In young children, anesthesia with sevoflurane has been shown to be accompanied by behavioral changes during recovery known as emergence agitation (EA) or emergence delirium (ED) [1,2]. EA is characterized by a period of restlessness, agitation, inconsolable crying, disorientation, delusions, and hallucinations with impaired cognition and memory [3].

ED is sometimes defined as a type of EA occurring in a few patients who experience delirium with psychomotor agitation [1,2]. The incidence of EA after sevoflurane anesthesia in children is reported to vary between 10%–80%. This depends on the definition of EA used and the duration of monitoring after emergence from anesthesia [4]. Sevoflurane and desflurane anesthesia have been shown to be risk factors for EA in children. However, the precise etiology is still unclear [4,5].

Many pharmacological interventions including propofol can minimize the incidence of EA following sevoflurane anesthesia [6–8]. A continuous infusion of propofol has been used throughout the procedure to reduce the incidence of EA in some studies [6,7]. A single bolus dose of 1 mg/kg of propofol has also been used with contrasting results [8–11]. Transition to propofol after sevoflurane anesthesia by repeated bolus administration has also been used to reduce EA [12]. An infusion of 3 mg/kg of propofol over 3 min following sevoflurane anesthesia has been found to be promising in children undergoing magnetic resonance imaging scans [13].

In this randomized trial, we aimed to determine whether transition to propofol over 3 min after cessation of sevoflurane anesthesia would reduce the incidence of EA in children undergoing inguinal hernia repair.

This prospective, single-blinded, randomized controlled trial was registered at www.clinicaltrials.gov (NCT03179293). The study was carried out at the Assiut University Paediatric Hospital, Egypt, between July 2017 to June 2018. The study protocol was approved by the Assiut Medical School Ethical Review Board (Ethical Committee N: IRB 17100561). Written informed consent was obtained from parents of all patients before inclusion in the study.

Children aged 1–12 years, American Society of Anesthesiology class I or II, assigned for elective inguinal hernia repair were included in the study. Exclusion criteria were patient or parent refusal, known allergy to propofol or egg products, a family history of malignant hyperthermia, operating time longer than 60 minutes, performance of any other procedure under the same anesthetic, other co-morbidities or congenital anomalies, neurological or psychiatric illness, use of sedatives, or the need for rapid sequence induction.

We assessed 70 children for eligibility. Five patients were excluded as the duration of surgery was > 60 min. The remaining 65 children were allocated into two groups. One patient was excluded from the final analysis due to laryngospasm requiring treatment (Fig. 1). Baseline characteristics were comparable in both groups (Table 1).

The overall incidence of EA according to the PEAD scale was 15.6% in the propofol group and 81.3% in the control group (P < 0.001). According to the Watcha scale, the incidence of EA was 15.6% in the propofol group and 78.13% in the control group (P < 0.001) (Table 2).

Peak PAED scores were also significantly lower in the propofol group (P < 0.001). The emergence times were significantly longer in the propofol group (P < 0.001); however, there was no significant difference in the duration of stay in the PACU between groups.

Fig. 2 depicts peak PAED scores with lower PAED scores in the propofol group (median score = 9) than the control group (median score = 16) during stay in the PACU.

Figs. 3 and 4 represent the percentage of children who experienced EA (PAED score > 12, Watcha score ≥ 3) during the first 30 minutes after emergence. In contrast to the control group, no patient in the propofol group experienced EA beyond 15 minutes of emergence.

In the present study, the simple transition to an infusion of propofol 3 mg/kg for 3 minutes following the cessation of sevoflurane anesthesia was associated with a significant decrease in the incidence, severity, and duration of EA.

The etiology of EA is not clearly understood. Yasui et al. [14] described an increase in noradrenaline release following sevoflurane exposure in the preoptic area of the brain in rats, especially in the locus coeruleus. It has been suggested that this leads to disorientation in the early stages of recovery, resulting in the agitation component of EA. EA may also occur due to rapid elimination of inhalational agents including sevoflurane, desflurane, and isoflurane [15] or due to postoperative pain [16]; transition to propofol may offset this effect and prevent EA.

Many clinical studies have shown that EA following sevoflurane anesthesia in children is a widespread problem and is significantly higher than with other inhalational anesthetics or propofol-based anesthesia [17–23]. Some studies have shown that the incidence of EA after total intravenous anesthesia with propofol ranges from 0%–9% compared to 23%–46% after sevoflurane anesthesia [19,20,24]. However, this technique is more complicated with the requirement for infusion pumps and complex dose adjustment.

In a previous report, propofol 2 mg/kg was administered at the beginning of surgery during sevoflurane anesthesia to reduce the incidence of EA. However, no decrease in the incidence of EA was observed with this technique [25]. A bolus dose of 1 mg/kg of propofol was administered at the completion of surgery to prevent EA, with contrasting results. Five studies showed benefit [8,9,26–28] with this technique while three studies showed no effect [10,11,29].

A preliminary trial with propofol 3 mg/kg administered over 3 minutes after cessation of sevoflurane anesthesia revealed a decrease in the incidence of EA according to the PAED and Watcha scales. However, this study was performed in children undergoing magnetic resonance imaging scan [30]. To our knowledge, ours is the only study that investigated the simple technique of transition to 3 mg/kg propofol administered over 3 minutes after cessation of sevoflurane anesthesia in an operative setting with associated stress factors.

Our study showed a significant decrease in the incidence of EA after sevoflurane anesthesia in the propofol group compared to control subjects. Overall, a decrease in the agitation scores was observed with propofol administration. This implies a decrease in the severity of EA with improvement in the quality of emergence leading to improved parental and staff satisfaction. Despite the significant increase in emergence times, there was no significant increase in the duration of PACU stay. Hence, the statistically significant decrease in EA also resulted in improved clinical outcomes. The delay in emergence in our study would suggest that the mechanism of action of transitioning to propofol may be the facilitation of sevoflurane washout prior to the point of emergence.

Our study has several limitations. First, we did not assess the safety of propofol use including airway complications and apnea. Second, the study subjects underwent inguinal hernia repair under caudal epidural analgesia combined with intravenous paracetamol; hence our results may not be generalizable to other surgical settings. Our results need to be validated in different surgical settings.

Another limitation of our study is the use of the PAED scale to detect EA. Sikich and Lerman [3] who devised this scale did not define a cut-off point that denotes the presence of EA. Cut-off scores ≥ 10 and ≥ 16 have been studied without strong justification for the scale used [1]. Aouad et al. [9] studied 80 children aged between 2–6 years who underwent strabismus surgery under sevoflurane anesthesia. A PAED score of 10 or less was considered acceptable in their study. Patients who were administered propofol 1 mg/kg intravenously at the end of surgery had a 19.5% incidence of EA compared to 47.2% in the placebo group. While the intervention group in this study experienced a significantly lower incidence of EA, the effect was less than optimal. The time to recovery was also longer in the propofol group compared to the control group. When a PAED score of 16 was considered acceptable, Abu-Shahwan [8] noticed a 4.8% incidence of EA with propofol 1 mg/kg compared to 26.8% in the control group. We speculate that acceptance of a higher PAED score in this study may have resulted in a significantly better outcome compared to the study by Aouad et al. The aforementioned studies using propofol [8,9] documented a high incidence of EA using a cut off of 10 points on the PAED score; the incidence was much lower using a cut off of 16 points. Clearly, a different approach towards the assessment of EA needs to be evaluated. In the present study, using a cut off of 12 on the PEAD scale and 3 on the Watcha scale for the assessment of EA after transition to propofol 3 mg/kg, we observed a lower incidence of EA in the propofol group (81.3% vs. 15.6%, P < 0.001 using the PAED scale and 78.1% vs. 15.6%, P < 0.001 using the Watcha scale). We used the Watcha scale in addition to the PAED scale because it is much simpler and easy to measure and may increase overall accuracy of assessment.

In conclusion, the transition to propofol 3 mg/kg for 3 minutes following sevoflurane anesthesia significantly decreased the incidence, severity, and duration of EA. Although there was an increase in emergence times, there was no increase in the duration of PACU stay after surgery.