Diazepam is nowadays used very rarely because of its long half-life of 25–30 h, according to a first survey in Germany its application was stated only by around 8 % of respondents [5]. In contrast, about 80 % of respondents use midazolam [5]. Its pharmacological advantages are a shorter half-life (1,5–3 h), a better retrograde amnesia and a higher water solubility. Compared to diazepam patient tolerance and sedation efficiency are consistently well [1]. A combination of benzodiazepines and opioids (predominantly midazolam plus meperidine) was used in about one third of all gastroenterologist for colonoscopy [5]. The use in diagnostic endoscopy of the upper digestive tract (e.g., diagnostic EUS), however, is considered obsolete [5]. The advantage of a combination of benzodiazepines and opioids at endoscopy is highly controversial and can be seen rather negatively according to recent data [1, 2].

With respect to the introduction of the short-acting propofol (plasma half-life of 7–8 min) legal and personnel requirements are still the center of discussion. Numerous guidelines for sedation in endoscopy [1–3, 9–16], were counted in the last years worldwide, their recommendations for the use of propofol are summarized in Table 28.3. All guidelines are uniform on the assumption that the endoscopist themselves can not perform the endoscopic procedure, patient monitoring and propofol administration at the same time. Demanded is therefore a further, independent assistant person not involved in the endoscopic procedure. This can be a qualified caregiver (so called NAPS, “nurse administered propofol sedation”) a second specialist in internal medicine or a gastroenterologist (usually called “ gastroenterologist -directed propofol sedation, G-DPS ”) or an anesthesia team (“monitored anesthesia care, MAC”). While in principle all guidelines advocate “MAC”, they differed with respect to the recommendation when the gastroenterologist should consider the use of anesthesia mandatory (Table 28.3). Provided however, is that the patient is monitored according to the rules of science and that for any incident the necessary personal and instrumental equipment is given [1–3]. The qualification of medical and non-medical personnel should be maintained by regular participation in structured training curricula as developed on national [17] or European basis [18] ensuring to comply with this legal requirement in different countries and again to meet the personal requirements postulated. This is of course not only in addition to the use of propofol, but also for the use of other substances for sedation or analgesia. While currently only isolated special training guidelines exist on premedication and management of emergency situations, it could have shown that specific training courses, such as those based on simulators, might improve physicians’ confidence in handling emergency situations [19].

Propofol has the distinct advantage of sedation compared to benzodiazepines, that the effect occurs much faster [20] and patients recover more rapidly [21–30]. This also applies to the regeneration of psychomotor functions, when propofol was compared with gastroscopy or colonoscopy with a combination of midazolam and meperidine by using a driving simulator [23]. Similar results were published by a study from Japan comparing propofol with midazolam for EGG [31]. The possible improvement of diagnostic accuracy in the EGD appears an additional advantage of propofol compared to midazolam in a randomized controlled trial by Meining et al. [32].

Regarding sedation with propofol the investigators evaluated the patients acceptability and tolerance for both the gastroscopy and colonoscopy, as well as for the ERCP in comparison to benzodiazepines as better [24, 25] or equally well [21, 33, 34]. In particular, for interventional examinations such as ERCP also a significantly better patient cooperation could have been shown [26, 27, 35]. Especially with the use of propofol for interventional studies one has to consider that this is not entirely without risk, as shown by data on a risk factor analysis [36]. Of 9547 patients who received propofol sedation, during the interventional upper endoscopy (EGD, ERCP, EUS) over a period of 6 years, 3151 patients received propofol as mono-sedation and the remaining 6396 a combination of propofol and midazolam. There were a total of 135 serious complications, premature termination of the procedure had to be made in 1.4 %. In 40 patients (0.4 %) a short-term mask ventilation and in nine patients (0.09 %) endotracheal intubation was necessary, another eight patients (0.08 %) had to be monitored on the ICU. Four patients died (mortality rate 0.03 %), in three cases potentially sedation associated side effects must be considered. As independent risk factors for the occurrence of cardiorespiratory complications emergency examinations and a higher dose of propofol were identified [36].

Both during the investigation and in the post-intervention phase under midazolam elderly patients are at increased risk of documented hypoxemia [37, 38]. In the elderly, it is therefore appropriate to reduce the dose of midazolam [39]. This also suggests the modification of the recommendations of the American Society of Gastroenterology for older patients undergoing gastrointestinal endoscopy [40]. In addition, substances with low cumulative dose [40, 41] should be preferred. The careful use of propofol due to its pharmacokinetic is even safe for high-risk patients aged over 85 years, as we could have shown in a randomized trial comparing propofol with a combination of midazolam and meperidine for ERCP [37]. Similar findings were obtained in a study by Heuss et al. [42]. Since cardio-respiratory events tend to occurred more frequently, there need to be increased care in these patients. Patients with liver cirrhosis are another risk group, where a hepatic encephalopathy may increase under midazolam [43–46]. This can lead to an unforeseen anesthetic stage during sedation, a prolonged wake up period with reduced psychomotor skills. In randomized controlled trial comparing midazolam with propofol for sedation in patients with liver cirrhosis undergoing interventional EGD the mentioned side effects did not occur when using the ultra-short acting propofol [46].

The propofol dose required can significantly be reduced “co-induction” with small amounts of midazolam (usually 2–3 mg) [30, 47], what particularly in often prolonged interventional endoscopic procedures is possible and valid only for this. For short duration procedures, most of which are diagnostic, should be dispensed with a co-induction with midazolam, because the savings effect is only insignificantly. In addition, the advantage of rapid psychomotor recovery when using propofol as a single agent [23] and the associated possibility of faster release should not be forgiven (discharge from the recovery room to the ward or outpatient examinations) in these cases.

Cordruwisch et al. [48] performed sedation in 64 patients, who underwent two successive, prolonged (>30 min) endoscopic examinations following up each other. In the first procedure sedation was performed with propofol and in the subsequent examination with a combination of midazolam and propofol. The combination had the advantage of a considerable saving effect of 59 % propofol. However post-interventional wake up was twice as long as in the propofol mono-sedation group (8 min versus 4 min). Van Natta et al. [49] examined in another randomized study 200 patients who received sedation with propofol either alone, propofol plus fentanyl, propofol plus midazolam or midazolam plus fentanyl. By combining process thereby moderate sedation with a shorter recovery time was reached. On the other hand correspondingly higher doses were required with sole administration of propofol, which induced a higher sedation depth leading to a substantially longer recovery time.

Akcaboy et al. [50] studied in a randomized trial in 100 patients during a colonoscopy, the sole administration of the short-acting analgesic drug remifentanil compared to mono-sedation with propofol. It was shown that remifentanil achieved an adequate sedation, amnesia and compared to propofol better analgesia. An increased incidence of nausea and vomiting during the recovery reduced this advantage, however, significantly. Moermann et al. [51] investigated the additive dose of remifentanil for sedation with propofol in a randomized double-blind trail in 50 relatively healthy patients (ASA I and II) undergoing colonoscopy. The combination of remifentanil and propofol showed significantly more often a decrease in blood pressure and oxygen saturation. By administering remifentanil the dose of propofol required could have been reduced, however, the recovery time under propofol mono-sedation was significantly shorter (p <0.01) and patients significantly more satisfied (p <0.01).