Sedation in endoscopy: better administered by the anesthetist than the gastroenterologist?

Sedation in endoscopy: better administered by the anesthetist than the gastroenterologist?

Till Wehrmann (Wiesbaden)

Gastrointest Endosc 2017; 85: 101-108

Patient safety during sedation with anesthesia professionals during routine upper endoscopy and colonoscopy: an analysis of 1,38 million procedures.
JJ Vargo, PJ Niklewski, JL Williams, JF Martin, DO Faigel.

This article has an accompanying continuing medical education activity on page e19. Learning Objective: Upon completion of this test, successful learners will be able to: (1) define post-ERCP pancraetitis; (2) discuss the efficacy of rectal indomethacin to prevent post-ERCP pancre-atitis; (3) identify risk of UGI bleeding among ERCP patients given rectal indomethacin.

Backround & Aims

Sedation for GI endoscopy directed by anesthesia professionals (ADS) is used with the intention of improving throughput and patient satisfaction. However, data on its safe-ty are sparse because of the lack of adequately powered, randomized controlled trials comparing it with endoscopist-directed sedation (EDS). This study was intended to determine whether ADS provides a safety advantage when compared with EDS for EGD and colonoscopy.


This retrospective, nonrandomized, observational cohort study used the Clinical Out-comes Research Initiative National Endoscopic Database, a network of 84 sites in the United States composed of academic, community, health maintenance organization, military, and Veterans Affairs practices. Serious adverse events (SAEs) were defined as any event requiring administration of cardiopulmonary resuscitation, hospital or emer-gency department admission, administration of rescue/reversal medication, emergency surgery, procedure termination because of an adverse event, intraprocedural adverse events requiring intervention, or blood transfusion.


There were 1,388,235 patients in this study that included 880,182 colonoscopy pro-cedures (21% ADS) and 508,053 EGD procedures (23% ADS) between 2002 and 2013. When compared with EDS, the propensityadjusted SAE risk for patients recei-ving ADS was similar for colonoscopy (OR, .93; 95% CI, .82-1.06) but higher for EGD (OR, 1.33; 95% CI, 1.18-1.50). Additionally, with further stratification by American Society of Anesthesiologists (ASA) class, the use of ADS was associated with a higher SAE risk for ASA I/II and ASA III subjects undergoing EGD and showed no difference for either group undergoing colonoscopy. The sample size was not sufficient to make a conclusion regarding ASA IV/V patients.


Within the confines of the SAE definitions used, use of anesthesia professionals does not appear to bring a safety benefit to patients receiving colonoscopy and is associated with an increased SAE risk for ASA I, II, and III patients undergoing EGD. (Gastrointest Endosc 2017;85:101-8.)

What you need to know

In the German-speaking countries (and also in Scandinavia), the majority of endoscopic procedures (usually diagnostic) that were carried out in the 1960s and 1970s were conducted without sedation. The first attempts to perform endoscopic retrograde cholangiopancreatography (ERCP) in Germany were also often without sedation — in heroic patients. During the 1970s, sedation with benzodiazepines (plus opiates as well if appropriate) became the standard, at least for therapeutic endoscopy. Those who carried out their own first attempts at ERCP with this regimen will well remember the (retrospectively) indescribable conditions that resulted from the combination of an inexperienced ECRPist and this sedation regimen! As I had an inner drive to do more and more ERCPs myself — including successful procedures (instead of having to call the senior physician because the endoscopy nurse thought the patient had now had enough …) — the introduction of propofol sedation at a few institutions in Germany came just in time for me. The patient did not have to suffer so much any more (or at worst from pancreatitis afterwards), and since very rapid completion of the ERCP was no longer necessary, the success curve improved. The introduction of propofol sedation provided the basis without which the development of highly invasive methods such as transgastric pancreatic interventions at the end of the 1990s would never have become possible. Anesthetists were left out at that time (with the exception of a very few institutions in Germany — e.g., the University of Mannheim — there was no “tradition” for them to be involved in gastroenterological endoscopy), and they only raised a warning finger to point out that propofol was an anesthetic agent and was therefore reserved for anesthetists alone (as specified in the specialist information for the drug at the time). But this argument was not really convincing, since at that time there was quite substantial usage of propofol in Germany by internal-medicine specialists, due to the long-term sedation of patients in internal-medicine intensive-care units. It was consequently not difficult to achieve a precedent verdict at the Federal Court of Justice in 1996 that permitted the use of propofol by internal-medicine specialists. That was the breakthrough — although not all endoscopists were convinced about it at even at that time (see the comments on the Endo-Updates at the time).

When in Rome, do as the Romans do: in the Anglo-American countries, even unsedated gastroscopy procedures were a rarity in the 1970s. Premedication, often with an individual “cocktail,” with all sorts of sedatives and analgetic agents, was the standard. There as well, however, anesthetists — at least in most states in the USA, in Canada, and the UK — were left out. However, the introduction of propofol into anesthesia (away from “inhalation anesthesia” and toward intravenous anesthesia) and into intensive-care wards (with long-term sedation of ventilated patients) took longer than in Europe, and American gastroenterologists’ interest in propofol was initially roused only by the first randomized studies from Europe [1, 2]. However, since interventional endoscopy in particular received generous remuneration in the USA and sedation procedures were also reimbursed — in contrast to the situation in Germany, for example — there was tremendous resistance from American anesthetists from the very start [3]. The process came to a head in 2010 with a policy decision by the funding institutions not to reimburse propofol sedation by gastroenterologists any longer — so that the great majority of endoscopists were forced either not to offer any more propofol-based sedation or to make use of anesthetic services. In our neighboring country of France, by contrast, developments proceeded quite differently. Unsedated endoscopies were already a minority of cases even in the 1970s, but sedation in hospitals was always administered by anesthetists, who accompanied the introduction of increasingly interventional endoscopy with propofol-supported general anesthesia. Resistance from anesthesiologists also developed in other countries in Europe in which sedation qualified for reimbursement, leading in 2011 to the withdrawal of the relevant specialist societies [4] from a European Society of Gastrointestinal Endoscopy (ESGE) guideline that had been developed jointly with them a year earlier [5].

With the rapid spread of propofol sedation in endoscopy in Germany starting after 2000, the majority of diagnostic endoscopies, even in the outpatient field, were then carried out with propofol [6]. As with the new introduction of any method, this led to a slight increase in complications, which in case of legal disputes were exclusively assessed and evaluated by anesthetists. The Board of the German Society for Gastroenterology, Digestive and Metabolic Diseases (DGVS) therefore decided in 2006 to commission a joint guideline with German anesthetists on the topic [7]. This joint guideline led to a slightly better legal “arms balance” before the courts, although some of the standards set by anesthetists were scarcely welcomed with any enthusiasm by many gastroenterologists. However, the numbers of court cases resulting from complications of sedation have been declining since 2008 — even though sedation is in the meantime administered in nearly 95% of cases [8].

For anesthetists in almost every country in the world, administering sedation during endoscopic procedures was unfamiliar territory and it held no particular attraction for them, even for the practitioners’ own careers. In countries in which the main burden of endoscopies had to be borne by anesthetists (France, for example), therapeutic endoscopies were therefore carried out with total intravenous anesthesia (TIVA), since deep sedation with spontaneous respiration seemed impractical and dangerous to the anesthetists (despite the fact that many studies demonstrated the opposite, although only published in the gastroenterological literature). This ultimately led to the scientific interest in this setting on the part of anesthetists worldwide remaining very limited. In Germany, however, this proved to be fortunate, as it led to the first joint guideline [7] remaining mainly in the hands of gastroenterologists due to the much more extensive literature on the topic published by gastroenterologists rather than anesthetists. Even up to the end of 2013, there were still only very few anesthesiological case series on sedation or anesthesia in endoscopy patients, including only a few tens of thousands of patients — in comparison with gastroenterological series that included more than a million cases [9].

The present study by John Vargo et al. [10] reports a retrospective analysis of data from 2002 to 2013 in the U.S. registry of the Clinical Outcomes Research Initiative (CORI) National Endoscopic Database. The registry included data for 880,182 colonoscopies and 508,053 gastroscopies that were carried out at university and non-university hospitals, as well as outpatient institutions. The data were separated according to whether sedation was carried out by the endoscopy team (in just under 78% of cases) or by an anesthetic team (approx. 22% of cases). The statistical analysis used a statistical model developed by Rubin in 1983, known as the “propensity score analysis,” which is intended to reduce the distortion of retrospectively obtained results resulting from covariate influences (bias) and hence permit retrospective quasi-randomization (with very large data volumes). The results for the two sedation strategies (endoscopist-led vs. anesthetist-led sedation) were evaluated relative to the occurrence of adverse events (AEs). Adverse events included a need for resuscitation measures, admission to emergency or intensive-care units, emergency surgery, premature termination of the procedure, mask ventilation, drug-administered adjustment of the depth of sedation, or administration of blood transfusions. The problems were thus not only sedation-related, but also typical sequelae of endoscopy (bleeding, perforation, etc.). In addition, very severe adverse events (such as resuscitation, etc.) were thus treated as equivalent to comparatively minimal problems (administration of benzodiazepine antagonists or opiate antagonists, short periods of mask ventilation).

The rate of AEs during gastroscopy with anesthetist-led sedation was 0.39%, higher than with endoscopist-led sedation at 0.32%. By contrast, the rate of AEs during colonoscopy was lower in the anesthetist group (0.20%) than in the endoscopist group (0.28%). The overall mortality (n = 10) was one in 138,824 and was similar with the two sedation regimens, both in gastroscopy and colonoscopy. Overall, six of the deaths were classed as potentially sedation-associated, with five in the gastroscopy group and one in the colonoscopy group.
In the traditional multivariate analysis, ASA class IV/V vs. ASA I–III was the most significant risk factor for the occurrence of AEs (odds ratio 5.85 in colonoscopy and 7.0 in gastroscopy). By contrast, there was only a moderate increase in risk with anesthetist-led sedation in comparison with endoscopist-led sedation in gastroscopy (OR 1.34) and only a minimal increase with colonoscopy (OR 1.18). In the propensity score analysis, the risk with anesthetist-led sedation in gastroscopy was significantly increased (OR 1.33, range 1.18 to 1.50), while it was equivalent in colonoscopy (OR 0.93, range 0.82 to 1.06). The reason for the increased risk with gastroscopy in the anesthetist group was mainly a markedly increased requirement for airways management, at 0.14% versus only 0.02% in the endoscopist-led sedation group.
The study thus certainly shows clearly that sedation in simple endoscopy procedures is at least no safer when it is anesthetist-led than when it is administered by gastroenterologists, even though it probably involves much higher resource costs. Does that mean everything is clear?

From the German point of view, the answer is unfortunately: not entirely. From our viewpoint, the study is “comparing apples with oranges.” In the endoscopist-led sedations, hardly any propofol-based sedation was used (2.7% of cases), while the sedation regimen in the anesthetist-led group is not described at all. In addition, the composition of the sedation teams is not reported (sedation by a non-anesthesia nurse, by the physician conducting the endoscopy, by a second gastroenterologist, by the anesthetist with or without a nurse, only by an anesthesia nurse, etc.). Essentially, the study thus probably reflects the current situation in the United States: propofol administration by anesthetists versus benzodiazepine with or without an opiate administered by gastroenterologists. Since in addition every single AE was evaluated as equally “severe,” it is not surprising that the AEs in the gastroenterologist group predominantly involved drug antagonists, while those in the anesthetist group mainly involved airways management — two events that cannot necessarily be regarded as equally severe in clinical terms.

Another study in the United States [11], based on reimbursement data for over 3 million colonoscopy procedures, also reports a higher rate of complications (in this case associated not so much with sedation, but rather with more frequent perforations or postoperative bleeding) when an anesthesia service is used — which at least does not contradict the results presented by Vargo et al. [10]. In addition, a smaller but prospective and randomized study in Portugal [12] provides evidence that propofol sedation by the gastroenterologist during colonoscopy is at least not inferior to sedation by the anesthetist in relation to safety.

A study published simultaneously in the American Journal of Gastroenterology by the same research group as the article being discussed here needs to be seen in the same context [13]. It is based on the group of patients at their own hospital. Between 2010 and 2013, 1176 ERCPs and EUS examinations were carried out under sedation with benzodiazepines/opiates administered by the gastroenterologist (60% of cases), or with propofol administered by the anesthetist (40% of cases). The retrospective analysis showed that with the same complication rates, premature abandonment of the examination was documented in 13% of cases with the gastroenterologist and in 9% of the anesthetist group. In the propensity score analysis that was also used here, it was possible to match 417 cases, which clearly showed that in the gastroenterologist-led group, sedation problems led to the procedure being abandoned significantly more often (OR 8.9) than technical endoscopic problems (OR 1.2). Does this have anything to do with anesthetists versus gastroenterologists? Probably not, as there is good evidence in the literature for an increase in the efficiency of therapeutic endoscopies (particularly ERCP) with propofol sedation in comparison with benzodiazepine/opiate sedation [1, 2, 7].

Conclusions: what do these data show?

The mortality associated with sedated gastroscopies and colonoscopies is approximately one in 100,000 to one in 150,000 in ASA I–III patients; this is now confirmed not only in registry studies but also in numerous other German and international registry data (and these data can be used for patient information). Complications of sedation represent the majority of causes of acute mortality (so that unsedated endoscopy is certainly worth considering). The general complication rate is independent of whether propofol or other drugs are administered. The use of an anesthesia service, of whatever sort, does not provide any safety advantages in this setting, and (apart from possible economic considerations) on the basis of the current research it may even be associated with disadvantages in relation to patient safety (although the plausibility of the data is weak here — see above).
Larger amounts of data are only available for gastroenterologist-led sedation in relation to interventional endoscopy in ASA I–III patients receiving propofol sedation, but the few data available for anesthetists show similar results. It cannot currently be determined whether involving an anesthetist here is advantageous. This leaves patients with ASA ≥ III, those with airways risks, and patients undergoing prolonged highly interventional procedures — in whom the involvement of an anesthetist should be considered, according to the current German guideline [9] . However, no valid data of any sort are available here either.
My personal preference would certainly be for sedation to be provided by a well-trained and highly motivated physician with good equipment and staff, quite independently of the specific medical discipline he or she might belong to.


  1. Wehrmann T, Kokabpick S, Lembcke B, et al. Efficacy and safety of intravenous propofol sedation during routine ERCP: a prospective, controlled study. Gastrointest Endosc 1999; 49: 677-683
  2. Jung M, Hoffmann C, Kiesslich, R et al. Improved sedation in diagnostic and therapeutic ERCP: propofol is an alternative to midazolam. Endoscopy 2000; 32: 233-238
  3. Graber RC. Propofol in the endoscopy suite: an anesthesiologist´s perspective. Gastrointest Endosc 1999; 49: 803-806
  4. Periel A. Non-anaesthesiologists should not be allowed to administer propofol for procedural sedation: a consensus statement of 21 European Societies of Anaesthesia. Eur J Anaesthsiol 2011; 28: 580-584
  5. Dumenoceau JM, Riphaus A, Apricio JR, et al. European Society of Gastrointestinal Endoscopy, European Society of Gastroenterology and Endoscopy Nurses and Associates, and the European Society of Anaesthesiology Guideline: Non-anaesthesiologist administration of propofol for GI-endoscopy. Endoscopy 2010; 42: 960 – 974
  6. Riphaus A, Rabofski M, Wehrmann T. Endoscopic sedation and monitoring practice in Germany: Results from the first nationwide survey. Z Gastroenterol 2010; 48 :392-397
  7. Riphaus A, Wehrmann T, Weber B et al. S3-Guidelines – Sedation for Gastrointestinal Endoscopy. Endoscopy 2009; 41: 787-815
  8. Riphaus A, Geist F, Wehrmann T, et al.: Endoscopic sedation an monitoring practice in Germany: Re-evaluation from the first nation-wide survey 3 years after the implementation of an evidence and consent based national guideline. Z Gastroenterol 2013; 51: 1082 -1088
  9. Riphaus A, Wehrmann T, Hausmann J et al. Update S3-guideline: “sedation for gastrointestinal endoscopy” 2014 (AWMF-register-no. 021/014). Z Gastroenterol 2016; 54: 58-95
  10. Vargo JJ, Niklewski PJ, Williams JL, et al. Patient safety during sedation by anaesthesia professionals during routine upper endoscopy and colonoscopy: an analysis of 1,38 million procedures. Gastrointest Endosc 2017; 85: 101-108
  11. Wernli KJ, Brenner AT, Rutter CM et al.. Risks associated with anesthesia services during colonoscopy. Gastroenterology 2016; 150: 888-894
  12. Ferreira AO, Torres J, Barias E et al. Non-anesthesiologist administered propofol sedation for colonoscopy is safe in low-risk patients: results of a non-inferiority randomized, controlled trial. Endoscopy 2016; 48: 747-753
  13. Buxbaum J, Roth N, Motanedi N al. Anesthesist-directed sedation favors success of advanced endoscopic procedures. Am J Gastroenterol 2017; 112: 290-296

Related Posts

Scroll to Top