Practice Patterns in Non–Operating Room Sedation

Sedation practice in the non–operating room setting varies widely throughout the world. This is exemplified by routine diagnostic GI endoscopy. In the United States, it is frequently assumed most patients want sedation for endoscopy. A study of U.S. gastroenterologists found that more than 98% of routine endoscopies are performed with sedation and approximately 25% of these providers now use propofol sedation. In addition, 68% of gastroenterologists using conventional sedation would consider using propofol if it were easily available. When propofol was used, it was provided by anesthesia-trained personnel in almost 90% of cases, with the remainder being directed by the endoscopist. Large regional variations are seen in the use of propofol from a low of 7% in the Northeast to a high of 37% in the Mid-Atlantic region. This variation appears to be driven by local reimbursement policies for sedation. In addition, an increasing number of more complex non–operating room procedures are performed in the United States. These factors have led to a dramatic increase in the need for high-quality sedation.

A review of sedation practice in 1998 revealed that in much of the world outside the United States and United Kingdom, the norm was for unsedated routine endoscopy. However, by 2006 this standard had begun to change in Europe; an observational study of 6004 patients undergoing colonoscopy found that 17% of patients had no sedation, 53% had moderate sedation, and 30% had deep sedation. More recently, Ladas et al collected data regarding the sedation rate for EGD and colonoscopy in Europe, North America, Asia, Africa, and Australia. No data were available from the last three continents, so questionnaire surveys were mailed to gastroenterologists in those locations. For upper GI endoscopy, rates of sedation varied in European countries from less than 20% to more than 75%; in the United States and Australia, over 98% of upper GI endoscopies were performed with sedation; in Asian countries many areas performed upper GI endoscopies without sedation, although some centers reported a 100% sedation rate. Finally, the surveys showed that 25% to 50% of patients received sedation in African countries. The authors note that the survey respondents reported that in as many as 46% of the countries, monitoring was not available in the majority of endoscopy suites; this lack of monitoring could be a factor in the low sedation rate. The rates of sedation for colonoscopy were shown to be quite variable among countries and among centers within each country. Sedation medication varied as well, although it was noted that propofol has been used in each country that responded to the survey. These data support an increasing expectation and use of sedation for endoscopy worldwide.

Propofol Use by Nonanesthesiologists

Propofol is used extensively in non–operating room sedation practice by both anesthesia-trained and nonanesthesiologist providers. Non–operating room settings where propofol frequently is given by nonanesthesiologists include the GI endoscopy suite, the electrophysiology laboratory, and, to a lesser degree, both diagnostic and interventional radiology areas. Propofol is a unique intravenous medication available as an oil-in-water emulsion for induction or maintenance of moderate-to-deep sedation and general anesthesia through facilitation of inhibitory neurotransmission mediated by gamma–aminobutyric acid (GABA). It is highly lipid-soluble, which results in an onset of action that is rapid; awakening from a single bolus dose is very rapid as well, because of a short initial redistribution half-life of 2 to 8 minutes. Propofol is cleared partly via hepatic conjugation and renal elimination of those inactive metabolites, but it also likely has a component of extrahepatic metabolism as the clearance of propofol exceeds hepatic blood flow. Propofol has many effects on the organ systems. It decreases arterial blood pressure as a result of a drop in systemic vascular resistance, cardiac contractility, and preload. It is a profound respiratory depressant and causes dose-dependent apnea while it inhibits hypoxic ventilatory drive and depresses the normal response to hypercarbia at sub–general anesthetic doses. It has antipruritic properties as well as antiemetic effects.

Propofol is preferred by many endoscopists and patients and has been shown in multiple studies to be associated with a faster induction of sedation and faster full recovery time, higher postprocedure patient satisfaction, and quicker anticipated return to baseline function. Dewitt et al demonstrated these findings in a prospective randomized, single-blinded trial of 80 consecutive patients who were randomized to receive either sedation with nurse-administered propofol sedation or midazolam and meperidine for EUS procedures. Nayar et al performed a retrospective analysis comparing outcomes of 1000 EUS procedures performed with anesthesia-assisted propofol sedation (propofol deep sedation) and 1000 EUS procedures performed under conventional sedation using midazolam and meperidine titrated to moderate sedation and found that sedation time, induction time, and intraprocedural time were significantly shorter in the propofol group. Randomized trials by Cohen, Sipe et al, and Ulmer et al that compared endoscopist-directed propofol (EDP) with traditional sedation with opioids and benzodiazepines in patients undergoing endoscopy or colonoscopy found that EDP resulted in faster onset of sedation, faster recovery, and similar or better patient satisfaction.

Not surprisingly, the use of propofol by non–anesthesia trained providers has led to some controversy. Anesthesiologists are concerned that propofol can induce general anesthesia and apnea very rapidly and that nonanesthesiologists lack sufficient airway expertise to rescue patients who receive propofol. The package insert for propofol states that propofol should be administered only by persons trained in rescue from unintended general anesthesia. At least 12 states in the United States have laws or regulations regarding nursing practice that prevent the administration of propofol to nonintubated patients. However, propofol has been and continues to be administered in the GI endoscopy suite without anesthesia-trained personnel. Several protocols have been described, but the generic term for this practice is nonanesthesiologist administration of propofol (NAAP). In the gastroenterology literature, it is also frequently referred to as EDP. Nonanesthesiologists believe that propofol is safe if used according to the principles that underlie safe procedural sedation with any agent—namely, slow titration to the desired level of sedation and awareness of individual variability in pharmacological response.

Two protocols for EDP sedation exist: nurse-administered propofol sedation (NAPS) and balanced propofol sedation (BPS). In NAPS a trained registered nurse’s sole responsibilities are patient monitoring and the administration of propofol, titrated to deep sedation. BPS combines several drugs in small doses, including fentanyl, midazolam, and propofol. Both the nurse and endoscopist participate in all dosing decisions that involve propofol. The nurse has primary responsibility for monitoring the patient, but the nurse also may perform brief tasks such as assisting with tissue acquisition. Both methods of sedation have been studied extensively, and gastroenterologists claim that these practices are safe for many patients.

Anesthesiologists counter that anesthesia care in remote locations, such as the GI endoscopy suite, poses a significant risk for the patient. This risk has been shown with data from the ASA closed claims database, with risks to the patient related to oversedation and ventilatory failure. The severity of these injuries was greater than those occurring within the operating room, and more than half of all sedation-related complications resulted in death. In addition, data from closed claims indicate that risk for complications increases when the number of agents used to provide sedation is greater than two. This may call into question the use of BPS by nonanesthesiologists, especially when a dedicated sedation provider is not used.

Propofol Use by Nonanesthesiologists

Propofol is used extensively in non–operating room sedation practice by both anesthesia-trained and nonanesthesiologist providers. Non–operating room settings where propofol frequently is given by nonanesthesiologists include the GI endoscopy suite, the electrophysiology laboratory, and, to a lesser degree, both diagnostic and interventional radiology areas. Propofol is a unique intravenous medication available as an oil-in-water emulsion for induction or maintenance of moderate-to-deep sedation and general anesthesia through facilitation of inhibitory neurotransmission mediated by gamma–aminobutyric acid (GABA). It is highly lipid-soluble, which results in an onset of action that is rapid; awakening from a single bolus dose is very rapid as well, because of a short initial redistribution half-life of 2 to 8 minutes. Propofol is cleared partly via hepatic conjugation and renal elimination of those inactive metabolites, but it also likely has a component of extrahepatic metabolism as the clearance of propofol exceeds hepatic blood flow. Propofol has many effects on the organ systems. It decreases arterial blood pressure as a result of a drop in systemic vascular resistance, cardiac contractility, and preload. It is a profound respiratory depressant and causes dose-dependent apnea while it inhibits hypoxic ventilatory drive and depresses the normal response to hypercarbia at sub–general anesthetic doses. It has antipruritic properties as well as antiemetic effects.

Propofol is preferred by many endoscopists and patients and has been shown in multiple studies to be associated with a faster induction of sedation and faster full recovery time, higher postprocedure patient satisfaction, and quicker anticipated return to baseline function. Dewitt et al demonstrated these findings in a prospective randomized, single-blinded trial of 80 consecutive patients who were randomized to receive either sedation with nurse-administered propofol sedation or midazolam and meperidine for EUS procedures. Nayar et al performed a retrospective analysis comparing outcomes of 1000 EUS procedures performed with anesthesia-assisted propofol sedation (propofol deep sedation) and 1000 EUS procedures performed under conventional sedation using midazolam and meperidine titrated to moderate sedation and found that sedation time, induction time, and intraprocedural time were significantly shorter in the propofol group. Randomized trials by Cohen, Sipe et al, and Ulmer et al that compared endoscopist-directed propofol (EDP) with traditional sedation with opioids and benzodiazepines in patients undergoing endoscopy or colonoscopy found that EDP resulted in faster onset of sedation, faster recovery, and similar or better patient satisfaction.

Not surprisingly, the use of propofol by non–anesthesia trained providers has led to some controversy. Anesthesiologists are concerned that propofol can induce general anesthesia and apnea very rapidly and that nonanesthesiologists lack sufficient airway expertise to rescue patients who receive propofol. The package insert for propofol states that propofol should be administered only by persons trained in rescue from unintended general anesthesia. At least 12 states in the United States have laws or regulations regarding nursing practice that prevent the administration of propofol to nonintubated patients. However, propofol has been and continues to be administered in the GI endoscopy suite without anesthesia-trained personnel. Several protocols have been described, but the generic term for this practice is nonanesthesiologist administration of propofol (NAAP). In the gastroenterology literature, it is also frequently referred to as EDP. Nonanesthesiologists believe that propofol is safe if used according to the principles that underlie safe procedural sedation with any agent—namely, slow titration to the desired level of sedation and awareness of individual variability in pharmacological response.

Two protocols for EDP sedation exist: nurse-administered propofol sedation (NAPS) and balanced propofol sedation (BPS). In NAPS a trained registered nurse’s sole responsibilities are patient monitoring and the administration of propofol, titrated to deep sedation. BPS combines several drugs in small doses, including fentanyl, midazolam, and propofol. Both the nurse and endoscopist participate in all dosing decisions that involve propofol. The nurse has primary responsibility for monitoring the patient, but the nurse also may perform brief tasks such as assisting with tissue acquisition. Both methods of sedation have been studied extensively, and gastroenterologists claim that these practices are safe for many patients.

Anesthesiologists counter that anesthesia care in remote locations, such as the GI endoscopy suite, poses a significant risk for the patient. This risk has been shown with data from the ASA closed claims database, with risks to the patient related to oversedation and ventilatory failure. The severity of these injuries was greater than those occurring within the operating room, and more than half of all sedation-related complications resulted in death. In addition, data from closed claims indicate that risk for complications increases when the number of agents used to provide sedation is greater than two. This may call into question the use of BPS by nonanesthesiologists, especially when a dedicated sedation provider is not used.

Safety of Propofol in the Non–Operating Room Setting

Many studies have sought to define the frequency of sedation-related complications and adverse events with propofol sedation. A prospective analysis by Cote et al found that the rate of airway modifications in patients given propofol by a certified registered nurse anesthetist with a goal of deep sedation during advanced endoscopy (ERCP, EUS, and small bowel enteroscopy) was 14.3% (114 of 799 patients). Airway modifications included chin lift, modified face mask ventilation, and nasal airway placement. In this study, hypoxemia (blood oxygen saturation [Sp o ₂ ] <90%) occurred in 12.8%, hypotension requiring vasopressors in 0.5%, and early procedure termination in 0.6%. No patients required bag-mask ventilation or endotracheal intubation. Other than the inability to give informed consent, there were no exclusion criteria. The authors found that body mass index (BMI), male sex, and ASA class 3 or higher were independent predictors of airway modifications. They concluded that propofol can be safely used for advanced endoscopic procedures when administered by a trained professional. They proposed that the highest-risk groups should be managed by professionals trained in advanced airway interventions and that lower-risk populations can be managed safely by professionals with less-intensive airway training (gastroenterologist-supervised, nurse-administered propofol).

Complications of NAAP versus the use of standard moderate sedation for endoscopy have been evaluated. Two recent reviews demonstrated that the rate of sedation-related adverse events during endoscopy performed under sedation with a benzodiazepine and opioid combination versus propofol is comparable. The study by Dewitt et al of 80 patients had no major complications (intubation, hospitalization, permanent neurologic impairment, or death) in patients receiving midazolam/meperidine or propofol, but the patient population included only ASA class 1 or 2 adults and was small. The study of EUS by Nayar et al had a total of six complications in 1000 persons in the propofol group, including perforation, aspiration, three patients with apnea requiring intubation, and hypotension. The only complication was prolonged somnolence in the group receiving moderate sedation, which received midazolam/meperidine.

Sharma et al retrospectively reviewed 324,737 patients who were sedated by endoscopists using traditional opioid/benzodiazepine sedation and reported a rate of 8 cardiopulmonary deaths per 100,000 patients and 9.3 cardiorespiratory complications in every 1000 cases. In comparison, Rex et al collected the largest summary of the experience of endoscopy performed with EDP using multiple reports in the literature. This review included the outcome of 646,080 endoscopic procedures and revealed 11 endotracheal intubations, no permanent neurologic injuries, and 4 deaths, all in patients with ASA status 3 or higher. The main significant complication was a rate of bag-valve-mask ventilation of only 0.1%. Limitations of this study include data that were self-reported by centers or endoscopists, some centers that collected safety data retrospectively, no tracking of late-occurring sedation complications, and a heterogeneous patient population. These two large evaluations appear to indicate that complications of NAAP for endoscopy are similar or less than those encountered during traditional sedation.

Wehrmann and Riphaus analyzed 9547 interventional endoscopic procedures performed over a 6-year period in which propofol sedation was by nonanesthetist physicians who were certified in internal medicine, intensive care, and cardiorespiratory resuscitation who were not involved in the procedures. Procedures included 5347 interventional procedures using upper GI endoscopy, 3937 ERCPs, and 236 EUSs. They found a 1.4% rate (135 in 9547) of adverse sedation-related events, defined as those that caused premature termination of the procedure, bag-valve-mask ventilation, endotracheal intubation, or unexpected intensive care admission. Forty patients required assisted ventilation (0.4%); 9 patients required endotracheal intubation (0.09%); 28 were admitted to the intensive care unit (0.3%); and 4 patients died, 3 potentially related to sedation-related side effects (0.03%). Independent risk factors were emergency procedures and total propofol use greater than 100 mg. These rates of complications are significantly higher than those found in other studies of EDP and certainly lead one to proceed with caution in more advanced endoscopic procedures. In addition, although NAAP was used for these patients, the use of highly trained physicians solely responsible for the sedation is a higher level of care than provided in many proposed NAAP protocols.

At this time, propofol remains the gold standard for procedural sedation. The use of propofol as a sole agent for procedural sedation often requires deep sedation for most patients because propofol has no inherent analgesic properties. BPS attempts to use a combination of sedative agents, including propofol, to attain the advantages of propofol while achieving a state of only moderate sedation. BPS versus propofol alone or opioid/benzodiazepine sedation has been compared in many studies. BPS has decreased the rate of bag-mask ventilation during upper endoscopic procedures from approximately 1 in 300 procedures using only propofol to approximately 1 in 1000 procedures. It has been shown that it is possible to titrate EDP administration to moderate sedation if propofol is given with small doses of opioids and benzodiazepines rather than used as a single agent. Patel et al showed that endoscopists using opioids/benzodiazepines produce deep sedation in more than half of patients, which is more often than with BPS. Clarke et al found no deaths or endotracheal intubations in more than 28,000 procedures over 7 years using a combination of propofol, fentanyl, and midazolam, targeting moderate sedation. A randomized control trial by van Natta and Rex compared the use of propofol titrated to deep sedation to various combinations of propofol, fentanyl, and/or midazolam titrated to moderate sedation for colonoscopy. This study involved 200 patients, randomized to one of four treatment arms: propofol alone titrated to deep sedation; propofol and fentanyl; propofol and midazolam; or propofol, midazolam, and fentanyl, all titrated to moderate sedation. Combining propofol with midazolam, fentanyl, or both allowed a significant reduction in propofol use, without an impact on patient satisfaction or postprocedural recovery. A slight reduction was seen in preprocedural efficiency (the time between initiation of sedation and when the patient is ready for colonoscopy). It appears that combination therapy with propofol can achieve effective moderate sedation for colonoscopy, but caution is warranted because this study was small. Further study into combination therapy targeted to moderate sedation for routine endoscopy is needed, especially in terms of safety.

Propofol use in the non–operating room setting is increasing in Europe. Propofol use was evaluated in Italy by Repici et al. They assessed the efficacy and safety of BPS administered by nonanesthesiologists for colonoscopy. They included 1593 patients with ASA class 1 or 2 with a Mallampati score of less than 2,who did not have obstructive sleep apnea, and who received one dose of midazolam and low-dose propofol. Moderate sedation was achieved in 98% of patients, and deep sedation occurred in 2%. No serious adverse events were observed. The authors conclude that BPS provides the endoscopist the ability to minimize the dose of propofol needed to target moderate sedation and that if patients are carefully selected, BPS can be safely administered by nonanesthesiologist endoscopists.

Switzerland may have the greatest national experience with widespread NAAP practice. Heuss et al studied the trends on the practice of endoscopic sedation in this country over the last 20 years. They note that because propofol use is not restricted to anesthesiologists, the country has to some extent become a testing ground for the broad application of NAAP for GI endoscopy. The authors sent a questionnaire to all active board-certified Swiss gastroenterologists in January 2010. They had a 78% response rate and received data regarding 263,370 GI endoscopies (EGDs, colonoscopies, EUSs, and ERCPs). More than 80% of their procedures were performed with sedation, and 73% of respondents used propofol without an anesthesiologist present. The overall sedation-related complication rate was 0.14%, and the mortality rate was 0.0018%, with four deaths. These results are similar to the morbidity and mortality rates (0.18% and 0.0014%, respectively) reported in data from 2003 in Switzerland. The authors conclude that their data support the position that the presence of only a single nurse for both administering propofol and carrying out duties relating to endoscopy is safe and that an anesthesiologist need not be present when propofol is being used.

Although the vast majority of research regarding sedation with propofol in the non–operating room setting has been done in the GI suite, NAAP has also been evaluated for use in cardiac electrophysiology procedures, but the data are much more limited. Sayfo et al describe the outcome of proceduralist-directed, nurse-administered propofol sedation (PDNAPS) in patients undergoing implantable cardioverter-defibrillator (ICD) procedures. This was a retrospective review of 672 procedures; any ICD-related procedure was included, and no control group was included. Patients were excluded if an anesthesiologist was present, if they were intubated before the procedure, or if they did not receive propofol. Propofol was administered by a registered nurse, guided by the electrophysiologist; other medications (midazolam, fentanyl, or others) were used at the discretion of the electrophysiologist. The outcomes included serious adverse events, such as procedural death, unexpected transfer to an ICU, respiratory failure requiring intubation or bag-mask ventilation, and hypotension requiring vasoconstrictor or inotrope support, as well as nonserious adverse events, such as hypotension requiring fluid resuscitation and hypoxemia requiring augmented respiratory support with nonrebreather mask, oral airway, or jaw lift. They found that 10% of patients had serious adverse events and 38.7% had nonserious adverse events. No procedural-related deaths occurred. Independent predictors of serious adverse events included longer procedure duration and biventricular implant. The authors conclude that proceduralist-directed, nurse administered propofol sedation (PDNAPS) for shorter ICD procedures, including single- and dual-chamber implants, generator changes, and defibrillation threshold testing have acceptable rates of serious adverse events and manageable nonserious adverse events and should be considered for further study. However, biventricular implants and other complex procedures and procedures such as lead repositioning that have variable duration should be done with an anesthesiologist.

Kottkamp et al evaluated whether BPS, titrated to achieve deep sedation, was feasible and safe for catheter ablation of atrial fibrillation. Sedation was initiated with midazolam, fentanyl, and propofol in 650 patients in Germany and Switzerland, given by a nurse under the supervision of an electrophysiologist; notably, the nurse did not have other responsibilities. The patients were ASA class 1 (25%), 2 (40%), or 3 (35%). They had no major sedation-related complications and no endotracheal intubations, and all patients maintained stable heart rate, invasive arterial blood pressure, and oxygenation. They concluded that deep sedation by nonanesthesiologists was safe and that consultation of an anesthesiologist was not necessary for this procedure.

Very limited data exist on comparison of NAAP to anesthesiologist-directed care. A recent randomized controlled trial evaluated whether there was a difference in endoscopist-administered versus anesthetist-administered propofol in terms of safety and patient satisfaction. Although the study included only 90 ASA class 1 or 2 patients, the authors found that patient satisfaction scores were slightly higher in the endoscopist group, despite actually having higher pain scores in that group. They also found that the occurrence of minor sedation-related complications (hypotension requiring increased crystalloid and desaturation of below 95% for 30 seconds) was higher in the anesthetist group. The study was quite limited by the small number of patients, and using this study to conclude that it is safer to undergo propofol sedation with an endoscopist may not be advisable because the overall rate of serious complications in patients with ASA class 1 and 2 undergoing routine endoscopy is so low. Furthermore, their definition of “complications” was not consistent with events that many anesthetists would consider complications.

A small observational study published in 2009 attempted to evaluate and compare the quality of sedation provided for upper EUS using anesthesiologist-directed propofol versus standard moderate sedation provided by the endoscopy team. Quality was determined using expert video analysis of the procedural sedation using a validated scoring system. The study included 50 patients with 25 in each group and found that 52% of patients receiving moderate sedation exhibited an uncontrolled patient state (significant undersedation or oversedation) at some time during the procedure versus 28% of the propofol group. Of more significance, patients receiving midazolam and an opioid spent 7.1% of the procedure in an uncontrolled patient state, whereas patients in the propofol group experienced an uncontrolled state approximately 1% of the procedure time. Overall efficiency was also considerably better in the propofol group; these patients experienced significantly less in-hospital and at-home nausea and vomiting and fell back to baseline status more quickly. Finally, patient satisfaction was improved in the propofol group: 60% thought the procedure was better than anticipated versus 21% in the standard group. It would be interesting to have similar data on NAAP versus anesthesiologist-directed care, but this type of information is currently unavailable.

Many studies have been done to date supporting the safety of nonanesthesiologist administered propofol. The available data suggest that rates of complications are no greater for NAAP compared to moderate sedation using a benzodiazepine/opioid combination and that NAAP seems to be safe for specific procedures, such as routine GI endoscopy, with appropriately trained personnel and appropriate patient selection. However, certain patients (e.g. those with ASA class 3 status or greater, chronic opioid use, elevated BMI) and advanced procedures (e.g., EUS, ERCP, biventricular pacemakers) may be at higher risk and warrant anesthesiologist-directed care. In addition, it is difficult to come to clear conclusions regarding the use of NAAP in other settings such as electrophysiology or radiology because the data are limited. Despite the apparent safety of NAAP, questions remain: Is it beneficial from an efficiency or quality standpoint to use anesthesiologists or anesthetists in the non–operating room setting? Should anesthesia professionals be involved in overseeing or training nonanesthesiologists providing sedation? What type of presence should anesthesia providers have in a stand-alone endoscopy setting where sedation is provided but immediate access to a “code” or “STAT airway” team may not be available? These questions become more important as NAAP appears to be safe and pressure to provide this type of care may become more common. If this practice is to be continued, careful patient and procedure selection is necessary and must be clearly defined.