The study flow diagram is presented in Figure 1. We identified 1,350 citations through database and targeted searching. Following title and abstract screening, the full-text records of 82 documents were assessed for eligibility. Thirty-two documents met all inclusion criteria (Appendix 2).18-49 Five of these documents were updates of previously published guidelines18,23,34 or position statements43,45 – only the most recent versions were included20,24,27,44,46; 2 citations were considered companion documents for each other and treated as a single result.38,39 Therefore, 26 documents were included in the final synthesis, including 19 guidelines19-22,24-33,35-37,39,40 and 7 position statements.41,42,44,46-49
Characteristics of Included Documents
A summary of the characteristics of the included documents is provided in Supplemental Table 1. Though most documents addressed GI endoscopy procedures in general, four documents specifically addressed colonoscopy.33,35,39,40 Some organizations developed more than one document addressing different topics related to GI endoscopy practice, including 7 documents from the American Society of Gastrointestinal Endoscopy (ASGE),21,26-28,32,42,49 2 documents from the Spanish Society of Gastrointestinal Endoscopy (SSGE),31,33 and 2 documents from the Society of Gastroenterology Nurses and Associates (SGNA)41,46
Quality Assessment
Scores on the AGREE II domains were generally low (Table 1). Among guidelines, the highest scoring domains were scope and purpose (average score = 53%) and clarity of presentation (average score = 47%); the lowest scoring domain was applicability (average score = 9%). Position statements had similar scores for the clarity of presentation domain (average score = 45%) but lower scores in the scope and purpose domain (average score = 27%). The majority (n = 22) of included documents scored <60% in the rigor of development domain. Of these, 15 were guidelines19-22,25-33,35,37 and 7 were position statements.41,42,44,46-49 Only 9 documents24,26,27,29,31,33,36,40,49 provided a grade or level of evidence in support of their recommendations, making it difficult to assess the evidentiary base for the reported recommendations. When used, recommendation and evidence grading systems varied across documents, making cross-document comparisons challenging.
Recommended Agents for Sedation
Twelve documents provided guidance on the choice of sedatives for GI endoscopy (Table 2).19,21,22,24,26,27,31,33,35,36,40,47
Two documents made recommendations about specific agents for moderate sedation. Both the German Society for Gastroenterology, Digestive and Metabolic Diseases (GSGMD)36 and the SSGE31 stated that midazolam is the preferred benzodiazepine for moderate sedation based on systematic reviews of cohort studies.
With respect to the administration of propofol, documents were not consistent in their recommendations. Documents from the GSGMD,36 European Society of Gastrointestinal Endoscopy (ESGE) & European Society of Gastrointestinal Endoscopy Nurses and Associates (ESGENA),24 and the Danish Secretariat for Reference Programmes for Gastroenterology, Surgery, and Anaesthetics (DSRPGSA)19 recommended propofol monotherapy. In contrast, a guideline from the SSGE31 recommended that patients receiving propofol be pre-medicated with midazolam to reduce the total dose of and adverse events associated with propofol; the ESGE/ESGENA24 guideline recommended pre-medication with midazolam only in select cases (Table 2). The GSGMD36 recommended administration by intermittent boluses, whereas the ESGE/ESGENA24 recommended intermittent bolus or perfusor systems, such as target-controlled or patient-controlled infusion systems. The ESGE/ESGENA guideline24 also suggested against the use of pharyngeal anesthesia for patients undergoing upper GI endoscopy under propofol sedation.
There was no consensus across documents on the optimal sedating agents (i.e. benzodiazepine +/- opioid vs. propofol) or targeted depth of sedation (i.e. moderate vs. deep) (Table 2). Based on high-quality evidence, documents from the SSGE31 and ESGE/ESGENA24 stated that moderate sedation provides high patient satisfaction for GI endoscopy but that deep sedation is preferred for complex procedures (e.g. EUS, ERCP). Although propofol can be targeted to moderate sedation (e.g. with use of balanced propofol sedation, which combines propofol with a benzodiazepine and opioid22,27,50), documents from the ASGE27 and the Canadian Association of Gastroenterology (CAG)47 specifically stated that the combination of an opioid and benzodiazepine is adequate for routine endoscopy. In contrast, documents from the GSGMD36 and SSGE31,33 expressed a preference for propofol over benzodiazepines. However, several documents, including guidelines from the GSGMD,36 ASGE,21,27 SSGE,31 and European Commission (EC),40 made recommendations for tailoring the agent and depth of sedation to the patient, generally based on low-quality evidence (Table 2). A document from the French Society of Digestive Endoscopy (FSDE)35 stated that general anesthesia should be used for all patients undergoing colonoscopy; however, this document did not define general anesthesia, nor did it specify which agents were recommended for use.
Personnel Capable of Administering Sedation
Nineteen documents (14 guidelines19-22,24,27,29-33,35-37 and 5 position statements41,44,47-49) provided recommendations regarding the types of healthcare professionals capable of administering sedation for routine GI endoscopy. These recommendations are summarized in Table 3. Few documents detailed the level of evidence or the strength of the recommendations (n = 6)24,27,31,33,36,49 (Supplemental Table 2).
Five documents, from 3 organizations, provided recommendations relevant to the administration of moderate sedation.20,21,27,32,41 Documents from the ASGE21,27,32 and SGNA41 supported nurse-administered moderate sedation with supervision from a physician. A guideline from the GESA,20 however, stated that an “appropriately trained medical practitioner,” who is not the endoscopist, is required to administer intravenous sedation.
Recommendations for administration of deep sedation were provided in documents from the ASGE,21 Gastroenterological Society of Australia (GESA),20 and SGNA.41 Although all three documents recommended that an anesthesia professional be involved in the administration of deep sedation, documents varied in the strength of their recommendations and the suggested personnel – the SGNA recommended that involvement of an anesthesiologist be considered for patients undergoing deep sedation41; the ASGE suggested that anesthesia professionals could include an anesthesiologist, a certified registered nurse anesthetist, or an anesthesiology assistant, depending on institutional policies21; and the GESA stated that an anesthesiologist or other appropriately trained and credential medical specialist must be present when deep sedation is used.20
Recommendations specific to the administration of propofol were provided in 13 documents19,20,22,24,27,29,31,33,36,44,47-49 and varied considerably. Documents from the CSGNA,44 GESA,20 and British Society of Gastroenterology (BSG)29 did not support nurse-administered propofol sedation; the GESA20 and BSG29 further recommended that propofol be administered by an anesthesiologist or a second, appropriately trained, medical practitioner who is not the endoscopist.20 In contrast, documents from the ESGE,24 DSRPGSA,19 ASGE,49 and Italian Society of Digestive Endoscopy (ISDE)48 specifically focused on non-anesthesiologist administered propofol (NAAP) sedation and 5 additional documents made recommendations regarding cases in which NAAP sedation would be appropriate.22,31,33,36,47 These documents generally stated that NAAP sedation is safe in appropriately selected patients22,31,33,47,48 and may improve efficiency of the endoscopy unit.31,49 The GSGMD,36 DSRPGSA,19 SSGE,31 and CAG47 specifically stated that propofol administration by nurses, under the direction of physicians, is safe for low-risk patients and documents from the SSGE,31,33 ASGE,49 and ISDE48 stated that involvement of an anesthesiologist for low-risk patients undergoing propofol sedation is not cost-effective. To ensure safety, two documents recommended that an anesthesiologist be readily available when NAAP sedation is used.19,31 Furthermore, most documents discussing propofol sedation provided recommendations for circumstances that would necessitate administration of propofol by an anesthesiologist. Important factors in the decision of whether to involve an anesthesiologist included the patient’s American Society of Anesthesiologists (ASA) class19,24,31,33,36,47,48; Mallampati class, presence of facial abnormalities, or other factors suggestive of difficult intubation or ventilation19,24,31,47,48; patients with other high risk medical conditions19,31; long and complex procedures19,24,31,47,48; and other risk factors, including individuals with previous problems with sedation, uncooperative patients, and chronic narcotic users (Table 3).19,24,48
Similar recommendations were made for the involvement of an anesthesiologist for sedation practice in general (Table 3). Unique recommendations relating to sedation practice in general included a recommendation from the ASGE21 that licensed practical nurses and unlicensed assistive personnel are not qualified to administer sedation and a recommendation from the FSDE35 that non-anesthesiologist intravenous sedation (not otherwise specified) not be used outside of clinical trials.
Personnel Responsible for Monitoring Sedated Patients
Recommendations for the healthcare personnel required for monitoring sedated patients were discussed in 17 documents (12 guidelines19-22,24,25,27,30-32,36,37 and 5 position statements41,44,46-48); few documents detailed the level of evidence or the strength of these recommendations (n = 3)24,31,36 (Supplemental Table 3).
For moderately sedated patients, there was consensus among documents from the American Gastroenterological Association (AGA),22 ASGE,21,27,32 SGNA,41 and the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES)30 that a single nurse is capable of both monitoring a moderately sedated patient and performing brief, interruptible tasks. A position statement from the Canadian Society of Gastroenterology Nurses and Associates (CSGNA)44 recommended that two health professionals be present in the endoscopy suite when moderate sedation is being used but did not define who these health professionals could be. There was also consensus among documents from the ASGE21,32 and SGNA41 that a second assistant be available to assist the endoscopist in complex procedures (e.g. difficult polypectomy)21,32,41 or severely ill patients,41 allowing the nurse administering sedation to focus on monitoring the patient. Guidelines from the ASGE21,32 stated that these second assistants could be registered nurses, licensed practical nurses, or unlicensed assistive personnel.
Five documents provided recommendations for patients undergoing deep sedation. Documents from the AGA,22 ASGE,21,32 and SGNA41 consistently recommended that when deep sedation is used, the individual monitoring the sedated patient should not have any other responsibilities. This necessitates an additional individual to assist the endoscopist with technical aspects of the procedure. Two guidelines suggested that an anesthesia professional be present to monitor the deeply sedated patient.21,37 A guideline from the Austrian Society of Gastroenterology and Hepatology (ASGH)37 suggested involving an anesthesiologist for patients who may require endotracheal intubation. The ASGE21 recognized that many institutions require an anesthesia professional for administration of deep sedation and recommended that this individual also monitor the patient during the procedure.
Two documents20,25 made recommendations for monitoring patients under general anesthesia; however, recommendations in these documents differed. A guideline from Joint Advisory Group on Gastrointestinal Endoscopy (JAG)25 recommended that an anesthesiologist be present to monitor patients under general anesthesia, whereas a guideline from GESA20 recommended that both an anesthesiologist and an individual dedicated to assisting the anesthesiologist be present.
Documents providing recommendations for monitoring patients sedated with propofol were generally in agreement that use of propofol sedation requires an individual dedicated to monitoring the patient who has no other responsibilities, necessitating a second individual to assist the endoscopist with the procedure.19,24,27,36,47,48 However, a document from the SSGE31 recommended that patient and procedure complexity be considered when determining whether an individual dedicated to monitoring sedation is needed. This document stated that basic endoscopic procedures on ASA class I-II patients do not require dedicated sedation staff but that complex therapeutic procedures or procedures performed on higher risk individuals (ASA >III) be staffed by individuals solely dedicated to monitoring the sedated patient. Two documents further recommended that a physician be present and available from the time of propofol administration to when the patient wakes up19 or is ready for discharge,27 but were not specific as to who this physician could be.
Documents providing guidance for sedation practices in general made a range of recommendations (Supplemental Table 3), including that a minimum of one nurse is required for endoscopy with sedation22,46; nurses are capable of monitoring sedating patients and performing brief, interruptible tasks46; sedated endoscopy requires an individual solely dedicated to monitoring the sedated patient20,36; additional staff are required for complex procedures or endoscopy performed on high-risk patients30,36,37,44,46; and when sedation is provided by anesthesia personnel, an individual responsible for assisting the endoscopist, and possibly the anesthesia professional, is needed.20,21,32,44,46 Uniquely, the CSGNA44 stated that a second Registered Nurse, Licensed Practical Nurse, or Registered Practical Nurse is required for “therapeutic” procedures; however, this document did not define the procedures considered to be therapeutic.
Skills and Training Required to Administer Sedation and Monitor Sedated Patients
Seventeen documents (12 guidelines19-22,24,27,28,30,31,33,36,37 and 5 position statements41,44,47-49) provided varying recommendations regarding the skills and training required for individuals involved in procedural sedation for endoscopy, often without a grade or level of evidence stated (Supplemental Table 4). Most recommendations for moderate sedation referred specifically to nurses. Recommendations included formal training in procedural sedation,44 knowledge of the sedatives used and their reversal agents,41,44 an understanding of airway management,44 skills to rescue patients who enter deeper levels of sedation than intended,44 and the ability to manage other complications.41,44 Guidelines from the AGA22 and ASGE28 also recommended that physicians involved in these procedures be able to rescue patients from deeper levels of sedation than intended. The CSGNA44 further recommended that endoscopy nurses working in hospitals have at least basic cardiac life support training and those working in private endoscopy clinics have advanced cardiac life support (ACLS) training; this document did not require ACLS training for nurses working within institutions with code response teams.
Documents addressing deep sedation similarly recommended knowledge of the medications used,41 the ability to rescue patients from a deeper level of sedation than intended (i.e. general anesthesia),28 and skills in advanced airway management and the management of cardiorespiratory complications.22,28,41 The SGNA specifically recommended that both nurses and physicians involved in deep sedation have skills in ACLS.41
Prior to involvement in non-anesthesiologist administered propofol sedation, documents recommended formal training in propofol administration.19,22,24,27,47-49 The ESGE24 further stated that intensive care or anesthesia experience for the physician directing propofol sedation is desirable. Additional recommendations for propofol administration included basic resuscitation skills,19,49 skills in managing complications,19,27,47 and skills in airway management.19,27,28,47-49 Documents from CAG,47 ASGE,27,49 and ISDE48 recommended ACLS training and the ESGE24 recommended that if the individual administering propofol has ACLS training a life support team does not need to be rapidly available.
Recommendations for sedation practice not tied to a specific level of sedation were similar, including recommendations for formal training in sedation,20,27,30,31,36 knowledge of the agents being used,21,22,28,31,36,37 the ability to recognize and manage complications and rescue patients from deeper than intended levels of sedation,20-22,27,28,31,37 skills in basic resuscitation21,22,30,31,36 and airway management,20,22,31,36 and ACLS training.20-22,30,33 Uniquely, the GSGMD recommended that the physician responsible for sedation generally have intensive care medicine experience.36
Equipment Required to Monitor Sedated Patients
Equipment recommendations are summarized in Table 4. The equipment most consistently recommended for monitoring all sedated patients (regardless of level targeted or use of propofol) included non-invasive blood pressure monitoring and pulse oximetry.19-22,24,27,30,31,33,36,37,39,44,47,48 Documents referring to moderate sedation and documents that did not specify the level of sedation generally recommended electrocardiography only for select cases. These cases included patients with cardiac or pulmonary disease,22,30,33,36,39,44 elderly patients,39 or prolonged procedures.39 A guideline from Cancer Care Ontario (CCO)39 also recommended the same factors to be considered when determining whether electrocardiography is needed for patients undergoing deep sedation. Five documents suggested routine use of electrocardiography for patients undergoing propofol sedation,19,27,29,37,47 whereas two documents recommended selective use, in particular for patients with cardiac24,48 and/or pulmonary disease.24 Recommendations for capnography varied. For patients undergoing moderate sedation, a guideline from the ASGE21 and a joint statement from the ASGE, American College of Gastroenterology (ACG), and AGA42 both stated that there was insufficient data to recommend routine use of capnography. For patients undergoing deep sedation, two documents from the ASGE21,27 stated that capnography may be considered. For patients specifically undergoing sedation with propofol, there was no consensus on the use of capnography – use was recommended for all patients by the BSG,29 recommended to be considered by the ASGE,27 and recommended only in select cases by the ESGE24; statements that routine use is not supported were made by CAG47 and ASGE.49 Similarly, documents that did not specify the level of sedation also differed in recommendations for capnography, which was recommended for use by the CSGNA44 and SSGE,31 while the GSGMD36 and GESA20 stated capnography may be considered. The AGA22 and ESGE24 did not recommend routine use of the bispectral index (BIS)/electroencephalography (EEG) during moderate sedation or NAAP, respectively; the GSGMD36 stated that a benefit to EEG monitoring has not been demonstrated for sedated patients (no specific level of sedation identified).