Secondary Logo

Journal Logo

Pediatric Anesthesia: Research Report

The Effect of Intranasal Fentanyl on the Emergence Characteristics After Sevoflurane Anesthesia in Children Undergoing Surgery for Bilateral Myringotomy Tube Placement

Finkel, Julia C., MD*†,; Cohen, Ira T., MD*†,; Hannallah, Raafat S., MD*†,; Patel, Kantilal M., PhD†,; Kim, Michelle S., MD*,; Hummer, Kelly A., RN*,; Choi, Sukgi S., MD†‡,; Pena, Maria, MD†‡,; Schreiber, Simeon B., MD†‡,; Zalzal, George, MD†‡

Author Information
doi: 10.1097/00000539-200105000-00016
  • Free


Bilateral myringotomy and insertion of pressure equalization tubes (BMT) is a common pediatric surgical procedure. Anesthesia for children undergoing BMT is typically administered via a face mask without establishing IV access. Even though the duration of BMT is brief, and the procedure is not perceived to be very painful, children undergoing BMT often exhibit pain-related behavior (agitation) and/or may verbally complain of discomfort in the postanesthesia care unit (PACU) (1,2). Orobello et al. (2) demonstrated that pain after BMT is present on awakening and subsides after 45–60 min. Watcha et al. (1) reported that, when an inhaled anesthetic technique is not supplemented by an analgesic, up to 76% of children undergoing BMT required pain relief in the early postoperative period.

Sevoflurane is rapidly becoming the most commonly used inhaled anesthetic for the induction of anesthesia in children. Its pleasant smell and low blood gas solubility coefficient allow for rapid and smooth induction. For brief procedures, such as BMT, sevoflurane is usually continued for the total duration of surgery. This has been shown by Lapin et al. (3) to decrease both recovery and discharge times when compared with halothane in these children. Unfortunately, emergence delirium or postoperative agitation occur in up to 67% of patients receiving unsupplemented sevoflurane for BMT surgery.

Fentanyl is a short-acting opioid analgesic that has sedative effects. Fentanyl (2.5 μg/kg IV) reduces the incidence of emergence delirium agitation in patients who have received sevoflurane and/or desflurane for adenoidectomy with or without BMT surgery in children (4). Unfortunately, because patients undergoing BMT surgery alone do not have an IV line established as a result of the brief nature of the procedure, IV fentanyl is seldom administered.

The use of intranasal fentanyl administration has a comparable analgesic effect to that of IV administration in an adult population (5). A recent study by Galinkin et al. (6) demonstrated that intranasal fentanyl 2 μg/kg achieved satisfactory blood levels for analgesia.

This double-blinded, placebo-controlled study was designed to evaluate nasally administered fentanyl as a possible technique to decrease agitation/discomfort after sevoflurane anesthesia for BMT, without compromising the rapid emergence characteristics of sevoflurane.


After we obtained IRB approval and informed parental consent, 150 ASA physical status I and II patients, 6 mo-5 yr of age, scheduled for routine BMT surgery, were studied. No premedication was used. All patients had anesthesia induced with sevoflurane (8%) in a 60%/N2O/O2 gas mixture and maintained with a 2%–3% inspired concentration via a Mapleson system and were permitted to breathe spontaneously. Patients were assigned to one of two treatment groups or to the Control group by a computer-generated randomization table. Study patients received intranasal fentanyl 1 or 2 μg/kg. The 1-μg/kg dose was diluted by 50% to make it equal in volume to the 2-μg/kg dose. The Control group received a volume of saline equal to that of the 2-μg/kg dose. The study drug was prepared in a 1-mL syringe for each patient by a member of the research team who was not involved in the care or observation of the patient being studied. The nasal solution was dripped slowly (at least 1 min after the induction to ensure adequate depth of anesthesia) half inside each nostril, with the head turned to the side so that the liquid stays in contact with the lateral surface of the nasal cavity and does not drip into the nasopharynx. (We had observed in a pilot series of patients that the depth of anesthesia immediately after the induction was insufficient to tolerate nasal instillation of fentanyl, which occasionally dripped into the larynx and produced coughing or laryngospasm.) All patients also received acetaminophen (40 mg/kg) rectally after the induction to ensure analgesia after discharge. Standard monitoring, including precordial stethoscope, electrocardiogram, blood pressure cuff, pulse oximeter, and end-tidal gas measurement, were used during anesthesia. The surgeon was asked to describe the condition of the middle ear (worst side) on a scale of 1–4 as described by Davis et al. (7) (1 = no fluid; 2 = serous fluid; 3 = pus; and 4 = thick tenacious mucus-glue ear).

At the completion of surgery (insertion of the second pneumatic equalization-tube), the volatile anesthetic and N2O were discontinued simultaneously, and patients were transferred to the PACU, breathing 100% O2. The emergence from anesthesia and recovery in the PACU were evaluated by the research nurse who was blinded to the treatment group used. The quality of emergence was evaluated by using the 4-point agitation/discomfort scale described by Watcha et al. (1) (1=calm; 2=crying, but can be consoled; 3= crying and cannot be consoled; and 4= agitated and thrashing around). A score of zero was assigned if the child was asleep. The objective pain scale was used to assess the need for analgesia (8). Recovery criteria were met when a Steward recovery score of 6 was achieved (9), at which time the children were united with their parents, and patients were evaluated by using the 4-point agitation/discomfort scale. Patients were offered popsicles or juice, but were not required to consume them before discharge. Acetaminophen (10 mg/kg orally) was used for rescue analgesia if needed (for an objective pain score of > 5 of 10). Presence or absence of vomiting was recorded. Time to meet discharge criteria from the PACU and short-stay recovery unit to home were determined by the blinded research nurse. Parents were contacted by phone 24 h after discharge to follow up on the incidence of pain, agitation, and vomiting at home.

Sample size calculations were based on the primary outcome agitation score and secondary outcome variables, emergence time to first event: cough, extubation, eye opening, purposeful movement, and recovery. A maximum sample size of 50 patients per treatment group was determined for the recovery time by a power analysis based on t-test and adjusting for nonparametric Wilcoxon-Mann-Whitney test (5%) and attrition protection (5%). This sample size also allowed the t-test to detect an effect size, δ = abs (mean A-mean B)/ς = standard deviation, > 0.75 in the agitation score of 1-μg/kg or 2-μg/kg Fentanyl groups versus Placebo Control with an α of 0.01 and a β of 0.2 (power of 80%). Here, the α is the Type I error.

The study data were summarized for each treatment group by means, standard deviation, and ranges for continuous variables and for ordinal variables by medians and interquartile ranges. The Kruskal-Wallis test of significance was used for the continuous and/or ordinal data. The Kruskal-Wallis test is a nonparametric test, and it is similar to the F test of the analysis of variance. The Kruskal-Wallis test circumvents the assumptions, required by the analysis of variance that means of the department variable come from a Gaussion distribution and the same variance within each group. For the pairwise comparison contrasting treatments versus Placebo group, the two-sample t-test and/or the nonparametric Wilcoxon-Mann-Whitney test were used. For nominal and/or categorical variables, the χ2 test and/or Fisher’s exact test was used. The multiple regression analysis was used for controlling the effect of ear condition on the agitation score. A P of value < 0.05 was considered significant.


There were no significant differences among the three groups regarding age, weight, operating surgeon, duration of anesthesia, or ear condition (Table 1). Recovery and discharge times were not significantly different. The agitation score at the time patients were reunited with their parents was significantly reduced in the 2-μg/kg Fentanyl group as compared with the Control (P = 0.012 and P = 0.004, Table 2). No significant differences, however, were found when comparing the agitation scores between the two Fentanyl-Treated groups. The incidence of vomiting was 0% in the Placebo group, 3.9% in the 1-μg/kg and 12% in the 2-μg/kg group (P < 0.025 2-μg/kg group versus placebo), though there was no statistically significant difference between the 1-μg/kg and 2-μg/kg groups (P = 0.16). Twenty-four–hour follow-up revealed no differences in analgesic requirements or vomiting as reported by parents at home (Table 2). A χ2 analysis examining the presence or absence of agitation appears in Table 3. The ear condition was positively associated with agitation scores. The Pearson correlation coefficient was r = 0.208 (P = 0.012) and Spearman correlation coefficient was r = 0.201 (P = 0.016) between the ear condition and agitation score. To control for this confounding effect of the ear condition, we performed the multiple regression analysis, and the results were similar (Table 4). After controlling for ear condition, fentanyl 2 μg/kg was significantly (P < 0.05) effective in reducing agitation scores.

Table 1
Table 1:
Demographic Data, Duration of Anesthesia Administration, and Ear Condition
Table 2
Table 2:
Quality of Recovery and Discharge Data
Table 3
Table 3:
χ2 Test Analysis: Agitation/Pain as Either 1 = Present or 0 = Not
Table 4
Table 4:
Multiple Regression Analysis to Control for Ear Condition: Agitation Score


Pain after BMT surgery usually occurs on awakening, and gradually subsides within 45–60 minutes (2). Therefore, a rational approach to providing analgesia, and preventing emergence agitation in these children requires that an adequate analgesic drug level must be present before emergence. Rectal acetaminophen is an effective analgesic, but recent evidence indicates that a larger dose than is commonly administered is required (up to 60 mg/kg rectally) (10), onset 60–90 minutes, and the peak effect is not reached until two–three hours after rectal administration (11). So, even if the drug is given in the operating room after the induction of general anesthesia, but before the brief surgery, the onset of analgesia will not coincide with emergence from anesthesia.

Fentanyl, an adjunct analgesic commonly used during the administration of anesthesia to reduce the incidence of pain, is an effective analgesic that has sedative effects. Fentanyl has a short onset time and duration of approximately one hour. It is used frequently, administered IV, during anesthesia administration in children undergoing surgical procedures expected to be associated with postoperative discomfort. Unfortunately, children who undergo BMT surgery do not require insertion of an IV line because of the extremely brief nature of the procedure and because fluid replacement is not necessary. Fentanyl, however, can be administered IM. Although effective, this approach is not widely used because of concerns about bleeding, pain, and swelling associated with IM injections in children.

The highly vascularized and large surface area of the nasal cavity allows for rapid absorption of drugs that are lipophilic and have a low molecular weight, which produces serum concentrations similar to those achieved with IV administration. In a study by Galinkin et al. (6), 2 μg/kg of intranasal fentanyl resulted in a serum concentration level that was above the minimum effective concentration for postoperative analgesia in adults.

The onset of analgesia after intranasal fentanyl appears to be dependent on the volume as well as the dose administered. In a study done in adults by Striebel et al. (12), intranasal fentanyl was administered by a metered device spray delivering six consecutive administrations of 4.7 μg each for a total dose of 27 μg with a 27-μg IV bolus of fentanyl. The mean times to onset and peak analgesic effect were slower in the intranasal (16.0 ± 12.6 minutes and 26.3 ± 15 minutes, respectively) versus the IV group (10.8 ± 9 minutes and 20.2 ± 12 minutes, respectively). In a subsequent study by the same group, Toussaint et al. (13) compared intranasal versus IV fentanyl. The intranasal fentanyl was given as a single bolus of 25 μg, and the IV bolus was 17.5 μg, to account for the 71% bioavailability of intranasal fentanyl (14). The onset of analgesia and peak effect was comparable. In each of these studies, nasal fentanyl was well accepted, nonirritating, and even in awake patients, produced side effects that were not different from the IV groups.

Several other analgesic regimens have been examined for their use in pediatric BMT patients. Bennie et al. (15) determined that nasal butorphanol 25 μg/kg reduced rescue analgesic requirements from 53% in the Placebo group to 7% after the administration of a halothane anesthetic. This dose did result in significant sedation, which resolved after more than 30 minutes in the PACU. The analgesic onset of nasal butorphanol is within 15 minutes with the peak effect occurring within one–two hours in adults, and so, unlike fentanyl, its peak effect occurs after the major painful stimulus has occurred (16). Davis et al. (7) found IV ketorolac with midazolam premedication to be effective, but it requires IV access, which is not otherwise indicated in these children. Watcha et al. (1) compared orally administered acetaminophen 10 mg/kg, ketorolac 1 mg/kg, and placebo given before a halothane anesthetic and found 76% of the patients in the Placebo group required rescue medication versus 30% in the Ketorolac group, which represented a significant reduction but not a desirable endpoint. This probably reflects the fact that oral administration does not allow for the timely establishment of a therapeutic blood level.

The 12% incidence of vomiting found with nasal fentanyl 2 μg/kg, although significantly more frequent than the 0% found in the Placebo group, is comparable to the incidence of emesis in other pediatric BMT studies (2,7,15,17) and is half that reported by Galinkin et al. (6) (24%). The impact of age on emesis is unclear in these series containing both preschool- and school-aged children because the data were not stratified by age. It may be that the relatively small incidence of vomiting found in this study may be the result of the young age of our patients (less than years) and may be further mitigated by delaying oral intake (18).

Given the nonirritating qualities and rapid onset of nasal fentanyl, its use as a rescue analgesic in BMT patients who are still agitated despite receiving 40 mg/kg of rectal acetaminophen or other analgesics, should be explored. This selective therapeutic versus routine preemptive use of the drug may have the possible advantage of an overall decreased incidence of vomiting in this population.

In conclusion, this study shows that in unpremedicated patients aged six months to five years undergoing BMT, intranasal fentanyl 2 μg/kg significantly decreased agitation and did not prolong recovery or discharge times, thus preserving the rapid emergence and recovery seen with sevoflurane anesthesia.


1. Watcha MF, Ramirez-Ruiz M, White PF, et al. Perioperative effects of oral ketorolac and acetaminophen in children undergoing bilateral myringotomy. Can J Anaesth 1992; 39: 649–54.
2. Orobello PW, Park RI, Wetzel RC, et al. Phenol as an adjuvant anesthetic for tympanostomy tube insertion. Int J Pediatr Otorhinolaryngol 1991; 21: 51–8.
3. Lapin SL, Auden SM, Goldsmith LJ, Reynolds AM. Effects of sevoflurane anaesthesia on recovery in children: a comparison with halothane. Paed Anaesth 1999; 9: 299–304.
4. Cohen IT, Hannallah RS, Hummer KA. Fentanyl prevents emergence agitation following desflurane anesthesia in children. Anesthesiology 1999; 91: 1300.
5. Striebel HW, Koenigs D, Kramer J. Postoperative pain management by intranasal demand-adapted fentanyl titration. Anesthesiology 1992; 77: 281–5.
6. Galinkin JL, Fazi LM, Cuy RM, et al. Use of intranasal fentanyl in children undergoing myringotomy and tube placement during halothane and sevoflurane anesthesia. Anesthesiology 2000; 93: 1378–83.
7. Davis PJ, Greenberg JA, Gendelman M, Fertal K. Recovery characteristics of sevoflurane and halothane in preschool-aged children undergoing bilateral myringotomy and pressure equalization tube insertion. Anesth Analg 1999; 88: 34–8.
8. Hannallah RS, Broadman LM, Belman AB, et al. Comparison of caudal and ilioinguinal/iliohypogastric nerve blocks for control of post-orchiopexy pain in pediatric ambulatory surgery. Anesthesiology 1987; 66: 832–4.
9. Steward DJ. A simplified scoring system for the postoperative recovery room. Can J Anaesth 1975; 22: 111–3.
10. Korpela R, Korvenoja, Meretoja OA. Morphine-sparing effect of acetaminophen in pediatric day-case surgery. Anesthesiology 1999; 91: 442–7.
11. Montgomery CJ, McCormack JP, Reichert CC, Marsland CP. Plasma concentrations after high-dose (45mg) rectal acetaminophen in children. Can J Anaesth 1995; 42: 982–6.
12. Striebel HW, Pommerening J, Rieger A. Intranasal fentanyl titration for postoperative pain management in an unselected population. Anaesthesia 1993; 48: 753–7.
13. Toussaint S, Maidl J, Schwagmeier R, Striebel HW. Patient-controlled intranasal analgesia: effective alternative to intravenous PCA for postoperative pain relief. Can J Anaesth 2000; 47: 299–302.
14. Streibel HW, Kramer J, LuhmannI, et al. Pharmacokinetics of intranasal fentanyl [German]. Der Schmerz 1993; 7: 122–5.
15. Bennie RE, Boehringer LA, Dierdorf SF, et al. Transnasal butorphanol is effective for postoperative pain relief in children undergoing myringotomy. Anesthesiology 1998; 89: 385–90.
16. Stadol NS (butorphanol tartrate) nasal spray [package insert]. Princeton, NJ: Bristol-Myers Squibb Co.; 1996.
17. Bean-Lijewski JD, Stinson JC. Acetaminophen or ketorolac for post myringotomy pain in children? A prospective, double-blinded comparison. Paediatr Anesth 1997; 7: 131–7.
18. Schreiner MS, Nicholson SC, Martin T, Whitney L. Should children drink before discharge from day surgery? Anesthesiology 1992; 76: 528–33.
© 2001 International Anesthesia Research Society