Ultrasound-Guided Lumbar Plexus Block Reduces Emergence Agitation in Children Undergoing Hip Surgery: A Prospective Randomized Controlled Trial

Emergence agitation (EA) is a common and challenging postoperative problem in children emerging from general anesthesia. It is associated with self-injury, increases stress on healthcare team members and postoperative maladaptive behavioral changes. However, no completely effective prevention has been found for EA. Pain is considered to be an important contributor to EA. Ultrasound-guided lumbar plexus block is a safe and effective anesthetic technique that can provide satisfactory pain relief in pediatric hip surgery. We aim to investigate the effect of ultrasound-guided lumbar plexus block on emergence agitation in children undergoing hip surgery.


Results
The incidence of EA was signi cantly lower in Group Block than in Group Control [13.3% vs. 43.3%, odds ratio (OR) 0.201, 95% con dence interval (CI) 0.082to 0.496, p<0.001]. Group Block had a lower incidence of severe EA than Group Control [3.3% vs. 18.3%, odds ratio (OR) 0.154, 95% con dence interval (CI) 0.032 to 0.727, p=0.019]. CHEOPS was lower in Group Block than in Group Control [mean (95%CI) Background Emergence agitation (EA) in young children, especially preschool-aged children, is a common and challenging problem in the early postoperative period, characterized by a series of presentations including non-purposeful movement, inconsolability, restlessness, thrashing and agitation [1].The incidence varies from 10 to 80% [2]. Although EA is self-limiting and lasts for a short time about 30 min, EA causes selfinjury of children, increases stress on healthcare team members [3,4] and even leads to postoperative maladaptive behavioral changes [1,5,6],such as sleep disturbances, attention seeking and temper tantrums. The detailed mechanism remains unknown under this phenomenon. Consequently, it is recommended that EA should be considered as a 'vital sign' and has become heightened interest [1].
Various strategies have been proposed for the prevention of EA, including some pharmacological intervention and non-pharmacological strategies [1]. The most favorable prevention method is not currently available [7].Many studies have found that different anesthetic techniques may in uence the incidence and severity of EA [1,[8][9][10]. Nerve block has been shown to be bene cial in emergence agitation risk reduction [11].Ultrasound-guided nerve block has become increasingly popular in pediatric orthopedic surgery in recent years, since it increases the safety and provides effective analgesia [12,13].
Moreover, pain is considered as an important factor in EA [1]. Taken together, it is expected that ultrasound-guided nerve block has bene t on EA. To date, none of studies has assessed the effect of ultrasound-guided lumbar plexus block on EA.
Therefore, in this prospective, randomized and controlled study, we aimed to evaluate whether ultrasoundguided lumbar plexus block combined with general anesthesia would reduce the incidence of EA in children undergoing elective hip surgery.

Ethics approved and registration
This prospective, single-center, randomized, controlled trial was approved by the Ethics Committee of Shanghai Sixth People's Hospital a liated to Shanghai Jiao Tong University and written informed consent was obtained from parents or guardians. The protocol was registered at the Chinese Clinical Trial Registry (http://www.chictr.org.cn/, Principal investigator: Hui Zhang, Registration number: ChiCTR-INR-17011525, Date of registration: 30/05/2017) and published [14]. The study complied with the Declaration of Helsinki and was monitored by the Good Clinical Practice (GCP) Unit at Shanghai Sixth People's Hospital a liated to Shanghai Jiao Tong University. The study was conducted from May 2017 to February 2020in Shanghai Sixth People's Hospital a liated to Shanghai Jiao Tong University.

Inclusion and exclusion criteria
We screened a total of 126 children, aged 1 to6yr with the ASA physical status I or II scheduled for osteotomy for developmental dislocation of the hip. Exclusion criteria included contraindication for lumbar plexus block, developmental delays, neurological or psychiatric disease, local infection at the needle entry point, coagulopathy, nerve injury, allergy to anesthetics and study medications, and lack of parental consent for the child's participation in the study.

Randomization and blinding
Randomization was based on computer-generated allocation and a randomization number was concealed in an opaque envelope. All participants were randomly assigned to either ultrasound-guided lumbar plexus block group (Group Block) or control group (Group Control) at a 1:1 allocation ratio after induction of general anesthesia. Anesthesiologist who performed intraoperative anesthetic care knew the assignment. Patients, parents, surgeons, assessment investigators, the medical staff who provided postoperative care in the post-anesthesia care unit (PACU), data collectors, and statisticians were all blinded to the group allocation. The nerve block details were recorded, kept in a sealed envelope and opened when medically necessary and at the beginning of statistical analysis.
Patients in Group Control received a bolus injection of fentanyl 1μg·kg −1 intravenously before skin incision.
Patients in Group Block received ultrasound-guided lumbar plexus block before skin incision. Ultrasound scans of lumbar plexus were carried out using the S-Nerve TM Ultrasound System (Sonosite Inc., Bothell, WA, USA) in patients assigned to Group Block. A linear array (6~13MHz) transducer was used to perform lumbar plexus block with the longitudinal approach. All nerve blocks in this study were performed by the same experienced anesthesiologist, who was quali ed for ultrasound-guided lumbar plexus block. After induction, every patient assigned to Group Block was placed in the lateral decubitus position with the operative side upwards and the hips and knees exed in order to enlarge interspinous spaces and provide better image of lumbar plexus. When transducer was placed 1.5~2 cm lateral to the midline of the patient's back and parallel to the long axis of the spine at the level L3-4, the lumbar transverse processes produced a 'trident sign' [15]. Because the psoas major muscle is anterior to the transverse processes, the lumbar plexus was distinguished within the major psoas muscle. With ultrasound guidance, the block needle was advanced cautiously perpendicular to the skin until the needle tip was located 1-1.5 cm below the space between the transverse processes. All patients received 0.2% ropivacaine 1ml·kg −1 (Naropin The success of lumbar plexus block was de ned as follows:(i) visualization of the needle tip in the right position, the spread of local anesthetic around the lumbar plexus nerve;(ii)an increase of no more than 15% in heart rate, blood pressure in response to the skin incision;(iii)an increase of no more than 25% in heart rate, blood pressure during the operation [17]. Patients in Group Block were considered as failed blocks once one of the above criteria could not be met. The surgery was performed at least 15 min after the nerve block.
Anesthesia was maintained with 60% nitrous oxide, 40% oxygen and 2% end-tidal sevo urane. The concentration of sevo urane was added by 0.5% if the values of heart rate, blood pressure increased more than 15% from the baseline values during the operation. Conversely, the concentration of sevo urane was reduced by 0.5% if the values of heart rate, blood pressure decreased more than 15% from the baseline values. If the increase was more than 25% from the baseline values, additional bolus of fentanyl 1μg·kg −1 was administered. Patients recorded as a failure or a inadequate lumbar plexus block were followed up. Intention-to-treat analysis was used to analyze data including failed block.

Intraoperative anesthetic management and postoperative assessments
During the operation, if the heart rate reduced to less than 50 beats per min, atropine 10μg·kg −1 was used.
Ephedrine 0.1mg·kg −1 was used to treat hypotension when blood pressure decreased more than 25% from the baseline value. At the end of the operation, sevo urane and nitrous oxide were discontinued and oxygen ow was increased to 6l·min −1 . Patients were transferred to PACU and monitored until the end of the study.
Intraoperative data collection included heart rate, blood pressure, respiratory rate, concentration of sevo urane, and fentanyl dose. The average value of end-tidal sevo urane concentration (EtSev%) was calculated from all values recorded every 5 min throughout the operation.
After extubation and emergence, a blinded well-trained observer performed all assessments. EA was assessed by PAED scale[18] at 0, 5, 10, 20, and 30 min after emergence. We de ned EA as PAED score≥10 [18,19]. Severe EA was de ned as PAED score≥13[18] and treated with a bolus of fentanyl 0.5μg·kg −1 [8,18]. The primary outcome was the incidence of EA 30 min after emergence from anesthesia.
Emergence time (from discontinuation of sevo urane to the rst response to a simple verbal command), extubation time (from the end of anesthesia to extubation), intraoperative collective fentanyl rescue dose, adverse effects and complications were recorded.
At the end of study, that is, 30 min after emergence, the patients were given nurse controlled analgesia (NCA) with morphine. Bolus dose is 10 μg·kg −1 , lockout interval 30 min and background infusion is 10 μg·kg −1 ·h −1 .

Statistical Analysis
The incidence of EA for pediatric hip surgery was approximately 42% in our pilot study. Therefore, a minimum sample size of 60 patients in each group would have 90% power to detect an absolute reduction of 15% in the incidence of EA in the intervention group at a signi cance level of 0.05, considering a possible dropout rate of 10%.
Categorical variables such as the incidence of EA and severe EA were expressed as frequencies (percentage) and analyzed with Chi-square test or Fisher's exact test. One-way Repeated-measures ANOVA or Generalized estimated equation (GEE)was applied to analyze data for PAED and CHEOPS according to the test for normality and homogeneity of variance. Continuous variables with a normal distribution were summarized with mean (standard deviation, SD) and were analyzed with an independent-sample t test.
The level of signi cance for each test was set at α=0.05, 2 tailed. All data were subjected to the Kolmogorov-Smirnov test for normality. All data were analyzed with SPSS 23.0 (IBM SPSS Statistics for Windows, Version 23.0; IBM Corp., Armonk, NY). Intention-to-treat strategy was performed in every case.

Results
A total of 120 subjects were randomly assigned to two groups (n=60 per group) and completed the study (Fig. 1). All children in Group Block received an effective lumbar-plexus block, as evidenced by the fact that all criteria for a successful block were met in each case. Table 1 shows patients baseline demographics and clinical characteristics. There were no signi cant differences between groups.  There was no case treated with atropine or ephedrine. In Group Control, two cases of blood pressure decreased more than 15% and less than 25% from the baseline values. The concentration of sevo urane was reduced by 0.5%. No signi cant difference was observed in the incidence of postoperative adverse events between the two groups( Table 2).There was no case of local anesthetic intoxation and nerve damage in Group Block. Emergence time and extubation time were signi cantly shorter in Group Block than in Group Control (Table 2).

Discussion
This prospective, randomized and controlled study revealed that ultrasound-guided lumbar plexus block could decrease the incidence and severity of EA effectively and offer better analgesia in children aged 1-6 yr undergoing elective hip surgery. Furthermore, ultrasound-guided lumbar plexus block reduced the use of fentanyl and sevo urane.
The exact etiology of EA remains unclear [1]. Amongst numerous factors associated with EA, pain plays an important role in EA. Adequate analgesia can contribute to a decrease of EA. According to several reports, nerve block can provide satisfactory analgesia [12,13,15,21].The effect of nerve block on EA remains uncertain since previous studies have not been consistent [1].Some previous studies have reported nerve block can reduce the incidence of EA [8,9]. The primary reasoning behind its bene t was providing satisfactory postoperative analgesia and the evident reduction in sevo urane [8]. In contrast, Ohashi et al. [17] found ilioinguinal/iliohypogastric block does not affect the incidence of EA. Ohashi et al. [17] attributed such result to the fact that the type of operation was minimally invasive and they thought ilioinguinal/iliohypogastric block would be more useful for invasive surgery. In this study, we chose osteotomy in children with developmental dislocation of the hip which is not only a common pediatric procedure but also associated with intense pain. Most previous studies about the effect of nerve block on EA were limited in short-term procedures with minimal to moderate postoperative pain and several super cial nerve [8, 9,17]. Our study aims to assess the effect of lumbar plexus block, a relatively deep procedure, on EA after long-term surgery. Ultrasound technology has revolutionized pediatric anesthesia, and ultrasound-guided nerve block has increased in popularity in pediatrics because of its e cacy [16,22]. The results demonstrated when ultrasound-guided lumbar plexus block was applied, the incidence and severity of EA in children undergoing hip surgery were indeed reduced. Meanwhile, Group Block had lower pain scores and less total fentanyl consumption than did Group Control. According to previous reports, nerve block decreased the incidence of EA by providing effective analgesic properties [8,9]. Our results and above literature are consistent. Patients need fentanyl rescue were signi cantly fewer in Group Block compared to Group Control. We assumed that better analgesic effect of ultrasound-guided lumbar plexus block might be a major factor reducing the incidence of EA. Emergence time and extubation time in Group Block were shorter than in Group Control. The reason may be lumbar-plexus block reduces the use of fentanyl.
Distinguishing between pain and emergence agitation is challenging [1]. Pain may remain a confounder in studies of EA in patients who are likely to experience postoperative pain. We use PAED which is currently the standard to diagnose EA in children [11].Considering fentanyl is considered as a rst-line agent with respect to PAED [11], we use fentanyl as rescue medication for agitation [8,18]. To minimize the confounding effect of pain on EA, we used su cient fentanyl to ensure adequate analgesia in both groups. Patients with a CHEOPS score of >6were considered to have pain [23].Our results showed mean CHEOPS scores were lower than 6 in both groups, indicating that patients received effective pain relief and patients of EA were experiencing EA instead of pain.
The PAED score was validated to re ect the presence of EA. Generalized estimated equation analysis for PAED showed the lumbar plexus block was associated with a decrease in PAED score. Our study showed a downward trend of PAED after emergence, which is in agreement with the study of Frederick et al [24]. PAED scores peaked at emergence and declined to below 5 at 30 min after emergence in both groups, this result con rmed that EA is self-limited and lasts for a short time about 30 min.
The use of sevo urane has been associated with a high incidence of EA [11]. Data from clinical studies suggest that intrinsic effect of sevo urane is associated with EA [25][26][27] and dose reduction of sevo urane might be a major contributor for reducing the incidence of EA [19]. As a consequence, reducing the use of sevo urane is strongly considered in children with a high risk of EA [11].The average value of end-tidal sevo urane concentration in Group Block was signi cantly lower than in Group Control in our study. Ultrasound-guided lumbar plexus block reduced the amount of sevo urane. Thus, the decreased amount of sevo urane could be another mechanism by which lumbar plexus block reduces the incidence of EA. Recent studies [28,29] have reported that nerve blockage might attenuate postoperative in ammation to improve postoperative cognitive impairment. This could also be a potential mechanism of effect of lumbar plexus block on EA in our study, which needs to be de ned in further studies.
Whatever the mechanism of lumbar plexus block on EA, lumbar plexus block provides better control of pain and reduction of sevo urane use and hence ameliorates EA.
Although bene ts of ultrasound-guided lumbar plexus block on EA were achieved by this prospective, randomized study, there are several limitations in our study. Our study was limited to 30 minutes after emergence because EA lasts for a short time about 30 minutes mostly. Further study is required to elucidate if lumbar plexus block could improve long-term effect especially maladaptive behavior [1]. The second limitation is that the target lumbar plexus would be more precise if ultrasound guidance was combined with nerve stimulation. However, this limitation may be offset somewhat because the ultrasound landmarks of lumbar plexus in children are more easily identi ed when compared with adults.
To this end, we drew up a standard procedure of nerve block and could ensure satisfactory regional block in our study.

Conclusion
In conclusion, our ndings indicate that the incidence and severity of EA in children undergoing hip surgery were decreased when applying ultrasound-guided lumbar plexus block. Lumbar plexus block is a potential practicable technique for the prevention of EA in pediatrics after general anesthesia. CONSORT owchart. EASevere