Sedation in ERCP

Defining the Continuum of Sedation

Sedation is typically characterized using the American Society of Anesthesiologists (ASA) Continuum of Sedation, which defines four discrete levels of sedation ( Table 6.1 ). Depth is most frequently defined by patient responsiveness to voice, light tactile stimulation, and painful stimulation during the procedure. However, the corresponding cardiopulmonary sequelae of this degree of awareness do not directly translate into the probability of sedation-related AEs. In moderate (also known as “conscious”) sedation, patients may be sleeping but will have purposeful response to tactile stimuli, yet may not respond to voice. In patients who are deeply sedated, this response occurs only after repeated or painful stimuli. Patients who do not respond to painful stimuli even if they are breathing on their own are by definition under general anesthesia. Monitored anesthesia care (MAC) is a term often used when talking about sedation. MAC does not describe the level of sedation; it just means that a trained anesthesia provider was involved in the care and the administration of drugs. In reality, a patient's level of sedation rarely meets only one of these definitions during the course of endoscopy, and these levels actually represent a continuum. The amount of sedation administered to achieve moderate sedation often inadvertently leads to deep sedation. Similarly, patients who are targeted for deep sedation often meet criteria for general anesthesia.

Many patients undergoing ERCP often require deep sedation, as opposed to the light or moderate sedation that is usually adequate for colonoscopy or esophagogastroduodenoscopy. ERCP procedures are typically longer in duration and require less spontaneous patient movement to achieve technical success. In a study serially assessing sedation depth during ERCP, 85% of patients met criteria for deep sedation during a segment of the procedure. Consequently, the ASA recommends that the sedation provider be adequately trained in rescue maneuvers commensurate with one level of sedation higher than the intended target. Therefore patients targeted for deep sedation should be managed by a provider who is trained in the administration of general anesthesia. This would include management in bag mask ventilation, laryngeal mask airway placement, and endotracheal intubation. The Centers for Medicaid & Medicare Services (CMS) have endorsed this recommendation, releasing a clarification letter to their policy on hospital anesthesia services in 2010 after the major gastrointestinal (GI) societies in the United States made a concerted effort to endorse nonanesthesiologist-administered propofol for low-risk patients undergoing standard endoscopy.

The current options for sedation in ERCP can be simplified into two categories: endoscopist-administered sedation and anesthesiologist-administered sedation. Computerized sedation systems that incorporate real-time patient feedback have been evaluated for standard endoscopic procedures but have not been investigated for patients undergoing ERCP. Because propofol can be administered only by anesthesia providers in the United States, endoscopist-administered sedation implies moderate sedation using conventional agents such as the combination of a benzodiazepine (e.g., midazolam) and an opiate (e.g., fentanyl or meperidine). Of note, benzodiazepine can be reversed by flumazenil and opioids by naloxone in the event of oversedation. Anesthesiologists may choose between general anesthesia with endotracheal intubation at the onset of the procedure and general anesthesia with use of a nasal cannula and having the patient breathe spontaneously during the procedure. In the latter scenario, patients are typically sedated using a propofol-based regimen, with a goal of achieving deep sedation or general anesthesia. Endoscopists increasingly prefer anesthesia-administered sedation for all endoscopic procedures. The growing role of propofol in endoscopic practice is reflected in epidemiologic data demonstrating a consistent increase in anesthesia-administered sedation. The overuse of anesthesia services for colonoscopy is under increased scrutiny. With a greater emphasis on cost-effectiveness in health care, judicious use of anesthesia will mandate an improved preprocedure risk assessment; this is particularly important in ERCP, where the potential for sedation-related complications is highest.

Initially approved for the induction and maintenance of anesthesia, propofol (2,6-diisopropylphenol) has become an increasingly popular sedative for endoscopic procedures because of its rapid onset of action (30 to 45 seconds) and short duration of effect (4 to 8 minutes). In the United States, propofol is currently restricted to anesthesiologists and some emergency medicine physicians because of its relative potency, lack of an antagonist, and potential for rapid change in the depth of sedation from moderate sedation to general anesthesia. Nevertheless, in a meta-analysis of 12 trials of propofol sedation during routine endoscopy, endoscopic ultrasonography (EUS), and ERCP, the overall rate of cardiopulmonary AEs was lower than that of standard combination opiate–benzodiazepine regimens.

There are several unique characteristics of ERCP compared with other endoscopic procedures that may accentuate the benefits of propofol. Specifically, ERCPs tend to be longer in duration and require sustained patient cooperation in order to achieve technical success. Longer procedures require higher cumulative doses of benzodiazepines and opiates to maintain moderate sedation, which translates into longer recovery times.

Defining Sedation-Related Complications (Adverse Events)

Adverse events (AEs) specifically related to sedation are usually classified in the literature by objective criteria such as oxygen desaturation, aspiration, laryngospasm, apnea, hypotension, arrhythmia, myocardial ischemia, and need for airway rescue maneuvers or reversal agents. Mortality data related to sedation in endoscopy are sparse, particularly in ERCP. The risk of death is probably close to 0.03% for patients undergoing standard endoscopy using conventional sedation regimens. Fewer studies track the frequency of airway rescue maneuvers, such as a chin lift, nasal trumpet insertion, and transient positive pressure (i.e., bag-mask) ventilation. These may be performed as a preventive maneuver in anticipation of hypoxemia or apnea and reflect the importance of having a sedation provider experienced in airway rescue. Rates of conversion from nasal cannula to endotracheal intubation during administration of intravenous propofol during ERCP are less defined, although a study of 528 patients undergoing ERCP with MAC reported a 3% incidence of unplanned endotracheal intubation. In a cohort study that included all endoscopic procedures performed during a 5-year period, patients with a higher ASA class ( Table 6.2 ) and those undergoing ERCP were more likely to require reversal agents such as flumazenil or naloxone.