In developed countries, the majority of routine, low-risk endoscopic procedures are currently performed with some form of sedation.1 Gastrointestinal (GI) endoscopy is an inherently uncomfortable procedure, with the potential for abdominal pain, cramping and bloating during colonoscopy,2 and gagging, retching and choking during esophagogastroduodenoscopy (EGD). The use of sedation allows for a more thorough and relaxed procedure, with higher polyp detection rates and procedures completion rates.3 A sedated, comfortable patient not only enhances endoscopist satisfaction,4 but is associated with a higher likelihood of the patient being willing to undergo the procedure again. The American Society of Anesthesiology (ASA) has published definitions for different levels of sedation (Table 1);5 specifically, at the level of “moderate sedation,” the patient maintains ventilatory and cardiovascular function and is able to make purposeful responses to verbal or tactile stimuli, while at “deep sedation,” patients cannot be easily aroused but is still able to respond to noxious or insistent stimuli. Airway support may be required for deep sedation. To allow the procedure to proceed smoothly, sedatives are usually dosed to achieve moderate sedation (benzodiazepine and opiate sedation) or deep sedation (monitored anesthesia care with propofol). In a minority of cases, full “general anesthesia” (with intubation of the patient and the use of anesthetic agents such as nitrous oxide or ketamine) is required because of patient characteristics or procedural complexity. Since full general anesthesia is not commonly used for routine GI endoscopy, it will not be discussed further in this review.

Although sedation for endoscopy is becoming widespread, there is still interest in non-sedated endoscopy, even in countries like the United States where almost all GI procedures are performed with sedation. Non-sedated endoscopy offers several potential advantages−lower cost, wider availability, less risk, higher efficiency and decreased post-procedural impairment, allowing for patients to drive or return to work immediately afterwards. There continue to be attempts to develop unsedated EGD using ultrathin transnasal endoscopes,6 and unsedated colonoscopy using water immersion techniques.7 Motivated patients can successfully undergo unsedated endoscopy,8 but the vast majority of patients in developed countries prefer and demand sedation, and the overall trend in GI endoscopy is moving towards more potent, not less potent, forms of sedation.910

Initial forms of sedation for GI endoscopy involved moderate sedation using such agents as midazolam, diazepam, pethidine, fentanyl, remifentanil and/or meperidine. Midazolam and diazepam are intravenous benzodiazepines, with powerful sedative, anxiolytic, hypnotic and amnestic effects. Fentanyl, remifentanil and meperidine are intravenous opioid analgesics, with only minimal sedative effects. Typically, moderate sedation is induced by a combination of a benzodiazepine sedative and an opioid analgesic, which have complementary and synergistic effects with each other. Midazolam and fentanyl (MF) have now become the most commonly used moderate sedation agents in the United States,11 and many other countries.12 Even though some studies reported similar efficacy for midazolam and diazepam,13 midazolam is now usually favored over diazepam because of its faster onset of action and better safety profile,1415 while fentanyl has now supplanted meperidine because the former is associated with more rapid onset and clearance, with less nausea.1617 Midazolam is a water-soluble compound prepared in an acidic solution. After introduction into the bloodstream, it reconfigures to a lipophilic structure at physiological pH, rapidly diffusing across the blood-brain barrier into the CNS. The usual total dose for GI endoscopy is 2 to 10 mg. Fentanyl, a synthetic opioid agonist with powerful analgesic but only mild sedative properties, is highly lipid soluble and 80 times more potent than morphine. The usual total dose is 50 to 200 µg. Both agents have a relatively fast onset of action when given intravenously and a short duration of effect (Table 2). Effective reversal agents, naloxone for opioids and flumazenil for benzodiazepines, are readily available. In addition to MF, ancillary agents such as diphenhydramine or droperidol are sometimes used. These agents potentiate the actions of the primary benzodiazepine/opioid regimen, and may be needed for patients who are unusually difficult to sedate. The use of droperidol has been limited by its cardiac risks, especially in patients with a history of prolonged QT intervals or electrolyte abnormalities.

Monitoring during sedated endoscopy typically involves assessment of the electrocardiogram (EKG) pattern, blood pressure, pulse oximetry and patient responsiveness;17 in many units, capnography is also used, particularly with propofol sedation. Experimental methods such as bispectral index monitoring, a form of electroencephalography, are not widely available as yet.

MF sedation has been found to be effective for routine GI endoscopy.18 The safety profile for MF compares quite favorably to that of propofol. Studies have shown that the cardiopulmonary complication rates with MF sedation are generally low, between 0.05% to 0.10% for EGDs and colonoscopies.17 A recent meta-analysis of 27 studies (many of them randomized controlled trials) found that propofol was associated with similar risks of hypoxia (OR, 0.82; 95% CI, 0.63−1.07), and hypotension (OR, 0.92; 95% CI, 0.64−1.32) as MF or other traditional sedative agents; for non-advanced procedures, propofol was slightly less likely to cause complications (OR, 0.61; 95% CI, 0.38−0.99).19

The administration of propofol by anesthesia professionals in order to induce deep sedation in unintubated patients is termed “monitored anesthesia care (MAC)” and is one of the most common sedation methods for GI procedures in North America and Europe. Propofol (2,6-diisopropyl phenol) is a short-acting agent with sedative, amnestic and hypnotic properties. It is not an analgesic, but has synergistic effects when given with opioids or benzodiazepines. The target entity is the type A γ-aminobutyric acid receptor. A typical preparation consists of a 1% solution, prepared in a mixture of soybean oil, glycerol and egg lecithin, hence it is contraindicated in patients with soy or egg allergy. Propofol, a highly lipophilic compound, is 98% plasma-protein bound, and is metabolized in the liver by conjugation to glucuronide sulfate to produce water-soluble metabolites excreted by the kidneys. The time from injection to onset of sedation is 30 to 60 seconds, and its duration of effect is 4 to 8 minutes, although its duration of peak effect is considerably shorter (often 1−2 minutes). Its pharmacokinetics do not change in patients with renal or liver failure (Table 2). Because it reduces cardiac output and systemic vascular resistance, it often causes hypotension, while its sedative effects can lead to hypoventilation. Propofol can also cause pain at the injection site, but it is rare for phlebitis to occur.

Because of these issues with MAC, there have been attempts to investigate nurse-administered propofol sedation (NAPS).47 The NAPS dosing protocol varies between centers and countries. In the United States, specially trained nurses typically administer small boluses of propofol at frequent intervals to achieve moderate sedation. In most cases, an initial bolus of 30 to 50 mg is given through a rapidly running intravenous line, followed by boluses of 10 to 20 mg every 30 to 60 seconds.48 The use of NAPS instead of MAC for low-risk patients undergoing routine GI procedures seems to represent “low lying fruit” for cost savings and better resource allocation,49 but NAPS has remained one of the most controversial aspects of the GI field.3435 European and American GI guidelines endorse the use of NAPS for low-risk GI procedures, provided that practitioners undergo proper training.505152 However, these guidelines are at odds with those issued by the ASA, which states that clinicians who use propofol “should be qualified to rescue patients from any level of sedation including general anesthesia.”53

Patient controlled sedation (PCS) is analogous to the patient-controlled analgesia used in hospitals, where patients can control their own dosage of pain medications. PCS has been looked at in several studies and found to be effective compared against MAC66 and conventional MF-type sedation.676869 However, PCS is not widely used because most patients are reluctant to take charge of their own sedation.

The SEDASYS® System (Ethicon Endo-Surgery, Inc., Cincinnati, OH, USA) was approved in the United States in 2013 for providing moderate sedation for ASA class 1 and 2 patients undergoing routine EGD and colonoscopy. Despite considerable media attention, there still seem to be some misconceptions about this technology. SEDASYS® is not a “robotic anesthesiologist.” It is a system with built-in safeguards to allow trained non-anesthesiologist physicians on-label access to propofol. Also, SEDASYS® is not a “closed-loop” system. Depending on patient vital signs, the system may restrict, suspend, decrease or stop the propofol infusion, but any increase in drug delivery must be initiated by the clinician. Finally, SEDASYS® is not a “target controlled” system and does not control delivery based on the propofol concentration in the plasma or at the site of drug effect.

In our unit, the universal enthusiasm for non-anesthesiologist-administered propofol sedation on the part of nurses, doctors and patients motivated us to develop so-called nurse-administered propofol continuous infusion sedation (NAPCIS), which is a propofol delivery method that replicates the capabilities of CAPS by using programmable intravenous fluid infusion pumps commonly available in the hospital (Alaris® pump module; Becton Dickinson, Franklin Lakes, NJ, USA). Patient selection criteria for NAPCIS are similar to those for CAPS; thus, NAPCIS is reserved for relatively healthy patients undergoing non-advanced, elective, outpatient GI procedures.

Many of the factors that pose as obstacles to the adoption of NAPS and CAPS also apply to NAPCIS as well. These include various regulatory, political and medicolegal issues. However, the most important obstacles are probably financial. At least in the United States, the current insurance coverage system makes MAC more financially attractive than NAPS or NAPCIS for providers,37 thus explaining the dramatic increase in MAC use while the use of non-anesthesia propofol sedation remains uncommon.10 However, with the anticipated advent of bundled payments for screening and surveillance colonoscopy,868788 non-anesthesiologist administered propofol sedation may become more attractive. Colonoscopy is an ideal candidate for bundling because it is a well-defined episode of care with little variability; elective EGD may also be an appropriate candidate. Bundling would involve payment of a prenegotiated amount that would cover the costs of the index procedure, any sedation or anesthesia, any pathologic analysis of biopsies, any repeat procedures (within a short period) performed because of poor bowel preparation or complications such as post-polypectomy bleed, and any secondary examinations such as virtual colonoscopy in cases of incomplete colonoscopy. Other methods of cost control, such as reference pricing, are also being investigated.8990 There is a delicate balance between cost control and maintenance of quality, making this an area of active research.

In conclusion, endoscopic sedation continues to evolve. In developed countries, it is expected that patients will demand more potent forms of sedation in the future, and thus propofol is likely to assume greater importance. Anesthesia-administered propofol sedation is effective and safe, but is limited by access and cost. CAPS is effective but unfortunately no longer available; nevertheless, non-anesthesia administered propofol sedation continues to be possible in the form of NAPS and NAPCIS. NAPS is already widely used in many European countries. The introduction of bundled payments may eventually make NAPS or NAPCIS the modality of choice in the United States. The roles played by these modalities will be determined by ongoing research on their efficacy and safety, as well by regulatory, medicolegal and financial factors.