Share this article on:

Midazolam-Ketamine Combination for Moderate Sedation in Upper GI Endoscopy

Motamed, Farzaneh*; Aminpour, Yasaman*; Hashemian, Hesam; Soltani, Alireza E.; Najafi, Mehri*; Farahmand, Fatemeh*

Journal of Pediatric Gastroenterology & Nutrition: March 2012 - Volume 54 - Issue 3 - p 422–426
doi: 10.1097/MPG.0b013e3182323c75
Original Articles: Gastroenterology

Objectives: The aim of the study was to compare the quality of sedation with 3 different sedation regimens in upper gastrointestinal endoscopy (UGIE) in pediatric patients.

Methods: One hundred fifty consecutive children who underwent UGIE were randomly assigned to 1 of the 3 medication regimens. Patients in group A (n = 49) received placebo. Forty-five minutes after the placebo was given, repeated intravenous (IV) doses of 0.1 mg/kg midazolam were administered titrated to achieve a level of deep sedation. Patients in group B (n = 51) received oral ketamine instead of placebo, and patients in group C (n = 50) received oral fentanyl instead of placebo with the same methodology and sedation endpoint.

Results: The mean dose of midazolam administered in group B patients was remarkably lower compared with that of groups A and C. Patients in group B showed less distress in IV line placement and separation from parents, higher comfort level, more endoscopist satisfaction, and higher sedation depth compared with groups A and C. The recovery time was significantly shorter in group B. All of the 3 regimens were safe. All of the complications were managed successfully.

Conclusions: Our data suggest that synergistic sedation with oral ketamine and IV midazolam for UGIE in children is a suitable and safe sedation. The higher rate of vomiting in group B in contrast to previous studies must be caused mainly by the oral route of ketamine administration.

*Department of Gastroenterology, Tehran University of Medical Sciences

Department of Ophthalmology, Farabi Hospital, Tehran University of Medical Sciences

Department of Anesthesiology, Tehran University of Medical Sciences, Tehran, Iran.

Address correspondence and reprint requests to Hesam Hashemian, Department of Ophthalmology, Farabi Hospital, Tehran University of Medical Sciences, Tehran, Iran (e-mail:

Received 7 June, 2011

Accepted 10 August, 2011

The authors report no conflicts of interest.

Endoscopy has become more and more common for evaluation of gastrointestinal (GI) disorders. In the United States, adult endoscopies have increased 4-fold in the last 15 years (1) and a similar situation probably exists for children (2). Endoscopic procedures are frequently performed outside the operating room to minimize costs and maximize availability (3). Pediatric upper GI endoscopy (UGIE) can be completed with no sedation, intravenous (IV) sedation, or with general anesthesia (1,4,5).

Although there is consensus on the use of general anesthesia for some patients (6–8), IV sedation proved to be a safe and effective method for endoscopy sedation in children (9–11). Moderate sedation (formerly known as conscious sedation) is a depression of consciousness in which patients respond purposefully to verbal commands. Airway is maintained without intervention (12).

Various classes of drugs have been used for moderate sedation including the benzodiazepines (eg, midazolam, lorazepam, diazepam), opioids (eg, morphine, fentanyl, meperidine), and sedative-hypnotics (eg, chloral hydrate, ketamine) (4,6,13). Inhaled agents (sevoflurane and nitrous oxide) and IV propofol have also been investigated (14,15). The most commonly used medications are benzodiazepines (16). IV midazolam alone proved to be relatively safe (17) but not efficient enough in pediatric GI endoscopy because most endoscopists prefer using midazolam together with other drugs (14).

One of the most intriguing anesthetic agents is ketamine. All of the reports agree on its safety and efficacy for endoscopic sedation in children. Its synergistic effect reduces requirements for narcotics (6,13). Ketamine is contraindicated for infants younger than 3 months and patients with established psychosis (11). The pain threshold increases 30 minutes after oral ketamine. Oral administration increases the metabolite norketamine, which may have contributed to the analgesic effect (18). Fentanyl has been used alone or in combination with other agents (6,13). Oral fentanyl has been used in other studies (19) and proved to be safe and efficient. Combinations of drugs in different classes typically have synergistic effects (7,17).

The safety and efficacy of individual protocols have been proven in single drug studies, but not enough experience exists to determine the best sedation protocol for pediatric endoscopy (1,13). A few studies have compared different protocols (1). The specific aims of the present study were to compare the frequency of complications and adequacy of sedation between 3 different sedation protocols in children undergoing diagnostic upper endoscopic procedures.

Back to Top | Article Outline


One hundred fifty children, 1 to 18 years old, were included in the study after ethics committee approval and written informed consent from the parents. The present prospective study was approved by the institutional review board of the Faculty of Medicine, Tehran University.

All of the patients undergoing outpatient GI endoscopy in our institute during the study period were considered eligible. The following exclusion criteria were applied: general anesthesia administration; emergency endoscopies and interventional procedures; children with a history of allergies to benzodiazepines, ketamine, fentanyl, or to their components; cardiovascular, respiratory, metabolic, or neurologic impairments, malignancies, or renal diseases; children with previous complications with IV sedation; children with ASA grade (American Society of Anesthesiologists’ Physical Status Classification) >II (20).

Three board-certified endoscopists performed all of the procedures. All of the procedures were under the direct observation of a pediatric anesthesiologist. With the exception of a registered nurse, who prepared the study medication and monitored the patients, all of the staff was blinded to the sedation regimen.

Topical pharyngeal anesthesia with 10% lidocaine spray was administered before IV sedation in all of the patients. Patients were randomly assigned to 1 of the 3 sedation protocols by using a table of random numbers.

Patients in group A (placebo-midazolam) received orally 5 mL of normal saline. Twenty minutes later, an IV line was inserted. Twenty-five minutes after the IV line placement (45 minutes after the oral intake), patients received intravenously 0.1 mg/kg of midazolam and, if needed, repeated to achieve the level of moderate sedation up to the maximum recommended total dose of 5 mg or 0.3 mg/kg (6).

Patients in group B (midazolam-ketamine) received orally 5 mg/kg of ketamine (6,21). Ketamine was given by mouth in the form of the injectable solution, which the child was asked to swallow. Other stages were the same as group A, and IV doses of midazolam were administered the same as group A. Patients in group C (midazolam-fentanyl) received orally 2 μg/kg of fentanyl (6,21). Fentanyl was given by mouth in the form of the injectable solution. Other stages were the same as group A and B. In all of the patients, midazolam was titrated to achieve the level of moderate sedation. The drugs were diluted with normal saline. The duration of the procedure was calculated from the first injection of the midazolam to the final withdrawal of the gastroscope.

Modified Ramsay sedation scoring was used to evaluate sedation by the endoscopist (16). The modified Ramsay score is on a scale from 1 to 6, depending on the level of the patient's cooperation and response. A modified Ramsay score of R5 (deep sedation: when patient responds only to noxious stimulations) was considered a contraindication for further doses of midazolam.

Heart rate, systolic arterial pressure, peripheral oxygen saturation, respiratory rate, and Ramsey sedation scores (16) of all of the patients were recorded at baseline, after the first injection of midazolam, and every 5 minutes during the procedure by the anesthesiologist. All of the patients received oxygenation (2 L/min) through a nasal cannula.

Adverse events were recorded, including hypotension (decrease in blood pressure by 20% from baseline and below normal for age), hypertension (increase in blood pressure by 20% from baseline and above normal for age), bradycardia (decrease in heart rate by 30% from baseline and below normal for age), and hypoxia (oxygen desaturation with peripheral oxygen saturation <90%).

An endoscopy nurse recorded the following time points for all patients: the time the first IV sedation was administered, the time the endoscope, inserted was the time of the endoscope withdrawal, the time the patient exited procedure room, and the time the patients were discharged from the hospital.

The ease of the IV line placement was assessed by the pediatrician, based on a 3-point scale: 1 = unable to insert an IV line because of patient's agitation; 2 = able to insert an IV line but patient is uncooperative (severely agitated or crying); and 3 = able to insert an IV line with ease, and patient is cooperative (6). The separation from the parents was assessed by the pediatrician, based on the behavior scoring system modified by Wilton et al (15). This scoring system was graded from 1 = patient is agitated to 5 = patient is asleep. At the end of the procedure, the gastroenterologist was asked to rate the ease of the procedure on a 3-point scale (1 = impossible, 2 = adequate, 3 = easy). The time of recovery was assessed using the REACT score, a modified postanesthesia care scale (13). It is calculated using respiration and circulation parameters, consciousness, temperature, and the patient's activity, and ranges from 0 to 10. In the present study, each patient was discharged from the procedure room when the REACT score was equal to 10. The time to reach a REACT score of 10 was also recorded. The patient's evaluation of the procedure (1 = extremely uncomfortable, 2 = slightly uncomfortable, 3 = no discomfort) was obtained in all of the patients ages 6 years or older by the pediatrician 24 hours after the procedure.

Before discharge, the child was asked by the pediatrician what he or she recalled about the procedure. The pediatrician called patients 24 hours after the procedure to evaluate any late-onset adverse effect.

Continuous data were compared with the Student t test and categorical variables were compared using the χ2 test. P ≤ 0.05 was considered statistically significant. The statistical computer package SPSS version 16 (SPSS Inc, Chicago, IL) was used.

Back to Top | Article Outline


A total of 49 patients were included in group A, 51 in group B, and 50 in group C. Demographic data are presented in Table 1. Groups A, B, and C were comparable in all of the demographic characteristics.

The recovery time from sedation, as examined by the REACT score, was remarkably shorter in group B patients compared with other groups (Table 2). The mean duration from the time of IV injection to procedure end was significantly lower in group B (Table 2).

Inadequate sedation was more common in group A (10.2%) and group C (8.0%) compared with group B (3.9%), the differences were statistically significant (P < 0.005 for both). The mean dose of midazolam administered in groups A and C was significantly higher compared with that of group B (Table 2). The number of patients who did not need additional midazolam was 13 in group A (26.5%), 38 in group B (74.5%), and 24 in group C (48%). The difference between group B and the other 2 groups was statistically significant (P < 0.005 for both).

The IV line placement (assessed by 3-point scale) was easier in group B patients compared with other groups (Table 2). The level of pain induced by the IV line placement was judged by patients older than 6 years and was significantly lower in group B compared with groups A (P = 0.001) and C (P = 0.009) (Table 2).

Separation from the parents was easier in group B compared with the other groups (Table 2). Endoscopist evaluation of patients’ sedation was significantly better in group B compared with the other 2 groups. (Table 2)

Overall, 39 patients (26%) experienced adverse events. Hypoxia occurred in 11 (7.3%) and upper airway obstruction occurred in no patients. Cardiovascular complications arose in 10 (6.7%) and consisted only of hypotension.

There were no bradycardia and no serious complications during or after the endoscopy. All of the adverse events were easily managed. There were no statistically significant differences in the overall complication rate between all 3 groups. Hypersalivation, laryngospasm, hallucinations, or emergence reactions were not seen in any patient. In general, dizziness (30 patients, 20%), nausea (15 patients, 10%), and hypoxia (11 patients, 7.3%) were the most frequently observed complications. Dizziness, vomiting, and cough were more frequently reported in group B compared with the other 2 groups, although this difference was statistically significant only in comparing group B with group C for diplopia and vomiting (P values 0.004 and 0.002, respectively).

More patients in group B achieved a modified Ramsay level 4 or 5 of sedation. There was a significant difference in the quality of sedation in group B compared with groups A and C (both at baseline and 30 minutes later) as shown in Table 2.

More than 4 children were asked about procedure recall. From 131 patients, 40 (30%) remembered it completely, 55 (41.98%) remembered part of it, and 36 (27.4%) remembered nothing. The difference was not significant between either group.

Back to Top | Article Outline


To the best of our knowledge, this is the only prospective, randomized study in which the same sedative agents (midazolam) given alone or in combination with other agents (oral ketamine or fentanyl) were compared. It is also the first study to use oral ketamine and fentanyl as a premedication.

The primary outcome in the present study indicated that lower dose of midazolam was needed in the midazolam-ketamine group, compared with other 2 groups. The secondary outcomes were easier and less painful IV line placement, the ease of separation from parents, and the deeper level of sedation in the midazolam-ketamine group.

Thus far, 2 studies have used oral midazolam as a premedication to IV sedation in pediatric UGIE (6,13). The present study is the first to compare the efficacy and safety of orally administered fentanyl and ketamine as a premedication to IV midazolam for l sedation in children undergoing endoscopy.

The low rate of inadequate sedation (<4%) in group B indicates that outpatient GI endoscopy can be completed with IV sedation with midazolam and ketamine in appropriately selected patients, although manual restraints used for some patients may have caused underestimation of the sedation failure rate.

Based on the Ramsey score, group B experienced more comfortable and deeper sedation. Adequate sedation was more frequent with midazolam-ketamine. The mean duration from the time of IV injection to procedure end was comparable with other studies (22).

Group B had shorter recovery time and shorter injection to end of procedure time. Recovery times for all of the groups were comparable to recovery times reported in previous studies (14,23,24). One advantage of our study was calculation of REACT score as a discharge criterion to determine recovery time.

The rate of adverse events in our study was relatively high (26%). We detected a relatively high overall rate of adverse events in 26% of procedures. Our study was prospective, and we reported all of the minor events. This explains the higher rate of complication in the present study. All of the complications were successfully managed, and all of the procedures were safely completed. If only serious adverse events were measured, the adverse event rate was zero.

Balsells et al (25), in a retrospective study, reported minor complications in 0.35% of patients; however, their complication criteria were not as strict as in our study. Mamula et al (14) reported a 20% complication rate, comparable with our study. It should be noted that patients with serious comorbidities had been excluded from the study. We emphasize that to improve safety, only appropriate patients should be selected for moderate sedation.

Overall, cardiopulmonary complications account for >50% of the major complications during endoscopy (20), which are often related to hypoxia. In our study, cardiovascular and respiratory adverse events accounted for 6.7% and 7.3%, respectively. Kennedy et al (26) reported the frequency of hypoxia as significantly higher in the fentanyl-midazolam group (25%) than in the ketamine-midazolam group (6%; P < 0.001) (27).

In our study, the rate of hypoxia in group B was lower than group C (3.9% vs 8%), although the difference was not statistically significant. Endoscopes can potentially compress and obstruct airway (6), so respiratory complications may be common during UGIE.

Hallucination and emergence reactions are the most frequently discussed adverse effects related to ketamine (19). Petrack (28) analyzed all of the data from 1990 to 1995 on emergence related to ketamine and reported their incidence from 0% to 9%. Emergence phenomena has been reported to occur more commonly when ketamine is used as a single agent, when given in large doses, if administered rapidly (<1 minute), and when there is excessive auditory and visual stimulation during recovery (29,30). In our study, emergence phenomenon was not seen perhaps because of low dosage, long absorption time through oral route, and addition of midazolam.

According to previous studies, 8.5% of children may develop emesis with ketamine, usually late in the recovery phase (28). In the present study, vomiting was seen in 9 (17.6%) patients in group B. Vomiting was not observed in other groups, except for 1 patient in group A. This high rate of vomiting in the ketamine group (17.6%) was not mentioned in previous studies with IV ketamine. It must be an adverse effect related to the oral route of ketamine. All 3 drug combinations appear to provide safe and effective sedation for UGIE in children.

Two previous studies (26,30) compared the midazolam-ketamine combination with other regimens and observed lower distress scores during the procedure and higher physician satisfaction rates. These data are consistent with our study with the ketamine-midazolam regimen.

Ketamine has many advantages that make it an ideal sedative (28,31,32). It has a short duration (15–30 minutes) of action and rapid recovery (28,31,32). For children, it is a reliable anxiolysic, analgesic, and amnesic drug, while protecting airway reflexes and the cardiovascular system (28,31,32). In addition, ketamine dosing has a wide safe margin in comparison with narcotics and benzodiazepines (28,31–33). However, all sedative agents, including ketamine, have the potential for serious adverse effects.

In conclusion, our data suggest that synergistic sedation with an oral dose of ketamine combined with midazolam presents advantages in terms of better quality and depth of sedation, fewer sedation failures, faster recovery time, reduced benzodiazepine requirements, easier IV line placement, easier separation from the parents, less pain induced by the IV line placement, and greater patient comfort.

Back to Top | Article Outline


1. Scottish Intercollegiate Guidelines Network. SIGN guideline 58: safe sedation of children undergoing diagnostic and therapeutic procedures. Pediatr Anesth 2008;18:11–2.
2. Sury MRJ, Smith JH. Deep sedation and minimal anesthesia. Pediatr Anesth 2008; 18:18–24.
3. Dilos BM. Anesthesia for pediatric airway endoscopy and upper gastrointestinal endoscopy. Int Anesthesiol Clin 2009;47: 55–62.
4. Chuang E, Zimmerman A, Neiswender KM, et al. Sedation in pediatric endoscopy. Gastrointest Endosc Clin N Am 2001; 11:569–584.
5. Amornyotin S, Aanpreung P. Clinical effectiveness of an anesthesiologist-administered intravenous sedation outside of the main operating room for pediatric upper gastrointestinal endoscopy in Thailand. Int J Pediatr 2010 [Epub Aug 2].
6. Hargrove CB, Ulshen MH, Shub MD. Upper gastrointestinal endoscopy in infant: diagnostic usefulness and safety. Pediatrics 1984; 74:828–831.
7. Annequin D, Carbajal R, Chauvin P, et al. Fixed 50% nitrous oxide oxygen mixture for painful procedures: a French survey. Pediatrics 2000; 105:E47.
8. Gilger MA. Sedation for pediatric GI endoscopy. Gastrointest Endosc 2007;65:211–2.
9. Koh JL, Black DD, Leatherman IK, et al. Experience with an anesthesiologist interventional model for endoscopy in a pediatric hospital. J Pediatr Gastroenterol Nutr 2001; 33:314–318.
10. Gilger MA. Gastroenterologic endoscopy in children: past, present, and future. Curr Opin Pediatr 2001; 13:429–434.
11. Mistry RB, Nahata MC. Ketamine for conscious sedation in pediatric emergency care. Pharmacotherapy 2005; 25:1104–1111.
12. Molleston J, Alonso E, Oliva-Hemker M. Sedation and anesthesia in pediatric endoscopy: one size does not fit all. J Pediatr Gastroenterol Nutr 2007; 44:295–297.
13. Tolia V, Peters JM, Gilger MA. Sedation for pediatric endoscopic procedures. J Pediatr Gastroenterol Nutr 2000; 30:477–485.
14. Mamula P, Markowitz JE, Neiswender K, et al. Safety of intravenous midazolam and fentanyl for pediatric GI endoscopy: prospective study of 1578 endoscopies. Gastrointest Endosc 2007; 65:203–210.
15. Michaud L. Sedation for diagnostic upper gastrointestinal endoscopy: a survey of the Francophone Pediatric Hepatology, Gastroenterology, and Nutrition Group. Endoscopy 2005; 37:167–170.
16. Martinez JL, Sutters KA, Waite S, et al. A comparison of oral diazepam versus midazolam, administered with intravenous meperidine, as premedication to sedation for pediatric endoscopy. J Pediatr Gastroenterol Nutr 2002; 35:51–58.
17. Verhage J, Mulder CJ, Willekens FL. Intravenous midazolam sedation in pediatric diagnostic upper digestive endoscopy. A prospective study in a general hospital. Rom J Gastroenterol 2003; 12:273–276.
18. Grant IS, Nimmo WS, Clements JA. Pharmacokinetics and analgesic effects of i.m. and oral ketamine. Br J Anaesth 1981; 53:805–810.
19. Tosun Z, Aksu R, Guler G, et al. Propofol-ketamine vs propofol-fentanyl for sedation during pediatric upper gastrointestinal endoscopy. Pediatr Anesth 2007; 17:983–988.
20. Lowrie L, Weiss AH, Lacombe C. The pediatric sedation unit: a mechanism for pediatric sedation. Pediatrics 1998; 102:e30.
21. Paspatis GA, Charoniti I, Manolaraki M, et al. Synergistic sedation with oral midazolam as a premedication and intravenous propofol versus intravenous propofol alone in upper gastrointestinal endoscopies in children: a prospective, randomized study. J Pediatr Gastroenterol Nutr 2006; 43:195–199.
22. Lightdale JR, Valim C, Newburg AR, et al. Efficiency of propofol versus midazolam and fentanyl sedation at a pediatric teaching hospital: a prospective study. Gastrointest Endosc 2008;67:1067–75.
23. Menigaux C, Fletcher D, Dupont X, et al. The benefits of intraoperative small-dose ketamine on postoperative pain after anterior cruciate ligament repair. Anesth Analg 2000; 90:129–135.
24. Guignard B, Coste C, Costes H, et al. Supplementing desflurane-remifentanil anesthesia with small-dose ketamine reduces perioperative opioid analgesic requirements. Anesth Analg 2002; 95:103–108.
25. Balsells F, Wyllie R, Kay M, et al. Use of conscious sedation for lower and upper gastrointestinal endoscopic examinations in children, adolescents, and young adults: a twelve-year review. Gastrointest Endosc 1997; 45:375–380.
26. Kennedy RM, Porter FL, Miller JP, et al. Comparison of fentanyl/midazolam with ketamine/midazolam for pediatric orthopedic emergencies. Pediatrics 1998; 102:956–963.
27. Montes RG, Bohn RA. Deep sedation with inhaled sevoflurane for pediatric outpatient gastrointestinal endoscopy. J Pediatr Gastroenterol Nutr 2000; 31:41–46.
28. Petrack EM. Ketamine. Clin Ped Emerg Med 2000; 1:281–284.
29. Chuang E, Wenner WJ Jr, Piccoli DA, et al. Intravenous sedation in pediatric upper gastrointestinal endoscopy. Gastrointest Endosc 1995; 42:156–160.
30. Gilger MA, Spearman RS, Dietrich CL, et al. Safety and effectiveness of ketamine as a sedative agent for pediatric GI endoscopy. Gastrointest Endosc 2004; 59:659–663.
31. Green SM, Li J. Ketamine in adults: what emergency physicians need to know about patient selection and emergence reactions. Acad Emerg Med 2000; 7:278–281.
32. Green SM, Rothrock SG, Lynch EL, et al. Intramuscular ketamine for pediatric sedation in the emergency department: safety profile in 1,022 cases. Ann Emerg Med 1998; 31:688–697.
33. Varadarajulu S, Eloubeidi MA, Tamhane A, et al. Prospective randomized trial evaluating ketamine for advanced endoscopic procedures in difficult to sedate patients. Aliment Pharmacol Ther 2007; 25:987–997.

fentanyl; ketamine; midazolam; upper GI endoscopy

Copyright 2012 by ESPGHAN and NASPGHAN