Washington, USA
Background and Aims: Antithrombotic therapy among patients undergoing GI procedures is frequently encountered and can impact the procedure and patient outcomes. The American Society for Gastrointestinal Endoscopy (ASGE) guidelines help to manage these medications before endoscopy depending on the patient’s clinical status and the type of GI procedure. However, currently there is no readily available electronic tool that can assist in decision-making regarding preprocedural management of these agents. Our aim was to develop an electronic application, endoscopy þ aid (ENDOAID), to help manage antithrombotic agents before endoscopy and to perform a validation study to test its accuracy. Methods: ENDOAID, a web-based application, was developed using JavaScript software (Oracle Corporation, Redwood Shores, Calif, USA) based on an algorithm to categorize patients and procedures into low and high risk as outlined in the updated ASGE guidelines published in 2016. Once pertinent information regarding a patient’s clinical status and the procedure are entered, the application generates recommendations for the management of antithrombotic agents based on their cardiovascular risk and published ASGE guidelines. We performed a validation study with 52 patients who were referred to endoscopy and were taking antithrombotic agents. The patients were divided into groups of 5, and in the simulation each patient had 4 procedures. Different GI procedures were used in the simulation for each group of patients to ensure the entire spectrum of procedures were covered for analysis. Every simulation was then run through ENDOAID. The results from ENDOAID were compared with recommendations based on ASGE guidelines. The latter was derived by consensus between 2 endoscopists (the criterion standard). The personnel using the ENDOAID and those using the ASGE guidelines were different to avoid bias. Any clinical scenario that was unclear or not clearly outlined in ASGE guidelines was discussed with expert endoscopists for a final decision. We evaluated ASGE recommendations and calculated concordance rates between guidelines and ENDOAID results. The Pearson correlation coefficient (r) was calculated to assess the correlation between ENDOAID results to guidelines. Results: There was a total of 208 simulated encounters, including 26 procedures. Initial concordance between ENDOAID recommendations and the criterion standard was seen in 206 encounters (99.03%). The 2 encounters that needed further review occurred among patients with Factor V Leiden mutation and deep vein thrombosis from antiphospholipid antibody syndrome and who were undergoing high-risk procedures that had ambiguous guidelines. ENDOAID suggested consultations with an expert before the elective procedure. This suggestion was agreed on by expert endoscopist consensus. Thus, ENDOAID showed a 100% concordance with the ASGE guideline for managing antithrombotics. There was a high degree of correlation (r Z .996, P < .01) between ENDOAID results with ASGE. Conclusions: We have developed and validated an easy-to-use web-based application that can help in periprocedural management of antithrombotics. Such an application has the potential to simplify the management of these agents and potentially prevent procedural delays, cancellations, or unnecessary consults. (Gastrointest Endosc 2019;-:1-7.)
(footnotes appear on last page of article)
www.giejournal.org
Volume
-,
No.
-
: 2019 GASTROINTESTINAL ENDOSCOPY 1
Web-based electronic application to manage antithrombotic agents before endoscopy
Anticoagulants and antiplatelet agents (APAs), also referred to collectively as antithrombotic agents, are widely prescribed to reduce the risk of thromboembolic events in patients with conditions such as atrial fibrillation, acute coronary syndrome and recent percutaneous coronary intervention, deep vein thrombosis (DVT), hypercoagulable states, and endoprostheses and have established benefits for these patients.1 The rising prevalence and incidence of atrial fibrillation currently requires about 2% of the population in developed countries to be on long-term anticoagulation, with levels reaching 8% to 10% in elderly patients.2 Periprocedural management of antithrombotic agents is a commonly encountered clinical dilemma for gastroenterologists. Annually, 10% of patients on anticoagulants require temporary discontinuation of therapy for surgical or other invasive elective procedures, which in many cases are represented by the GI endoscopy.3 Balancing the risk of periprocedural/postprocedural bleeding with that of a thromboembolic event by the continuation or discontinuation of antithrombotics, respectively, can be a daunting challenge for the endoscopist. Conventional GI procedures are divided into high risk (HR) and low risk (LR) based on their potential to induce bleeding, as per American Society for Gastrointestinal Endoscopy (ASGE) guidelines. HR GI procedures are associated with a higher possibility for bleeding that requires an intervention, such as hospitalization, transfusion, endoscopic treatment, or surgery.4 Similarly, the risk of a cardiovascular event on temporary discontinuation of antithrombotics is also divided into HR and LR. The possibility of a thromboembolic event related to the temporary interruption of an antithrombotic therapy for a GI procedure depends on the indication for the therapy and individual patient characteristics. For example, patients with nonvalvular atrial fibrillation and a CHADSVASC (C, Congestive heart failure history, H, Hypertension history, A, Age, D, Diabetes history, S, Stroke history, Vasc, Vascular disease history) score 2 have a high risk for thromboembolism.5 Furthermore, identifying scenarios at HR for thrombosis, such as patients on APAs for drugeluting coronary stents 12 months, can be challenging.6 Patients with venous thromboembolism and the presence of underlying thrombophilia are at higher risk of future thromboembolic risk.7 Comprehensive knowledge of these different clinical scenarios and the recommendations for managing antithrombotics for this wide array of possibilities is often overwhelming and time-consuming for gastroenterologists. Often, patients are referred back to the prescribing physician to obtain recommendations, thereby delaying procedures and adding to the workload of the prescriber. Procedure cancellations and delays can also result from unnecessary continuation of these agents, which negatively impacts the efficiency of the endoscopy block. ASGE has published evidence-based guidelines that help to manage these medications before endoscopy 2 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2019
Nutalapati et al
depending on the patient’s clinical status and the type of GI procedure.7 This is a current standard of care. However, currently there are no readily available tools or electronic applications that can assist in decision-making regarding periprocedural management of these antithrombotic agents. Resorting to ASGE guidelines for every patient using either a paper or electronic format may cause delays in decision-making. It also complicates the process for nonphysician providers who are triaging patients before endoscopy becasue those providers have to either consult an endoscopist frequently or schedule procedures without careful, efficient planning. We aimed to develop an electronic application, endoscopy þ aid (ENDOAID), to help manage antithrombotic agents before endoscopy and performed a validation study to test its performance and accuracy.
METHODS ENDOAID is an electronic application, currently webbased URL (Uniform Resource Locator) that we have developed using JavaScript software for use on personal computers, laptops, smartphones, iPads, or any electronic devices with Internet capability. ENDOAID was built using data and recommendations provided in the updated ASGE 2016 guidelines for the management of antithrombotic agents in patients undergoing GI endoscopy.7 Because ENDOAID does not require input of patient identifiers, there is no risk of HIPAA violation. ENDOAID uses an algorithm to categorize patients and procedures into LR and HR based on the risk of cardiovascular or thromboembolic events related to the interruption of antithrombotic agents and the risk of bleeding associated with the procedures. Although the procedures are characterized as LR and HR based on the risk of bleeding associated with them, this categorization requires pertinent information regarding a patient’s clinical status and the procedure to be entered into the application. Based on ASGE guidelines, the application includes an array of common and some rare clinical conditions that require patients to be on antithrombotic agents. Along with this, all the different GI procedures listed in the ASGE guidelines were incorporated. The application also includes a list of commonly used antithrombotic agents. After all pertinent information is entered by selecting the different options, the application generates recommendations for the management of antithrombotics (Figs. 1 and 2) Furthermore, while categorizing patients into LR versus HR, the application specifically divides them into risk for venous thromboembolism and risk for cardiovascular events based on their respective clinical conditions. The application generates the following recommendations: continue antithrombotics without interruption during the GI procedure, discontinue all antithrombotics and restart after completion of GI procedure, discontinue www.giejournal.org
Nutalapati et al
Web-based electronic application to manage antithrombotic agents before endoscopy
Figure 1. Pictorial demonstration of home page, clinical conditions, medications page, procedure selection page, and any additional information page. Available at: http://www.endoaid.net/. ENDOAID, endoscopy þ aid.
Figure 2. Recommendations page. Available at: http://www.endoaid.net/patient_report.
all antithrombotics with bridge therapy, and defer the procedure until completion of antithrombotics (except aspirin [ASA]) or after a consultation with an expert based on the procedure and the clinical condition. www.giejournal.org
Finally, the half-life of the respective antithrombotic agents is provided to ensure discontinuation at the appropriate time. All details pertinent to clinical scenario and antithrombotic agents are added to the required Volume
-,
No.
-
: 2019 GASTROINTESTINAL ENDOSCOPY 3
Web-based electronic application to manage antithrombotic agents before endoscopy
TABLE 1. Demographics and comorbidities Characteristic
Value 66.3 12.35
Age, y Gender Male
30/52 (57.7)
Female
22/52 (42.3)
Comorbidities Atrial fibrillation
39/52 (75)
Diabetes mellitus
17/52 (32.7)
Congestive heart failure
6/52 (11.5)
Hypertension
28/52 (53.8)
Deep vein thrombosis
9/52 (17.3)
Pulmonary embolism
5/52 (9.6)
Transient ischemic attack
3/52 (5.7)
Peripheral vascular disease/coronary artery disease
25/52 (48)
Antiphospholipid antibody syndrome
1/52 (1.9)
Factor V Leiden
1/52 (1.9)
Aortic valve replacement
4/52 (7.7)
Drug-eluting stent
6/52 (11.5)
Bare metal stent
1/52 (1.9)
Values are mean standard deviation or n/N (%).
information in the application. The application then generates a specific set of directions for each antithrombotic agent that will lead to recommendations (results) for the clinical question. To demonstrate the accuracy of the application’s recommendations for managing antithrombotics, we performed a pilot validation study. We selected consecutive patients who were on antithrombotic agents and who were referred for various elective GI procedures to the Department of Gastroenterology at the University of Kansas Medical Center between January 2017 and August 2017. After selection, patients were divided in groups of 5. Each patient in a group was simulated to undergo 4 different GI procedures, thereby covering most GI endoscopic procedures. Different GI procedures were used in the simulation for each group of patients to ensure the entire spectrum of procedures was covered to avoid ascertainment bias. Selection of procedures was made initially to cover 2 LR and 2 HR procedures among the 4 different procedures each patient was simulated to undergo. After all GI procedures were included in at least 1 simulation, they were repeated with more emphasis on including HR procedures given their higher risk of bleeding and therefore increased the complexity in the management of antithrombotics. Data input in ENDOAID was made by 2 different investigators (V.N. and M.D.). The results from ENDOAID were compared with recommendations for managing antithrombotic agents based on ASGE guidelines. The latter were manually derived by consensus between 2 endoscopists 4 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2019
Nutalapati et al
(S.P. and V.K.) and were considered the criterion standard against which the output from ENDOAID was compared. The personnel entering the data in the ENDOAID and those manually using ASGE guidelines to generate the recommendations were different to avoid any measurement bias. Any clinical scenario that was unclear, rare, or not readily available in ASGE guidelines were referred for discussion with expert endoscopists for a final decision (M.O. and A.R.). This study was approved with a waiver of consent by the Institutional Review Board at the University of Kansas Medical Center. We evaluated ASGE recommendations and calculated concordance rates between guidelines and ENDOAID results. Pearson’s correlation coefficient (r) was calculated to assess correlation between ASGE recommendations and results from the ENDOAID application. All statistical analysis was performed using Microsoft Excel 2016 (Microsoft, Redmond, Wash, USA).
RESULTS Fifty-two consecutive patients who were referred and scheduled for elective outpatient GI endoscopy procedures while being on anticoagulants, APAs, or both between January 2017 and August 2017 were included in the analysis. The mean age was 66.3 years, with 30 men and 22 women. The population was representative of 13 different comorbidities. Table 1 delineates further demographics and comorbidity details. After dividing patients into groups of 5, there were 208 simulated procedures including 4 for each patient. Twentysix GI procedures were included covering most endoscopic procedures. Details and distribution are presented in Table 2. Patients were on a variety of different antithrombotics, with 5 different anticoagulant agents and 3 different APAs. Their distribution is detailed in Table 3. Of the 208 simulated procedures, 119 (57%) were HR procedures and 89 (43%) were LR procedures. The recommendation regarding management of antithrombotics that was manually derived by consensus between 2 endoscopists (criterion standard) who provided their recommendations based on the ASGE 2016 guidelines are presented in Table 4. Of the 208 simulations, a high concordance rate between the ENDOAID recommendations and the criterion standard was observed in 206 encounters (99.03%). The 2 encounters that needed further review occurred among patients with Factor V Leiden mutation and DVT from APLA and were simulated to undergo HR procedures, which have more ambiguous guidelines given the rarity of the clinical scenarios. ENDOAID suggested a consultation with a hematologist before an elective procedure as an additional recommendation after considering them to be HR for venous thromboembolism. The manual review (by V.N. and M.D.) recommended further discussion with expert endoscopists (A.R. and M.O.). The latter gave www.giejournal.org
Nutalapati et al
Web-based electronic application to manage antithrombotic agents before endoscopy
TABLE 2. Distribution of procedures
Procedures
TABLE 3. Classifications and specific medications No. of simulations
Low risk (n Z 89)
n/N (%) Antiplatelets only
5/52 (9.6)
Antiplatelets and anticoagulants
18/52 (34.6) 29/52 (55.7)
EGD
10
Anticoagulants only
Barrett’s ablation
10
Rivaroxaban (Janssen Pharmaceuticals, Titusville, NJ)
13/52 (25)
EUS without FNA
5
3/52 (5.8)
Argon plasma coagulation
10
Dabigatran (Boehringer Ingelheim Pharmaceuticals, Ingelheim am Rhein, Germany)
Colonoscopy with biopsy sampling
10
Coumadin (Bristol-Myers Squibb, New York, NY)
15/52 (29)
Flexible sigmoidoscopy
10
Apixaban (Bristol-Myers Squibb, New York, NY)
11/52 (21)
Diagnostic balloon-assisted enteroscopy
5
Enoxaparin (Sanofi-Aventis, Bridgewater, NJ)
3/52 (5.8)
Push enteroscopy
5
Clopidogrel (Bristol-Myers Squibb, New York, NY)
7/52 (13.4)
Capsule endoscopy
6
Ticagrelor (AstraZeneca, Cambridge, UK)
2/52 (3.8)
ERCP/papillary balloon dilation without sphincterotomy
5
Aspirin
ERCP/stent (biliary) without sphincterotomy
5
ERCP/stent (pancreatic) without sphincterotomy
5
17/52 (32.7)
TABLE 4. Procedure and recommendations based on ASGE guidelines
High risk (n Z 119) Bougie dilation
7
Procedure and recommendations
ESD
9
High-risk procedures
119/208 (57)
EMR
11
Low-risk procedures
89/208 (43)
Colonoscopy with polypectomy
11
Continue antithrombotic
88/208 (42)
ERCP w/sphincterotomy
11 10
Discontinue anticoagulation without bridge
27/208 (13)
Variceal bleeding management PEG
10
Discontinue anticoagulation with bridge
77/208 (37)
Consult specialist/defer until completion of antithrombotic therapy
16/208 (7.7)
PEJ
5
Pneumatic dilation
10
Therapeutic balloon-assisted enteroscopy
10
EUS/FNA
10
Tumor ablation
5
Cystogastrostomy
6
Ampullary resection
5
ESD, Endoscopic submucosal dissection; PEJ, percutaneous endoscopic jejunostomy.
a similar recommendation of consultation with a hematology expert also. Finally, there was 100% concordance between the management of antithrombotics recommended by ENDOAID and the manual recommendations by ASGE guidelines. Therefore, the accuracy of ENDOAID in this set of patients undergoing various endoscopic procedures was 100%. Correlation coefficient (r) was .996 with P < .01 suggestive of a high degree of correlation between the ASGE recommendations and results derived from the application.
DISCUSSION The accurate recommendations for managing antithrombotics before endoscopic procedures relates to balancing the risk of peri- and postprocedural bleeding www.giejournal.org
n/N (%)
ASGE recommendations were derived manually. ASGE, American Society for Gastrointestinal Endoscopy.
versus the risk of thromboembolic events with the continuation and interruption of therapy, respectively. For this purpose the endoscopic procedures are divided into LR and HR for bleeding, whereas patients are divided into HR and LR of a thromboembolic event when antithrombotic therapy is interrupted. Over the years, major societies like the ASGE, the British Society of Gastroenterology, and the European Society of Gastrointestinal Endoscopy have published guidelines for managing patients on antithrombotics undergoing endoscopic procedures. For the endoscopist, remembering the details and intricacies of these guidelines can be cumbersome and difficult, thereby making accurate recommendations a labor-intensive process. Given recent advances and widespread availability of smartphones and the Internet, innovative ways to implement guidelines need to be developed. In the field of endoscopy, data show that bowel preparation for colonoscopy can be improved with a smartphone application.8 Similarly, if a computer-based application can make Volume
-,
No.
-
: 2019 GASTROINTESTINAL ENDOSCOPY 5
Web-based electronic application to manage antithrombotic agents before endoscopy
Nutalapati et al
TABLE 5. Concordance between ASGE recommendations and ENDOAID ASGE recommendations ENDOAID
Continue antithrombotic Continue antithrombotic
Discontinue anticoagulation without bridge
Discontinue anticoagulation Consult specialist/defer until with bridge completion of antithrombotic therapy
88/88 (100)
Discontinue anticoagulation without bridge
27/27 (100)
Discontinue anticoagulation with bridge
75/77 (97.4) ENDOAID added consultation recommendation in patients with antiphospholipid antibody syndrome and Factor V Leiden
Consult specialist/defer until completion of antithrombotic therapy
16/16 (100)
Values are n/N (%). ASGE, American Society for Gastrointestinal Endoscopy; ENDOAID, endoscopy þ aid.
accurate recommendations for managing antithrombotics in an automated fashion, then this task can become less taxing. We used the most recent ASGE recommendations for the development of the ENDOAID application. ENDOAID as an application was developed to ensure easy applicability and availability. The recommendations made by ENDOAID showed an initial concordance of 99% with those made by 2 endoscopists following the ASGE guidelines (Table 5). The scenarios where the recommendations needed further discussion occurred among patients with Factor V Leiden and patients with DVT and APLA simulated undergoing HR procedures. These procedures were EMR for the patient with Factor V Leiden and ampullary resection for the patient with APLA. ASGE recommended considering these conditions as medium risk/HR for venous thromboembolism. ENDOAID provided recommendations to consult with hematologists in addition to ASGE guidelines given the rarity of the scenarios and lack of specific knowledge regarding these conditions and recommendations.7 This recommendation was agreed on after consultation with 2 expert endoscopists, thereby yielding a 100% concordance between ENDOAID recommendations and those derived manually following ASGE guidelines. In addition, the high correlation between guideline recommendations and ENDOAID results further validates the findings. A special mention must be made regarding patients with conditions such as DVT/pulmonary embolism 3 months or patients with drug-eluting stents 12 months and those with bare metal stents <1 month. In these situations ENDOAID recommended to defer the HR procedure until completion of the antithrombotic regimen and consult with a specific specialist, such as a cardiologist, regarding periprocedural management of these agents if there is clinical necessity to perform the procedure. For patients with 6 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2019
rare clinical conditions (eg, APLA, protein C/S deficiency, or Factor V Leiden) with unclear procedure recommendations, ENDOAID generates the recommendation for consultation with the necessary specialty like hematology to prevent delays and adverse events.7 These specialists are better versed with the nuances of these clinical conditions than a gastroenterologist, and their input is invaluable. ENDOAID is limited in that it is a reflection of the guidelines on which the application was built. A limitation of building an application from such guidelines is the challenge of categorizing some procedures to a certain risk level. For example, PEG tube is listed under HR procedure in the ASGE guidelines with the caveat that it is LR if the patient is on single APA, either aspirin or thienopyridines. In this iteration of the application we considered PEG tube as a HR procedure based on further documentation in the ASGE guidelines that recommended discontinuation of thienopyridines at least 5 to 7 days before HR endoscopic procedure or switching to ASA monotherapy and continuing until the thienopyridine can be safely resumed. Furthermore, most patients on thienopyridines are usually also on aspirin therapy. ASGE does not categorize every procedure (eg, enteral stent placement, ERCP with ductoscopy), which led to some procedure exclusions. Finally, we were unable to input all possible clinical scenarios, such as patients with cancer, because giving accurate recommendations in such scenarios would need extensive clinical history input, making the application usage cumbersome. The application may also need updating whenever newer iterations of AGSE guidelines are published. The application cannot completely incorporate and adjudicate on all rare clinical scenarios and cannot be used in emergent situations. Further validation of the application needs to be performed with the help of prospective studies. www.giejournal.org
Nutalapati et al
Web-based electronic application to manage antithrombotic agents before endoscopy
Future updates can include incorporation of medication stoppage based on the creatinine clearance as well as any changes in future recommendations by ASGE. The use of the ENDOAID application is based on access to electronic devices and Internet service. Because most hospitals and providers based in the United States have access to such resources, this limitation is redundant but should still be considered. A similar application was developed by James et al9 for periprocedural management of antithrombotics. However, their study was a quality improvement initiative for fellow education. It was not a validation study like ours where we assessed the concordance and accuracy of the application recommendations with ASGE guidelines. In conclusion, we successfully developed and validated an electronic application that can help in periprocedural management of antithrombotics. ENDOAID can easily be used on multiple devices, such as personal computers and smartphones. It is a user-friendly application with the potential to simplify the management of antithrombotic agents. If the accuracy observed in this study is further validated by others, the application has the potential to be more broadly used for the management of antithrombotics. It may even be used by ancillary staff at the time of scheduling endoscopic procedures to assist with decisions regarding management of antithrombotic agents.
4. Eisen GM, Baron TH, Dominitz JA, et al. Guideline on the management of anticoagulation and antiplatelet therapy for endoscopic procedures. Gastrointest Endosc 2002;55:775-9. 5. Lip GY, Nieuwlaat R, Pisters R, et al. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the Euro Heart Survey on atrial fibrillation. Chest 2010;137:263-72. 6. Jneid H, Anderson JL, Wright RS, et al. 2012 ACCF/AHA focused update of the guideline for the management of patients with unstable angina/non-ST-elevation myocardial infarction (updating the 2007 guideline and replacing the 2011 focused update): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2012;60:645-81. 7. ASGE Standards of Practice Committee; Acosta RD, Abraham NS, Chandrasekhara V, et al. The management of antithrombotic agents for patients undergoing GI endoscopy. Gastrointest Endosc 2016;83: 3-16. 8. Desai M, Nutalapati V, Bansal A, et al. Use of smart phone application to improve quality of bowel prep for colonoscopy: a systematic review and meta-analysis. Endosc Int Open 2019;7:E216-24. 9. James TW, Strauss AT, Mathews SC. Fellow education improved through mobile clinical decision support application: a multi-center approach involving peri-procedural antithrombotic use. Gastroenterology 2018;155:2014-5.
Abbreviations: APA, antiplatelet agent; APLA, antiphospholipid antibody syndrome; ASGE, American Society for Gastrointestinal Endoscopy; DVT, deep vein thrombosis; ENDOAID, endoscopy þ aid; HR, high risk; LR, low risk. DISCLOSURE: All authors disclosed no financial relationships relevant to this publication.
REFERENCES 1. Holbrook A, Schulman S, Witt DM, et al. Evidence-based management of anticoagulant therapy: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidencebased clinical practice guidelines. Chest 2012;141:e152S-84S. 2. Chugh SS, Havmoeller R, Narayanan K, et al. Worldwide epidemiology of atrial fibrillation: a Global Burden of Disease 2010 Study. Circulation 2014;129:837-47. 3. Douketis JD, Spyropoulos AC, Spencer FA, et al. Perioperative management of antithrombotic therapy: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 2012;141:e326S-50S.
www.giejournal.org
Copyright ª 2019 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 https://doi.org/10.1016/j.gie.2019.06.015 Received February 10, 2019. Accepted June 7, 2019. Current affiliations: Department of Gastroenterology, University of Kansas Medical Center, Kansas City, Kansas, USA (1), Department of Computer Sciences, Clairvoyant LLC, Phoenix, Arizona, USA (2), Department of Gastroenterology, Swedish Medical Center, Seattle, Washington, USA (3). Reprint requests: Amit Rastogi, MD, Department of Gastroenterology, University of Kansas Medical Center, Kansas City, KS 66160. Tel: 816-8614700.
Volume
-,
No.
-
: 2019 GASTROINTESTINAL ENDOSCOPY 7