Articles

Neurodevelopmental outcome at 2 years of age after general

anaesthesia and awake-regional anaesthesia in infancy (GAS):
an international multicentre, randomised controlled trial
Andrew J Davidson, Nicola Disma, Jurgen C de Graaff, Davinia E Withington, Liam Dorris, Graham Bell, Robyn Stargatt, David C Bellinger,
Tibor Schuster, Sarah J Arnup, Pollyanna Hardy, Rodney W Hunt, Michael J Takagi, Gaia Giribaldi, Penelope L Hartmann, Ida Salvo, Neil S Morton,
Britta S von Ungern Sternberg, Bruno Guido Locatelli, Niall Wilton, Anne Lynn, Joss J Thomas, David Polaner, Oliver Bagshaw, Peter Szmuk,
Anthony R Absalom, Geoff Frawley, Charles Berde, Gillian D Ormond, Jacki Marmor, Mary Ellen McCann, for the GAS consortium*

Summary
Background Preclinical data suggest that general anaesthetics affect brain development. There is mixed evidence from Lancet 2016; 387: 239–50
cohort studies that young children exposed to anaesthesia can have an increased risk of poor neurodevelopmental Published Online
outcome. We aimed to establish whether general anaesthesia in infancy has any effect on neurodevelopmental October 25, 2015
http://dx.doi.org/10.1016/
outcome. Here we report the secondary outcome of neurodevelopmental outcome at 2 years of age in the General
S0140-6736(15)00608-X
Anaesthesia compared to Spinal anaesthesia (GAS) trial.
This online publication has been
corrected. The corrected version
Methods In this international assessor-masked randomised controlled equivalence trial, we recruited infants younger first appeared at thelancet.com
than 60 weeks postmenstrual age, born at greater than 26 weeks’ gestation, and who had inguinal herniorrhaphy, on Jan 14, 2016
from 28 hospitals in Australia, Italy, the USA, the UK, Canada, the Netherlands, and New Zealand. Infants were See Comment page 202
randomly assigned (1:1) to receive either awake-regional anaesthesia or sevoflurane-based general anaesthesia. *See appendix for a full list of
Web-based randomisation was done in blocks of two or four and stratified by site and gestational age at birth. Infants study investigators

were excluded if they had existing risk factors for neurological injury. The primary outcome of the trial will be the Anaesthesia and Pain
Management Research Group,
Wechsler Preschool and Primary Scale of Intelligence Third Edition (WPPSI-III) Full Scale Intelligence Quotient Murdoch Childrens Research
score at age 5 years. The secondary outcome, reported here, is the composite cognitive score of the Bayley Scales of Institute, Melbourne, VIC,
Infant and Toddler Development III, assessed at 2 years. The analysis was as per protocol adjusted for gestational age Australia (A J Davidson MD,
at birth. A difference in means of five points (1/3 SD) was predefined as the clinical equivalence margin. This trial is M J Takagi PhD,
P L Hartmann BPsych,
registered with ANZCTR, number ACTRN12606000441516 and ClinicalTrials.gov, number NCT00756600. G Frawley MBBS,
G D Ormond MSc); Melbourne
Findings Between Feb 9, 2007, and Jan 31, 2013, 363 infants were randomly assigned to receive awake-regional Children’s Trials Centre,
anaesthesia and 359 to general anaesthesia. Outcome data were available for 238 children in the awake-regional group Murdoch Childrens Research
Institute, Melbourne, VIC,
and 294 in the general anaesthesia group. In the as-per-protocol analysis, the cognitive composite score (mean [SD]) Australia (A J Davidson);
was 98·6 (14·2) in the awake-regional group and 98·2 (14·7) in the general anaesthesia group. There was equivalence Department of Anaesthesia and
in mean between groups (awake-regional minus general anaesthesia 0·169, 95% CI –2·30 to 2·64). The median Pain Management, The Royal
duration of anaesthesia in the general anaesthesia group was 54 min. Children’s Hospital, Melbourne,
VIC, Australia (A J Davidson,
G Frawley); Department of
Interpretation For this secondary outcome, we found no evidence that just less than 1 h of sevoflurane anaesthesia in Paediatrics, University of
infancy increases the risk of adverse neurodevelopmental outcome at 2 years of age compared with awake-regional Melbourne, Melbourne, VIC,
anaesthesia. Australia (A J Davidson,
R W Hunt PhD); Department of
Anesthesia, Istituto Giannina
Funding Australia National Health and Medical Research Council (NHMRC), Health Technologies Assessment- Gaslini, Genoa, Italy
National Institute for Health Research UK, National Institutes of Health, Food and Drug Administration, Australian (N Disma MD, G Giribaldi MD);
Department of Anaesthesia,
and New Zealand College of Anaesthetists, Murdoch Childrens Research Institute, Canadian Institute of Health
Wilhelmina Children’s Hospital,
Research, Canadian Anesthesiologists’ Society, Pfizer Canada, Italian Ministry of Heath, Fonds NutsOhra, and UK University Medical Center
Clinical Research Network (UKCRN). Utrecht, Utrecht, Netherlands
(J C de Graaff PhD); Department
of Anesthesia, Montreal
Introduction with several different general anaesthesia drugs, are
Children’s Hospital, Montreal,
Substantial preclinical evidence exists that describes how greater with longer exposure, and are less severe in older Canada (D E Withington BM);
general anaesthesia drugs change brain development in animals.2,8 The clinical relevance of these findings is Department of Anesthesia,
young animals.1 These changes include accelerated unknown and much debated.11–14 McGill University, Montreal,
Canada (D E Withington);
apoptosis and other effects such as changes to dendritic In human beings, there is conflicting evidence for an
Paediatric Neurosciences
morphology.2–5 Findings have also shown that exposure association between exposure to anaesthesia in early Research Group, Fraser of
to general anaesthesia in young animals is associated childhood and adverse long-term neurodevelopmental Allander Unit (L Dorris DClinPsy),
with long-term cognitive and behavioural changes.3,6,7 outcome; however, confounding restricts any assumption Department of Anaesthesia
(G Bell MBChB, N S Morton MD),
These effects have been described in various species of causality.15–30 Young children who receive anaesthesia
Royal Hospital for Children,
including non-human primates.7–10 The changes are seen are inevitably having surgery or an investigative procedure.

www.thelancet.com Vol 387 January 16, 2016 239

 Articles

Glasgow; Mental Health and

Wellbeing, University of Research in context
Glasgow, Glasgow, UK (L Dorris);
School of Psychological Science, Evidence before this study neurodevelopmental outcome. We used the best measure of
La Trobe University, Victoria, We searched MEDLINE and Cochrane controlled trial register neurodevelopment available to assess 2-year-old children, and
VIC, Australia (R Stargatt PhD); (last search done on Sept 18, 2015) for original research and noted strong evidence for equivalence between the use of
Child Neuropsychology,
Murdoch Childrens Research
meta-analyses describing the association between anaesthesia awake-regional anaesthesia and just less than 1 h of general
Institute, Melbourne, VIC, exposure in early life and neurodevelopmental outcome. We anaesthesia. However, it should be noted that this was an
Australia (R Stargatt, M J Takagi); used combinations of the search terms “anesthesia”, and “child analysis of a secondary outcome with the primary outcome
Department of Neurology development”, or “learning disorders”. The search found no planned at 5 years of age, and in view of the limited sensitivity
(D C Bellinger PhD,
J Marmor MEd), Department of
randomised trials but several cohort studies. Several reviews of developmental assessment at 2 years of age, this trial does
Psychiatry (D C Bellinger), have concluded that there is an association between anaesthesia not provide the definitive answer.
Boston Children’s Hospital, in childhood and neurodevelopmental outcome. Findings of
Harvard Medical School, Implications of all the available evidence
two meta-analyses have shown an association between
Boston, MA, USA; Department Although there are some limitations that should be noted
of Environmental Health,
anaesthesia in children and a range of neurodevelopmental
when interpreting the trial, the randomised prospective design
Harvard T H Chan School of outcomes. All reviews and meta-analyses acknowledge the
adds substantially to the weight that should be given to the
Public Health, Boston, MA, USA weaknesses of the cohort studies; including strong likelihood of
(D C Bellinger); Clinical results compared with the mixed results found in previous
confounding, bias, heterogeneous populations at times of
Epidemiology and Biostatistics cohort studies. However, reassessment at an older age is
Unit (T Schuster PhD,
exposure, and heterogeneous outcome measures, some of
necessary before definitive conclusions can be drawn. The
S J Arnup MBiostat), Neonatal which are poorly defined or insensitive. All reviews conclude that
trial does not rule out the possibility that longer or many
Research Group (R W Hunt), causation cannot be established or excluded.
Murdoch Childrens Research exposures to anaesthesia in early childhood can cause
Institute, Melbourne, VIC, Added value of this study neurodevelopmental changes. Further research is needed to
Australia; National Perinatal We report results from the first randomised controlled trial address these questions.
Epidemiology Unit, Clinical
Trials Unit, University of
assessing the effect of general anaesthesia in infancy on
Oxford, Oxford, UK
(P Hardy MSc); Department of
Added risk of poor neurodevelopmental outcome might anaesthesia versus general anaesthesia in infants
Neonatal Medicine, The Royal
Children’s Hospital, Melbourne, be due to the underlying pathology, comorbidity, or other undergoing inguinal herniorrhaphy. The trial was done at
Australia (R W Hunt); perioperative risk factors. These results have prompted 28 hospitals in Australia, Italy, the USA, the UK, Canada,
Department of Anesthesiology recommendations to consider delaying surgery in infancy the Netherlands, and New Zealand. Institutional review
and Pediatric Intensive Care,
and there have been several calls for more research to board or ethics committee approval was obtained at each
Ospedale Pediatrico ‘Vittore
Buzzi’, Milan, Italy (I Salvo MD); address this important issue.12,13,31 site and written consent obtained from the child’s parents
University of Glasgow, Glasgow, In view of the many potential confounding factors, a or guardians. A summary of the protocol is available online.
UK (N S Morton); School of randomised trial is the best study design to establish
Medicine and Pharmacology,
University of Western Australia,
whether anaesthesia exposure in early childhood causes Participants
Perth, WA, Australia long-term neurodevelopmental changes. Fortuitously Eligibility criteria included infants up to 60 weeks
(Prof B S von Ungern there are two established anaesthetic techniques for postmenstrual age (ie, gestational age at birth plus
Sternberg PhD); Department of inguinal herniorrhaphy in infancy; awake-regional and chronological age) scheduled for unilateral or bilateral
Anaesthesia and Pain
Management, Princess
sevoflurane-based general anaesthesia. Therefore, we inguinal herniorraphy born at greater than 26 weeks’
Margaret Hospital for Children, undertook a randomised controlled trial comparing gestation. Exclusion criteria included any contraindication
Perth, WA, Australia neurodevelopmental outcome in children who were for either anaesthetic technique, a history of congenital
(Prof B S von Ungern Sternberg); randomly assigned to receive either awake-regional or heart disease requiring surgery or pharmacotherapy,
Department of Anesthesia,
Ospedale Papa Giovanni XXIII,
sevoflurane-based general anaesthesia for inguinal mechanical ventilation immediately before surgery,
Bergamo, Italy herniorrhaphy in early infancy: the General Anaesthesia known chromosomal abnormalities or other known
(B G Locatelli MD); Department compared to Spinal anaesthesia (GAS) trial. The primary acquired or congenital abnormalities that might affect
of Paediatric Anaesthesia and outcome for the trial will be the Wechsler Preschool and neurodevelopment, previous exposure to volatile general
Operating Rooms, Starship
Children’s Hospital, Auckland
Primary Scale of Intelligence Third Edition (WPPSI-III) anaesthesia or benzodiazepines as a neonate or in the
District Health Board, Auckland, Full Scale Intelligence Quotient score at age 5 years. As a third trimester in utero, any known neurological injury
New Zealand (N Wilton MBBS); secondary outcome, we also planned a priori to assess such as cystic periventricular leukomalacia or grade
Department of Anesthesiology, neurodevelopmental outcome at age 2 years. In this paper three or four intraventricular haemorrhage, any social or
University of Washington,
Seattle, WA, USA (A Lynn MD);
we report all secondary outcomes at 2 years of age. Data geographical factor that might make follow-up difficult
Department of Anesthesia, from the trial relating to post-anaesthesia apnoea and (eg, planned house move, homelessness, no telephone
University of Minnesota, success of regional block have been published elsewhere.32,33 communication available), or having a primary language
Minneapolis, MN, USA
at home in a region where neurodevelopmental tests are
(J J Thomas MD); Department of
Anesthesiology, Children’s Methods not available in that language. We identified eligible
Hospital Colorado and Study design infants from operating room schedules or at pre-
University of Colorado School of In this observer-blind, international, multisite, randomised, admission clinics and recruited in the clinic or in the
Medicine, Aurora, CO, USA
controlled, equivalence trial, we assessed awake-regional preadmission areas of the operating floor.

240 www.thelancet.com Vol 387 January 16, 2016

 Articles

Randomisation and masking were not informed of the group allocation but were told if (Prof D Polaner MD);
A 24 h web-based randomisation service was managed by they asked. The psychologists and paediatricians who did Department of Anaesthesia,
Birmingham Children’s
the Data Management and Analysis Centre, Department the assessment were masked to group allocation. Once Hospital, Birmingham, UK
of Public Health, University of Adelaide, Australia. their assessment was completed they were asked to (O Bagshaw FRCA); Department
Participants were randomly assigned (1:1) to receive either indicate if they were aware of group allocation. of Anesthesiology, Children’s
general anaesthesia or awake-regional anaesthesia. Medical Centre Dallas, Dallas,
TX, USA (P Szmuk MD);
Randomisation was done in blocks of two or four and Procedures Department of
stratified by site and gestational age at birth: 26–29 weeks The awake-regional group received either an awake-spinal Anaesthesiology, University
and 6 days, 30–36 weeks and 6 days, and 37 weeks or more. anaesthetic, an awake-caudal anaesthetic, or a combined Medical Centre Groningen,
The anaesthetist was aware of group allocation. Parents spinal-caudal anaesthetic according to institutional Groningen University,
Groningen, Netherlands
(Prof A R Absalom MBChB); and
4023 patients screened for eligibility 3301 excluded Department of Anesthesiology,
1085 children met predefined exclusion criteria Perioperative and Pain
1084 surgeon or anaesthesiologist or both Medicine, Boston Children’s
declined participation Hospital, Harvard Medical
728 parent or guardian did not consent School, Boston, MA, USA
404 other reasons for not randomising (Prof C Berde MD,
722 randomly assigned (largely logistical reasons) M E McCann MD)
Correspondence to:
Dr Andrew J Davidson,
Anaesthesia and Pain
363 assigned to awake-regional anaesthesia 359 assigned to general anaesthesia Management Research Group,
Murdoch Childrens Research
Institute, The Royal Children’s
2 misrandomised 1 withdrawal of consent Hospital, Flemington Road,
Parkville, Victoria 3052, Australia
[email protected]
361 in intention-to-treat analysis 358 in intention-to-treat analysis See Online for appendix
For the protocol see http://www.
thelancet.com/protocol-
74 violated protocol 2 violated protocol
reviews/09PRT-9078
5 surgery cancelled 2 surgery cancelled
69 general anaesthetic required

14 lost to follow-up, no assessment 1 lost to follow-up, no assessment

54 Bayley measurement completed 1 Bayley measurement completed

6 partial assessment,
2 child uncooperative
1 child unwell
1 visual defect
1 hearing defect
1 reason unknown

287 in per-protocol analysis 356 in per-protocol analysis

37 lost to follow-up, no assessment 47 lost to follow-up, no assessment

12 partial assessment 15 partial assessment

2 administration error 4 administration error
2 child uncooperative 1 child uncooperative
1 acquired brain injury 1 child unwell
7 cause for incomplete assessment 3 visual defect
unknown 1 developmental delay
5 cause for incomplete assessment
unknown

238 complete case 294 complete case

Figure: Trial profile

www.thelancet.com Vol 387 January 16, 2016 241

 Articles

protocols. Spinal anaesthesia was done with 0·2 mL/kg other forms of sedation avoided. If the awake-regional
0·5% isobaric bupivacaine with a minimum volume of anaesthesia was ineffective then a general anaesthesia
0·5 mL. Because isobaric bupivacaine was unavailable at was done with sevoflurane, and if the child became
some sites, other agents were used (in the USA, unsettled intraoperatively, sevoflurane was given to
0·13 mL/kg of hyperbaric 0·75% bupivacaine, and in supplement the awake-regional anaesthesia. Both were
the UK 0·2 mL/kg 0·5% levobupivacaine). Caudal regarded as protocol violations.
anaesthesia was done with up to a total dose of 2·5 mg/kg The general anaesthesia group received sevoflurane
of 0·25% bupivacaine. In the UK, 0·25% levobupivacaine for induction and maintenance in a mix of air and
was used. In the USA, if surgery was likely to take longer oxygen. The concentration of sevoflurane was left to the
than 1 h, some patients were given a loading dose of 3% discretion of the anaesthetist, as was choice of airway
chloroprocaine (1 mL/kg in divided doses of no more device, ventilation technique, and use of any
than 0·25 mL/kg per 15 s) via a caudal cannula and then neuromuscular blocking agents. No opioid or nitrous
an infusion of 1–2 mL/kg per h. Ilioinguinal and field oxide was allowed. A caudal, ilioinguinal–iliohypogastric
blocks could also be done. The total dose of bupivacaine or field block with bupivacaine could be done in both
did not exceed 2·5 mg/kg. In the awake-regional group, groups to provide postoperative analgesia. Oral or
oral sucrose was used to settle the child if needed and all intravenous acetaminophen could also be given. Heart

RA group as per GA group as per RA group intention GA group intention to

protocol (N=287) protocol (N=356) to treat (N=361) treat (N=358)
Baseline demographics
Sex, male 232 (81%) 304 (85%) 294 (82%) 306 (86%)
Chronological age at surgery (days) 68·9 (31) 71·1 (32) 70·1 (32) 71·0 (32)
Postmenstrual age at surgery (days) 317·2 (32) 319·7 (32) 318·3 (33) 319·5 (32)
Weight of child at surgery (kg) 4·2 (1·1) 4·3 (1·1) 4·2 (1·1) 4·3 (1·1)
Pregnancy and birth details
Postmenstrual age at birth (days) 248·2 (29) 248·6 (27) 248·3 (29) 248·6 (27)
Prematurity (born <37 weeks’ gestation) 160 (56%) 195 (55%) 198 (55%) 196 (55%)
Birthweight (kg) 2·3 (0·9) 2·3 (0·9) 2·4 (0·9) 2·3 (0·9)
Z score for birthweight –0·68 (1·3) –0·69 (1·3) –0·66 (1·2) –0·69 (1·3)
Apgar score at 1 min 9 (7–9) 8·5 (7–9) 9 (7–9) 9 (7–9)
Apgar score at 5 min 9 (9–10) 9 (9–10) 9 (9–10) 9 (9–10)
One of a multiple pregnancy 52 (18%) 61 (17%) 62 (17%) 62 (17%)
Mother received partial course antenatal steroids 16 (6%) 19 (5%) 20 (6%) 19 (5%)
Mother received complete course antenatal steroids 95 (33%) 98 (28%) 114 (32%) 98 (28%)
Mother diagnosed with chorioamnionitis 10 (4%) 12 (3%) 11 (3%) 12 (3%)
Prolonged rupture of the membranes (>24 h) 28 (10%) 34 (10%) 32 (9%) 34 (10%)
Mother diagnosed with pre-eclampsia 50 (17%) 68 (19%) 60 (17%) 68 (19%)
Sepsis during pregnancy 36 (13%) 50 (14%) 43 (12%) 50 (14%)
Mode of delivery of birth
Cephalic vaginal 135 (47%) 157 (44%) 169 (47%) 157 (44%)
Breech vaginal 1 (<1%) 6 (2%) 3 (1%) 6 (2%)
Compound vaginal 2 (1%) 4 (1%) 3 (1%) 4 (1%)
Caesarean section 149 (52%) 189 (53%) 185 (51%) 191 (53%)
Caesarean section and mother went into labour 42 (15%) 58 (16%) 52 (14%) 59 (16%)
Mother exposed to nitrous oxide during delivery 48 (18%) 62 (18%) 61 (18%) 62 (18%)
IVH 7 (2%) 6 (2%) 8 (2%) 6 (2%)
IVH grade 1 5 (2%) 6 (2%) 5 (2%) 6 (2%)
IVH grade 2 2 (1%) 0 2 (1%) 0
Retinopathy of prematurity 17 (9%) 16 (6%) 30 (8%) 16 (6%)
Hearing defects detected by perinatal screening 7 (3%) 10 (3%) 8 (3%) 10 (3%)
PDA diagnosed 23 (8%) 21 (6%) 27 (8%) 21 (6%)
PDA never treated 9 (3%) 9 (3%) 11 (3%) 9 (3%)
PDA treated with non-steroidal anti-inflammatory drugs 14 (5%) 10 (3%) 16 (4%) 10 (3%)
(Table 1 continues on next page)

242 www.thelancet.com Vol 387 January 16, 2016

 Articles

rate, blood pressure, oxygen saturation, and (where Communicative Development Inventory: Words and
applicable) expired sevoflurane concentrations were Sentences (MacArthur-Bates).35 The MacArthur-Bates is a
recorded every 5 min. parent informant measure that assesses expressive
Serum glucose was measured after anaesthetic language in children aged 16–30 months. We also
induction. There were rescue protocols for hypoglycaemia, recorded demographic data, family history, and medical
hypotension, and hypoxaemia. If the blood pressure fell history, and did a brief physical and neurological
more than 20% below baseline, an intravenous bolus fluid examination. The physical examination included
was given plus vasoactive drugs if deemed necessary. anthropometric measurements such as length, weight,
Hypoglycaemia (blood sugar <3·0 mmol/L) was treated and arm and head circumference. The neurological
with a bolus of 5 mL/kg of 10% dextrose. Oxygen by face examination included cranial nerve examination, posture
mask in the awake-regional arm and an increased FiO2 in assessment, and the muscle strength, tone, and reflexes
the general anaesthesia group was used at the discretion of the arms and legs.
of the anaesthetist to maintain arterial oxygen saturation All study data were sent to the Murdoch Children’s
higher than 95%. Research Institute in Melbourne, Australia. All forms
Assessments were undertaken within 2 months either were checked for data quality by trained research assistants
side of 2 years of age (corrected for prematurity). The and double checked by a research assistant who was not
assessment took about 2 h to complete. A trained involved in the primary data collection or entry. An
psychologist administered the Bayley-III. 34 The Bayley-III independent data safety monitoring committee met every
has cognitive, language, and motor scales. The cognitive 6 months during recruitment. Summary data by allocation
scale includes tasks assessing attention, memory, were presented to the committee. There were no formal
sensorimotor development, exploration, concept form- interim analyses of neurodevelopmental outcome.
ation, and simple problem solving. The language scale
assesses expressive and receptive skills, and the motor Statistical analysis
scale assesses fine and gross motor skills. Parents The main outcome for the analysis at 2 years of age was
completed the Bayley-III Social-Emotional and Adaptive prespecified to be the composite cognitive score of the
Behaviour Questionnaires and the MacArthur-Bates Bayley-III. The hypothesis (as stated in the protocol)

RA group as per GA group as per RA group intention GA group intention to

protocol (N=287) protocol (N=356) to treat (N=361) treat (N=358)
(Continued from previous page)
Familial demographics
Primary language(s) only spoken* 252 (88%) 305 (86%) 311 (86%) 307 (86%)
Maternal age at birth >21 years 273 (96%) 339 (95%) 339 (95%) 341 (95%)
Family structure two caregivers together, at birth 261 (91%) 324 (91%) 328 (91%) 326 (91%)
Maternal education
Completed tertiary studies 150 (52%) 171 (48%) 181 (51%) 171 (48%)
Continuing tertiary studies 50 (17%) 67 (19%) 68 (19%) 67 (19%)
Completed year 11 or 12 62 (22%) 83 (23%) 77 (22%) 84 (24%)
Did not complete year 11 25 (9%) 33 (9%) 32 (9%) 34 (10%)
Anaesthesia details
Blood glucose concentration (mmol/L) 5·4 (4·7–6·1) 5·5 (4·8–6·4) 5·4 (4·7–6·2) 5·5 (4·8–6·4)
Rescue glucose given intravenously 2 (1%) 4 (1%) 2 (1%) 4 (1%)
Haemoglobin (g/100 mL) 10·3 (2·1) 10·2 (2·0) 10·3 (2·1) 10·2 (2·0)
Need for fluid bolus for hypotension 15 (5%) 59 (17%) 21 (6%) 59 (17%)
Vasoactive drugs given (including atropine) 4 (1%) 17 (5%) 6 (2%) 17 (5%)
Duration of surgery (min) 26·0 (19·0–35·0) 28·0 (20·0–40·0) 28·0 (20·0–38·0) 28·0 (20·0–40·0)
Duration of sevoflurane exposure (min) NA 54·0 (41·0–70·0) 42·0 (31·0–62·5)† 54·0 (41·0–70·0)
End tidal sevoflurane concentration (%) NA 2·6 (0·7) 2·3 (0·8)† 2·6 (0·7)
Total concentration per h NA 2·6 (1·1) 1·9 (1·0)† 2·6 (1·1)
Any significant apnoea to 12 h postoperatively‡ 6 (2%) 15 (4%) 10 (3%) 15 (4%)

Data are n (% of non-missing data) or mean (SD), median (IQR) unless otherwise stated. RA=awake-regional anaesthesia. GA=general anaesthesia. IVH=intraventricular
haemorrhage. PDA=patent ductus arteriosus. *The primary language spoken at home is the primary language in each country that the Bayley was done (eg, Italian in Italy).
†For those cases that received sevoflurane. ‡Significant apnoea defined as a pause in breathing for >15 s or >10 s if associated with oxygen saturation <80% or bradycardia
(20% decrease in heart rate).

Table 1: Baseline demographics

www.thelancet.com Vol 387 January 16, 2016 243

 Articles

was that the composite cognitive score of the Bayley-III

RA group as GA group as RA group GA group measured at 2 years of age in infants who are
per protocol per protocol intention to intention to anaesthetised for inguinal herniorraphy is equivalent
(N=287) (N=356) treat (N=361) treat (N=358)
when using general anaesthesia compared with
Assessment details awake-regional anaesthesia. The components of the
Location of 2-year assessment at hospital 204 (96%) 240 (94%) 250 (95%) 241 (94%) Bayley-III are reported as scaled scores and as composite
Family demographics at 2 years scores. The five composite scores (cognitive, language,
Paid employment is main family income 222 (90%) 267 (88%) 274 (90%) 268 (88%) motor, adaptive behaviour, and social-emotional scales)
Family structure, two caregivers living 226 (91%) 274 (90%) 277 (90%) 275 (90%) are standardised to have a mean of 100 and an SD of 15
together
in the reference population. The subscales (eg, fine
Number of children at home
motor scale) are reported as scaled scores, with a mean
1 88 (36%) 118 (39%) 115 (37%) 118 (39%) of 10 and an SD of 3. The other secondary outcomes for
2 109 (44%) 120 (40%) 131 (43%) 121 (40%) this analysis are the language, motor, social-emotional,
3 37 (15%) 43 (14%) 45 (14%) 43 (14%) and adaptive behaviour scores from the Bayley-III and
>3 14 (6%) 22 (7%) 17 (6%) 22 (7%) the age-adjusted Vocabulary Production Score from
Birth order the MacArthur-Bates. Published normative scores were
1 123 (50%) 161 (53%) 154 (50%) 161 (53%) used at all sites with forms and instructions translated
2 87 (35%) 90 (30%) 107 (35%) 91 (30%) locally. Diagnosis of cerebral palsy was another
>2 37 (15%) 52 (17%) 46 (15%) 52 (17%) prespecified secondary outcome
Corrected age at assessment (weeks) 108·9 (13·0) 108 (9·8) 108·7 (12·5) 108 (9·8) Because this is an equivalence study, the outcome was
Events since original anaesthesia analysed on an as-per-protocol basis to ensure a
Number of hospitalisations since inguinal conservative estimate in the direction of non-equivalence.
herniorrhaphy operation Equivalence was defined a priori if the 95% confidence
0 172 (69%) 206 (68%) 210 (68%) 207 (68%) interval of the difference in means lies within minus
1 51 (20%) 64 (21%) 69 (22%) 64 (21%) five and plus five points. Intention-to-treat analyses were
2 14 (6%) 18 (6%) 16 (5%) 18 (6%) also planned. Analyses were adjusted for categories of
>2 6 (2%) 8 (3%) 8 (3%) 8 (3%) gestational age at birth (182–209 days; 210–258 days;
Number of anaesthetics since inguinal ≥259 days).
herniorrhaphy operation
The sample size was based on the primary outcome
1 34 (14%) 36 (12%) 42 (14%) 36 (12%)
for the GAS trial; the 5-year follow-up WPPSI-III
2 5 (2%) 6 (2%) 6 (2%) 6 (2%) Full Scale Intelligence Quotient score. Assuming an
>2 4 (2%) 4 (1%) 4 (1%) 4 (1%) expected difference of one standardised score point and
Child had a head injury that involved the 7 (3%) 4 (1%) 7 (2%) 4 (1%) a 90% chance that a 95% CI will exclude a difference of
loss of consciousness
more than five points (the largest difference acceptable
Child has an acquired brain injury 1 (0%) 1 (0%) 1 (0%) 1 (0%)
to show equivalence), the trial would need 598 infants.
Child has any malformations
Enrolling roughly 720 participants would allow for 10%
Cardiac 0 4 (1%) 0 4 (1%)
loss to follow-up and 10% with a major protocol violation.
CNS 3 (1%) 1 (<1%) 3 (1%) 1 (<1%)
We used multiple imputation with chained equations
Genitourinary 6 (2%) 4 (1%) 8 (3%) 4 (1%)
to impute missing outcome data in the analysis of all
Genetic condition 1 (<1%) 0 1 (<1%) 0
outcomes.36 The following prespecified variables were
Respiratory 0 1 (<1%) 0 1 (<1%) used as predictor variables within the imputation
Skeletal 4 (2%) 11 (4%) 4 (1%) 11 (4%) approach: anaesthesia group, country, sex, gestational
Cleft lip or palate 1 (<1%) 0 1 (<1%) 0 age at birth, standardised Z score for birthweight,
Craniofacial 2 (1%) 0 2 (1%) 0 mother received antenatal steroids, mother diagnosed
Child has any chronic illness 42 (17%) 43 (14%) 50 (16%) 43 (14%) with chorioamnionitis, intraventricular haemorrhage,
Child had any prescribed medication for 43 (17%) 50 (16%) 93 (17%) 59 (19%) maternal age, maternal education, rescue glucose given
2 months or longer
intravenously, need for fluid bolus for hypotension,
Child had febrile seizures after the hernia 8 (3%) 9 (3%) 10 (3%) 9 (3%)
repair
vasoactive drugs given for hypotension, duration of
surgery, dose of sevoflurane (concentration multiplied
Child had other seizures after the hernia 1 (<1%) 4 (1%) 1 (<1%) 4 (1%)
repair by h), significant postoperative apnoea, corrected age at
The child has had an intervention for 46 (19%) 55 (18%) 54 (18%) 55 (18%) assessment, any more anaesthetic exposures since the
neurodevelopmental issues since the inguinal herniorraphy, any malformations, any chronic
inguinal herniorrhaphy operation illness, any prescribed medication for 2 months or
Speech therapy 22 (9%) 27 (9%) 28 (9%) 27 (9%) longer, total length of any readmission to hospital, any
Physiotherapy 22 (9%) 27 (9%) 26 (8%) 27 (9%) interventions for neurodevelopmental problems,
(Table 2 continues on next page) diagnosis of cerebral palsy, any other neurological
abnormality.

244 www.thelancet.com Vol 387 January 16, 2016

 Articles

For the purpose of sensitivity analysis, effect estimates

RA group as GA group as RA group GA group
were computed using best and worst case imputation per protocol per protocol intention to intention to
scenarios. Furthermore, effect estimates and CIs based on (N=287) (N=356) treat (N=361) treat (N=358)
inverse probability of censoring weighting were reported.37 (Continued from previous page)
Risk ratios with 95% CIs were reported for the Occupational therapy 9 (4%) 12 (4%) 12 (4%) 12 (4%)
proportion of individuals that fell below one and two SDs Psychology 1 (<1%) 6 (2%) 1 (<1%) 6 (2%)
of the composite cognitive score. Risk ratios were Developmental medicine or early 8 (3%) 7 (2%) 9 (3%) 7 (2%)
generated using generalised linear models for a binomial intervention
distributed response variable using a log link Child attends play group or child care on a 147 (60%) 177 (58%) 186 (61%) 178 (58%)
(binomial log-linear regression). These analyses were regular basis
not prespecified in the study protocol (post-hoc analyses). Physical examination
All analyses were done in Stata (version 13). Height (cm) 86·6 (5·5) 86·9 (4·9) 86·4 (5·2) 86·9 (4·9)
This trial is registered with ANZCTR, number Weight (kg) 12·6 (2·0) 12·6 (1·9) 12·6 (2·0) 12·6 (1·9)
ACTRN12606000441516, ClinicalTrials.gov, number Head circumference (cm) 49·1 (2·1) 48·8 (2·2) 49·0 (2·0) 48·8 (2·2)
NCT00756600, and the UK Clinical Research Network Arm circumference (cm) 16·4 (2·0) 16·1 (1·8) 16·4 (2·0) 16·1 (1·8)
(UKCRN), number 12437565.
Data are n (% of non-missing data) or mean (SD), unless otherwise stated. RA=awake-regional anaesthesia. GA=general
anaesthesia.
Role of the funding source
The funders of the study had no role in study design, Table 2: 2-year descriptive statistics demographic data
data collection, data analysis, data interpretation, or
writing of the report. All authors had access to the data and the MacArthur-Bates data for each group. For the
and AJD, GDO, and Suzette Sheppard were responsible cognitive composite score, we noted evidence for
for submitting the manuscript. AJD made the final equivalence in means between the awake-regional
decision to submit the paper for publication. anaesthesia and general anaesthesia groups in both the
as-per-protocol and the intention-to-treat analyses using
Results multiple imputation to account for missing outcome
Between Feb 9, 2007, and Jan 31, 2013, we recruited data (awake-regional minus general anaesthesia: 0·169,
722 infants from 28 hospitals in Australia, the USA, the 95% CI –2·30 to 2·64 for the as-per-protocol analysis
UK, Italy, the Netherlands, Canada, and New Zealand and 0·256, –2·06 to 2·57 for the intention-to-treat
(appendix p4). There were two misrandomisations and analysis). These results were consistent with the findings
one withdrawal of consent leaving 361 in the intention- of the complete case analyses (awake-regional minus
to-treat analysis in the awake-regional anaesthesia group general anaesthesia 0·458, 95% CI –2·02 to 2·94 for the
and 358 in the general anaesthesia group (figure). Table 1 as-per-protocol analysis and 0·430, –1·90 to 2·76 for the
summarises demographic data for each group at baseline intention-to-treat analysis). There was also evidence for
and table 2 summarises demographic data at 2 years. equivalence between groups in the composite motor
There were 74 protocol violations in the awake-regional scores, composite language scores, and the composite
anaesthesia group (five due to surgery being cancelled adaptive behaviour scores (table 4). The results were
and 69 received some sevoflurane or other general consistent in both as-per-protocol and intention-to-treat
anaesthesia) and two violations in the general anaesthesia analyses, and when using complete case and multiple
group (surgery cancelled). imputation. With mean differences of one and two score
Follow-up was from March 5, 2009, to March 6, 2015. points (multiple imputation and complete case analysis
47 families were lost to follow-up in the general for as per protocol and intention to treat) and upper 95%
anaesthesia group and 52 in the awake-regional confidence interval limits exceeding the prespecified five
anaesthesia group. Of those lost to follow-up, some point equivalence margin, evidence for equivalence with
reason for non-attendance was gained in 19 and in only regard to the social-emotional composite scale of the
one case was non-attendance due to developmental delay Bayley-III was not compelling. There was no evidence
(this child was in the awake-regional arm). Of those that for a difference between groups in MacArthur-Bates
attended for assessment, the cognitive scale of the scores (table 4).
Bayley-III was completed by 292 in the awake-regional The appendix shows results of the inverse probability
group and 295 in the general anaesthesia group (figure). weighting and worst case imputation scenarios for
Very few children were unable to complete the Bayley-III missing data (appendix pp 5–6). The worst case scenario
due to developmental delay or other recognised reason results represent theoretical boundaries to what extent
for cognitive impairment. In 97% of cases the psychologist the actual effect estimates could have been affected by
and paediatrician were unaware of group allocation at the selective dropout. However, both multiple imputation
time of assessment (appendix p5). analysis as well as inverse probability weighting showed
Table 3 shows the Bayley-III cognitive, language, consistent robustness of the study findings with regard
motor, social-emotional, and adaptive behaviour scores, to data missingness.

www.thelancet.com Vol 387 January 16, 2016 245

 Articles

RA group as per protocol GA group as per protocol RA group intention to treat GA group intention to treat
Cognitive
Cognitive, scaled score 238, 9·7 (2·8) 294, 9·6 (2·9) 292, 9·7 (2·8) 295, 9·6 (2·9)
Cognitive, composite score 238, 98·6 (14·2) 294, 98·2 (14·7) 292, 98·6 (14·2) 295, 98·2 (14·6)
Language
Receptive language, scaled score 236, 8·7 (2·9) 285, 8·6 (2·9) 287, 8·8 (2·9) 286, 8·6 (2·9)
Expressive language, scaled score 235, 9·3 (2·9) 290, 9·3 (3·0) 287, 9·4 (2·9) 291, 9·3 (3·0)
Language, composite score 235, 94·6 (15·4) 285, 94·0 (15·6) 286, 94·9 (15·5) 286, 94·0 (15·6)
Motor
Fine motor, scaled score 234, 10·5 (2·7) 287, 10·4 (2·7) 287, 10·6 (2·8) 288, 10·4 (2·7)
Gross motor, scaled score 234, 8·8 (2·4) 279, 8·7 (2·6) 285, 8·9 (2·5) 280, 8·7 (2·6)
Motor, composite score 232, 98·3 (13·2) 274, 97·9 (13·4) 283, 98·9 (13·5) 275, 97·8 (13·4)
Social-emotional
Social-emotional, scaled score 218, 9·5 (3·8) 267, 9·1 (3·7) 267, 9·5 (3·8) 268, 9·1 (3·7)
Social-emotional, composite score 218, 97·4 (19·0) 267, 95·4 (18·3) 267, 97·4 (19·2) 268, 95·4 (18·3)
Adaptive behaviour
Communication scaled score 233, 9·7 (2·9) 291, 9·6 (2·9) 288, 9·8 (2·9) 292, 9·6 (2·9)
Community use scaled score 233, 9·8 (2·8) 291, 9·9 (2·7) 288, 9·9 (2·8) 292, 9·8 (2·7)
Functional pre-academics scaled score 233, 9·0 (3·0) 291, 9·2 (2·9) 288, 9·1 (3·0) 292, 9·2 (2·9)
Home living scaled score 233, 9·9 (2·8) 291, 10·1 (2·7) 288, 9·9 (2·9) 292, 10·1 (2·7)
Health and safety scaled score 233, 9·0 (2·8) 291, 9·3 (2·7) 288, 9·0 (2·9) 292, 9·3 (2·7)
Leisure scaled score 233, 9·4 (3·0) 291, 9·9 (2·8) 288, 9·5 (3·1) 292, 9·9 (2·8)
Self-care scaled score 233, 6·8 (2·6) 291, 6·6 (2·5) 288, 6·8 (2·6) 292, 6·6 (2·5)
Self-direction scaled score 233, 9·7 (3·2) 291, 10·0 (3·2) 288, 9·8 (3·2) 292, 10·0 (3·2)
Social scaled score 233, 9·3 (2·9) 291, 9·5 (2·8) 288, 9·4 (2·9) 292, 9·5 (2·8)
Motor scaled score 233, 9·8 (3·2) 291, 10·0 (2·9) 288, 9·9 (3·3) 292, 10·0 (2·9)
Adaptive behaviour composite score 233, 93·1 (15·6) 291, 94·3 (14·7) 288, 93·4 (16·1) 292, 94·3 (14·7)
MacArthur-Bates percentile score 195, 32·4 (27·9) 247, 34·7 (28·7) 240, 33·6 (28·0) 247, 34·7 (28·7)

Data are n, mean (SD). RA=awake-regional anaesthesia. GA=general anaesthesia.

Table 3: Descriptive statistics Bayley-III and MacAthur-Bates scores by group

Overall, only a few children had a diagnosis of cerebral found an association between exposure to anaesthesia
palsy, hearing or visual impairment, or specific in early childhood and increased risk of poor
behavioural diagnoses such as autism spectrum disorder neurodevelopmental outcome.16–18,20–24,28 Although this
(table 5). The event rate was too low for any meaningful association fits with preclinical animal data, it could
comparative analysis. There was no evidence for a also be explained by the confounding effects of surgery,
difference between groups in the proportion of children pathology, or comorbidity. Conversely, some cohort
one or two SDs below the age mean on the cognitive studies have found no evidence for an association.25–27
composite score (appendix pp 6–7). These studies have limited ability to rule out a
Details of adverse events during and immediately after link between anaesthesia and neurodevelopmental
anaesthesia have been reported previously.32 outcome because of a reliance on outcome measures,
such as school grade, which might not detect subtle
Discussion effects, or because their broad inclusion criteria
We noted strong evidence for equivalence between include children exposed to anaesthesia at an older age
awake-regional anaesthesia and general anaesthesia in when the risk might be less. The heterogeneity of the
infancy in terms of neurodevelopmental outcome at cohort studies also makes it difficult to analyse the
2 years of age. Equivalence was shown in many domains effects of duration of exposure, type of anaesthetic
of neurodevelopmental assessment and the 95% CIs fell drugs used, or doses or combination of drugs used.
within a third of an SD, well inside our predefined The above limitations inherently limit the capacity for
boundaries of clinical equivalence. cohort studies to establish the link between exposure
There are no previous randomised trials assessing to anaesthesia and neurodevelopmental outcome.
the effect of anaesthesia in infancy on long-term These limitations highlight the importance of
neurodevelopmental outcomes. Previous cohort methodologically robust and adequately powered trials
studies have found mixed results.19 Some studies have such as this one.31

246 www.thelancet.com Vol 387 January 16, 2016

 Articles

Difference Difference 95% CI for RA group as GA group as RA group GA group

in RA–GA* in SE difference in RA–GA per protocol per protocol intention to intention to
(N=287) (N=356) treat (N=361) treat (N=358)
Cognitive composite score
APP multiple imputation 0·169 1·26 –2·30 to 2·64 Child has a hearing defect

APP complete case 0·458 1·26 –2·02 to 2·94 Conductive 9 (3%) 6 (2%) 9 (2%) 6 (2%)

ITT multiple imputation 0·256 1·18 –2·06 to 2·57 Sensorineural 0 3 (1%) 1 (<1%) 3 (1%)

ITT complete case 0·430 1·19 –1·90 to 2·76 Hearing aid 1 (<1%) 3 (1%) 2 (1%) 3 (1%)

Language composite score Legally blind (<6/60 in both eyes) 1 (<1%) 0 1 (<1%) 0

APP multiple imputation 1·146 1·39 –1·59 to 3·88 Cerebral palsy 1 (<1%) 4 (1%) 1 (0%) 4 (1%)

APP complete case 0·628 1·37 –2·07 to 3·32 Autism spectrum disorder 2 (1%) 0 2 (1%) 0

ITT multiple imputation 1·454 1·32 –1·14 to 4·05 Data are n (% of non-missing data). RA=awake-regional anaesthesia. GA=general anaesthesia.
ITT complete case 0·942 1·30 –1·61 to 3·49
Table 5: 2-year non-psychometric outcome data
Motor composite score
APP multiple imputation 0·598 1·20 –1·77 to 2·97
APP complete case 0·410 1·19 –1·92 to 2·74 spectrum disorder, cerebral palsy, and visual or hearing
ITT multiple imputation 0·143 1·13 –1·08 to 3·37 defects. This trial was not powered to detect differences
ITT complete case 1·031 1·14 –1·20 to 3·26 in these diagnoses or events, and as expected we noted a
Social-emotional composite score low event rate in both groups. At 2 years of age it is
APP multiple imputation 1·005 2·09 –3·12 to 5·13 difficult to accurately diagnose the presence of disorders
APP complete case 2·012 1·70 –1·32 to 5·35 such as autism spectrum disorder, or to accurately assess
ITT multiple imputation 1·183 2·03 –2·82 to 5·19 vision and hearing, and some children could still have
ITT complete case 2·015 1·62 –1·17 to 5·20 undiagnosed neurological or neurobehavioural
Adaptive behaviour composite score disorders.
APP multiple imputation –0·893 1·34 –3·52 to 1·73 Data from most preclinical studies suggest that
APP complete case –1·223 1·33 –3·83 to 1·38 prolonged exposure to general anaesthesia is necessary
ITT multiple imputation –0·502 1·28 –3·03 to 2·02 before injury is seen (usually at 2 or 3 h).8 However,
ITT complete case –0·830 1·28 –3·34 to 1·68 changes have been noted with 1 h of exposure.38 In this
MacArthur-Bates percentile score trial, the median sevoflurane exposure was 54 min in the
APP multiple imputation –1·811 3·06 –7·85 to 4·23 general anaesthesia group and hence the results are
APP complete case –2·359 2·71 –7·69 to 2·98 consistent with most preclinical data. The trial is an
ITT multiple imputation –0·544 2·87 –6·20 to 5·11 important adjunct to these data because translating doses
ITT complete case –1·113 2·57 –6·17 to 3·94 and exposures from animals to human beings is uncertain,
and shorter duration of exposure could still have clinically
RA=awake-regional anaesthesia. GA=general anaesthesia. APP=as per protocol. relevant effects that cannot be detected in animal models.
ITT=intention to treat. *Adjusted for gestational age at birth.
In human cohorts, some researchers have found an
Table 4: Between-group comparisons in Bayley-III and MacArthur-Bates association with a single short exposure,17,24 whereas
scores others have only found an association after longer or
several exposures.22 There was no increase in learning
In this analysis we chose the cognitive scale of the disabilities in infants and toddlers exposed to 2 h or less
Bayley-III as the main outcome of interest. Changes of general anaesthesia in one study;22 anaesthetic
recorded in preclinical studies tend to be diffusely exposure was less than 90 min in 365 (61%) of
distributed over several brain regions. Such diffuse 593 exposed patients. This finding highlights that most
changes are most likely to have an effect on general anaesthetics in young children are of fairly brief duration.
cognition. An internal audit of anaesthetic duration in infants at
The results of two recent studies have shown that Boston Children’s Hospital showed that 53% of
whereas children exposed to anaesthesia had similar anaesthetics done in babies younger than 12 months of
school grades, those exposed had an increased risk of age were less than 2 h in duration. Thus, with regards to
not sitting the tests.26,28 This finding raises the possibility duration of exposure, our results are probably relevant to
that a subpopulation of exposed children might have roughly half the anaesthetics given to infants.
significant neurodevelopmental delay. To investigate this The finding of equivalence after short exposure does
possibility, we compared the proportion of children in not rule out the possibility that longer exposure to
each group that scored two SDs below the age mean on anaesthetics might have an effect on neurodevelopment.
the composite cognitive score. We noted no difference; Further trials are needed before any assumptions can be
however, in view of the limited power of this analysis, made about the effect of prolonged anaesthesia exposure
equivalence cannot be assumed. We have also reported in infancy. Results of some studies have also shown
the number of children with the diagnosis of autism a stronger association between several anaesthesia

www.thelancet.com Vol 387 January 16, 2016 247

 Articles

exposures and adverse outcome than with a single In this study, more than 80% of participants were male.
exposure.20,30 This situation might be the result of a It is well recognised that sex can have an effect on recovery
greater effect of confounding; inevitably, children who from brain injury. The effect is variable and depends on
undergo many procedures are more likely to have chronic the nature of the injury and outcome measured, although
disease. Our trial cannot address the possible increased generally greater effects are recorded in males and indeed
toxic effects with multiple exposures. the neurotoxic effect of anaesthesia on rodents has been
Our trial has several limitations. Awake-regional shown to be greater in males.41 Thus, the finding of
anaesthesia inevitably has a failure rate. As this was an equivalence in our trial with a preponderance of males
equivalence trial, we took the as-per-protocol analysis to makes it unlikely that equivalence would not also be
be the most conservative analysis, assuming that shown in females.
treatment failure would bias toward no difference. In In this trial, sevoflurane was used without other
view of the possibly contentious nature of this general anaesthetics. We chose a sevoflurane-only
assumption, we planned a priori to undertake a secondary anaesthetic because this reflects common practice for
intention-to-treat analysis. We noted no measureable anaesthesia for inguinal herniorrhaphy, and the
differences between the as-per-protocol and intention-to- preclinical effects of sevoflurane have been clearly
treat analyses, implying no bias was introduced by described. Some preclinical data have suggested that
treatment failure. In this study there was a loss to combinations of general anaesthetics might be more
follow-up of almost 14%. This, along with awake-regional injurious, and thus our trial cannot shed light on the
anaesthesia failure, led to an appreciable amount of possibility that an effect might be seen if other agents
missing data; however, both the multiple imputation are added.3 Finally, the MacArthur-Bates score is
analysis and the inverse probability weighting showed dependent on parental report and hence might be open
consistent robustness of the findings. to bias. Additionally, the standardisation data are of
Another limitation is that although the Bayley-III is varying degrees of validation across different languages.
a well validated assessment method of current In conclusion, this trial found strong evidence that
development, early neurobehavioural assessment of exposure of just less than 1 h to a sevoflurane general
children is not a perfect predictor of long-term outcome anaesthesia in infancy does not increase the risk of
because of the substantial variability in developmental adverse neurodevelopmental outcome at 2 years of age.
timing in young children. Although Bayley-III has a Although not definitive, this is the strongest clinical
stronger correlation with intelligence quotient at age evidence to date that sevoflurane general anaesthesia in
5 years than earlier versions of the test, it was not infancy does not result in substantial neurotoxicity.
designed to assess a broad range of cognitive functions. Contributors
Cognitive skills emerge and differentiate over childhood AJD was involved in study design and concept, conduct, data coordination,
and a more detailed neuropsychological assessment is contribution to the statistical analysis plan, data interpretation, writing and
coordinating drafts of the report and revising it critically, and approving
needed at a later date to identify mild or circumscribed the version to be published. ND was involved in study design and conduct,
deficits in cognitive functions such as executive skills data acquisition and coordination, data interpretation, and revising the
and memory.39,40 Therefore, it is important that the report critically. JCdG was involved in the coordination and supervision of
children be reassessed later in their development to data collection, data analysis and interpretation, contribution to the
statistical analysis plan, revised the report, and approved the final report as
confirm the results and to more thoroughly assess submitted. DEW was involved in study design and conduct, data
multiple domains of cognition. Children in this trial acquisition and coordination, data interpretation, and revising the report
are undergoing assessment at 5 years of age and the critically. LD contributed to protocol development, data collection,
results should be known after 2018. statistical plan, statistical analysis, data interpretation, and writing of the
report. GB was involved in study conduct, data coordination, and writing
It is important to note that this study reports the and reviewing the report. RS was the lead neuropsychologist and, along
results of a secondary outcome. This analysis of the with DCB and RWH, was involved in study design, concept, conduct, data
secondary outcome was prespecified in the study interpretation, and critically revising the report. TS and SJA were involved
protocol; however, the study was not specifically in interim analyses, contribution to the statistical analysis plan, data
interpretation, and revising the report critically. PH was involved in study
powered for the secondary outcome and thus it should design, study conduct, interim analyses, contribution to the statistical
be interpreted with caution and not regarded as analysis plan, data interpretation, and editing of the report. MJT
definitive. The analysis of the secondary outcome was contributed to the statistical analysis, data interpretation, and preparation
planned because of the recognition that there was of the report. GG and PLH were involved in study conduct, data
acquisition, data interpretation, and revising the report critically. IS,
growing concern over the issue of neurotoxicity and BSvUS, BGL, NW, AL, JJT, DP, OB, PS, ARA, and JM were involved in
existing evidence to guide practice was inherently study conduct, data acquisition, and coordination and revising the report
limited, and although the 2-year assessment was not critically. NSM and MEM were involved in study design, concept, and
conduct, data coordination, data interpretation, writing the report, and
definitive, it would still provide higher quality evidence
revising it critically. GF and CB were involved in study design and concept,
than that which existed up to now. The 2-year study conduct, data acquisition, contribution to data interpretation, and
assessment was also planned because of concerns over revising the report critically. GDO was involved in study conduct, data
the feasibility of maintaining the cohort for the longer- acquisition and coordination, contribution to the statistical analysis plan,
and revising the report.
term follow-up.

248 www.thelancet.com Vol 387 January 16, 2016

 Articles

Declaration of interests 6 Paule MG, Li M, Allen RR, et al. Ketamine anesthesia during the
The views expressed in this publication are those of the author(s) and first week of life can cause long-lasting cognitive deficits in rhesus
not necessarily those of the MRC, NHS, NIHR or the Department of monkeys. Neurotoxicol Teratol 2011; 33: 220–30.
Health UK. We declare no competing interests. The full list of members 7 Raper J, Alvarado MC, Murphy KL, Baxter MG. Multiple anesthetic
of the GAS consortium, the Trial Steering Committee and Data Safety exposure in infant monkeys alters emotional reactivity to an acute
Monitoring Committee are listed in the appendix (pp 1–3). stressor. Anesthesiology 2015; published online Sept 2. DOI:10.1097/
ALN.0000000000000851
Acknowledgments 8 Slikker W Jr, Zou X, Hotchkiss CE, et al. Ketamine-induced
All hospitals and centres were generously supported by departmental neuronal cell death in the perinatal rhesus monkey. Toxicol Sci
funding. In addition to this funding, specific grants received for this study 2007; 98: 145–58.
are as follows: Australia: The Australian National Health & Medical 9 Brambrink AM, Back SA, Riddle A, et al. Isoflurane-induced
Research Council (Project Grants #1002906 and 491226); Australian and apoptosis of oligodendrocytes in the neonatal primate brain.
New Zealand College of Anaesthetists (Project Grant #11/021 and 07/012); Ann Neurol 2012; 72: 525–35.
Murdoch Childrens Research Institute; Victorian Government’s 10 Brambrink AM, Evers AS, Avidan MS, et al. Isoflurane-induced
Operational Infrastructure Support Program. US: National Institute of neuroapoptosis in the neonatal rhesus macaque brain.
Health (Project Grant # NIH 1-R01 hD06 1136-01A1); Food and Drug Anesthesiology 2010; 112: 834–41.
Administration (Project Grant #FDA-SOL-08-SAFEKIDS-Clin 002-Project 11 Hansen TG. Anesthesia-related neurotoxicity and the developing
2). Italy: Italian Ministry of Heath, Young Researchers Grant (Project Grant animal brain is not a significant problem in children.
#167/GR-2009-1476067). Netherlands: Fonds NutsOhra grant #1305-144 Paediatr Anaesth 2015; 25: 65–72.
and Vrienden WKZ 16.13.038. Canada: Canadian Institute of Health 12 Psaty BM, Platt R, Altman RB. Neurotoxicity of generic anesthesia
Research (Project Grant #MCT-98031), Canadian Anesthesiologists’ agents in infants and children: an orphan research question in
Society, Pfizer Canada Inc. UK: Health Technologies Assessment–National search of a sponsor. JAMA 2015; 313: 1515–16.
Institute for Health Research UK (Project Grant # 07/01/05). We are also 13 Rappaport BA, Suresh S, Hertz S, Evers AS, Orser BA. Anesthetic
grateful to the following individuals who contributed to the running of the neurotoxicity—clinical implications of animal models. N Engl J Med
2015; 372: 796–97.
study: Mark Fagjman (Department of Anaesthesia, Monash Medical
Centre, Melbourne, VIC, Australia); Daniela Tronconi (Department of 14 Crosby G, Davis PJ. General anesthesia in infancy is associated with
learning disabilities-or not. Anesth Analg 2013; 117: 1270–72.
Anesthesia, Istituto Giannina Gaslini, Genoa, Italy); David C van der Zee
(Department of Pediatric Surgery, Wilhelmina Children’s Hospital, 15 Davidson AJ. Anesthesia and neurotoxicity to the developing brain:
the clinical relevance. Paediatr Anaesth 2011; 21: 716–21.
University Medical Centre Utrecht, Netherlands); Jan Hulscher
(Department of Surgery, University Medical Center Groningen, Groningen 16 DiMaggio C, Sun LS, Ing C, Li G. Pediatric anesthesia and
neurodevelopmental impairments: a Bayesian meta-analysis.
University, Groningen, Netherlands); Michael Rivkin (Department of
J Neurosurg Anesthesiol 2012; 24: 376–81.
Neurology, Children’s Hospital Boston, Boston, MA, USA);
17 Ing C, DiMaggio C, Whitehouse A, et al. Long-term differences in
Michelle Sadler-Greever (University of Washington, Seattle Children’s
language and cognitive function after childhood exposure to
Hospital, Department of Anesthesia and Pain Medicine, Seattle, WA, anesthesia. Pediatrics 2012; 130: e476–85.
USA); Debra Faulk (Children’s Hospital Colorado and University of
18 Ing CH, DiMaggio CJ, Malacova E, et al. Comparative analysis of
Colorado School of Medicine, Department of Anesthesiology, Colorado, outcome measures used in examining neurodevelopmental effects of
CO, USA); Greta Wilkening (Neurosciences Institute, Children’s Hospital early childhood anesthesia exposure. Anesthesiology 2014; 120: 1319–32.
Colorado, CO, USA); Edward Goldson (Department of Pediatrics, Division 19 Lei SY, Hache M, Loepke AW. Clinical research into anesthetic
of Developmental-Behavioral Pediatrics, Children’s Hospital Colorado, neurotoxicity: does anesthesia cause neurological abnormalities in
Colorado, USA). Danai Udomtecha and Sarah Titler (The University of humans? J Neurosurg Anesthesiol 2014; 26: 349–57.
Iowa Hospital, Department of Anesthesia, Iowa, IA, USA); Roxana Ploski 20 Flick RP, Katusic SK, Colligan RC, et al. Cognitive and behavioral
and Alan Farrow-Gilespie (Children’s Medical Centre Dallas, Department outcomes after early exposure to anesthesia and surgery. Pediatrics
of Anesthesiology, University of Texas Southwestern Medical Center, 2011; 128: e1053–61.
Dallas and Children’s Medical Center at Dallas and Outcome Research 21 Sprung J, Flick RP, Katusic SK, et al. Attention-deficit/hyperactivity
Consortium, Dallas, TX, USA); Timothy Cooper (Division of disorder after early exposure to procedures requiring general
Developmental Medicine and the Center for Child Development, Monroe anesthesia. Mayo Clin Proc 2012; 87: 120–29.
Carell Jr Children’s Hospital at Vanderbilt, Nashville, TN, USA); 22 Wilder RT, Flick RP, Sprung J, et al. Early exposure to anesthesia
Elizabeth Card (Perioperative Clinical Research Institute, Vanderbilt and learning disabilities in a population-based birth cohort.
University Medical Center, Nashville, TN, USA); Wendy Boardman Anesthesiology 2009; 110: 796–804.
(Dartmouth-Hitchcock Medical Center, Department of Anesthesiology, 23 DiMaggio C, Sun LS, Li G. Early childhood exposure to anesthesia
NH, USA); Theodora K Goebel (Department of Anesthesiology and and risk of developmental and behavioral disorders in a sibling
Critical Care Children’s Hospital of Philadelphia, Philadelphia, PA, USA). birth cohort. Anesth Analg 2011; 113: 1143–51.
24 DiMaggio C, Sun LS, Kakavouli A, Byrne MW, Li G. A retrospective
References cohort study of the association of anesthesia and hernia repair
1 Jevtovic-Todorovic V, Absalom AR, Blomgren K, et al. Anaesthetic surgery with behavioral and developmental disorders in young
neurotoxicity and neuroplasticity: an expert group report and children. J Neurosurg Anesthesiol 2009; 21: 286–91.
statement based on the BJA Salzburg Seminar. Br J Anaesth 2013;
25 Bartels M, Althoff RR, Boomsma DI. Anesthesia and cognitive
111: 143–51.
performance in children: no evidence for a causal relationship.
2 Briner A, Nikonenko I, De Roo M, Dayer A, Muller D, Vutskits L. Twin Res Hum Genet 2009; 12: 246–53.
Developmental stage-dependent persistent impact of propofol
26 Hansen TG, Pedersen JK, Henneberg SW, et al. Academic
anesthesia on dendritic spines in the rat medial prefrontal cortex.
performance in adolescence after inguinal hernia repair in infancy:
Anesthesiology 2011; 115: 282–93.
a nationwide cohort study. Anesthesiology 2011; 114: 1076–85.
3 Jevtovic-Todorovic V, Hartman RE, Izumi Y, et al. Early exposure to
27 Hansen TG, Pedersen JK, Henneberg SW, Morton NS,
common anesthetic agents causes widespread neurodegeneration
Christensen K. Educational outcome in adolescence following
in the developing rat brain and persistent learning deficits.
pyloric stenosis repair before 3 months of age: a nationwide cohort
J Neurosci 2003; 23: 876–82.
study. Paediatr Anaesth 2013; 23: 883–90.
4 Sanchez V, Feinstein SD, Lunardi N, et al. General anesthesia
28 Block RI, Thomas JJ, Bayman EO, Choi JY, Kimble KK, Todd MM.
causes long-term impairment of mitochondrial morphogenesis and
Are anesthesia and surgery during infancy associated with altered
synaptic transmission in developing rat brain. Anesthesiology 2011;
academic performance during childhood? Anesthesiology 2012;
115: 992–1002.
117: 494–503.
5 Istaphanous GK, Ward CG, Nan X, et al. Characterization and
29 Bong CL, Allen JC, Kim JT. The effects of exposure to general
quantification of isoflurane-induced developmental apoptotic cell
anesthesia in infancy on academic performance at age 12.
death in mouse cerebral cortex. Anesth Analg 2013; 116: 845–54.
Anesth Analg 2013; 117: 1419–28.

www.thelancet.com Vol 387 January 16, 2016 249

 Articles

30 Wang X, Xu Z, Miao CH. Current clinical evidence on the effect of 36 van Buuren S. Multiple imputation of discrete and continuous data
general anesthesia on neurodevelopment in children: an updated by fully conditional specification. Stat Methods Med Res 2007;
systematic review with meta-regression. PLoS One 2014; 9: e85760. 16: 219–42.
31 Davidson AJ, Becke K, de Graaff J, et al. Anesthesia and the 37 Seaman SR, White IR. Review of inverse probability weighting for
developing brain: a way forward for clinical research. dealing with missing data. Stat Methods Med Res 2013; 22: 278–95.
Paediatr Anaesth 2015; 25: 447–52. 38 Johnson SA, Young C, Olney JW. Isoflurane-induced
32 Davidson AJ, Morton NS, Arnup SJ, et al, and the General neuroapoptosis in the developing brain of nonhypoglycemic mice.
Anesthesia compared to Spinal anesthesia (GAS) Consortium. J Neurosurg Anesthesiol 2008; 20: 21–28.
Apnea after Awake Regional and General Anesthesia in Infants: 39 Harris SR, Langkamp DL. Predictive value of the Bayley mental
The General Anesthesia Compared to Spinal Anesthesia scale in the early detection of cognitive delays in high-risk infants.
Study-Comparing Apnea and Neurodevelopmental Outcomes, J Perinatol 1994; 14: 275–79.
A Randomized Controlled Trial. Anesthesiology 2015; 123: 38–54. 40 Bode MM, D’Eugenio DB, Mettelman BB, Gross SJ. Predictive
33 Frawley G, Bell G, Disma N, et al, and the General Anesthesia validity of the Bayley, Third Edition at 2 years for intelligence
compared to Spinal anesthesia (GAS) Consortium. Predictors of quotient at 4 years in preterm infants. J Dev Behav Pediatr 2014;
Failure of Awake Regional Anesthesia for Neonatal Hernia Repair: 35: 570–75.
Data from the General Anesthesia Compared to Spinal Anesthesia 41 Lee BH, Chan JT, Kraeva E, Peterson K, Sall JW. Isoflurane
Study-Comparing Apnea and Neurodevelopmental Outcomes. exposure in newborn rats induces long-term cognitive dysfunction
Anesthesiology 2015; 123: 55–65. in males but not females. Neuropharmacology 2014; 83: 9–17.
34 Bayley N. Bayley scales of infant and toddler development. 3rd ed.
San Antonio: Harcourt Assessment Inc, 2006.
35 Fenson L, Marchman VA, Thal DJ, Dale PS, Reznic JS, Bates E.
MacArthur-Bates Communicative Development Inventories: User’s
Guide and Technical Manual 2nd edn. Baltimore Brookes
Publishing, 2007.

250 www.thelancet.com Vol 387 January 16, 2016