Oral transmucosal fentanyl citrate (OTFC) is the only medication with a Food and Drug Administration labeled indication for preoperative sedation in children. (1) At the time of this study, three doses were available, containing fentanyl citrate 200, 300, or 400 micro g in a sweet, raspberry-flavored lozenge attached to a plastic stick. While previous studies have suggested an optimal OTFC dose of between 15 and 20 micro g/kg, such use was accompanied by a high incidence of postoperative nausea and vomiting, and occasional postoperative respiratory depression requiring naloxone [1-3]. Preoperative vomiting has been noted in several studies, especially at doses higher than 20 micro g/kg [1,3,4]. Accordingly, the labeled dose in healthy children was decreased to 10-15 micro g/kg.
(1) Physicians Desk Reference. Montvale, NJ: Medical Economics, 1996.
In previous studies of OTFC, children of all ages were combined in the data analysis. Since young children are the patients most likely to require preoperative sedation, and because the recommended dose is less than the optimal dose, we designed this Phase IV study to investigate the safety and efficacy of OTFC 15 micro g/kg in children aged two to six years. In addition, we wished to determine whether a propofol infusion, after mask induction of anesthesia, would reduce the incidence of postoperative vomiting to that seen when intravenous (IV) fentanyl was administered intraoperatively.
With institutional review board approval and informed, written parental consent, we enrolled 19 ASA physical status I and II patients aged 2 through 6 yr undergoing brief (30-120 min) elective surgical procedures associated with postoperative pain. Children undergoing strabismus repair were excluded, as were those taking preoperative opioids or weighing less than 13.3 kg. Only parents who responded affirmatively that their children were anxious about their surgery were asked to allow their children to participate in the study.
Baseline sedation, apprehension, and cooperation were scored (Table 1), and respiratory rate and SpO2 were measured. All scores were assigned by the clinical research nurses involved in the study (ATM, JBL). Parents also scored their child's anxiety (1 = none, 2 = minimal, 3 = some, 4 = moderate, 5 = severe) and ease of separation (1 = poor, 2 = not very good, 3 = OK, 4 = very good, 5 = excellent).
Patients were randomized to receive either OTFC or an identical placebo consisting of the sweetened matrix without the opioid (placebo lozenge). The formulation of OTFC (200, 300, or 400 micro g fentanyl citrate) that provided as close to 15 micro g/kg without exceeding this limit was selected. The pharmacy provided a coded lozenge and a syringe containing either fentanyl citrate (50 micro g/mL) or saline for intraoperative use in the placebo lozenge and OTFC patients, respectively. Investigators and patients were blinded to the identity of the lozenge and the contents of the syringe.
The children were given instructions to suck and not to chew the lozenge, and were continuously monitored for compliance by a study investigator. Heart rate and saturation (by pulse oximetry) and respiratory rate were measured and recorded every 5 min once ingestion of the lozenge began. Sedation, apprehension, and cooperation scores were assessed every 10 min until the induction of anesthesia (Table 1). A separation score was measured when the patient was taken from the parents to go to the operating room (OR). Patients were separated from their parents between 30 and 60 min after the start of lozenge ingestion, according to when the OR became available.
After arrival in the OR, monitors were applied and induction of general anesthesia with nitrous oxide, halothane, and oxygen commenced. Halothane was introduced at an inspired concentration of 0.5% and increased by 0.5% every two to three breaths to a maximum of 5%, or until the patient lost the eyelash reflex (induction time). The induction time and the maximum halothane concentration reached were recorded. An induction cooperation score was assessed (Table 1). After obtaining IV access, vecuronium 0.1 mg IV was given and the halothane was discontinued. Propofol 1.5 mg/kg was given IV and an endotracheal tube was placed. Anesthesia was maintained with a 2:1 mixture of nitrous oxide to oxygen and a propofol infusion was started at 150 micro g [centered dot] kg-1 [centered dot] min-1. The infusion was decreased by 25 micro g [centered dot] kg-1 [centered dot] min-1 approximately every 10 min if the patient was judged clinically to be adequately anesthetized. Fifteen minutes before the anticipated end of surgery, the propofol infusion was discontinued and 1 to 2 micro g/kg fentanyl (in the placebo lozenge group) or the equivalent volume of saline (in the OTFC group) was given from the previously described blinded syringe. Neither antiemetics or nonsteroidal antiinflammatory drugs were given intraoperatively. Ilioinguinal-iliohypogastric nerve blocks were administered in children undergoing herniorrhaphy or hydrocele repair. During surgery, 20-30 mL/kg of crystalloid were administered.
At the end of surgery, the stomach was decompressed with a 14 Fr suction catheter; the aspirated volume was not measured. Residual neuromuscular blockade was reversed with IV neostigmine, 50 micro g/kg, and atropine, 20 micro g/kg. Patients were tracheally extubated when they were breathing regularly, moving purposefully, and demonstrating clinical evidence of adequate muscle strength.
The children were transported to the postanesthesia care unit (PACU) on their sides and breathing room air. On arrival, vital signs and SpO2 were determined, then blow-by oxygen was administered until the children awakened spontaneously. No attempts were made to accelerate awakening by stimulation, and fluid intake was not encouraged in the PACU. Aldrete and behavior scores (Table 1) were assessed at 5-min intervals for the first 30 min after spontaneous awakening, and vital signs and pulse oximetry were recorded at 5 to 15 min from the time of arrival in the PACU. Children were considered ready for discharge from the PACU when their Aldrete score reached 9. Parents were given a questionnaire and contacted 24-48 h later to determine the incidence of nausea and vomiting after discharge from the PACU. They were also asked to describe the extent of their child's preoperative anxiety, how well their child handled separation from them, and whether they thought their child had received the active drug or the placebo.
Episodes of vomiting, retching, and respiratory depression were recorded in the pre- and postoperative periods. The occurrence of preoperative pruritus was also noted.
Quantitative data between the two groups were compared using analysis of variance, while descriptive score data were compared using the Mann-Whitney U-test. Incidence comparisons were made using Fisher's exact test. P < 0.05 was required to claim statistical significance. Values are reported as the mean +/- SD. The study was originally planned to enroll 40 patients, but the protocol was terminated after 19 patients because of an unacceptable incidence of preoperative vomiting. Statistical implications of stopping the study early are addressed in the Discussion.
Nineteen patients were enrolled in this study, 10 of whom received OTFC, and 9 of whom received the placebo lozenge. There were no significant differences between the OTFC and placebo groups in age (4.6 +/- 1.3 vs 5.2 +/- 1.1 yr, respectively), weight (21.1 +/- 3.3 vs 22.5 +/- 6.6 kg), physical status (PS) (7 PS I and 3 PS II versus 7 PS I and 2 PS II), or type of surgery (six tonsillectomy/adenoidectomy, four hernias versus four tonsillectomy/adenoidectomy, three hernias, one popliteal cyst). Baseline sedation, apprehension, and cooperation scores were similar between the groups (Figure 1).
The lozenges were readily accepted, and all children consumed 100% of the offered dose. None of the children chewed the lozenge. The dose of fentanyl in the OTFC group was 14.7 +/- 1.7 micro g/kg. There were no significant differences between the OTFC and the placebo group in the time to consume the lozenge (13 +/- 4 vs 14 +/- 4 min, respectively), the time from the start of ingestion to separation from the parents (49 +/- 9 vs 52 +/- 9 min), the induction time (110 +/- 68 vs 121 +/- 31 s), the duration of surgery (41 +/- 8 vs 44 +/- 9 min), or the time from extubation until PACU discharge criteria were satisfied (82 +/- 32 vs 68 +/- 29 min). Patients who had been given the placebo lozenge received fentanyl 1.2 +/- 0.3 micro g/kg IV during the procedure.
Children who received OTFC had a significant increase in sedation 45 min from the start of ingestion, while the placebo lozenge patients maintained their initial level of alertness and activity (Figure 1). Most children were only mildly apprehensive at baseline, and this apprehension diminished equally in both groups over time. Cooperation was very good in almost all patients and this continued throughout the preoperative period in both groups. There was no difference in the parental separation score between the OTFC and placebo groups (1.6 +/- 0.8 vs 1.8 +/- 1.3, respectively), with 18 of 19 scores being rated as good or excellent. Cooperation at induction was equally good between the groups (1.7 +/- 0.7 vs 1.6 +/- 0.9, respectively). Induction times were similar in the two groups, although there was a trend for loss of the eyelash reflex to occur at a lower inspired halothane concentration in the OTFC group (3.5% +/- 1.9% vs 4.7% +/- 0.8%; P = 0.09).
During the preoperative period, the lowest SpO2 observed in the OTFC group was 95% versus a low value of 96% in the placebo group. No children required supplemental oxygen in the preoperative period. The lowest respiratory rate recorded was 12 breaths/min noted in a patient who had received OTFC. In the OR just before induction, there were no statistically significant differences noted in heart rate, respiratory rate, blood pressure, or SpO2. On arrival to the PACU, these variables were also not significantly different between the two groups (Table 2).
Preoperative pruritus was more common in patients who received OTFC than in those who received the placebo lozenge (9/10 vs 0/9, respectively; P < 0.001). These patients exhibited mild facial pruritus, usually manifested by occasional rubbing of the nose. No children appeared to be distressed by this, and treatment was not required. Preoperative pruritus was not observed in the children who received the placebo lozenge. Postoperative pruritus was not observed.
Preoperative vomiting of a copious amount of clear, pink-tinged fluid occurred in 3 of 10 patients who received OTFC (P = not significant). No placebo lozenge patients vomited or complained of preoperative nausea. In one patient, vomiting took place in the holding area in the presence of the parents. In the two other patients, vomiting occurred in the OR just as anesthesia was to be induced. In one patient who vomited in the OR, it was elected to delay the procedure because she continued to complain of nausea. The patient was returned to her room and brought back to the OR several hours later, at which time she had an uneventful induction of anesthesia by mask. The other two patients who vomited preoperatively were observed for a period of time in the OR and had no further episodes of emesis or complaints of nausea. An inhalational induction of general anesthesia was then uneventfully completed.
There was a trend for a higher incidence of postoperative vomiting in the PACU in the OTFC group than in the placebo lozenge group (5/10 vs 1/9; P = 0.09). However, in the first 24 h after discharge from the PACU, the incidence of vomiting was equivalent (4/10 vs 3/9, respectively).
One patient receiving OTFC developed rigidity during the induction of anesthesia which was relieved by the administration of vecuronium. The episode lasted about 2 min, during which time the lowest SpO2 recorded was 85%. Ventilation was possible (as assessed by the continued presence of a capnogram trace), although an inspiratory pressure of approximately 35 cm H2 O was required.
We were able to reach by telephone 9 of 10 parents in the OTFC group and 9 of 9 patients in the placebo group. Of the nine patients who had received OTFC, parents of six thought they had received the active drug, while the other three believed they had received the placebo. Five of eight parents whose children had received the placebo thought they had received the active drug, while the other three correctly surmised that they were given the placebo. There were no statistical differences between OTFC and placebo groups in the parents' perceptions of their children's baseline level of anxiety (2.8 +/- 1.1 vs 2.4 +/- 1.2, respectively), their anxiety after ingestion of the lozenge (2.3 +/- 1.0 vs 1.9 +/- 1.1), or the ease with which their children separated from them (4.9 +/- 1.0 vs 5.0 +/- 1.0).
There were no differences between the OTFC group and the placebo lozenge group in the time from extubation to spontaneous awakening in the PACU, or to time until ready for discharge from the PACU.
Although the children who received OTFC were more sedated 45 minutes after the start of ingestion than those who received the placebo lozenge, OTFC did not result in a smoother separation from the parents or better cooperation during induction of anesthesia. An explanation for this finding may be that the baseline level of apprehension was low in both groups, and the presence of the staff attending to the preoperative monitoring served as a further calming influence. Thus the "placebo" group actually received a form of treatment (attention plus a pleasant tasting lozenge).
In the OTFC dose we studied (approximately 15 micro g/kg), there was no difference in either SpO2 or respiratory rate between the two groups. However, the size of this study is inadequate to conclude that patients receiving this dose should not be monitored for possible adverse respiratory events. Wherever OTFC is used, we recommend having the facilities to supply supplemental oxygen by bag and mask and the ready availability of naloxone, should either be required.
The 30% incidence of preoperative vomiting we noted was of significant concern, and caused us to prematurely terminate this study halfway through the intended enrollment. In each case where a patient vomited, we suspected on clinical grounds that the patient had received OTFC. The study monitors at Abbott Laboratories (N. Chicago, IL) were informed of these incidents, and after the third incident, it was jointly decided to terminate the study short of the intended enrollment of 40 patients. Although we could not measure the amount of the vomitus (since it ended up on the floor), we estimate that it was at least several ounces. This material was clear and pinktinged, and presumably consisted of the dissolved lozenge along with swallowed oral secretions. In a previous study of gastric residual volumes after OTFC administration, patients receiving OTFC or a placebo lozenge had larger gastric residual volumes than children not ingesting a lozenge (14.6, 15.6, and 7.6 mL, respectively) . The amount of vomitus we estimated appears to be considerably larger than was measured by aspiration. This may be because only patients with the largest residual volumes vomit, resulting in a volume larger than the average residual volume measured by aspiration. There are also methodological problems with measuring residual gastric volume by aspiration that lead to underestimation of the stomach contents .
In a previous study, preoperative nausea (without vomiting) was noted in 2 of 12 patients after 15-20 micro g/kg OTFC, and vomiting occurred in 1 of 12 patients who received 20-25 micro g/kg . In another study, 6 of 20 children with congenital heart disease experienced preoperative vomiting after 20-25 micro g/kg OTFC was administered . In none of these studies did any child who received a placebo or control premedication vomit.
The reason for the high incidence of preoperative emesis we noted after a smaller dose of OTFC is not clear. Perhaps the combination of the opioid and the ride back to the OR on a gurney was sufficient to trigger vomiting. It was disconcerting that two of the three children who vomited did so without warning just as we were about to induce anesthesia. If this had happened a minute or so later, significant morbidity could have occurred.
Because we terminated the study on the occurrence of a preoperative vomiting event, the incidence of preoperative vomiting may be overstated by our study. However, the incidence we observed (30%) is similar to that reported previously [3,4]. Because of the small size of our study, the difference between groups in the incidence of preoperative vomiting did not reach statistical significance. However, clinical concerns about the safety of our patients made us consider it imprudent to complete the protocol. The small sample size also makes it possible that a real, but statistically insignificant difference might have been missed (Type II error). However, the small, statistically insignificant differences in parental separation and induction cooperation scores observed are not of clinical importance. For the observed increase of 14 minutes in the OTFC group for the time from extubation to discharge from the PACU to reach significance, power analysis suggests a study of 200 patients would be required. Thus, our study has insignificant power to exclude the possibility that use of OTFC may delay recovery.
Despite the use of a propofol infusion during maintenance of anesthesia, the incidence of postoperative vomiting in the OTFC group was greater than in the placebo group, where IV fentanyl was administered during the procedure. Thus, consideration should be given to administering an additional antiemetic drug during surgery if OTFC is used for preoperative sedation. Propofol infusion for maintenance of anesthesia after a mask induction does not appear to be effective in reducing postoperative vomiting to a clinically acceptable level.
A 65% to 100% incidence of mild preoperative pruritus (usually rubbing of the eye or nose) has been previously noted after OTFC [2-4], similar to our finding. This occurrence did not appear distressing either to our patients, or to patients with pruritus in other studies [3,7]. It has been suggested that facial pruritus may be a useful sign that heralds the onset of the opioid effect .
The patient we observed who developed rigidity during induction is the first patient reported with this complication after OTFC. This patient was induced with nitrous oxide, oxygen, and halothane, and rigidity occurred at loss of consciousness. We believe the difficulty we encountered in ventilating him was rigidity and not upper airway obstruction or laryngospasm for several reasons. First, we were able to move the chest and cause the exchange of small amounts of gas (as displayed on the capnogram), albeit at a greatly increased inspiratory pressure. Second, there were no inspiratory sounds indicative of laryngospasm. If the patient had had complete laryngospasm, we would not have been able to ventilate him at all. There was no audible evidence of partial laryngospasm. The rigidity resolved after administration of a muscle relaxant and without airway manipulation. One would not expect upper airway obstruction (e.g., from a posteriorly displaced tongue) to be relieved by muscle paralysis. Although rigidity is probably an uncommon occurrence after OTFC, it should be included in the differential diagnosis if one experiences difficulty in ventilating a patient during induction after its preoperative administration, and appropriate corrective measures instituted.
In summary, we determined that the use of oral transmucosal fentanyl citrate 15 micro g/kg in children aged two to six years was associated with a high incidence of preoperative vomiting. Although these children became more sedated than the children who received a placebo lozenge, the ease of parental separation and cooperation with a mask induction was not different between the two groups. We do not recommend 15 micro g/kg OTFC as a preoperative sedative in children six years old or younger.