Awareness During Anesthesia in Children: A Prospective Cohort Study

Transcription

1 Awareness During Anesthesia in Children: A Prospective Cohort Study Andrew J. Davidson, MBBS, GradDipEpiBiostats, FANZCA*, Grace H. Huang, BMedSci*, Caroline Czarnecki, BMedSci*, Margaret A. Gibson, BN, RN*, Stephanie A. Stewart, BN, RN*, Kris Jamsen, BSc, PGDip(Stats), and Robyn Stargatt, PhD, MAPS *Department of Anaesthesia and Pain Management, The Royal Children s Hospital, Parkville, Victoria, Australia; Department of Pharmacology, University of Melbourne, Melbourne, Victoria, Australia; Clinical Epidemiology and Biostatistics Unit, The Royal Children s Hospital, Parkville, Victoria, Australia; and Department of Psychology, The Royal Children s Hospital, Parkville, Victoria, Australia During routine adult anesthesia, the risk of awareness is 0.1% 0.2%. No recent studies have reported the incidence in children. Altered pharmacology and differing anesthesia techniques suggest that the incidence may differ in children. In this prospective cohort study, we determined the incidence of awareness during anesthesia in children. Eight-hundred-sixty-four children aged 5 12 yr who had undergone general anesthesia at The Royal Children s Hospital were interviewed on 3 occasions to determine the incidence of awareness. The awareness assessment was nested within a larger study of behavior change after anesthesia. Reports of suspected awareness were sent to four independent adjudicators. If they all agreed, a case was classified as true awareness. Twenty-eight reports were generated. There were 7 cases of true awareness, for an incidence of 0.8% (95% confidence interval, 0.3% 1.7%). Only one aware child received neuromuscular blockers, compared with 12% in the nonaware group. No aware child reported distress, and no substantial difference was detected in behavior disturbance between aware (20%) and nonaware (16%) children. The data provide some evidence that, like adults, children are also at risk of intraoperative awareness. Although the cause remains unclear, anesthesiologists should be alerted to the possibility of awareness in children. (Anesth Analg 2005;100:653 61) Awareness is the postoperative recall of events that occur during general anesthesia. It is a rare and well described complication of anesthesia in the adult surgical population, with an incidence of 0.1% 0.2% (1 5). In adults, awareness is often a distressing event and may have significant psychological ramifications (6). Very little is known about awareness in modern pediatric anesthesia. One study, performed in 1973, found an incidence of awareness in children of 5% (7). In this study, some children received, a neuromuscular blocker, and no volatile or continuous IV anesthetic the Liverpool technique. This technique has a high risk of awareness and is no longer Supported by the Australian Society of Anaesthetists, the University of Melbourne, and the Murdoch Childrens Research Institute. Accepted for publication October 29, Address correspondence and reprint requests to Andrew J. Davidson, MBBS, GradDipEpiBiostats, FANZCA, Department of Anaesthesia, Royal Children s Hospital, Flemington Rd., Parkville 3052, Victoria, Australia. Address to andrew.davidson@rch.org.au. DOI: /01.ANE used. Since then, several small studies in children have failed to elicit any cases of awareness (8 12). All of these studies enrolled small numbers of children, and this limited their power to determine the true incidence of awareness. Because of differing anesthesia requirements and techniques, it is possible that the incidence of awareness differs significantly in children (13), who, compared to adults, also have different expectations, fears, and ways of managing stressful events. It may be incorrect to extend to children conclusions about the effects of awareness in adults. The prevention of awareness continues to be a priority for anesthesiologists. A recent randomized controlled trial demonstrated that processed electroencephalogram (EEG) monitoring reduced the incidence of awareness in a particular high-risk adult population (14). Before routine EEG monitoring can be advocated for reduction of awareness in children, the nature, causes, implications, and incidence of awareness in children need to be known. Once this information is known, the potential role for EEG monitoring in children will be clearer, and relevant randomized controlled trials can be 2005 by the International Anesthesia Research Society /05 Anesth Analg 2005;100:

2 654 PEDIATRIC ANESTHESIA DAVIDSON ET AL. ANESTH ANALG AWARENESS IN CHILDREN 2005;100: planned. We performed a large, prospective cohort study to determine the incidence of awareness during routine pediatric anesthesia. Methods The awareness assessment was nested within a larger detailed study of predictors of behavior change in children after hospitalization and anesthesia. This article presents only the awareness data. Children aged 5 12 yr old having general anesthesia for any procedure were eligible. Children were excluded if they had a degree of developmental delay that precluded interview, were expected to be ventilated postoperatively, or could not understand English or if the anesthesia was for an out-of-hours emergency procedure. The Royal Children s Hospital Human Research Ethics Committee approved the protocol. Parents provided written consent before the child s enrollment in the study. Children were enrolled at The Royal Children s Hospital, Melbourne, between October 2002 and March This was a prospective cohort study. Eligible children were randomly chosen from the operating room lists. Each day a three-digit random number was generated. This number was compared with the last three digits of the hospital identification number of children scheduled for anesthesia. Children with the last three digits of their hospital identification number closest to the random number were approached in order of closeness. Up to seven children were approached each half-day. Ninety-five percent of parents approached enrolled their children. The child was enrolled on the day of the procedure, before anesthesia. During the consent process, the family was informed that the study was a survey of behavior change after anesthesia. After an in-depth explanation of the questionnaires needed for the behavior assessment, the family was also informed that there would be questions about the child s memories and experiences in the period around the anesthetic. Anesthesiologists were informed about the awareness study before it started. Within the department, the profile of the awareness assessment was then kept low. For any particular case, the anesthesiologist was not aware that a child had been selected for the study until after the anesthetic. Once the child was in the recovery room, the anesthesiologist was asked to provide details of anesthetic technique and a description of any critical incidents. Screening questions were asked at three time points to detect possible awareness. The first screening interview was during the hospital stay. For day-stay patients, this was performed just before discharge. For in-patients, the first interview was on either the same day or the next day if the child was unwilling or unable to talk on the day of the anesthetic. During this interview, the awareness screening questions were inserted among data collected for the behavior study. The questions are listed in Table 1. The same questions were used for all children. If the child answered no, then the screening questions were not repeated. No incentives were given for the children to answer, and no leading questions were asked during the screening interview. If the child answered no to Question 6, there was no further prompting. If the child answered yes, and if the screening interviewer thought there was a possibility that the answer represented awareness, then the principle investigator (AJD) was called as soon as possible to continue the interview. The screening questions were asked by one of four trained interviewers. A memory was classified as a dream if the child considered the memory to be a dream. The second and third sets of awareness screening questions were asked on the 3rd and 30th days, respectively, after the anesthetic. The parents asked these questions. They were given a set of questionnaires about their child s behavior and a specific set of questions to ask their children. These questions are listed in Table 1. To justify repeated questioning, the parents were told that a child s memories of their anesthetic might change with time. They were instructed to ask the questions exactly as they were written and to write down their child s reply verbatim. The principal investigator (AJD) interviewed all children who were identified as possible awareness cases from the screening interview. Initially openended, nonleading questions were asked. Direct questions were also asked to elicit the extent of the memories and consistency with the known environment. Children were asked to describe their memories in more detail. They were asked if they saw, heard, or felt anything; if they were surprised; or if they talked or did anything. Wherever possible, any leading question was matched with a false lead. For example, a child reports hearing noises. The interviewer asks, What noises? The child replies, Just noises. The interviewer asks, Did you hear beeping noises or music? The child replies, Beeping noises. The children were also asked if the memories upset them. The family was offered psychological support and given contact details for any further needs. They were contacted 3 and 30 days later and were again offered assistance as required. If the first report of potential awareness was in the second or third screening assessment, then the principal investigator contacted the family and spoke to the child by telephone. When any potential cases of awareness were identified, the anesthesiologist who performed the anesthetic was informed. Any comment by the anesthesiologist that was relevant to the possible awareness event was recorded.

3 ANESTH ANALG PEDIATRIC ANESTHESIA DAVIDSON ET AL ;100: AWARENESS IN CHILDREN Table 1. Structured Awareness Screening Interview Interview on Day 1 1. Were you upset, worried, or frightened about your 2. How upset, worried, or frightened were you? 3. What was the last thing you remember before the 4. How did the doctor make you go to sleep? 5. What is the next thing you remember after the 6. Did you have any dreams or feel or hear anything while you were having the 7. Were you sore after the 8. How much did it hurt? 9. What did you like least about having an 10. What did you like best about having an 11. Did someone talk to you about what the operation was going to be like? 12. Was that about the same as what happened? 13. How do you feel about the 14. Would you like to have known more? Questions asked by parents on Day 3 and Day What was the last thing you remember before the 2. How did the doctor make you go to sleep? 3. What is the next thing you remember after the 4. Did you have any dreams or feel or hear anything while you were having the For every potential case of awareness, a report was generated. This report included details of the procedure, the age of the child, any pertinent comments by the anesthesiologist, and many of the child s actual words. These reports were sent to four pediatric anesthesiologists in Australia and New Zealand. These adjudicators did not know how many children had been enrolled in the study. They classified each report as awareness, possible awareness, or not awareness. If all four adjudicators rated a case as awareness, then it was classified as a true awareness case. Behavior change was assessed with a modification of the 27-item Vernon Post Hospitalization Behavior Questionnaire (15). Prehospitalization behavior was assessed on the day of the procedure, and posthospitalization behavior was assessed 3 and 30 days postprocedure. To avoid repeated measures, the 3-day assessment was not used for this aspect of the study. The more clinically relevant 30-day assessment was used to assess the behavioral effect of awareness. A count of seven or more negative changes in behavior was considered a clinically significant deterioration in behavior. An estimate of the necessary sample size was based on the desire to confidently exclude the possibility of not detecting an incidence of awareness that was at least five times the incidence described in adults. The incidence in adults was taken to be 0.15%. An exact binomial test with a one-sided of 0.05 showed that 750 children would be needed to give a power of 81% to detect an incidence more than 0.75%. Allowing for an 80% interview success rate, at least 950 children were needed in the awareness component of the study. Because the inclusion criteria for the behavior study were broader than for the awareness study, 1250 children were enrolled into the behavior component of the study; we anticipated that at least 75% would be eligible for the awareness assessment. Data are presented in descriptive form. The numbers in the aware group were not sufficient for comparative analysis. When the characteristics of the awareness and nonawareness groups are presented, only the true awareness cases are included in the aware group. Results One-thousand-two-hundred-fifty children were recruited into the behavior study, of which 921 were eligible for the awareness study. Fifty-seven of these were then excluded from the awareness study. Two withdrew, 2 had their procedures canceled, and 53 could not be assessed for awareness at any time. The remaining 864 children had at least 1 assessment for awareness. Seven-hundred-fifty-six (88%) were assessed within 1 day of the procedure, 704 (81%) at Day 3, and 590 (68%) at 30 days. The predominant reason for missing the first assessment was discharge of the child before the interview. Assessments were missed on Day 3 or 30 if the parents did not return the questionnaire. The types of procedures performed are shown in Table 2. The principal investigator interviewed 28 children who had suspected awareness from the screening questions. Twenty-eight reports were generated and sent to the adjudicators. In 12 cases, at least 1 adjudicator reported the case as awareness. These 12 cases are briefly described in Table 3. There were 7 cases of true awareness (cases 1 7; all 4 adjudicators rated the case as awareness ), for an incidence of 0.8% (95% confidence interval, 0.3% 1.7%). From the 12 cases, there were 5 other cases for which at least 1 adjudicator classified the case as awareness (cases 8 12). For Case 8, three adjudicators classified the case as awareness, and one classified it as possible awareness. For Cases 9 and 10, two adjudicators classified the cases as awareness, and the other two classified it as possible awareness. For Cases 11 and 12, one adjudicator classified the case as awareness, whereas the other three classified it as possible awareness. Cases that were not classified as awareness were usually children with auditory memories of events that could have occurred in recovery and children with inconsistent or contradictory memories. Features of the awareness experience for the 12 cases in which at least 1 adjudicator classified the case

4 656 PEDIATRIC ANESTHESIA DAVIDSON ET AL. ANESTH ANALG AWARENESS IN CHILDREN 2005;100: Table 2. Details of Procedures Performed Under General Anesthesia Type of procedure No. Children Dental 43 (5%) Dermatology laser 22 (3%) Ear, nose, and throat 184 (21%) Gastroscopy or colonoscopy 105 (12%) General surgery and urology 216 (25%) Gynecology 1 ( 1%) Neurosurgery 4 ( 1%) Lumbar puncture or bone marrow aspirate 4 ( 1%) Ophthalmology 53 (6%) Orthopedic surgery 108 (13%) Plastic surgery 91 (11%) Radiology procedures 13 (2%) Bronchoscopy 3 ( 1%) Other 17 (2%) as awareness are shown in Table 4. Of the seven true-awareness cases, four cases were first identified in the first interview, one was first identified in the second interview, and two were first identified in the third interview. Of the seven true-awareness cases, only one child (Case 6) volunteered the awareness experience to the interviewer before the specific screening question. Another child (Case 7) first reported being aware to her brother. Auditory and tactile sensations were both reported. Of the seven trueawareness cases, four children reported mild pain, and one reported moderate pain. No child reported severe pain. Demographic data and details of the anesthesia for the 12 cases are listed in Table 5. To provide insight into the etiology of awareness in children, summary details of the seven true-awareness cases are compared with those of the nonaware population in Table 6. Use of sedative premedication, seniority of the anesthesiologist, mode of induction, method of maintenance of anesthesia, use of neuromuscular blocker, length of anesthesia, and occurrence of any critical incidents were recorded for all cases. No differences were detected, but the small number of awareness cases limited the power of any statistical comparisons between groups. In the nonaware group, details were not recorded for specific induction drugs or maintenance anesthetics. In our institution, the most commonly used induction drugs are propofol or sevoflurane, and the most commonly used maintenance drug is isoflurane (with nitrous oxide). The aware group reported dreaming more frequently and had longer hospital stays. Children who reported dreaming and awareness clearly differentiated the dream from the awareness memory. Compared with the nonaware children, no substantial increase was detected in behavior deterioration in the aware group. Five-hundred-ninety-two of the children assessed for awareness had a complete behavior assessment at baseline and 30 days. Five of the 7 children with true awareness had a complete behavior assessment at baseline and 30 days. Cases 1 and 3 did not return the 30-day behavior assessment, but both families were contacted by telephone. Case 1 was reported to be improving, having seen a psychologist. Case 3 was not reported by the family to be distressed or behaving abnormally. Of the five who had complete assessments, only one had a clinically significant deterioration in behavior. In the nonaware population, 93 (15.8%) children had a clinically significant deterioration in behavior. The behavior assessment included a measure of sleep disturbance (a symptom often reported after awareness in adults). In the aware group, one child had increased sleep disturbance, one had decreased sleep disturbance, and three had no change in their sleep pattern. At the time of the interview, no child reported being upset by the awareness memories. Only one family requested psychological referral. This child (Case 1) saw a psychologist, who confirmed that the child had been aware. The child s distress, however, was due to a newly diagnosed phobia unrelated to the awareness experience. Discussion In our pediatric anesthesia population, we found an incidence of awareness of 0.8%. Few previous studies have examined awareness in children. In 1973, an audit of 202 children aged 7 14 years found an incidence of 5% (7). Many of the children in that study who reported awareness had an anesthetic without a volatile drug. This technique, known as the Liverpool technique, has a frequent incidence of awareness and has since been abandoned. In 1988, two studies also using the Liverpool technique investigated the incidence of dreaming and awareness in children (8,9). No cases of awareness were recorded. The number of children in these studies was small, and the children were interviewed only once. Three further studies investigating implicit learning during anesthesia in children also interviewed children for awareness (10 12). No cases of awareness were found. These studies also enrolled very small numbers of children and interviewed the children only once. In this study, some features of awareness were similar to those found in studies of adult awareness. Children described auditory and tactile sensations in proportions similar to those described in adults (1). Children did not always report the awareness in the first interview. This is consistent with findings in adult studies in which awareness would have been missed if only one interview had been conducted (1,5). In this study, none of the children who reported awareness had received sedative premedication; however, the nonaware group also received premedication infrequently. In adults, sedative premedication has not

5 ANESTH ANALG PEDIATRIC ANESTHESIA DAVIDSON ET AL ;100: AWARENESS IN CHILDREN Table 3. Experiences of the 12 Cases for Which at Least 1 Adjudicator Classified the Memory as Awareness Patient No. Memory 1 Felt the tubes going into my ears. The right ear hurt. Heard a loud noise and beeping. When asked, the child reproduced a noise very similar to operating room suction. Tried to talk but couldn t let it out. Got freaked out. Felt a sort of rug covering me. The child later dreamt of a tortoise. 2 I could hear during my operation and people talking. The child described definitely hearing men talking for a long time about the operation and how it was going to be when it was finished. The child then went back to sleep again. There were no men in recovery. 3 Felt a jab in my hip. The child remembered going to sleep then something touching my bone and then a shiver went up my spine. The child later dreamt he was a stunt man. 4 Heard drill noises and beeping. The child also described trying to talk but could not. 5 Heard the saw. It sounded like a bee. When asked, the child accurately reproduced the sound of an orthopedic saw. The child described feeling pain in her leg and also hearing people talking. Chatter, chatter, chatter. Tried to block my ears and it worked. The child then described going back to sleep and later dreamt of a Mars chocolate bar. 6 Felt a steel plate being put into my mouth to open it up wide. 7 The child described falling asleep then felt a thing going down my throat, it hurt. Going down my neck, just in and out again. The child then described falling asleep again and dreaming of her cat. 8 I heard the doctors working. The child described going to sleep and then hearing male voices and hearing them putting stitches in. The child described feeling no pain and reported being pleased that the surgeons were doing their work. The surgeons had used local anesthetic infiltration before suturing. 9 The child clearly remembered the surgeon talking during the operation along with some female voices. The child also dreamt about Disneyland. 10 Heard a loud noise. The child described going to sleep and then felt warmth on her hand and a pillow sound as they put me on the table. The child also dreamt about monsters. 11 Felt the machine spinning him around. The child later remembered the spinning thing around him. 12 Felt the needles. The child tried to move his arms but couldn t. He also heard laughing, beeping, and babies crying. The account was inconsistent and became increasingly complex and fantastic. The child also dreamt of birds. Patients 1 7 were classified as true awareness, with all four adjudicators in agreement. For patients 8 12, at least one adjudicator, but not all, classified the case as awareness. been shown to reduce the incidence of awareness (1). No conclusions can be drawn from our study about premedication and awareness. In adult anesthesia, awareness is more common when muscle relaxation is used (1). Awareness occurred in our study despite the infrequent use of muscle relaxants. In our study, none of the children who were aware described severe pain. This is consistent with the finding that severe pain during awareness is more frequently described when muscle relaxants have been used (1). The incidence of dreaming in adult anesthesia has been reported to be between 1% and 40% (3,5,14,16,17). In our study, dreaming was more frequent in children who were aware. In adults, dreaming has also been associated with awareness (17). It is conceivable that dreaming is more common during light anesthesia. None of the children who were aware reported being upset by their memories. Similarly, the incidences of behavior change and sleep disturbance were similar to those in the nonaware population. The small number of children who were aware limits the power of these observations. It is also possible that children developed psychological problems later than 30 days or gave an understated assessment of their distress. Children may internalize traumatic events. One cause for distress after awareness is denial of the event by health care providers (18,19). Psychological consequences from the awareness experience in this study may have been reduced by the interest and understanding inherent to the interview process. Children without awareness find hospitalization and surgery a stressful event (20). This is evident from the finding that 15.8% of nonaware children had clinically significant behavior changes at 30 days. Memories of being awake and in pain may certainly be significant and frightening for some children. It is also possible that remembering only noises or voices may be a relatively minor concern compared with a child s fear of surgery and hospitalization. A child s expectations may also be different. Compared with adults, young children may not have the same degree of understanding of anesthesia. An adult expects to be unconscious with no memories. A young child may equate anesthesia with sleep and therefore have less faith that he or she will not wake up during the surgery. The incidence of awareness found in this study was more frequent than that reported in several studies during routine anesthesia in an adult population (1 5). When directly comparing this study with other studies, the differences in study design must be considered. Although many aspects of the design were based

6 658 PEDIATRIC ANESTHESIA DAVIDSON ET AL. ANESTH ANALG AWARENESS IN CHILDREN 2005;100: Table 4. Details of Awareness in Cases for Which at Least One Adjudicator Classified the Case as Awareness Case No. True awareness No. Adjudicators classifying case as awareness Dream as well as awareness Time of recall First 24 h Day 3 Day 30 Sensation Pain Comment 1 True 4 Yes Yes Yes Yes Tactile and Mild Patient moved auditory 2 True 4 No Yes a Yes Yes Auditory None 3 True 4 Yes Yes Yes Yes Tactile Moderate Difficult procedure 4 True 4 No No No Yes Auditory None 5 True 4 Yes Yes a Yes Yes Tactile and Mild auditory 6 True 4 No No Yes No Tactile None Volunteered memory 7 True 4 Yes No No Yes Tactile Mild Told her brother 8 No 3 No Yes Yes Yes Auditory No 9 No 2 Yes No Yes Auditory No 10 No 2 Yes Yes No No Tactile and No auditory 11 No 1 No Yes Yes Tactile No 12 No 1 Yes Yes Yes Tactile and auditory Mild a Child interviewed between Day 1 and Day 3 after the operation. Table 5. Details of Demographics and Anesthesia in Cases for Which at Least One Adjudicator Classified the Case as Awareness Case No. True awareness Sex Age ASA status Procedure Sedative premedication Induction drug Maintenance drug Other drugs Local anesthesia Neuromuscular blocker 1 True M 8y8mo I Eartubes No Propofol Halothane, 2 True M 6 y 11 mo I Pyeloplasty 3 True M 10 y3mo II Arthrogram hip 4 True F 6 y II Dental restorations 5 True F 6y5mo II Femoral osteotomy 6 True F 7y2mo I Tonsillectomy 7 True F 6 y II Gastroscopy 8 No F 6y4mo I Excision lesion on cheek 9 No M 5y6mo I Tonsillectomy Midazolam Propofol Isoflurane, 10 No F 5y2mo I Dermatology laser 11 No M 5y7mo I MRI 12 No M 6y6mo I Steroid injections Midazolam Sevoflurane Isoflurane, Morphine Infiltration No No Yes Morphine, ketamine Femoral nerve block No Morphine No No Infiltration No Morphine No No Midazolam No No IV MRI magnetic resonance imaging. on recent adult studies, several changes were necessary to suit the pediatric population. We conducted three screening interviews, as is now standard for awareness assessment. Rather than the investigator s telephoning all children, the parents conducted the second and third screening interviews. It was hoped that children would be more candid with their parents. Although parents were instructed to ask the

7 ANESTH ANALG PEDIATRIC ANESTHESIA DAVIDSON ET AL ;100: AWARENESS IN CHILDREN Table 6. Comparative Demographics of True Awareness and Nonaware Groups Variable Aware (n 7) Nonaware (n 857) Sex Male 3 (43%) 490 (57%) Female 4 (57%) 367 (43%) Age (yr) 7.3 (1.6) 8.6 (2.3) ASA status I 3 (43%) 633 (74%) II 4 (57%) 171 (20%) III 0 53 (6%) Seniority of anesthesiologist Consultant 7 (100%) 714 (83%) Fellow 0 62 (7%) Registrar 0 81 (9%) Sedative premedication (16%) Inhaled induction 6 (86%) 391 (46%) Total IV anesthesia 0 24 (3%) Relaxant anesthesia 1 (14%) 99 (12%) Duration of anesthesia 68 (69) (40) (min) Length of hospital stay (d) 2.7 (2.8) (2.0) 1 48 Critical incident during 0 25 (3%) anesthetic Dreaming 4 (57%) 90 (11%) Data are number (%) or mean (sd) range for duration of anesthesia and hospital stay. questions exactly and to report the answers verbatim, this less controlled situation raises the possibility of reporter bias. The screening questions were modified from the Brice interview (21), which has undergone several minor variations since it was originally described. The first three questions of the original Brice interview remain the core of all awareness interviews. We adapted the interview for children in a manner similar to that described by McKie and Thorp (7). In the Day 1 interview, we inserted the questions among other questions to reduce the emphasis on the awareness assessment. In the parent-guided interviews, we omitted the reference to worst memories in order to avoid unduly upsetting the child. The questions about dreaming and other experiences were combined (as originally described by Brice) to produce more openended questions and hence are less likely to suggest an answer. The question asking how the child went to sleep was added in order to focus the child on the time around the anesthetic. When comparing awareness studies, the method of judging awareness should also be considered. This varies widely among studies. Whether a child is reporting a dream, confabulating, or reporting a true event is a difficult, and, to some extent, subjective, decision. To avoid an investigator bias, we chose a more rigorous adjudication process. There are other limitations to this study. It was performed at only one center, and although a number of different anesthesiologists were involved, the practice at this center may not reflect the practice elsewhere. To reduce bias, an effort was made to keep the profile of this study low. However, as the study progressed, the anesthesia staff became more familiar with the study. It was not possible to completely blind the anesthesia staff, for ethical and medicolegal reasons. It was also useful to be able to gather information about the case from the involved anesthesiologist. It is possible that as the profile of the study gradually increased, the anesthesia staff changed their routine practice. The awareness assessment was nested within a study of behavior change. The nesting was done for several reasons. Primarily it was performed to reduce the emphasis on the assessment of awareness. The parents and children were assessed for awareness in the setting of a longer interview and while completing multiple questionnaires. It also allowed the interviewers to gain rapport with the family and to provide some data on behavioral change that would be useful for comparison if there were a large number of children who reported awareness. Finally, the behavior study provided data that were of interest and that will be described fully in a separate article. The most important limitation to this study is the suggestibility of children and the reliability of their reports. Young children do report events accurately if the correct interviewing techniques are used (22). The interviewer should be unbiased, use a neutral tone, and not use leading or suggestive questions, and no incentive should be given to the child (22). The interviewers for this study were aware of this. Answers to open-ended questions are the most reliable (23). If the child replied no to any memories in the screening interview, then there was no further probing for possible memories. Particular care was taken to avoid leading questions in the initial screening interview. Younger children are more likely to be suggestible, although all ages are susceptible to some extent (24). In our study, many of the reports rejected by the adjudicators were in younger children. A suggestible memory may have features of increasingly more bizarre detail, as demonstrated by Patient 12. Children s suggestibility increases with the number of interviews, and the first interview is usually the most accurate (24). Awareness is not always reported at the first interview. In our study, we performed three interviews, increasing the chance of detecting an awareness event but also increasing the risk of suggestibility. The parents performed the second and third interviews. It was possible that some parents did use leading or suggestive techniques despite the instructions to simply ask the questions as written and to

8 660 PEDIATRIC ANESTHESIA DAVIDSON ET AL. ANESTH ANALG AWARENESS IN CHILDREN 2005;100: write exactly what their child replied. The more detailed follow-up by the principal investigator also increased the risk of suggestibility creating untrue memories. This risk must be balanced against the need to assess the likelihood that a memory was indeed an accurate recall of events, and questions are needed to overcome a child s reluctance to discuss their memories. When children or adults are assessed for awareness, there is always the possibility that some of the reports are due to suggestibility. In our pediatric awareness assessment, the delayed reports in the younger children should be regarded with the most suspicion. Finally, when strict criteria are applied to avoid suggestibility, this inevitably increases the chance that a child who is reluctant to report the awareness will not be detected in the assessment. There are several potential explanations for why awareness may occur in children. Compared with adults, children require a larger concentration of anesthetic to achieve anesthesia. It is therefore possible that children may be more frequently underdosed. Potency of inhaled anesthetic drugs is measured in terms of the minimum alveolar anesthetic concentration (MAC) needed to produce an effect in 50% of the population. The concentration of volatile anesthetic required to prevent movement (MAC) increases as age decreases (13,25,26). When considering awareness, a more relevant measure of anesthesia potency is MACawake, the amount of anesthetic needed to prevent awakening. MAC-awake also increases as age decreases (27), and the ratio of MAC to MAC-awake does not change with age. There are, however, very few MAC-awake studies in children, and as with most MAC studies, the up-down technique is used. The up-down technique may underestimate the true variance in the population (28). The pharmacology of IV anesthetics is also different in children: larger doses are required to produce anesthesia in children (29). At our hospital, the use of induction rooms is routine. The anesthetic starts in a quiet, less threatening environment, and the child is transferred to the adjacent operating room after he or she is anesthetized. Outside the United States, induction rooms are widely used in pediatric anesthesia. Using induction rooms has potential advantages for the child. There is, however, the potential for the anesthesia to lighten during transfer. The anesthesia circuit is disconnected for several seconds during transfer, and the alveolar concentration of volatile anesthetics will decrease slightly if room air is inhaled. This effect is enhanced if the circuit in the operating room has not been primed with suitable concentrations of anesthetic. Although it cannot be certain, five of the seven cases of true awareness (Cases 1, 2, 4, 6, and 7) could have occurred shortly after transfer. This article may indicate the need for anesthesiologists to be more careful during transfer. However, it would be imprudent for anesthesiologists who do not use induction rooms to dismiss the findings of this article as irrelevant to their practice. The administration of anesthesia involves balancing multiple cues and priorities. Preventing awareness is one of those priorities. In adult anesthesia, awareness is well recognized as a significant complication. Since awareness is not well described in pediatric anesthesia, it is possible to speculate that although pediatric anesthesiologists never intend to increase the risk of awareness, they may be more likely to occasionally accept situations in which awareness is a greater risk. Given that awareness does occur in children, it is unclear why no recent case series have been published. Possible explanations might include the difficulty in verifying a child s account of an event or the tendency for children not to report the experience. They may not report the event either because they are traumatized or, conversely, because they do not attach the same significance to an awareness event that an adult might. In conclusion, we have demonstrated that awareness can occur in children. This is despite the infrequent use of muscle relaxants. The etiology of awareness in children remains unclear. We did not detect a substantial degree of distress or behavior disturbance associated with the awareness. However, because of the small numbers and limited behavioral analysis, we cannot exclude that some children are significantly affected by an awareness experience. Given that awareness does occur in children, further work is needed to examine the causes, consequences, and possibilities of applying new technologies to reduce this incidence. The first step in reducing awareness in children is to acknowledge that it occurs. We thank all members of The Royal Children s Hospital Department of Anesthesia; staff from the day surgical unit, presurgical unit, and recovery; and the Clinical Epidemiology and Biostatistics Unit for their assistance with this study. In particular, we acknowledge the assistance from other institutions of Drs. Tom Mohler, Ian Webb, Patrick Farrell, and Ian Chapman. References 1. Sandin RH, Enlund G, Samuelsson P, Lennmarken C. Awareness during anaesthesia: a prospective case study. Lancet 2000; 355: Myles PS, Williams DL, Hendrata M, et al. Patient satisfaction after anaesthesia and surgery: results of a prospective survey of 10,811 patients. Br J Anaesth 2000;84: Liu WH, Thorp TA, Graham SG, Aitkenhead AR. Incidence of awareness with recall during general anaesthesia. Anaesthesia 1991;46: Wennervirta J, Ranta SO, Hynynen M. Awareness and recall in outpatient anesthesia. Anesth Analg 2002;95:72 7.

9 ANESTH ANALG PEDIATRIC ANESTHESIA DAVIDSON ET AL ;100: AWARENESS IN CHILDREN 5. Sebel PS, Bowdle TA, Ghoneim MM, et al. The incidence of awareness during anesthesia: a multicenter United States study. Anesth Analg 2004;99: Lennmarken C, Bildfors K, Enlund G, et al. Victims of awareness. Acta Anaesthesiol Scand 2002;46: McKie BD, Thorp EA. Awareness and dreaming during anaesthesia in a paediatric hospital. Anaesth Intensive Care 1973;1: Hobbs AJ, Bush GH, Downham DY. Peri-operative dreaming and awareness in children. Anaesthesia 1988;43: O Sullivan EP, Childs D, Bush GH. Peri-operative dreaming in paediatric patients who receive suxamethonium. Anaesthesia 1988;43: Kalff AC, Bonke B, Wolters G, Manger FW. Implicit memory for stimuli presented during inhalation anesthesia in children. Psychol Rep 1995;77: Bonke B, Van Dam ME, Van Kleff JW, Slijper FM. Implicit memory tested in children during inhalation anaesthesia. Anaesthesia 1992;47: Standen PJ, Hain WR, Hosker KJ. Retention of auditory information presented during anaesthesia: a study of children who received light general anaesthesia. Anaesthesia 1987;42: Mapleson WW. Effect of age on MAC in humans: a metaanalysis. Br J Anaesth 1996;76: Myles PS, Leslie K, McNeil J, et al. Bispectral index monitoring to prevent awareness during anaesthesia: the B-Aware randomised controlled trial. Lancet 2004;363: Vernon DT, Schulman JL, Foley JM. Changes in children s behavior after hospitalization: some dimensions of response and their correlates. Am J Dis Child 1966;111: Hellwagner K, Holzer A, Gustorff B, et al. Recollection of dreams after short general anaesthesia: influence on patient anxiety and satisfaction. Eur J Anaesthesiol 2003;20: Ranta SO, Laurila R, Saario J, et al. Awareness with recall during general anesthesia: incidence and risk factors. Anesth Analg 1998;86: Cobcroft MD, Forsdick C. Awareness under anaesthesia: the patients point of view. Anaesth Intensive Care 1993;21: Osterman JE, van der Kolk BA. Awareness during anesthesia and posttraumatic stress disorder. Gen Hosp Psychiatry 1998; 20: Thompson RH, Vernon DT. Research on children s behavior after hospitalization: a review and synthesis. J Dev Behav Pediatr 1993;14: Brice DD, Hetherington RR, Utting JE. A simple study of awareness and dreaming during anaesthesia. Br J Anaesth 1970;42: Bruck M, Ceci SJ. The suggestibility of children s memory. Annu Rev Psychol 1999;50: Peterson C, Bell M. Children s memory for traumatic injury. Child Dev 1996;67: Ceci SJ, Huffman ML. How suggestible are preschool children? Cognitive and social factors. J Am Acad Child Adolesc Psychiatry 1997;36: Gregory GA, Eger EI II, Munson ES. The relationship between age and halothane requirement in man. Anesthesiology 1969;30: Lerman J, Sikich N, Kleinman S, Yentis S. The pharmacology of sevoflurane in infants and children. Anesthesiology 1994;80: Katoh T, Suguro Y, Ikeda T, et al. Influence of age on awakening concentrations of sevoflurane and isoflurane. Anesth Analg 1993;76: Paul M, Fisher DM. Are estimates of MAC reliable? Anesthesiology 2001;95: Eyres R. Update on TIVA. Paediatr Anaesth 2004;14:374 9.