- Email: [email protected]
Systematic review of the neurocognitive outcomes used in studies of paediatric anaesthesia neurotoxicity
British Journal of Anaesthesia, ▪ (▪): 1e19 (2018) doi: 10.1016/j.bja.2017.11.107 Advance Access Publication Date: xxx Review Article
REVIEW ARTICLE
Systematic review of the neurocognitive outcomes used in studies of paediatric anaesthesia neurotoxicity € hler1 and T. G. Hansen1,2 N. G. Clausen1,2,3,*, S. Ka 1
Department of Anaesthesia and Intensive Care, University Hospital Odense, J.B.Winsløwsvej 4, DK-5000 Odense, Denmark, 2University of Southern Denmark, Department of Clinical Research, Winsløwparken 19, DK-5000 Odense, Denmark and 3Children’s Hospital University of Zu ¨ rich, Department of Anaesthesia, Steinwiesstrasse 75, CH-8032 Zu¨rich, Switzerland *Corresponding author. Department of Anaesthesia and Intensive Care, University Hospital Odense, J.B.Winsløwsvej 4, DK-5000 Odense, Denmark. E-mail: [email protected].
Abstract Neurotoxicity of anaesthetics in developing brain cells is well documented in preclinical studies, yet results are conflicting in humans. The use of many and different outcome measures in human studies may contribute to this disagreement. We conducted a systematic review to identify all measures used to assess long-term neurocognitive outcomes following general anaesthesia (GA) and surgery in children. The quality of studies was assessed according to the Newcastle-Ottawa Scale (NOS) for observational studies. PubMed/MEDLINE, EMBASE, Cinahl, Web of Science, and the Cochrane Library were searched for studies investigating neurocognitive outcome after GA in children <18 yr. Sixtyseven studies were identified from 19 countries during 1990e2017. Most assessments were performed within cognition, sensory-motor development, academic achievement or neuropsychological diagnosis. Few studies assessed other outcomes (magnetic resonance imaging, serum-biomarkers, mortality, neurological examination, measurement of head circumference, impairment of vision). Rating according to the NOS rewarded a mean of six stars out of nine. Some concerns prevail regarding potential inter-rater variability because of equivocal description of rating criteria. Specific features such as stability over lifetime and inter-relations of outcomes (e.g. prediction of subsequent development or diagnosis of neuropsychological conditions) are discussed. The importance of validity and reliability of the various test instruments are described. The studies vary immensely in important characteristics. Future observational studies should be more consistent in the choice of study population, age at exposure, follow-up, indication for and type of surgery, and outcomes. Assessment of sensory-motor development seems feasible in young children (age <4 yr), intelligence/cognition in older children. Keywords: anaesthesia; general; child development; infant; review
Editorial decision: November 21, 2017; Accepted: November 21, 2017 © 2017 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved. For Permissions, please email: [email protected]
1
2
-
Clausen et al.
Confronted with the lack of definite answers, researchers continue to investigate whether anaesthetics cause harm to the developing brain in children. Evidence from preclinical studies on various species including non-human primates, a variety of anaesthetic drugs, modes of administration, and durations, substantiate this concern. In humans, an association between general anaesthesia (GA) in young children and subsequent impaired neurodevelopment remains to be properly established or rejected. Single and short exposures lasting <1 h are considered ‘safe’ regarding neurodevelopmental outcome.1 In a recent sibling-matched cohort study it was reported that children exposed once for inguinal hernia repair before age 3 yr showed no difference in intelligence quotient testing compared with their unexposed siblings.2 A Swedish population-based cohort study detected a low overall difference in academic achievement and intelligence after single and multiple exposures before age 4 yr for miscellaneous types of surgeries.3 These findings concur with the results of an international randomised trial comparing neurodevelopmental outcome at age 2 yr between infants exposed to a brief sevoflurane anaesthesia or an awake spinal-/caudal anaesthetic for inguinal herniorrhaphy; no difference was found between groups.4 Quite the opposite, some studies focusing on exposure in 3- or 4-yr-old children to anaesthesia for miscellaneous types of surgery indicate long-term effects on cognition, learning, and behavioural disorders.5,6 As an example, exposure to GA and surgery before age 3 yr was found to associate with a remarkably increased risk of being diagnosed with a developmental/behavioural disorder.7 Interestingly, various authorities interpret the existing results differently. While European scientific societies agree, that at present, there is no human evidence to support any change in anaesthetic practice in children and pregnant women,1 the Food and Drug Administration recently warned against anaesthesia in children <3 yr of age prolonged anaesthesia.8 Such different interpretations might have serious clinical implications for patients (e.g. postponement of surgery). The majority of studies within the field of anaesthesiarelated neurotoxicity are cohort studies, which knowingly are vulnerable to confounding. As a consequence, recent literature calls for the use of other modalities (e.g. neuroimaging and biomarkers), to evaluate neuro-apoptosis and neuro-inflammation.9 Furthermore, other factors that might mitigate morbidity after GA and surgery in children are increasingly acknowledged and subjected to intensive research.10e13 Despite this broadened search for answers, anaesthesia-related neurotoxicity has not yet been properly clinically defined. Consequently, it is unknown where clinical research should be focused. Multiple factors, for example, sample size; selection of exposed cohort; age at exposure; type of, and indication for, surgery; follow-up as well as type of outcome measures could likely influence results. We hypothesised that inconsistency in the choice of outcome measures, age at exposure, and differences in study quality are major confounding factors for divergent results obtained in human studies. The aims of this systematic review were to identify the type of outcomes used in studies investigating neurodevelopmental consequences of anaesthesia/surgery exposure in infancy and early childhood. Further, we wished to compare the studies with regard to other crucial factors and assess the methodological quality of these studies.
Methods The study protocol was published at PROSPERO (registration ID: CRD42016042450) and reported according to the PRISMA guidelines.14
Search strategy We searched PubMed/MEDLINE, EMBASE, Cinahl, Web of Science and the Cochrane Library (last search on June 16, 2017) using relevant terms concerning ‘general anaesthesia’ and ‘neurocognitive outcome’. Via the MESH database and the EMTREE thesaurus we identified relevant search terms. Additionally, the reference list of included studies was hand-checked for other potentially relevant publications. ‘Neurobehavioral outcome’ was part of the search. However, most articles on postoperative behaviour assessed temporary changes (e.g. emergency delirium and anxiety). These were not considered to be long-term results of anaesthesia-related neurotoxicity and articles using only behaviour as outcomes were not included in the content analyses. Librarians at the Medical Research Library, part of The University Library of Southern Denmark, Odense University Hospital, Denmark approved the construction of the final search. The exact search for the respective databases is available as Supplementary Material (Supplementary 1).
Study selection To be eligible for this review studies must report on: Age at exposure <18 yr Exposure to single/multiple GA or surgery (it is assumed, that surgery is only delivered with a concomitant GA) Evaluation of cognitive function after the exposure(s) Studies must be published in peer reviewed journals, identified by experts within the field, or have been cited by peer reviewed papers. We excluded studies that were not reported in English, evaluated neurotoxic effects of topically/locally administered anaesthetics, or studies conducted in animals. Two authors (N.G.C. and S.K.) independently assessed title, abstract and full text for eligibility using the Covidence software (Covidence systematic review software, Veritas Health Innovation, Melbourne Australia. Available at https://www. covidence.org).
Outcomes We identified all outcome measures used in studies investigating neurocognitive consequences of surgery and GA exposure in children (primary outcome). Further, the quality of studies regarding selection and comparability of study participants, as well as ascertainment of outcomes were evaluated based on the Newcastle-Ottawa Scale (NOS).15
Data extraction We recorded study ID, design, reason for anaesthesia (type of surgery or diagnostics), and specifics on anaesthetic procedure (type, drugs, doses, means of administration) if provided. We also recorded information on the population studied (study base, number of exposed/non-exposed individuals), age at exposure, male/female distribution, type of outcome, and age at assessment.
Neurocognitive outcomes in children after general anaesthesia
-
3
Fig 1. Illustration of selection process.
Quality assessment
Results
Study quality was evaluated using the NOS for quality assessment of non-randomised studies. The evaluation comprised three broad perspectives: selection of study groups, comparability of groups, and ascertainment of either exposure or outcome of interest. The scale uses a ‘star system’; each study can be accredited a maximum of nine stars if all requirements are met. Risk of bias evaluations of individual studies can be retrieved by contacting the corresponding author.
Study selection Our systematic literature review identified 18 011 titles after removal of duplicates. N.G.C. and S.K. independently screened titles and abstracts, eliminating irrelevant and double items. A total of 397 titles qualified for full text screening and finally, 67 studies were included in this review.2e7,16e75 The main reasons for excluding studies were: ‘Wrong intervention’: comparison of interventions other than or not related to anaesthesia
4
-
Clausen et al.
Table 1 List of different neurocognitive and sensorimotor function outcomes used to assess outcome in children exposed to general anaesthesia (alphabetic order) and the domain to which the test was assigned. *Individualised Educational Program (IEP) is not a psychometric test but the intervention applied for a child with special needs. The IEP is meant to address each child’s unique learning issues and include specific educational goals. It is a legally binding document, sanctioned by the United States federal law called the Individuals with Disabilities (IDEA). The IEP describes the goals the team (e.g. teachers, parents, school psychologists) sets for a child during the school year, as well as any special support needed to help achieve them. yFunctional Independence Measure is used in follow-up of rehabilitation Psychometric test ABAS-II AIMS e AVLT ASQ BDS BRIEF BSID-II BSID-III CAT CBCL CDI CELF CELF-E
Domain of testing Adaptive Behavior Assessment Systemd2nd Edition Alberta Infant Motor Scale Albert Einstein College of Medicine Neonatal Neurobehavioral Assessment Scale Rey Auditory Verbal Learning Test Ages & Stages Questionnaire Backward digit span test Behavior Rating Inventory of Executive Functions Bayley Scales of Infant and Toddler Developmentd2nd Edition Bayley Scales of Infant and Toddler Developmentd3rd Edition California Achievement Test Child Behavior Checklist Child Depression Inventory
DSM-ADH
Clinical Evaluation of Language Fundamentals Clinical Evaluation of Language Fundamentalsdexpressive language score Clinical Evaluation of Language Fundamentalsdreceptive language score Clinical Evaluation of Language Fundamentalsdtotal language ability The Child Health Questionnaire 50 Raven’s Colored Progressive Matrices Continuous Performance Test II Conners’ teacher Rating ScaledRevised California Verbal Learning TestdChildren Delis-Kaplan Executive Function Systems/Trail Making Subtests Diagnostic and Statistical Manual of Mental Disordersd4th Edition DSMdattention deficit hyperactivity scores
EDI FDS FSIQ G-TVPS GDQ GDS GMDS GMFCS HAWIVA-III GPT ICD-9
Early Development Instrument Forward digit span test Full scale intelligence-quotient Gardner Test of Visual-Perceptual Skills Revised General Developmental Quotient Gesell Developmental Schedule Griffiths Mental Development Scale General Motor Function Classification Score HannovereWechsler Intelligence Scale, 3rd Edition Grooved Pegboard Test International classification of Diseased9th Edition
ICD-9-CM 299.00
International Classification of Diseased9th Autistic Disorder Diagnoses International Classification of Diseases 9thdAttention Deficit Hyperactivity Disorder Individualised Educational ProgramdDisorders of Emotion and Behaviour Individualised Educational ProgramdSpeech and Language Kaufmann Assessment Battery for Children Kognitiver Enwicklungstest fu¨r das Kindergartenalter Learning Disability McCarron Assessment of Neuromuscular Development
CELF-R CELF-T CHQ50 CPM CPT-II CTRS-R CVLT-C DKEFS DSM-IV
ICD-9-CM 314.01 IEP-EBD IEP-SL K-ABC KET-KID LD MAND
Development Development Development Intelligence/cognition Development Intelligence/cognition Development Development intelligence Development intelligence Academic achievement Development Screening/diagnosis (psychiatric disorder) Development Development Development Development Development Intelligence/cognition Development Development Academic achievement Academic achievement Screening/diagnosis (psychiatric disorder) Screening/diagnosis (psychiatric disorder) Development Intelligence/cognition Intelligence/cognition Development Development Development Development Development Intelligence/cognition Intelligence/cognition Screening/diagnosis (psychiatric disorder) Screening/diagnosis (psychiatric disorder) Screening/diagnosis (psychiatric disorder) Academic achievement* Academic achievement* Intelligence/cognition Intelligence/cognition Screening/diagnosis Development
Continued
Neurocognitive outcomes in children after general anaesthesia
-
5
Table 1 Continued Psychometric test MacArthur-Bates CDI MRI NEPSY NEPSY-2-NL OLSAT OWLS PDMS e PIQ PPVT RDLS PSLE SDMT SON-R e SB-5 TCS TIQ TEA-Ch NL TMTdA VABS VIQ VMI Wallin BP WAMSE WASI WJ III WJ IIIdVisual Matching WeeFIM WPPSI-R WISC-III WISC-III-NL WISC-IV
Domain of testing McArthur Bates Communicative Development Inventory Magnetic resonance imaging Developmental NEuroPSYchological Assessment NDevelopmental Neuropsychological Assessment Batteryd2nd Edition, Dutch version OtiseLennon School Ability test Oral and Written Language Scales Peabody Developmental Motor Scales Phonemic verbal fluency test
Development Screening/diagnosis (somatic disorder) Academic achievement Development Cognition Academic achievement Development Intelligence/cognition
Performance Intelligence-Quotient Peabody Picture Vocabulary Test Reynell Developmental Language Scales Primary School Leaving Examination Symbol Digit Modality Test Semantic verbal fluency test Hogrefe/Snijders-Oomen Non-Verbal Intelligence TestdRevised Stanford
Intelligence/cognition Academic achievement Development Academic achievement Development Intelligence/cognition Intelligence/cognition
StanfordeBinet Intelligence Scalesd5th Edition Stroop Color and Word Test Total Cognitive Skills Total Intelligence Quotient Test of Everyday Attention for Children, Dutch version Trail Making Testdpart A The Vineland Adaptive Behavior Scale, 2nd Edition Verbal Intelligence-Quotient BeeryeBuktenica Developmental Test of Visual Motor Integration, 5th Edition Wallin B pegboard Western Australian Monitoring Standards in Education Wechsler Abbreviated Scale of Intelligence WoodcockeJohnson III WoodcockeJohnson testdVisual Matching Functional Independence Measure Wechsler Preschool and Primary Scales for Intelligence: Revised Wechsler Intelligence Scale for Childrend3rd Version Wechsler Intelligence Scale for Childrend3rd Version, Dutch version Wechsler Intelligence Scale for Childrend4th Edition
Intelligence/cognition Intelligence/cognition Intelligence/cognition Intelligence/cognition Intelligence/cognition
Academic achievement
Intelligence/cognition Development Intelligence/cognition Intelligence/cognition Development Academic achievement Intelligence/cognition Academic achievement Academic achievement Developmenty Intelligence/cognition Intelligence/cognition Intelligence/cognition Intelligence/cognition
Non-psychometric tests/means to assess cognitive, developmental outcome, or both, of children exposed to general anaesthesia Serum biomarkers of Brain derived neurotrophic factor (BDNF), S100B Neurological examination Surrogate measures of visual function (visual acuity, refractive error, thickness of retinal nerve fibre layer)
‘Wrong outcomes’: for example, emergency delirium shortly after procedure (i.e. within 30 days following exposure or temporary behavioural outcomes). One article was excluded, because these results were based on re-analysis of a previously described dataset.76 The process of selection is illustrated in Figure 1.
Neuron damage Neurologic deficit Sensorineural deficit
Study outcomes A variety of psychometric tests were used (Table 1). For simplicity, tests were allocated to four core domains: intelligence/cognition, development, achievement, and screening/ diagnosis. In the majority of studies, more than one test was used, and often more than one domain was tested.
6
Table 2 Summarised characteristics of studies included in the current review.*MRI scan at 7 days, BSID-III at 12 months of age;**MRI scan at 39e42 weeks (corrected age), BSID-II at 2 yr
-
Type of outcome
<4
16 yr
School grades Intelligence scores
Yes
<1
?
Developmental delay Psychiatric diagnoses
Yes
<3 yr
14e18 yr
Diagnoses Academic performance (single/group)
No
<7 yr
16 yr (mean)
Cognition (VIQ and WAIS) MRI scan Height þ head circumference GDMS Sensorineural deficits Others (weight, length, head circumference) School performance Psychiatric diagnoses
Study design
Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at exposure controls (month, yr) Yes/No
Glatz and colleagues3
Sweden 2017
Nationwide cohort study
No
No
Yes
Nestor and colleagues16
USA 2017
Retrospective single Yes centre cohort study
No
Hu and colleagues17 USA 2017
Retrospective birth- Yes cohort study
No
n¼33 512 (single exp.) n¼37 231 (multiple exp.) n¼63 (no surgery) n¼171 (1 surgery) n¼64 (>1 surgery) n¼465 (no surgery) n¼466 (1 surgery) n¼126 (>1 surgery) n¼48 (CLP) n¼19 (CP) n¼20 (CL)
Conrad and colleagues18
USA 2017
Retrospective birth cohort study
No
No
Birajdar and colleagues19
Australia 2017
Retrospective cohort study
Yes
Yes
n¼33
No
<5 days (mean)
12 months
Terushkin and colleagues20
USA 2017
Yes
Yes
n¼33
No
<4 yr
Variable (3.4e11.4 yr)
Lap and colleagues21
Netherlands 2017
Retrospective institutional cohort study Retrospective cohort study
No
Yes
n¼19
Yes
Newborn
Retrospective cohort study Multicentre, sibling matched cohort study
No
No
n¼127
No
<3 yr
Median exp/controls: 8.5/6.0yr 20 yr
Yes
Yes
n¼116
Yes
<3 yr
8e15 yr
Observational cohort study Randomised controlled trial
Yes
No
n¼21
No
<1 month
<7 yr
Yes
Yes
n¼358 (GA group) n¼361 (RA group)
No
<60 weeks (corrected age)
2 yr
Yazar and Australia 2016 colleagues22 2 Sun and colleagues USA 2016
Seltzer and colleagues23 Davidson and colleagues4
USA 2016 Australia 2016
Cognition (IQ, memory) Development (motor, vision, behaviour) Three surrogate measures of visual function IQ Tests of domain specific neurocognitive functions and behaviour Vineland-II BSID-III McArthureBates
Continued
Clausen et al.
Age at outcome assessment
Country of origin and publication yr
Study ID
Table 2 Continued Study ID
Study design
Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at controls exposure Yes/No (month, yr)
Age at outcome assessment
Type of outcome
O’Leary and colleagues24
Canada 2016
No
No
n¼28 366
Yes
Before age 5.7 yr
Denmark 2016
No
Yes
n¼549
Yes
Mean age 9.1 yr
5.7 yr (mean age at EDI) 15e16 yr
EDI
Djurhuus and colleagues25
Population-based, retrospective cohort study Retrospective nationwide cohort study
Clausen and colleagues75
Denmark 2016
Retrospective nationwide cohort study
No
No
n¼558
Yes
2 months to 2 yr
Poor Zamany Nejat Kermany and colleagues26
Iran 2016
Retrospective cohort study
Yes
Yes
n¼68
No
<3 yr
Hoffman and colleagues27 Hansen and colleagues28 Graham and colleagues29
USA 2016
No
No
n¼178
No
Variable
School grades non-attainment rates at national school exam 15e16 yr School grades non-attainment rates at national school exam 6 months after PVF test final exposure SVF test FDS test BDS test Variable BSID-III
No
Yes
n¼71
No
<1 month
<6 months
No
No
n¼4479
Yes
<4 yr
<6 yr
HAWIVA-III KETKID EDI scores
No
No
n¼40
Yes
<7days
2 months
BSID-II
Yes
Yes
n¼96
No
<12 months
4e5 yr
WPPSI-III
No
No
n¼415
Yes
<5 yr
6 yr
IQetest
Yes
No
n¼118
Yes
5e12 yr
Retrospective Yes cohort study Bidirectional cohort Yes study Prospective follow- No up study (single centre)
No
n¼35
Yes
1.5e5 yr
No
n¼15
Yes
0e2 yr
1 day þ 6 weeks Tests of domain specific postoperative neurocognitive functions (perception, memory, comprehension) Not stated CBCL DSM-IV 10e17 yr MRI-imaging
No
n¼21
Yes
<1 yr
34e39 months
Doberschuetz and colleagues30 Diaz and colleagues31 de Heer and colleagues32 Aun and colleagues33
Bakri and colleagues34 Taghon and colleagues35 Naguib and colleagues36
Observational cohort study Germany 2016 Observational cohort study Canada 2016 Retrospective matched cohort study Germany 2016 Ambi-directional matched cohort study USA 2016 Subgroup analysis as part of prospective study Netherlands Retrospective 2016 cohort study Hong Kong 2016 Prospective cohort study
Saudi Arabia 2015 USA 2015 USA 2015
IQ test Psychiatric diagnosis
-
Continued
Neurocognitive outcomes in children after general anaesthesia
Country of origin and publication yr
7
Table 2 Continued
8
Study ID
Hansen and colleagues38
Denmark 2015
Gano and colleagues39
USA 2015
Backeljauw and colleagues40
USA 2015
ckova and Petra colleagues41 Yan and colleagues42 Yin and colleagues43
Czech Republic 2015 China 2014
Morriss and colleagues44
USA 2014
China 2014
Ing and colleagues45 Australia 2014
Gaynor and colleagues46
USA 2014
Garcia Guerra and colleagues47 Cheng and colleagues48 Andropoulos and colleagues49 Ko and colleagues50
Canada 2014
Minutillo and colleagues51 Hansen and colleagues52
Australia 2013
USA & Canada 2014 USA 2014 Taiwan 2014
Denmark 2013
Type of outcome
<2 yr
>2 yr
Psychiatric diagnosis
Yes
Not specified
15e16 yr
Mortality School grades
n¼43
No
3e6 yr
WPPSI-III
No
n¼53
Yes
<42 weeks gestational age 1 day to 3.8 yr
Yes
Yes
n¼21
No*
8 days
Variable (128 þ / 44 months) 3e7 yr
Yes
Yes
n¼49
No
3e22 months
WPPSI-III OWLS MRI CHQ 50 test SON-R 2.5e7 yr test BSDI-II
Yes
Yes
n¼60
No*
7e13 yr
No
No
n¼2970
Yes
Variable
Retrospective cohort study
No
No
n¼321
Yes
<3 yr
Secondary analysis of data from prospective study Prospective followup study Prospective cohort study Retrospective cohort study Nationwide retrospective matched cohort study Retrospective cohort study Nationwide register-based follow-up study
No
No
n¼365
No
<6 months
Yes
Yes
n¼135
No
<6 weeks
54 months
No
No
n¼124
No
<3 yr
1 yr
Yes
No
n¼93
No
<30 days
No
No
n¼3293
Yes
<3 yr
7 days and 12 months* Variable
WPPSI-III FSIQ PIQ VIQ VMI BSID-II MRI5-imaging BSID-III MRI-imaging Psychiatric diagnosis
No
Yes
n¼82
Yes
<1 months
12 months
GMDS
No
Yes
n¼779
Yes
<3 months
15e16 yr
School grades
Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at exposure controls (month, yr) Yes/No
Nationwide retrospective matched cohort study Nationwide register-based follow-up study Cohort study
No
No
n¼5197
Yes
No
No
n¼228
Yes
No
Retrospective matched cohort study Prospective cohort study Prospective cohort study Observational prospective study (single centre) Retrospective cohort study
Yes
3 days after final operation 7þ90 days after Wechsler Memory scale exposure egraphic, verbal, visual, acoustic components 18e22 months BSID-III BSID-II GMFCS 10 yr SDMT, CPM MAND, CELF CELF-R, CELF-E CELF-T, PPVT 4e5 yr WPPSI-III School grades
Continued
Clausen et al.
Ko and colleagues37 Taiwan 2015
Age at outcome assessment
Study design
-
Country of origin and publication yr
Table 2 Continued Age at outcome assessment
Type of outcome
1 week and 1, 3, 6 months after surgery 12 yr
WPPSI-III
Country of origin and publication yr
Study design
Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at exposure controls (month, yr) Yes/No
Fan and colleagues53
China 2013
Yes
No
n¼100
No
<4 yr
Bong and colleagues54
Singapore 2013
Yes
No
n¼100
Yes
<1 yr
Yang and colleagues55 Andropoulos and colleagues56
South Korea 2012 USA 2012
Yes
Yes
n¼21
No
5e10 yr
Yes
Yes
n¼30
No
<1 month
Walker and colleagues57
Australia 2012
No
No
n¼300
Yes
<3 months
12 months
Sprung and colleagues58 Sananes and colleagues59
USA 2012
Single centre observational study Retrospective observational cohort study Prospective cohort study Prospective, observational cohort study Prospective population-based cohort study Retrospective cohort study Prospective followup study
Yes
No
n¼350
Yes
<2 yr
No
No
n¼131
No
<3 months
Rocha and colleagues60
Portugal 2012
No
Yes
n¼39
No
<1 month
Long and colleagues61
Australia 2012
No
No
n¼195
No
<8 weeks
24 months
BSID-III
Long and colleagues62
Australia 2012
No
No
n¼195
No
<8 weeks
4, 8, 12, 16 months
Alberta Infant Motor Scale
Ing and colleagues5
Australia 2012
Descriptive followup study (single centre) Prospective, longitudinal cohort study Prospective longitudinal cohort study Retrospective cohort study
Variable Psychiatric diagnosis (<19 yr) 8, 12, 24 months PDMS BSID-II ASQ Variable GMDS
No
No
n¼321
Yes
<3 yr
10 yr
Filan and colleagues63
Australia 2012
Observational cohort study
No
No
n¼30
Yes
Hansen and colleagues64
Denmark 2011
Yes
n¼2689
Yes
39e42 weeks (corrected age) and 2 yr** 15e16 yr School grades
Flick and colleagues65
USA 2011
No
n¼350
Yes
<2 yr
Various
Learning disability
DiMaggio and colleagues7
USA 2011
Nationwide No register-based follow-up study Population based Yes matched cohort study No Retrospective sibling-matched birth cohort study
26e41 weeks (corrected age) <1 yr
SDMT, CPM MAND, CELF CELF-R, CELF-E CELF-T, PPVT MRI imaging, BSID-II
No
n¼5284
Yes
<3 yr
z10 months
Psychiatric diagnosis
Study ID
1 day þ 1 month K-ABC postoperative 7 days þ 12 MRI months* BSID-III BSID-III
Neurocognitive outcomes in children after general anaesthesia
Canada 2012
PSLE
9
Continued
Table 2 Continued Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at exposure controls (month, yr) Yes/No
Age at outcome assessment
Type of outcome
Walker and colleagues57 Fan and colleagues66
Australia 2010
Population based cohort study Prospective study
No
Yes
n¼52
Yes
10e77 days
1 yr
BSID-III
Yes
No
n¼100
No
<4 yr
GDS
DiMaggio and colleagues67 Wilder and colleagues6
USA 2009
No
Yes
n¼383
Yes
12e48 months
1 week and 1, 3, 6 months postoperative Variable
Psychiatric diagnosis
Yes
No
n¼593
Yes
<4 yr
Variable
Learning disability
Majnemer and colleagues68
Canada 2009
No
No
n¼131
No
<6 months
Variable
Bartels and colleagues69
Netherlands 2009
No
No
n¼1143
Yes
Variable
12 yr
Various neurobehavioral and neurocognitive scales School grades
No
No
n¼368
No
Variable
<19 yr
CBCL
Yes
No
n¼23
No
4e18 yr
Variable
No
No
n¼54
Yes
Prospective No longitudinal study Prospective No longitudinal study
No
n¼30
Yes
2, 5, 8 yr Variable (any time during first hospitalisation) <30 days 3 yr
DSM-IV CBCL WPPSI-R Sensory function
No
n¼30
Yes
<30 days
USA 2009
Netherlands Kalkman and 2009 colleagues70 Kayaalp and Turkey 2006 colleagues71 The Victorian Infant Australia 1996 Collaboration72
Ludman and colleagues73 Ludman and colleagues74
England 1993 England 1990
Retrospective cohort study Retrospective, population based cohort study Prospective study
Register based national cohort study Retrospective cohort study Retrospective cohort study Prospective, longterm follow-up study
6 þ12 months
GMDS GMDS
yAbbreviations ASQ, Ages and Stages Questionnaire; BDS, Backward digit span test; BSID-II, Bayley Scales of Infant and Toddler Development, 2nd Edition, Bayley Scales of Infant Development; BSID-III, Bayley Scales of Infant and Toddler Development, 3rd Edition; CBCL, Child Behaviour Checklist; CELF, Clinical Evaluation of Language Fundamentals; CELF-E, Clinical Evaluation of Language Fundamentalsdexpressive language score; CELF-R, Clinical Evaluation of Language Fundamentalsdreceptive language score; CELF-T, Clinical Evaluation of Language Fundamentalsdtotal language ability; CHQ 50, The Child Health Questionnaire; CPM, Raven’s Coloured Progressive Matrices; DSM-IV Diagnostic and Statistical Manual of Mental Disorders, 4th Edition; EDI, Early Developmental Instrument; FDS, Forward digit span test; FSIQ, Full Scale Intelligence-Quotient; GDMS, General Decision Making Scale; GDS, Gesell Developmental Schedule; GMDS, Griffith’s Mental Development Scale; GMFCS, General Motor Function Classification Score; HAWIVA-III, Hannover-Wechsler Intelligence Scale, 3rd edition; IQ, Intelligence Quotient; K-ABC, Kaufmann Assessment Battery for Children (Korean Version); KET-KID, Kognitiver Entwicklungstest fu ¨ r das Kindergartenalter; MAND, McCarron Assessment of Neuromuscular Development; MRI, Magnetic Resonance Imaging; OWLS, Oral and Written Language; PDMS, Peabody Developmental Motor Scales; PIQ, Performance Intelligence-Quotient; PPVT, Peabody Picture Vocabulary Test; PSLE, Primary School Leaving Examination; PVF, Phonemic Verbal Fluency test; SDMT, Symbol Digit Modality Test; SON-R, 2.5e7yr test Hogrefe/Snijders-Oomen Non-Verbal Intelligence Test; SVF, Semantic Verbal Fluency test; VIQ, Verbal Intelligence-Quotient; WAIS Wechsler Adult Intelligence Scale; VMI Beery-Buktenica Developmental Test of Visual Motor Integration, 5th Edition; WPPSI-III, Wechsler Preschool and Primary Scale of Intelligence, 3rd Edition; WPPSI-R, Wechsler Preschool and Primary Scales for Intelligence: revised.
Clausen et al.
China 2010
-
Study design
10
Country of origin and publication yr
Study ID
Neurocognitive outcomes in children after general anaesthesia
‘Development’ was assessed in 35 studies. One-third of these used the Bayley Scales of Infant Development (BSID) (2nd or 3rd edition) or parts hereof. Tests evaluating ‘intelligence/ cognition’ were used in 21 studies, with seven of these being the Wechsler Preschool and Primary Scale of Intelligence. In 15 studies assessing ‘academics/achievement’, this was either the sole measure or the primary outcome. The domain of ‘diagnosis/screening’, covering neuropsychological disorders and learning disability (LD), was used as an outcome measure in 14 studies. Magnetic resonance imaging (MRI) was performed in seven studies. Eight studies assessed miscellaneous outcomes including: biomarkers, mortality, neurological examination, measurement of head circumference or impairment of vision: visual acuity, refractive error, and thickness of retinal nerve fibre layer. The main features of included studies are summarised in Table 2. Complete study characteristics (study identification, design, anaesthesia protocol, type of surgical/diagnostic procedure, cohort selection, number of exposed/non-exposed individuals, age at exposure, gender, age at follow-up, and outcome measures) are presented as Supplementary Material (Supplementary 2). Overall, studies originated from 19 different countries published from 1990 to 2017. The annual publication rate has increased over time, ranging from one publication in 1990 to 15 in 2016. The prevalent study designs were retrospective and prospective observational studies. Number of exposed individuals ranged from n¼15 to n¼37 231 (mean n¼1454). In more than half of the studies (38 out of 67) the outcome was compared between exposed and unexposed individuals. Anaesthetic drugs and procedures were specified in 28 (42%) studies. Outcome was assessed at one point in time in 53 (79%) studies. Age at exposure and outcome assessment ranged from less than 30 days to age 20 yr, respectively (mean¼37 months). For the youngest infants, age at exposure seemed to cluster around 1e3 months (Fig. 2). In many studies investigating outcome later than 2 yr of age, infants and toddlers were around 36e48 months old at time of exposure (Fig. 3).
Quality assessment Risk of bias evaluation according to NOS assessment tool for cohort studies is illustrated in Table 3. Overall, studies were rewarded three to eight stars (mean, six) out of a maximum of nine stars. As a result of a different design, one study was rewarded a ‘low risk of bias’ in three out of four domains according to the Cochrane risk of bias tool for randomised controlled trials.4 Exposed individuals are representative of the particular community in 57 of the studies. According to the NOS manual, ‘community’ is defined as the study population (of exposed individuals).15 Twenty-nine studies (43%) did not evaluate an outcome in non-exposed controls, which introduced selection bias. If ‘ascertainment of exposure’ was based on a civil register, this was judged of equal quality as hospital files or journals, even though the data were collected for administrative purposes.77 In total, 55 studies (82%) were labelled to have a low risk of bias in this category. In less than half of the studies (n¼22) the relevant outcome was measured before exposure. This is not considered an increased risk of bias if surgery and GA was conducted in early infancy as testing before exposure would have been nonfeasible.
-
11
Fig 2. Age at exposure <2 yr of age (mean¼5.2 months).
Fig 3. Age at exposure >2 yr of age (mean¼70 months).
As a result of the observational, prospective, or retrospective design of most studies internal validity could not easily be optimised during the design phase. Naturally, most studies adjust for potential confounding factors using stratification of relevant groups, propensity matching, or regression analysis. A majority of studies detected outcome by independent assessment or linkage of (hospital) records. Only few blinded their assessors. As 29 (43%) studies did not include nonexposed controls, blinding would not have added any additional value and ‘non-blinding’ is not considered a significant risk of bias in these studies. Whether length to follow-up was adequate, was evaluated individually in each study depending on the outcome measure. For instance, in studies assessing academic achievement at age 15e16 yr following exposure before age 4 yr, this was considered ‘adequate’, based on the assumption that adolescents possess all cognitive skills necessary to sit a standardised test. Additionally, as the BSID is validated for young ages, assessment of infants using this scale was considered adequate. This domain was rewarded a ‘low risk of bias’ in 53 (79%) studies. In comparison, only 37 (55%) studies were rewarded a star for ‘adequacy of follow-up’; either studies followed up on 70% or less of exposed individuals or did not provide a description of those lost.
Discussion Numerous studies have investigated long-term cognitive consequences of GA, surgery, or both, in neonates and infants. Judged by the increasing numbers of papers published each year, the interest in this topic is growing, presumably because
12
-
Clausen et al.
Table 3 Risk of bias evaluation; A-I represents rating categories according to Newcastla-Ottawa Scale for assessing quality of non-randomised studies. A, Representativeness of the exposed cohort; B, Selection of the non-exposed cohort; C, Ascertainment of exposure; D, Outcome of interest was not present at start of study; E, Comparability of cohorts - major factor controlled for; F, Comparability of cohorts - any additional factor controlled for; G, Assessment of outcome; H, Follow-up long enough for outcome to occur? I, Adequacy of follow-up of cohorts. Study ID 3
Glatz and colleagues Nestor and colleagues16 Hu and colleagues17 Conrad and colleagues18 Birajdar and colleagues19 Terushkin and colleagues20 Lap and colleagues21 Yazar and colleagues22 Sun and colleagues2 Seltzer and colleagues23 Davidson and colleagues4 O’Leary and colleagues24 Djurhuus and colleagues25 Clausen and colleagues75 Poor Zamany Nejat Kermany and colleagues26 Hoffman and colleagues27 Hansen and colleagues28 Graham and colleagues29 Doberschuetz and colleagues30 Diaz and colleagues31 de Heer and colleagues32 Aun and colleagues33 Bakri and colleagues34 Taghon and colleagues35 Naguib and colleagues36 Ko and colleagues37 Hansen and colleagues38 Gano and colleagues39 Backeljauw and colleagues40 ckova and colleagues41 Petra Yan and colleagues42 Yin and colleagues43 Morriss and colleagues44 Ing and colleagues45 Gaynor and colleagues46 Garcia Guerra and colleagues47 Cheng and colleagues48 Andropoulos and colleagues49 Ko and colleagues50 Minutillo and colleagues51 Hansen and colleagues52 Fan and colleagues53 Bong and colleagues54 Yang and colleagues55 Andropoulos and colleagues56 Walker and colleagues57 Sprung and colleagues58 Sananes and colleagues59 Rocha and colleagues60 Long and colleagues61 Long and colleagues62 Ing and colleagues5 Filan and colleagues63
A B
C D E
F
G H I
+ + + + + + + + +
+ + + + + +
+ + + + + + + + +
+ +
+ + + + + + + + +
+ + + + + + +
+ + + + + + +
+ + + + + + + +
+ + + + + +
+ + + +
+ + + +
+ + + +
+ + +
+ + + +
+ + +
+ + +
+ + + +
+ + +
+ +
+ + +
+ + +
+ + + +
+ + + +
+ + +
+ +
+ + + + + + + + + + + + +
+ + + + + + + + + + + +
+ + + + + + + + + + + + + + +
+ + + + + +
+ + + + + + + + +
+ + + + + + + + + + + + + + +
+ + + + + + + + + +
+ + + + + + + + + + + +
+ + + + + + + +
+ + + + + + + + + + + + + + + + + + +
+ + +
+ + + + + + +
+ + + +
+ + +
+ + + + + +
+ + + + +
+ + + + +
+ + + +
+ + + + + + +
+ + + +
+ + + + + + +
+ +
+ + + + + +
+ + + + + +
+ + + + + + + +
+ + +
Continued
Table 3 Continued Study ID
A B
C D E
F
G H I
Hansen and colleagues64 Flick and colleagues65 DiMaggio and colleagues7 Walker and colleagues57 Fan and colleagues66 DiMaggio and colleagues67 Wilder and colleagues6 Majnemer and colleagues68 Bartels and colleagues69 Kalkman and colleagues70 Kayaalp and colleagues71 The Victorian Infant Collaboration72 Ludman and colleagues73 Ludman and colleagues74 Representativeness of the exposed cohort Selection of the non-exposed cohort Ascertainment of exposure Outcome of interest was not present at start of study Comparability of cohortsdmajor factor controlled for Comparability of cohortsdany additional factor controlled for Assessment of outcome Follow-up long enough for outcomes to occur? Adequacy of follow-up of cohorts
+ + + + + + + + + + + +
+ + + + + + + +
+ + + + + + + + + +
+ + + + + + + +
+ + + + + + + + +
+ + + + + + + +
+ + + + + + +
+ + + + + + + + + +
+ + + + + + +
+ + + + + + + + + + + A B C D
E
F
G H I
of the lack of a robust and reproducible phenotype. Investigators have used a variety of specific assessment tools, or parts hereof, to evaluate different criteria within development, cognition, intelligence, academic achievement, and neuropsychological diagnosis. Comparatively, few studies have assessed outcomes such as neuroimaging,18,35,48,49,56,63 mortality, or serum-biomarkers.18,19,22,38,39,42,44,48 Previously, it has been questioned which criteria are suitable outcome measures in studies on long-term neurocognitive consequences following GA and surgery.78 So far, epidemiologic studies have shown contradictory results. Register-based studies are unable to demonstrate any causative associations. Although this is largely made up for by the opportunity to analyse large sample sizes, epidemiologic studies have been criticised and the call for randomised clinical trials has been repetitively stated. Based on the findings in this review, this criticism may not be entirely reasonable, as age at exposure has not been assessed systematically. Rather, focus has been on infants or young children between 1 and 3 months or 2 and 4 yr, respectively. This clustering can probably be explained by the assumption that development of the infant brain is most vulnerable to potential toxins during peak synaptogenesis of neurons (i.e. period from 3rd trimester to 12e24 months of age).79 In comparison, the effect of GA exposure during age 3 months to 2 yr or >4 yr has been investigated scarcely. As brain development in young children is dynamic and on-going, associations in other domains might have been missed. Furthermore, it may be reasonable to assume, that infants
Neurocognitive outcomes in children after general anaesthesia
undergoing surgery in the first months of life suffer from significant underlying conditions, which might confound the effect of exposure to GA on cognitive outcome. The studies seem to resemble each other regarding design (i.e. retrospective cohorts), but vary immensely in all other characteristics highlighted in this review (type of, and indication for, surgery; sample size; age at exposure and follow-up; as well as outcome measures).
Cognition and neurodevelopment Intelligence, covering several cognitive domains (memory, reasoning, spatial ability, executive function, and processing speed), is considered a relatively stable and constant trait throughout life.80 Cognition in male adolescents was tested in the Vietnam Experience Study, a cohort of soldiers involved in military service abroad,81 and these individuals were reassessed at middle-age. A high correlation between scores in young- and middle-aged individuals was found. In a Scottish birth cohort, intelligence was assessed in 11-yr-old children who were followed-up 66 yr later. It was concluded that mental abilities are stable across a human lifespan unless a person suffers severe diseases.80 Similar findings have been confirmed in a multitude of similar longitudinal studies regarding intelligence. Assessment of development has focused on the domains of language, perception, and motor development. It has long been recognised that infant neuro-behaviour follows an age specific pattern. In 1931, Gesell82 described the ‘Developmental morphology of infant behaviour pattern’ as ‘significant consistencies in the tempo, the forms and sequences of behaviour growth which point to the presence of a lawful developmental mechanic’. Unlike intelligence, abilities within these domains change over time. Development within the norm is evaluated based on whether the child reaches predefined ‘milestones’. Deviations from or delay in progression might be as a result of underlying disease or external factors.83 Interestingly, development and overall cognitive ability has been shown to correlate. In a 34-yr follow-up of a sample of 307 children identified in the Copenhagen Perinatal Cohort comprising 9125 deliveries from late 1959 to 1961, an association between several milestones of motor development and intelligence in adulthood was demonstrated.84 In a French mother-child cohort, it was shown that milestones before age 2 yr predicted intelligence in 5e6-yr-old children.85 In summary, sensorimotor development relates well to cognition/intelligence and psychiatric disorders. Further, as milestones are well defined, development can be assessed even in the youngest ages. At the same time, development can be influenced both positively and negatively by external factors (e.g. by interaction or with parents or severe diseases necessitating frequent treatment or hospitalisation, xenobiotics, etc.), which compromises the stability of development as an outcome measure. Intelligence is a stable factor across lifetime, but assessment is not feasible until the child has achieved some basic cognitive skills. Hence, assessing children younger than school age is not recommended.
Neuropsychological diagnosis LD, autism spectrum disorders (ASD), and attention deficit hyperactivity disorder (ADHD) have also been used as outcome measures in the current context.
-
13
LD comprises dyscalculia and dyslexia. Both disorders are common (with prevalence ~5% for each item). Their origin is multifactorial, and their clinical presentation heterogeneous and changing over time.86,87 LD has been associated with psychiatric diseases and maladaptive traits.88 However, if LD is diagnosed and supportive interventions appointed, deficits can be attenuated. Additionally, differences in scores of verbal intelligence and performance intelligence have been shown to predict LD.89 Importantly, a diagnosis of LD is not necessarily equivalent to lower intelligence scores. ASD is characterised by impaired neurodevelopment and considered to be a lifelong condition.90,91 Children with ASD have difficulties interacting and communicating socially and present a behavioural pattern shaped by strict routine.81 Severity of symptoms varies among individuals and with development,92 which aggravates the clinical diagnosis. However, the course of sensorineural development relates to the clinical presentation and hence predicts the seriousness of the condition.93 ADHD is a mental development disorder clinically presented by hyperactivity and impaired ability to focus and concentrate. ADHD has been linked with neurocognitive development.94 While the natural course of ADHD cannot be predicted by neurocognitive assessments, the causal pathway of the disorder is known to consist of multiple factors, including an inherited genetic disposition.95 The prevalence varies with gender, ethnicity, countries, states96 and there prevails some disagreement whether ADHD should be diagnosed by detection of core symptoms only or whether emotional and executive functions should be assessed as well.97 Several studies have used academic achievement (i.e. scores in standardised exams). Not surprising, intelligence and academic achievement are linked together.98,99 However, academic performance is influenced by many other factors unrelated to GA/surgical exposure,25 for example, altered relationships with peers at school and low self-esteem because of childhood obesity,100 as well as specific life-style factors (longer sleep duration101 and physical activity102). In a recent study by our group, achievement at the Danish national 9th grade exam was compared between 558 children undergoing surgery for isolated oral clefts and randomly selected controls. The type of oral cleft, parental age, and length of education were more important than number and timing of exposures.75
Miscellaneous outcomes Neuroimaging by MRI offers measurement of gross brain structures. MRI has shown promising results in the early (<2 yr of age) identification of ASD. Children who were diagnosed with ASD later in life presented an atypical pattern of connectivity of brain regions compared with non-exposed children; microstructural properties of white matter fibre tracts were characterised by fractional anisotropy and axial diffusivity. Further, children with high risk of ASD presented blunted trajectories consistent with reduced fractional anisotropy.103 Among studies identified in this review, few have used MRI to look for changes after exposure to GA in childhood.35,40,63 As was shown, GA itself could not be associated with any reduction in brain tissue, exposed children had lower scores of performance and language intelligence, and lower scores were associated with decreased volume of grey matter in posterior
14
-
Clausen et al.
brain regions.40 However, 53 of 81 types of surgery were ear, nose, throat (ENT) related procedures. This (particular) population of children has been suspected to suffer language and speech problems and achieve poor academic results because of prolonged hearing loss caused by, for example, otitis media.104,105 One study compared MRI scans of premature infants (born <30 weeks of gestation) and neonates with low birth weight (<1250 g), both groups either exposed or non-exposed to GA and surgery at 39e42 weeks corrected age. Additionally, development was assessed at age 2 yr. Brain volume was decreased and mental developmental index score (one of many items included in the BSID III) lowered in exposed children.35 As these results were based on 15 exposed and nonexposed individuals, they suffer from a comparatively small sample size. The serum-concentration of S100B following GA was assessed as a secondary outcome in one study.42 In traumatology, S100B was introduced as a marker of severity of brain damage after traumatic brain injury.106,107 It has also been used for follow-up of patients with malignant melanoma.108 However, S100B is not brain-tissue specific109,110 and its utility as a marker for potential neuronal damage following GA is unknown.
Methodological considerations Psychometric tests are used for commercial, research, and clinical purposes.111 Crucial for evaluating the quality of a test are reliability (consistency of measurements of a test) and validity (degree to which a test measures, what it claims to measure). Internal reliability reflects the interrelation between items contained in a test.112 Validity is a crucial feature of a test.113 Validity does not apply to the test itself but resembles a property of a test score. When a test has been ‘validated’ (i.e. from one language or country to another), its ability to measure the construct of interest (intelligence, behaviour, etc.) has been correlated with tests of the same construct (‘convergent validity’) or has shown poor correlation with a test of a different construct (‘discriminative validity’). Validity is not a static factor but must be considered in each specific study context (study population, test users, overall setting). For instance, concurrent validity of the Bayley Developmental Index and the Cognitive Adaptive Test/Clinical Linguistic and Auditory Milestone Scale (CAT/CLAMS) was determined in a primary care health setting assessing children 12- and 30-months-old without suspected developmental delay.113 The authors concluded that the CAT/CLAMS had reasonable concurrent validity with the Bayley Scales and was easy and time efficient to use by professionals (e.g. paediatricians) without special training.114,115 As shown in our data, sample sizes vary tremendously, with the lowest numbers being 1535 and 21.23,36,41,55 Small sample sizes are problematic, as they reduce the power of a study. This eventually reduces reliability of findings because of low probability of finding the true effects, reduced positive predictive value (i.e. the probability of a true association in case of a significant result) and ‘effect inflation’.116,117 Considering the great number of children exposed to surgery and GA each year worldwide, the effect size of potential neurotoxicity is most likely to be minimal, as the clinical presentation or potential phenotype has not yet been defined. Evaluation of study effect size is not within the scope of this review. Nevertheless, reduced power because of small sample
and effect size remains a major concern within the field of anaesthesia-related neurotoxicity research. Few studies (n¼22) have reported on ‘baseline’ values before exposure, that is, demonstrated that the outcome of interest (reduced intelligence, delay of development, etc.) was not present before exposure (Table 3). However, inclusion of individuals prone to neurodevelopmental delay might lead to overestimation of an adverse effect of GA/surgery on neurocognitive outcome. Several studies have evaluated a selected study group knowingly associated with impaired neurodevelopment.118,119 These comprised children with hypoplastic left heart syndrome undergoing the Norwood procedure,28 congenital diaphragmatic hernia, or gastroschisis.51,60 Children undergoing major surgery might suffer from other underlying diseases/conditions, which confound the effect estimate and blur a ‘true’ association. Confounding by indication and selection bias is a concern in 29 (43%) studies, which did not compare outcome between exposed and non-exposed controls. Similarly, the potential confounding by co-morbidity cannot be entangled in studies assessing outcome after exposure for miscellaneous types of surgery [n¼45 (67%)]: minor surgery, for instance laser surgery for vascular malformations20 or inguinal hernia repair,7,64 compared with major surgery, is less likely to associate with severe underlying disease. Some outcomes either vary with age (behaviour, development) or show a heterogenic clinical presentation (ASD, LD, ADHD). Detecting an impaired score within these ‘unstable’ domains might overestimate an effect size that the child might ‘catch up’ later. ‘Loss-to-follow-up’ or missing details of those lost are a concern in 29 (43%) studiesdwhether this loss underestimates or overestimates the true effect depends on the reason(s) why individuals cannot be assessed. As these studies either followed-up on less than 70% of respective study populations or do not account for reasons and characteristics of those lost, specifics cannot be eluded. In concordance with our findings, previously published data reported a loss-to-follow-up because of migration of approximately 30%. Within the framework of observational studies based on birth cohorts, increasing mobility was considered to represent a major challenge. However, the authors did not consider differences between migrants and non-migrant (i.e. prevalence of congenital defects, maternal age and level of education, compliance to attend prenatal visits, ethnicity or single parent status) to influence their results on the prevalence on LD in their specific cohort.120 This may be a significantly flawed approach.
Quality assessment using the NOS We chose (to use) the NOS for quality assessment of the included studies, as this scale has been recommended for cohort studies and endorsed by the Cochrane collaboration.121 Evaluator burden has been described as less compared with other tools because of ease of use and limited time-consumption.122 These findings do not coincide completely with our experience. Some decision rules for quality rating of domains are less comprehensive and prone to misinterpretations. Evaluating (e.g. ‘comparability’) turned out to be difficult. Neurodevelopment in children is recognised as a complex and multifactorial phenomenon. Focusing on a single most
Neurocognitive outcomes in children after general anaesthesia
important factor is frequently not possible. In studies evaluating academic achievements, ‘male gender’ was a major confounder, as boys perform poorer than girls. When studies investigated outcome after neonatal cardio-thoracic surgery, ‘underlying conditions’ also constituted a major confounder; the prognosis differed depending on the specific cardiac malformation necessitating surgery (for instance uni-ventricular vs bi-ventricular cardiac malformation). We followed the same rule of decision-making for studies investigating outcomes after miscellaneous surgeries, as the impact of underlying condition is assumed to be pivotal for development. Recently, 13 papers were reviewed and rated according to NOS as part of a meta-analysis, discussing outcome measurements after a single GA administered before 3 and 4 yr.123 All 13 papers were included in the current study. The overall number of stars rewarded varied slightly from ours. Because subgroup labelling is not stated, any variation in decision-making cannot be specifically compared. Based on these considerations, we agree that inter-rater reliability of the NOS is a concern.124 To the best of our knowledge, we identified all studies relevant to this review according to inclusion criteria. As a result of exclusion of papers not published in English, we might have missed relevant work.
Conclusion The effect(s) of GA (and surgery) in young children have been assessed by various outcome measures. This variability in test items used reflects the problem, that the clinical presentation of a potential damaging impact on young children’s brains following GA/surgery is not properly defined. Studies within this topic continue to be inconsistent regarding outcome measures as well as other important factors. Consequently, future observational studies are only considered meaningful if they are conducted in cohorts of considerable sample size and comparable regarding type of surgery, comorbidity, age at exposure, follow-up, indication for intervention, and outcome. As development of neurocognition is multifactorial, the investigation of a single causative factor might be too simplistic. Neurotoxicity of anaesthetics might be clinically detectable only if several factors occur in combination. Regarding the choice of outcome measurement, the prevailing use of development tests is considered feasible in infants. In older age groups (i.e. >6 yr), intelligence tests seem reasonable, because intelligence is a stable characteristic and covers cognitive functions, as do academic achievements. The majority of studies included in this review have a low to moderate risk of bias in the domains of selection, comparability, and outcome assessment. However, some concerns exist about the feasibility of NOS. To avoid confusion, it is recommended to specify quality-rating rules.
Authors’ contributions Designed and composed the review protocol: all authors Independently screened title and abstracts and full text of articles: N.G.C., S.K. Extracted data: N.G.C. Checked extractions for completeness and correctness: S.K. Conducted quality assessment of studies by the NewcastleOttawa Scale: N.G.C. Quality assessment of studies re-assessed: S.K.
-
15
Drafted the paper: N.G.C. Critically revised the article repetitively: S.K., T.G.H. Approved the final version of the paper: all authors
Acknowledgements We are grateful to Research Librarian, Mrs Eriksen MB at The Medical Research Library, University Library of Southern Denmark for her kind assistance with design of the elaborate database searches.
Declaration of interest The authors have no conflicts of interests to declare. No ethical approvals were required for conductance of this study.
Funding The study is part of N.G.C.’s PhD project, which was awarded with the European Society of Anaesthesia (ESA) MAQUET GRANT at the annual meeting of the European Society of Anaesthesia 2015. N.G.C.’s project is further supported by funds provided by the Department of Anaesthesia and Intensive Care Medicine, University Hospital Odense.
Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.bja.2017.11.107.
References 1. Hansen TG. Use of anaesthetics in young children: consensus statement of the European Society of Anaesthesiology, the European Society of Paediatric Anaesthesiology, the European Association of Cardiothoracic Anaesthesiology and the European Safe Tots Anaesthesia Research Initiative. Eur J Anaesthesiol 2017; 34: 327e8 2. Sun LS, Li G, Miller TL, et al. Association between a single general anesthesia exposure before age 36 months and neurocognitive outcomes in later childhood. JAMA 2016; 315: 2312e20 3. Glatz P, Sandin RH, Pedersen NL, Bonamy AK, Eriksson LI, Granath F. Association of anesthesia and surgery during childhood with long-term academic performance. JAMA Pediatr 2017; 171: e163470 4. Davidson AJ, Disma N, de Graaff JC, et al. Neurodevelopmental outcome at 2 years of age after general anaesthesia and awake-regional anaesthesia in infancy (GAS): an international multicentre, randomised controlled trial. Lancet 2016; 387: 239e50 5. Ing C, DiMaggio C, Whitehouse A, et al. Long-term differences in language and cognitive function after childhood exposure to anesthesia. Pediatrics 2012; 130: e476e85 6. Wilder RT, Flick RP, Sprung J, et al. Early exposure to anesthesia and learning disabilities in a populationbased birth cohort. Anesthesiology 2009; 110: 796e804 7. DiMaggio C, Sun LS, Li G. Early childhood exposure to anesthesia and risk of developmental and behavioral disorders in a sibling birth cohort. Anesth Analg 2011; 113: 1143e51
16
-
Clausen et al.
8. FDA U. FDA Drug Safety Communication: FDA review results in new warnings about using general anesthetics and sedation drugs in young and pregnant women. 2016. Available from: https://wwwfdagov/Drugs/DrugSafety/ucm532356htm [Accessed 10 June 2017] 9. Bilotta F, Evered LA, Gruenbaum SE. Neurotoxicity of anesthetic drugs: an update. Curr Opin Anaesthesiol 2017; 30: 452e7 10. Disma N, Mondardini MC, Terrando N, Absalom AR, Bilotta F. A systematic review of methodology applied during preclinical anesthetic neurotoxicity studies: important issues and lessons relevant to the design of future clinical research. Paediatr Anaesth 2016; 26: 6e36 11. Ringer SK, Ohlerth S, Carrera I, et al. Effects of hypotension and/or hypocapnia during sevoflurane anesthesia on perfusion and metabolites in the developing brain of pigletsda blinded randomized study. Paediatr Anaesth 2016; 26: 909e18 12. McCann ME, de Graaff J. Current thinking regarding potential neurotoxicity of general anesthesia in infants. Curr Opin Urol 2017; 27: 27e33 13. Weiss M, Hansen TG, Engelhardt T. Ensuring safe anaesthesia for neonates, infants and young children: what really matters. Arch Dis Child 2016; 101: 650e2 14. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and metaanalyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009; 151: W65e94 15. Wells GA, Shea B, O’Connell D, Peterson J, Welch V, Losos M, Tugwell P. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available from: http://wwwohrica/ programs/clinical_epidemiology/oxfordasp [Accessed 22 August 2016 & 3 March 2017]. 16. Nestor KA, Zeidan M, Boncore E, et al. Neurodevelopmental outcomes in infants undergoing general anesthesia. J Pediatr Surg 2017; 52: 895e900 17. Hu D, Flick RP, Zaccariello MJ, et al. Association between exposure of young children to procedures requiring general anesthesia and learning and behavioral outcomes in a population-based birth cohort. Anesthesiology 2017; 127: 227e40 18. Conrad AL, Goodwin JW, Choi J, Block RI, Nopoulos P. The relationship of exposure to anesthesia on outcomes in children with isolated oral clefts. J Child Neurol 2017; 32: 308e15 19. Birajdar S, Rao S, McMichael J. Neurodevelopmental outcomes of neonates undergoing surgery under general anesthesia for malrotation of intestines. Early Hum Dev 2017; 109: 32e6 20. Terushkin V, Brauer J, Bernstein L, Geronemus R. Effect of general anesthesia on neurodevelopmental abnormalities in children undergoing treatment of vascular anomalies with laser surgery: a retrospective review. Dermatol Surg 2017; 43: 534e40 21. Lap CC, Bolhuis SW, Van Braeckel KN, et al. Functional outcome at school age of children born with gastroschisis. Early Hum Dev 2017; 106e107: 47e52 22. Yazar S, Hewitt AW, Forward H, et al. Early anesthesia exposure and the effect on visual acuity, refractive error, and retinal nerve fiber layer thickness of young adults. J Pediatr 2016; 169: 256e9
23. Seltzer L, Swartz MF, Kwon J, et al. Neurodevelopmental outcomes after neonatal cardiac surgery: role of cortical isoelectric activity. J Thorac Cardiovasc Surg 2016; 151: 1137e42 24. O’Leary JD, Janus M, Duku E, et al. A population-based study evaluating the association between surgery in early life and child development at primary school entry. Anesthesiology 2016; 125: 272e9 25. Djurhuus BD, Hansen TG, Pedersen JK, Faber CE, Christensen K. School performance in cholesteatomaoperated children in Denmark: a nationwide population-based register-study. Acta Otolaryngol 2016; 136: 663e8 26. Poor Zamany Nejat Kermany M, Roodneshin F, Ahmadi Dizgah N, Gerami E, Riahi E. Early childhood exposure to short periods of sevoflurane is not associated with later, lasting cognitive deficits. Paediatr Anaesth 2016; 26: 1018e25 27. Hoffman GM, Brosig CL, Bear LM, Tweddell JS, Mussatto KA. Effect of intercurrent operation and cerebral oxygenation on developmental trajectory in congenital heart disease. Ann Thorac Surg 2016; 101: 708e16 28. Hansen JH, Rotermann I, Logoteta J, et al. Neurodevelopmental outcome in hypoplastic left heart syndrome: impact of perioperative cerebral tissue oxygenation of the Norwood procedure. J Thorac Cardiovasc Surg 2016; 151: 1358e66 29. Graham MR, Brownell M, Chateau DG, Dragan RD, Burchill C, Fransoo RR. Neurodevelopmental assessment in kindergarten in children exposed to general anesthesia before the age of 4 years: a retrospective matched cohort study. Anesthesiology 2016; 125: 667e77 30. Doberschuetz N, Dewitz R, Rolle U, Schlosser R, Allendorf A. Follow-up of children with gastrointestinal malformations and postnatal surgery and anesthesia: evaluation at two years of age. Neonatology 2016; 110: 8e13 31. Diaz LK, Gaynor JW, Koh SJ, et al. Increasing cumulative exposure to volatile anesthetic agents is associated with poorer neurodevelopmental outcomes in children with hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 2016; 152: 482e9 32. de Heer IJ, Tiemeier H, Hoeks SE, Weber F. Intelligence quotient scores at the age of 6 years in children anaesthetised before the age of 5 years. Anaesthesia 2017; 72: 57e62 33. Aun CS, McBride C, Lee A, et al. Short-term changes in postoperative cognitive function in children aged 5 to 12 years undergoing general anesthesia: a cohort study. Medicine 2016; 95: e3250 34. Bakri MH, Ismail EA, Ali MS, Elsedfy GO, Sayed TA, Ibrahim A. Behavioral and emotional effects of repeated general anesthesia in young children. Saudi J Anaesth 2015; 9: 161e6 35. Taghon TA, Masunga AN, Small RH, Kashou NH. A comparison of functional magnetic resonance imaging findings in children with and without a history of early exposure to general anesthesia. Paediatr Anaesth 2015; 25: 239e46 36. Naguib AN, Winch PD, Tobias JD, et al. Neurodevelopmental outcome after cardiac surgery utilizing cardiopulmonary bypass in children. Saudi J Anaesth 2015; 9: 12e8
Neurocognitive outcomes in children after general anaesthesia
37. Ko WR, Huang JY, Chiang YC, et al. Risk of autistic disorder after exposure to general anaesthesia and surgery: a nationwide, retrospective matched cohort study. Eur J Anaesthesiol 2015; 32: 303e10 38. Hansen TG, Pedersen JK, Henneberg SW, Morton NS, Christensen K. Neurosurgical conditions and procedures in infancy are associated with mortality and academic performances in adolescence: a nationwide cohort study. Paediatr Anaesth 2015; 25: 186e92 39. Gano D, Andersen SK, Glass HC, et al. Impaired cognitive performance in premature newborns with two or more surgeries prior to term-equivalent age. Pediatr Res 2015; 78: 323e9 40. Backeljauw B, Holland SK, Altaye M, Loepke AW. Cognition and brain structure following early childhood surgery with anesthesia. Pediatrics 2015; 136: e1e12 ckova I, Zach J, Borsky J, et al. Early and late opera41. Petra tion of cleft lip and intelligence quotient and psychosocial development in 3e7 years. Early Hum Dev 2015; 91: 149e52 42. Yan J, Li YR, Zhang Y, Lu Y, Jiang H. Repeated exposure to anesthetic ketamine can negatively impact neurodevelopment in infants: a prospective preliminary clinical study. J Child Neurol 2014; 29: 1333e8 43. Yin J, Wang SL, Liu XB. The effects of general anaesthesia on memory in children: a comparison between propofol and sevoflurane. Anaesthesia 2014; 69: 118e23 44. Morriss Jr FH, Saha S, Bell EF, et al. Surgery and neurodevelopmental outcome of very low-birth-weight infants. JAMA Pediatr 2014; 168: 746e54 45. Ing CH, DiMaggio CJ, Malacova E, et al. Comparative analysis of outcome measures used in examining neurodevelopmental effects of early childhood anesthesia exposure. Anesthesiology 2014; 120: 1319e32 46. Gaynor JW, Ittenbach RF, Gerdes M, et al. Neurodevelopmental outcomes in preschool survivors of the Fontan procedure. J Thorac Cardiovasc Surg 2014; 147: 1276e82. discussion 82e83 e5 47. Garcia Guerra G, Robertson CM, Alton GY, et al. Neurotoxicity of sedative and analgesia drugs in young infants with congenital heart disease: 4-year follow-up. Paediatr Anaesth 2014; 24: 257e65 48. Cheng HH, Wypij D, Laussen PC, et al. Cerebral blood flow velocity and neurodevelopmental outcome in infants undergoing surgery for congenital heart disease. Ann Thorac Surg 2014; 98: 125e32 49. Andropoulos DB, Ahmad HB, Haq T, et al. The association between brain injury, perioperative anesthetic exposure, and 12-month neurodevelopmental outcomes after neonatal cardiac surgery: a retrospective cohort study. Paediatr Anaesth 2014; 24: 266e74 50. Ko WR, Liaw YP, Huang JY, et al. Exposure to general anesthesia in early life and the risk of attention deficit/ hyperactivity disorder development: a nationwide, retrospective matched-cohort study. Paediatr Anaesth 2014; 24: 741e8 51. Minutillo C, Rao SC, Pirie S, McMichael J, Dickinson JE. Growth and developmental outcomes of infants with gastroschisis at one year of age: a retrospective study. J Pediatr Surg 2013; 48: 1688e96 52. Hansen TG, Pedersen JK, Henneberg SW, Morton NS, Christensen K. Educational outcome in adolescence following pyloric stenosis repair before 3 months of age: a nationwide cohort study. Paediatr Anaesth 2013; 23: 883e90
-
17
53. Fan Q, Cai Y, Chen K, Li W. Prognostic study of sevoflurane-based general anesthesia on cognitive function in children. J Anesth 2013; 27: 493e9 54. Bong CL, Allen JC, Kim JT. The effects of exposure to general anesthesia in infancy on academic performance at age 12. Anesth Analg 2013; 117: 1419e28 55. Yang HK, Chungh DS, Hwang JM. The effect of general anesthesia and strabismus surgery on the intellectual abilities of children: a pilot study. Am J Ophthalmol 2012; 153: 609e13 56. Andropoulos DB, Easley RB, Brady K, et al. Changing expectations for neurological outcomes after the neonatal arterial switch operation. Ann Thorac Surg 2012; 94: 1250e5. discussion 5e6 57. Walker K, Halliday R, Holland AJ, Karskens C, Badawi N. Early developmental outcome of infants with infantile hypertrophic pyloric stenosis. J Pediatr Surg 2010; 45: 2369e72 58. Sprung J, Flick RP, Katusic SK, et al. Attention-deficit/ hyperactivity disorder after early exposure to procedures requiring general anesthesia. Mayo Clin Proc 2012; 87: 120e9 59. Sananes R, Manlhiot C, Kelly E, et al. Neurodevelopmental outcomes after open heart operations before 3 months of age. Ann Thorac Surg 2012; 93: 1577e83 60. Rocha G, Azevedo I, Pinto JC, Guimaraes H. Follow-up of the survivors of congenital diaphragmatic hernia. Early Hum Dev 2012; 88: 255e8 61. Long SH, Galea MP, Eldridge BJ, Harris SR. Performance of 2-year-old children after early surgery for congenital heart disease on the bayley scales of infant and toddler development, third edition. Early Hum Dev 2012; 88: 603e7 62. Long SH, Harris SR, Eldridge BJ, Galea MP. Gross motor development is delayed following early cardiac surgery. Cardiol Young 2012; 22: 574e82 63. Filan PM, Hunt RW, Anderson PJ, Doyle LW, Inder TE. Neurologic outcomes in very preterm infants undergoing surgery. J Pediatr 2012; 160: 409e14 64. Hansen TG, Pedersen JK, Henneberg SW, et al. Academic performance in adolescence after inguinal hernia repair in infancy: a nationwide cohort study. Anesthesiology 2011; 114: 1076e85 65. Flick RP, Katusic SK, Colligan RC, et al. Cognitive and behavioral outcomes after early exposure to anesthesia and surgery. Pediatrics 2011; 128: e1053e61 66. Fan XC, Ye M, Li DZ, Shi Y, Xu Y. Cognitive function in congenital heart disease after cardiac surgery with extracorporeal circulation. World J Pediatr 2010; 6: 268e70 67. DiMaggio C, Sun LS, Kakavouli A, Byrne MW, Li G. A retrospective cohort study of the association of anesthesia and hernia repair surgery with behavioral and developmental disorders in young children. J Neurosurg Anesthesiol 2009; 21: 286e91 68. Majnemer A, Limperopoulos C, Shevell MI, Rohlicek C, Rosenblatt B, Tchervenkov C. A new look at outcomes of infants with congenital heart disease. Pediatr Neurol 2009; 40: 197e204 69. Bartels M, Althoff RR, Boomsma DI. Anesthesia and cognitive performance in children: no evidence for a causal relationship. Twin Res Hum Genet 2009; 12: 246e53 70. Kalkman CJ, Peelen L, Moons KG, et al. Behavior and development in children and age at the time of first anesthetic exposure. Anesthesiology 2009; 110: 805e12
18
-
Clausen et al.
71. Kayaalp L, Bozkurt P, Odabasi G, et al. Psychological effects of repeated general anesthesia in children. Paediatr Anaesth 2006; 16: 822e7 72. Group TVICS. Surgery and the tiny baby: sensorineural outcome at 5 years of age. The victorian infant collaborative Study group. J Paediatr Child Health 1996; 32: 167e72 73. Ludman L, Spitz L, Lansdown R. Intellectual development at 3 years of age of children who underwent major neonatal surgery. J Pediatr Surg 1993; 28: 130e4 74. Ludman L, Spitz L, Lansdown R. Developmental progress of newborns undergoing neonatal surgery. J Pediatr Surg 1990; 25: 469e71 75. Clausen NG, Pedersen DA, Pedersen JK, et al. Oral clefts and academic performance in adolescence: the impact of anesthesia-related neurotoxicity, timing of surgery, and type of oral clefts. Cleft Palate Craniofac J 2017; 54: 371e80 76. Ing C, Wall MM, DiMaggio CJ, et al. Latent class analysis of neurodevelopmental deficit after exposure to anesthesia in early childhood. J Neurosurg Anesthesiol 2017; 29: 264e73 77. Lynge E, Sandegaard JL, Rebolj M. The Danish national patient Register. Scand J Public Health 2011; 39: 30e3 78. Davidson AJ, Becke K, de Graaff J, et al. Anesthesia and the developing brain: a way forward for clinical research. Paediatr Anaesth 2015; 25: 447e52 79. Lagercrantz H. Connecting the brain of the child from synapses to screen-based activity. Acta Paediatr 2016; 105: 352e7 80. Deary IJ, Penke L, Johnson W. The neuroscience of human intelligence differences. Nat Rev Neurosci 2010; 11: 201e11 81. American Psychiatric Association. Diagnostic and statistical manual of mental Disorderse5. 5th Edn. Washington: DAPA; 2013 82. Gesell A. The developmental morphology of infant behavior pattern. Proc Natl Acad Sci U S A 1932; 18: 139e43 83. Diepeveen FB, Dusseldorp E, Bol GW, OudesluysMurphy AM, Verkerk PH. Failure to meet language milestones at two years of age is predictive of specific language impairment. Acta Paediatr 2016; 105: 304e10 84. Flensborg-Madsen T, Mortensen EL. Infant developmental milestones and adult intelligence: a 34-year follow-up. Early Hum Dev 2015; 91: 393e400 85. Peyre H, Charkaluk ML, Forhan A, Heude B, Ramus F. Do developmental milestones at 4, 8, 12 and 24 months predict IQ at 5e6 years old? Results of the EDEN motherchild cohort. Eur J Paediatr Neurol 2017; 21: 272e9 86. Lagae L. Learning disabilities: definitions, epidemiology, diagnosis, and intervention strategies. Pediatr Clin North Am 2008; 55: 1259e68 87. Dykens EM. Psychopathology in children with intellectual disability. J Child Psychol Psychiatr 2000; 41: 407e17 88. Emerson E. Prevalence of psychiatric disorder in children and adolescents with and without intellectual disability. J Intellect Disabil Res 2003; 47: 51e8 89. Drummond CR, Ahmad SA, Rourke BP. Rules for the classification of younger children with nonverbal learning disabilities and basic phonological processing disabilities. Arch Clin Neuropsychol 2005; 20: 171e82 90. American Psychiatric Association Task Force on DSM-IV, Diagnostic and statistical manual of mental disorders, DSMIV-TR, 4. edition, text revision. Washington, D.C.: American Psychiatric Association; 2000. xxxvii, 943 s.ill
91. Taylor MJ, Gillberg C, Lichtenstein P, Lundstrom S. Etiological influences on the stability of autistic traits from childhood to early adulthood: evidence from a twin study. Mol Autism 2017; 8: 5 92. Jones RM, Lord C. Diagnosing autism in neurobiological research studies. Behav Brain Res 2013; 251: 113e24 93. Lord C, Risi S, DiLavore PS, Shulman C, Thurm A, Pickles A. Autism from 2 to 9 years of age. Arch Gen Psychiatry 2006; 63: 694e701 94. van Lieshout M, Luman M, Twisk JW, et al. Neurocognitive predictors of ADHD outcome: a 6-year followup study. J Abnorm Child Psychol 2017; 45: 261e72 95. Thapar A, Cooper M, Eyre O, Langley K. What have we learnt about the causes of ADHD? J Child Psychol Psychiatr 2013; 54: 3e16 96. Cuffe SP, Moore CG, McKeown RE. Prevalence and correlates of ADHD symptoms in the national health interview survey. J Atten Disord 2005; 9: 392e401 97. Zhou X, Reynolds CR, Zhu J, Kamphaus RW, Zhang O. Evidence-based assessment of ADHD diagnosis in children and adolescents. Appl Neuropsychol Child 2017: 1e7 98. Cattell RB, Barton K, Dielman TE. Prediction of school achievement from motivation, personality, and ability measures. Psychol Rep 1972; 30: 35e43 99. Barton K, Dielman TE, Cattell RB. Personality and IQ measures as predictors of school achievement. J Educ Psychol 1972; 63: 398e404 100. Wang F, Veugelers PJ. Self-esteem and cognitive development in the era of the childhood obesity epidemic. Obes Rev 2008; 9: 615e23 101. Chaput JP, Gray CE, Poitras VJ, et al. Systematic review of the relationships between sleep duration and health indicators in school-aged children and youth. Appl Physiol Nutr Metab 2016; 41: S266e82 102. Poitras VJ, Gray CE, Borghese MM, et al. Systematic review of the relationships between objectively measured physical activity and health indicators in school-aged children and youth. Appl Physiol Nutr Metab 2016; 41: S197e239 103. Wolff JJ, Gu H, Gerig G, et al. Differences in white matter fiber tract development present from 6 to 24 months in infants with autism. Am J Psychiatr 2012; 169: 589e600 104. Roberts JE, Rosenfeld RM, Zeisel SA. Otitis media and speech and language: a meta-analysis of prospective studies. Pediatrics 2004; 113: e238e48 105. Fougner V, Korvel-Hanquist A, Koch A, et al. Early childhood otitis media and later school performanceda prospective cohort study of associations. Int J Pediatr Otorhinolaryngol 2017; 94: 87e94 106. Thelin EP, Jeppsson E, Frostell A, et al. Utility of neuronspecific enolase in traumatic brain injury; relations to S100B levels, outcome, and extracranial injury severity. Crit Care 2016; 20: 285 107. Dadas A, Washington J, Marchi N, Janigro D. Improving the clinical management of traumatic brain injury through the pharmacokinetic modeling of peripheral blood biomarkers. Fluids Barriers CNS 2016; 13: 21 108. Pham N, Fazio V, Cucullo L, et al. Extracranial sources of S100B do not affect serum levels. PLoS One 2010; 5(9): e12691. https://doi.org/10.1371/journal.pone.0012691 109. Steiner J, Schiltz K, Walter M, et al. S100B serum levels are closely correlated with body mass index: an important caveat in neuropsychiatric research. Psychoneuroendocrinology 2010; 35: 321e4
Neurocognitive outcomes in children after general anaesthesia
110. Sorci G, Riuzzi F, Arcuri C, et al. S100B protein in tissue development, repair and regeneration. World J Biol Chem 2013; 4: 1e12 111. Strauss ES, Elisabeth MS, Spreen O. A compendium of neuropsychological tests: administration, norms and commentary. Oxford: Oxford University Press; 2006 112. Cronbach LJ, Meehl PE. Construct validity in psychological tests. Psychol Bull 1955; 52: 281e302 113. Tieman BL, Palisano RJ, Sutlive AC. Assessment of motor development and function in preschool children. Ment Retard Dev Disabil Res Rev 2005; 11: 189e96 114. Wachtel RC, Shapiro BK, Palmer FB, Allen MC, Capute AJ. CAT/CLAMS. A tool for the pediatric evaluation of infants and young children with developmental delay. Clinical Adaptive Test/Clinical Linguistic and Auditory Milestone Scale. Clin Pediatr (Phila) 1994; 33: 410e5 115. Voigt RG, Brown 3rd FR, Fraley JK, et al. Concurrent and predictive validity of the cognitive adaptive test/clinical linguistic and auditory milestone scale (CAT/CLAMS) and the Mental Developmental Index of the Bayley Scales of Infant Development. Clin Pediatr (Phila) 2003; 42: 427e32 116. Button KS, Ioannidis JP, Mokrysz C, et al. Power failure: why small sample size undermines the reliability of neuroscience. Nat Rev Neurosci 2013; 14: 365e76 117. Ioannidis JP. Why most discovered true associations are inflated. Epidemiology 2008; 19: 640e8
-
19
118. Sistino JJ. Attention deficit/hyperactivity disorder after neonatal surgery: review of the pathophysiology and risk factors. Perfusion 2013; 28: 484e94 119. Bevilacqua F, Rava L, Valfre L, et al. Factors affecting short-term neurodevelopmental outcome in children operated on for major congenital anomalies. J Pediatr Surg 2015; 50: 1125e9 120. Katusic SK, Colligan RC, Barbaresi WJ, Schaid DJ, Jacobsen SJ. Potential influence of migration bias in birth cohort studies. Mayo Clin Proc 1998; 73: 1053e61 121. Tools for assessing methodological quality or risk of bias in non-randomized studies. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions; 2011. Version 5.1.0. Available from: www. cochrane-handbook.org. 122. Hartling L, Milne A, Hamm MP, et al. Testing the Newcastle Ottawa Scale showed low reliability between individual reviewers. J Clin Epidemiol 2013; 66: 982e93 123. Zhang H, Du L, Du Z, Jiang H, Han D, Li Q. Association between childhood exposure to single general anesthesia and neurodevelopment: a systematic review and metaanalysis of cohort study. J Anesth 2015; 29: 749e57 124. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010; 25: 603e5
Handling editor: J.G. Hardman
REVIEW ARTICLE
Systematic review of the neurocognitive outcomes used in studies of paediatric anaesthesia neurotoxicity € hler1 and T. G. Hansen1,2 N. G. Clausen1,2,3,*, S. Ka 1
Department of Anaesthesia and Intensive Care, University Hospital Odense, J.B.Winsløwsvej 4, DK-5000 Odense, Denmark, 2University of Southern Denmark, Department of Clinical Research, Winsløwparken 19, DK-5000 Odense, Denmark and 3Children’s Hospital University of Zu ¨ rich, Department of Anaesthesia, Steinwiesstrasse 75, CH-8032 Zu¨rich, Switzerland *Corresponding author. Department of Anaesthesia and Intensive Care, University Hospital Odense, J.B.Winsløwsvej 4, DK-5000 Odense, Denmark. E-mail: [email protected].
Abstract Neurotoxicity of anaesthetics in developing brain cells is well documented in preclinical studies, yet results are conflicting in humans. The use of many and different outcome measures in human studies may contribute to this disagreement. We conducted a systematic review to identify all measures used to assess long-term neurocognitive outcomes following general anaesthesia (GA) and surgery in children. The quality of studies was assessed according to the Newcastle-Ottawa Scale (NOS) for observational studies. PubMed/MEDLINE, EMBASE, Cinahl, Web of Science, and the Cochrane Library were searched for studies investigating neurocognitive outcome after GA in children <18 yr. Sixtyseven studies were identified from 19 countries during 1990e2017. Most assessments were performed within cognition, sensory-motor development, academic achievement or neuropsychological diagnosis. Few studies assessed other outcomes (magnetic resonance imaging, serum-biomarkers, mortality, neurological examination, measurement of head circumference, impairment of vision). Rating according to the NOS rewarded a mean of six stars out of nine. Some concerns prevail regarding potential inter-rater variability because of equivocal description of rating criteria. Specific features such as stability over lifetime and inter-relations of outcomes (e.g. prediction of subsequent development or diagnosis of neuropsychological conditions) are discussed. The importance of validity and reliability of the various test instruments are described. The studies vary immensely in important characteristics. Future observational studies should be more consistent in the choice of study population, age at exposure, follow-up, indication for and type of surgery, and outcomes. Assessment of sensory-motor development seems feasible in young children (age <4 yr), intelligence/cognition in older children. Keywords: anaesthesia; general; child development; infant; review
Editorial decision: November 21, 2017; Accepted: November 21, 2017 © 2017 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved. For Permissions, please email: [email protected]
1
2
-
Clausen et al.
Confronted with the lack of definite answers, researchers continue to investigate whether anaesthetics cause harm to the developing brain in children. Evidence from preclinical studies on various species including non-human primates, a variety of anaesthetic drugs, modes of administration, and durations, substantiate this concern. In humans, an association between general anaesthesia (GA) in young children and subsequent impaired neurodevelopment remains to be properly established or rejected. Single and short exposures lasting <1 h are considered ‘safe’ regarding neurodevelopmental outcome.1 In a recent sibling-matched cohort study it was reported that children exposed once for inguinal hernia repair before age 3 yr showed no difference in intelligence quotient testing compared with their unexposed siblings.2 A Swedish population-based cohort study detected a low overall difference in academic achievement and intelligence after single and multiple exposures before age 4 yr for miscellaneous types of surgeries.3 These findings concur with the results of an international randomised trial comparing neurodevelopmental outcome at age 2 yr between infants exposed to a brief sevoflurane anaesthesia or an awake spinal-/caudal anaesthetic for inguinal herniorrhaphy; no difference was found between groups.4 Quite the opposite, some studies focusing on exposure in 3- or 4-yr-old children to anaesthesia for miscellaneous types of surgery indicate long-term effects on cognition, learning, and behavioural disorders.5,6 As an example, exposure to GA and surgery before age 3 yr was found to associate with a remarkably increased risk of being diagnosed with a developmental/behavioural disorder.7 Interestingly, various authorities interpret the existing results differently. While European scientific societies agree, that at present, there is no human evidence to support any change in anaesthetic practice in children and pregnant women,1 the Food and Drug Administration recently warned against anaesthesia in children <3 yr of age prolonged anaesthesia.8 Such different interpretations might have serious clinical implications for patients (e.g. postponement of surgery). The majority of studies within the field of anaesthesiarelated neurotoxicity are cohort studies, which knowingly are vulnerable to confounding. As a consequence, recent literature calls for the use of other modalities (e.g. neuroimaging and biomarkers), to evaluate neuro-apoptosis and neuro-inflammation.9 Furthermore, other factors that might mitigate morbidity after GA and surgery in children are increasingly acknowledged and subjected to intensive research.10e13 Despite this broadened search for answers, anaesthesia-related neurotoxicity has not yet been properly clinically defined. Consequently, it is unknown where clinical research should be focused. Multiple factors, for example, sample size; selection of exposed cohort; age at exposure; type of, and indication for, surgery; follow-up as well as type of outcome measures could likely influence results. We hypothesised that inconsistency in the choice of outcome measures, age at exposure, and differences in study quality are major confounding factors for divergent results obtained in human studies. The aims of this systematic review were to identify the type of outcomes used in studies investigating neurodevelopmental consequences of anaesthesia/surgery exposure in infancy and early childhood. Further, we wished to compare the studies with regard to other crucial factors and assess the methodological quality of these studies.
Methods The study protocol was published at PROSPERO (registration ID: CRD42016042450) and reported according to the PRISMA guidelines.14
Search strategy We searched PubMed/MEDLINE, EMBASE, Cinahl, Web of Science and the Cochrane Library (last search on June 16, 2017) using relevant terms concerning ‘general anaesthesia’ and ‘neurocognitive outcome’. Via the MESH database and the EMTREE thesaurus we identified relevant search terms. Additionally, the reference list of included studies was hand-checked for other potentially relevant publications. ‘Neurobehavioral outcome’ was part of the search. However, most articles on postoperative behaviour assessed temporary changes (e.g. emergency delirium and anxiety). These were not considered to be long-term results of anaesthesia-related neurotoxicity and articles using only behaviour as outcomes were not included in the content analyses. Librarians at the Medical Research Library, part of The University Library of Southern Denmark, Odense University Hospital, Denmark approved the construction of the final search. The exact search for the respective databases is available as Supplementary Material (Supplementary 1).
Study selection To be eligible for this review studies must report on: Age at exposure <18 yr Exposure to single/multiple GA or surgery (it is assumed, that surgery is only delivered with a concomitant GA) Evaluation of cognitive function after the exposure(s) Studies must be published in peer reviewed journals, identified by experts within the field, or have been cited by peer reviewed papers. We excluded studies that were not reported in English, evaluated neurotoxic effects of topically/locally administered anaesthetics, or studies conducted in animals. Two authors (N.G.C. and S.K.) independently assessed title, abstract and full text for eligibility using the Covidence software (Covidence systematic review software, Veritas Health Innovation, Melbourne Australia. Available at https://www. covidence.org).
Outcomes We identified all outcome measures used in studies investigating neurocognitive consequences of surgery and GA exposure in children (primary outcome). Further, the quality of studies regarding selection and comparability of study participants, as well as ascertainment of outcomes were evaluated based on the Newcastle-Ottawa Scale (NOS).15
Data extraction We recorded study ID, design, reason for anaesthesia (type of surgery or diagnostics), and specifics on anaesthetic procedure (type, drugs, doses, means of administration) if provided. We also recorded information on the population studied (study base, number of exposed/non-exposed individuals), age at exposure, male/female distribution, type of outcome, and age at assessment.
Neurocognitive outcomes in children after general anaesthesia
-
3
Fig 1. Illustration of selection process.
Quality assessment
Results
Study quality was evaluated using the NOS for quality assessment of non-randomised studies. The evaluation comprised three broad perspectives: selection of study groups, comparability of groups, and ascertainment of either exposure or outcome of interest. The scale uses a ‘star system’; each study can be accredited a maximum of nine stars if all requirements are met. Risk of bias evaluations of individual studies can be retrieved by contacting the corresponding author.
Study selection Our systematic literature review identified 18 011 titles after removal of duplicates. N.G.C. and S.K. independently screened titles and abstracts, eliminating irrelevant and double items. A total of 397 titles qualified for full text screening and finally, 67 studies were included in this review.2e7,16e75 The main reasons for excluding studies were: ‘Wrong intervention’: comparison of interventions other than or not related to anaesthesia
4
-
Clausen et al.
Table 1 List of different neurocognitive and sensorimotor function outcomes used to assess outcome in children exposed to general anaesthesia (alphabetic order) and the domain to which the test was assigned. *Individualised Educational Program (IEP) is not a psychometric test but the intervention applied for a child with special needs. The IEP is meant to address each child’s unique learning issues and include specific educational goals. It is a legally binding document, sanctioned by the United States federal law called the Individuals with Disabilities (IDEA). The IEP describes the goals the team (e.g. teachers, parents, school psychologists) sets for a child during the school year, as well as any special support needed to help achieve them. yFunctional Independence Measure is used in follow-up of rehabilitation Psychometric test ABAS-II AIMS e AVLT ASQ BDS BRIEF BSID-II BSID-III CAT CBCL CDI CELF CELF-E
Domain of testing Adaptive Behavior Assessment Systemd2nd Edition Alberta Infant Motor Scale Albert Einstein College of Medicine Neonatal Neurobehavioral Assessment Scale Rey Auditory Verbal Learning Test Ages & Stages Questionnaire Backward digit span test Behavior Rating Inventory of Executive Functions Bayley Scales of Infant and Toddler Developmentd2nd Edition Bayley Scales of Infant and Toddler Developmentd3rd Edition California Achievement Test Child Behavior Checklist Child Depression Inventory
DSM-ADH
Clinical Evaluation of Language Fundamentals Clinical Evaluation of Language Fundamentalsdexpressive language score Clinical Evaluation of Language Fundamentalsdreceptive language score Clinical Evaluation of Language Fundamentalsdtotal language ability The Child Health Questionnaire 50 Raven’s Colored Progressive Matrices Continuous Performance Test II Conners’ teacher Rating ScaledRevised California Verbal Learning TestdChildren Delis-Kaplan Executive Function Systems/Trail Making Subtests Diagnostic and Statistical Manual of Mental Disordersd4th Edition DSMdattention deficit hyperactivity scores
EDI FDS FSIQ G-TVPS GDQ GDS GMDS GMFCS HAWIVA-III GPT ICD-9
Early Development Instrument Forward digit span test Full scale intelligence-quotient Gardner Test of Visual-Perceptual Skills Revised General Developmental Quotient Gesell Developmental Schedule Griffiths Mental Development Scale General Motor Function Classification Score HannovereWechsler Intelligence Scale, 3rd Edition Grooved Pegboard Test International classification of Diseased9th Edition
ICD-9-CM 299.00
International Classification of Diseased9th Autistic Disorder Diagnoses International Classification of Diseases 9thdAttention Deficit Hyperactivity Disorder Individualised Educational ProgramdDisorders of Emotion and Behaviour Individualised Educational ProgramdSpeech and Language Kaufmann Assessment Battery for Children Kognitiver Enwicklungstest fu¨r das Kindergartenalter Learning Disability McCarron Assessment of Neuromuscular Development
CELF-R CELF-T CHQ50 CPM CPT-II CTRS-R CVLT-C DKEFS DSM-IV
ICD-9-CM 314.01 IEP-EBD IEP-SL K-ABC KET-KID LD MAND
Development Development Development Intelligence/cognition Development Intelligence/cognition Development Development intelligence Development intelligence Academic achievement Development Screening/diagnosis (psychiatric disorder) Development Development Development Development Development Intelligence/cognition Development Development Academic achievement Academic achievement Screening/diagnosis (psychiatric disorder) Screening/diagnosis (psychiatric disorder) Development Intelligence/cognition Intelligence/cognition Development Development Development Development Development Intelligence/cognition Intelligence/cognition Screening/diagnosis (psychiatric disorder) Screening/diagnosis (psychiatric disorder) Screening/diagnosis (psychiatric disorder) Academic achievement* Academic achievement* Intelligence/cognition Intelligence/cognition Screening/diagnosis Development
Continued
Neurocognitive outcomes in children after general anaesthesia
-
5
Table 1 Continued Psychometric test MacArthur-Bates CDI MRI NEPSY NEPSY-2-NL OLSAT OWLS PDMS e PIQ PPVT RDLS PSLE SDMT SON-R e SB-5 TCS TIQ TEA-Ch NL TMTdA VABS VIQ VMI Wallin BP WAMSE WASI WJ III WJ IIIdVisual Matching WeeFIM WPPSI-R WISC-III WISC-III-NL WISC-IV
Domain of testing McArthur Bates Communicative Development Inventory Magnetic resonance imaging Developmental NEuroPSYchological Assessment NDevelopmental Neuropsychological Assessment Batteryd2nd Edition, Dutch version OtiseLennon School Ability test Oral and Written Language Scales Peabody Developmental Motor Scales Phonemic verbal fluency test
Development Screening/diagnosis (somatic disorder) Academic achievement Development Cognition Academic achievement Development Intelligence/cognition
Performance Intelligence-Quotient Peabody Picture Vocabulary Test Reynell Developmental Language Scales Primary School Leaving Examination Symbol Digit Modality Test Semantic verbal fluency test Hogrefe/Snijders-Oomen Non-Verbal Intelligence TestdRevised Stanford
Intelligence/cognition Academic achievement Development Academic achievement Development Intelligence/cognition Intelligence/cognition
StanfordeBinet Intelligence Scalesd5th Edition Stroop Color and Word Test Total Cognitive Skills Total Intelligence Quotient Test of Everyday Attention for Children, Dutch version Trail Making Testdpart A The Vineland Adaptive Behavior Scale, 2nd Edition Verbal Intelligence-Quotient BeeryeBuktenica Developmental Test of Visual Motor Integration, 5th Edition Wallin B pegboard Western Australian Monitoring Standards in Education Wechsler Abbreviated Scale of Intelligence WoodcockeJohnson III WoodcockeJohnson testdVisual Matching Functional Independence Measure Wechsler Preschool and Primary Scales for Intelligence: Revised Wechsler Intelligence Scale for Childrend3rd Version Wechsler Intelligence Scale for Childrend3rd Version, Dutch version Wechsler Intelligence Scale for Childrend4th Edition
Intelligence/cognition Intelligence/cognition Intelligence/cognition Intelligence/cognition Intelligence/cognition
Academic achievement
Intelligence/cognition Development Intelligence/cognition Intelligence/cognition Development Academic achievement Intelligence/cognition Academic achievement Academic achievement Developmenty Intelligence/cognition Intelligence/cognition Intelligence/cognition Intelligence/cognition
Non-psychometric tests/means to assess cognitive, developmental outcome, or both, of children exposed to general anaesthesia Serum biomarkers of Brain derived neurotrophic factor (BDNF), S100B Neurological examination Surrogate measures of visual function (visual acuity, refractive error, thickness of retinal nerve fibre layer)
‘Wrong outcomes’: for example, emergency delirium shortly after procedure (i.e. within 30 days following exposure or temporary behavioural outcomes). One article was excluded, because these results were based on re-analysis of a previously described dataset.76 The process of selection is illustrated in Figure 1.
Neuron damage Neurologic deficit Sensorineural deficit
Study outcomes A variety of psychometric tests were used (Table 1). For simplicity, tests were allocated to four core domains: intelligence/cognition, development, achievement, and screening/ diagnosis. In the majority of studies, more than one test was used, and often more than one domain was tested.
6
Table 2 Summarised characteristics of studies included in the current review.*MRI scan at 7 days, BSID-III at 12 months of age;**MRI scan at 39e42 weeks (corrected age), BSID-II at 2 yr
-
Type of outcome
<4
16 yr
School grades Intelligence scores
Yes
<1
?
Developmental delay Psychiatric diagnoses
Yes
<3 yr
14e18 yr
Diagnoses Academic performance (single/group)
No
<7 yr
16 yr (mean)
Cognition (VIQ and WAIS) MRI scan Height þ head circumference GDMS Sensorineural deficits Others (weight, length, head circumference) School performance Psychiatric diagnoses
Study design
Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at exposure controls (month, yr) Yes/No
Glatz and colleagues3
Sweden 2017
Nationwide cohort study
No
No
Yes
Nestor and colleagues16
USA 2017
Retrospective single Yes centre cohort study
No
Hu and colleagues17 USA 2017
Retrospective birth- Yes cohort study
No
n¼33 512 (single exp.) n¼37 231 (multiple exp.) n¼63 (no surgery) n¼171 (1 surgery) n¼64 (>1 surgery) n¼465 (no surgery) n¼466 (1 surgery) n¼126 (>1 surgery) n¼48 (CLP) n¼19 (CP) n¼20 (CL)
Conrad and colleagues18
USA 2017
Retrospective birth cohort study
No
No
Birajdar and colleagues19
Australia 2017
Retrospective cohort study
Yes
Yes
n¼33
No
<5 days (mean)
12 months
Terushkin and colleagues20
USA 2017
Yes
Yes
n¼33
No
<4 yr
Variable (3.4e11.4 yr)
Lap and colleagues21
Netherlands 2017
Retrospective institutional cohort study Retrospective cohort study
No
Yes
n¼19
Yes
Newborn
Retrospective cohort study Multicentre, sibling matched cohort study
No
No
n¼127
No
<3 yr
Median exp/controls: 8.5/6.0yr 20 yr
Yes
Yes
n¼116
Yes
<3 yr
8e15 yr
Observational cohort study Randomised controlled trial
Yes
No
n¼21
No
<1 month
<7 yr
Yes
Yes
n¼358 (GA group) n¼361 (RA group)
No
<60 weeks (corrected age)
2 yr
Yazar and Australia 2016 colleagues22 2 Sun and colleagues USA 2016
Seltzer and colleagues23 Davidson and colleagues4
USA 2016 Australia 2016
Cognition (IQ, memory) Development (motor, vision, behaviour) Three surrogate measures of visual function IQ Tests of domain specific neurocognitive functions and behaviour Vineland-II BSID-III McArthureBates
Continued
Clausen et al.
Age at outcome assessment
Country of origin and publication yr
Study ID
Table 2 Continued Study ID
Study design
Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at controls exposure Yes/No (month, yr)
Age at outcome assessment
Type of outcome
O’Leary and colleagues24
Canada 2016
No
No
n¼28 366
Yes
Before age 5.7 yr
Denmark 2016
No
Yes
n¼549
Yes
Mean age 9.1 yr
5.7 yr (mean age at EDI) 15e16 yr
EDI
Djurhuus and colleagues25
Population-based, retrospective cohort study Retrospective nationwide cohort study
Clausen and colleagues75
Denmark 2016
Retrospective nationwide cohort study
No
No
n¼558
Yes
2 months to 2 yr
Poor Zamany Nejat Kermany and colleagues26
Iran 2016
Retrospective cohort study
Yes
Yes
n¼68
No
<3 yr
Hoffman and colleagues27 Hansen and colleagues28 Graham and colleagues29
USA 2016
No
No
n¼178
No
Variable
School grades non-attainment rates at national school exam 15e16 yr School grades non-attainment rates at national school exam 6 months after PVF test final exposure SVF test FDS test BDS test Variable BSID-III
No
Yes
n¼71
No
<1 month
<6 months
No
No
n¼4479
Yes
<4 yr
<6 yr
HAWIVA-III KETKID EDI scores
No
No
n¼40
Yes
<7days
2 months
BSID-II
Yes
Yes
n¼96
No
<12 months
4e5 yr
WPPSI-III
No
No
n¼415
Yes
<5 yr
6 yr
IQetest
Yes
No
n¼118
Yes
5e12 yr
Retrospective Yes cohort study Bidirectional cohort Yes study Prospective follow- No up study (single centre)
No
n¼35
Yes
1.5e5 yr
No
n¼15
Yes
0e2 yr
1 day þ 6 weeks Tests of domain specific postoperative neurocognitive functions (perception, memory, comprehension) Not stated CBCL DSM-IV 10e17 yr MRI-imaging
No
n¼21
Yes
<1 yr
34e39 months
Doberschuetz and colleagues30 Diaz and colleagues31 de Heer and colleagues32 Aun and colleagues33
Bakri and colleagues34 Taghon and colleagues35 Naguib and colleagues36
Observational cohort study Germany 2016 Observational cohort study Canada 2016 Retrospective matched cohort study Germany 2016 Ambi-directional matched cohort study USA 2016 Subgroup analysis as part of prospective study Netherlands Retrospective 2016 cohort study Hong Kong 2016 Prospective cohort study
Saudi Arabia 2015 USA 2015 USA 2015
IQ test Psychiatric diagnosis
-
Continued
Neurocognitive outcomes in children after general anaesthesia
Country of origin and publication yr
7
Table 2 Continued
8
Study ID
Hansen and colleagues38
Denmark 2015
Gano and colleagues39
USA 2015
Backeljauw and colleagues40
USA 2015
ckova and Petra colleagues41 Yan and colleagues42 Yin and colleagues43
Czech Republic 2015 China 2014
Morriss and colleagues44
USA 2014
China 2014
Ing and colleagues45 Australia 2014
Gaynor and colleagues46
USA 2014
Garcia Guerra and colleagues47 Cheng and colleagues48 Andropoulos and colleagues49 Ko and colleagues50
Canada 2014
Minutillo and colleagues51 Hansen and colleagues52
Australia 2013
USA & Canada 2014 USA 2014 Taiwan 2014
Denmark 2013
Type of outcome
<2 yr
>2 yr
Psychiatric diagnosis
Yes
Not specified
15e16 yr
Mortality School grades
n¼43
No
3e6 yr
WPPSI-III
No
n¼53
Yes
<42 weeks gestational age 1 day to 3.8 yr
Yes
Yes
n¼21
No*
8 days
Variable (128 þ / 44 months) 3e7 yr
Yes
Yes
n¼49
No
3e22 months
WPPSI-III OWLS MRI CHQ 50 test SON-R 2.5e7 yr test BSDI-II
Yes
Yes
n¼60
No*
7e13 yr
No
No
n¼2970
Yes
Variable
Retrospective cohort study
No
No
n¼321
Yes
<3 yr
Secondary analysis of data from prospective study Prospective followup study Prospective cohort study Retrospective cohort study Nationwide retrospective matched cohort study Retrospective cohort study Nationwide register-based follow-up study
No
No
n¼365
No
<6 months
Yes
Yes
n¼135
No
<6 weeks
54 months
No
No
n¼124
No
<3 yr
1 yr
Yes
No
n¼93
No
<30 days
No
No
n¼3293
Yes
<3 yr
7 days and 12 months* Variable
WPPSI-III FSIQ PIQ VIQ VMI BSID-II MRI5-imaging BSID-III MRI-imaging Psychiatric diagnosis
No
Yes
n¼82
Yes
<1 months
12 months
GMDS
No
Yes
n¼779
Yes
<3 months
15e16 yr
School grades
Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at exposure controls (month, yr) Yes/No
Nationwide retrospective matched cohort study Nationwide register-based follow-up study Cohort study
No
No
n¼5197
Yes
No
No
n¼228
Yes
No
Retrospective matched cohort study Prospective cohort study Prospective cohort study Observational prospective study (single centre) Retrospective cohort study
Yes
3 days after final operation 7þ90 days after Wechsler Memory scale exposure egraphic, verbal, visual, acoustic components 18e22 months BSID-III BSID-II GMFCS 10 yr SDMT, CPM MAND, CELF CELF-R, CELF-E CELF-T, PPVT 4e5 yr WPPSI-III School grades
Continued
Clausen et al.
Ko and colleagues37 Taiwan 2015
Age at outcome assessment
Study design
-
Country of origin and publication yr
Table 2 Continued Age at outcome assessment
Type of outcome
1 week and 1, 3, 6 months after surgery 12 yr
WPPSI-III
Country of origin and publication yr
Study design
Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at exposure controls (month, yr) Yes/No
Fan and colleagues53
China 2013
Yes
No
n¼100
No
<4 yr
Bong and colleagues54
Singapore 2013
Yes
No
n¼100
Yes
<1 yr
Yang and colleagues55 Andropoulos and colleagues56
South Korea 2012 USA 2012
Yes
Yes
n¼21
No
5e10 yr
Yes
Yes
n¼30
No
<1 month
Walker and colleagues57
Australia 2012
No
No
n¼300
Yes
<3 months
12 months
Sprung and colleagues58 Sananes and colleagues59
USA 2012
Single centre observational study Retrospective observational cohort study Prospective cohort study Prospective, observational cohort study Prospective population-based cohort study Retrospective cohort study Prospective followup study
Yes
No
n¼350
Yes
<2 yr
No
No
n¼131
No
<3 months
Rocha and colleagues60
Portugal 2012
No
Yes
n¼39
No
<1 month
Long and colleagues61
Australia 2012
No
No
n¼195
No
<8 weeks
24 months
BSID-III
Long and colleagues62
Australia 2012
No
No
n¼195
No
<8 weeks
4, 8, 12, 16 months
Alberta Infant Motor Scale
Ing and colleagues5
Australia 2012
Descriptive followup study (single centre) Prospective, longitudinal cohort study Prospective longitudinal cohort study Retrospective cohort study
Variable Psychiatric diagnosis (<19 yr) 8, 12, 24 months PDMS BSID-II ASQ Variable GMDS
No
No
n¼321
Yes
<3 yr
10 yr
Filan and colleagues63
Australia 2012
Observational cohort study
No
No
n¼30
Yes
Hansen and colleagues64
Denmark 2011
Yes
n¼2689
Yes
39e42 weeks (corrected age) and 2 yr** 15e16 yr School grades
Flick and colleagues65
USA 2011
No
n¼350
Yes
<2 yr
Various
Learning disability
DiMaggio and colleagues7
USA 2011
Nationwide No register-based follow-up study Population based Yes matched cohort study No Retrospective sibling-matched birth cohort study
26e41 weeks (corrected age) <1 yr
SDMT, CPM MAND, CELF CELF-R, CELF-E CELF-T, PPVT MRI imaging, BSID-II
No
n¼5284
Yes
<3 yr
z10 months
Psychiatric diagnosis
Study ID
1 day þ 1 month K-ABC postoperative 7 days þ 12 MRI months* BSID-III BSID-III
Neurocognitive outcomes in children after general anaesthesia
Canada 2012
PSLE
9
Continued
Table 2 Continued Anaesthesia protocol stated Yes/No
Single Number of type of exposed surgery Yes/No
Non-exposed Age at exposure controls (month, yr) Yes/No
Age at outcome assessment
Type of outcome
Walker and colleagues57 Fan and colleagues66
Australia 2010
Population based cohort study Prospective study
No
Yes
n¼52
Yes
10e77 days
1 yr
BSID-III
Yes
No
n¼100
No
<4 yr
GDS
DiMaggio and colleagues67 Wilder and colleagues6
USA 2009
No
Yes
n¼383
Yes
12e48 months
1 week and 1, 3, 6 months postoperative Variable
Psychiatric diagnosis
Yes
No
n¼593
Yes
<4 yr
Variable
Learning disability
Majnemer and colleagues68
Canada 2009
No
No
n¼131
No
<6 months
Variable
Bartels and colleagues69
Netherlands 2009
No
No
n¼1143
Yes
Variable
12 yr
Various neurobehavioral and neurocognitive scales School grades
No
No
n¼368
No
Variable
<19 yr
CBCL
Yes
No
n¼23
No
4e18 yr
Variable
No
No
n¼54
Yes
Prospective No longitudinal study Prospective No longitudinal study
No
n¼30
Yes
2, 5, 8 yr Variable (any time during first hospitalisation) <30 days 3 yr
DSM-IV CBCL WPPSI-R Sensory function
No
n¼30
Yes
<30 days
USA 2009
Netherlands Kalkman and 2009 colleagues70 Kayaalp and Turkey 2006 colleagues71 The Victorian Infant Australia 1996 Collaboration72
Ludman and colleagues73 Ludman and colleagues74
England 1993 England 1990
Retrospective cohort study Retrospective, population based cohort study Prospective study
Register based national cohort study Retrospective cohort study Retrospective cohort study Prospective, longterm follow-up study
6 þ12 months
GMDS GMDS
yAbbreviations ASQ, Ages and Stages Questionnaire; BDS, Backward digit span test; BSID-II, Bayley Scales of Infant and Toddler Development, 2nd Edition, Bayley Scales of Infant Development; BSID-III, Bayley Scales of Infant and Toddler Development, 3rd Edition; CBCL, Child Behaviour Checklist; CELF, Clinical Evaluation of Language Fundamentals; CELF-E, Clinical Evaluation of Language Fundamentalsdexpressive language score; CELF-R, Clinical Evaluation of Language Fundamentalsdreceptive language score; CELF-T, Clinical Evaluation of Language Fundamentalsdtotal language ability; CHQ 50, The Child Health Questionnaire; CPM, Raven’s Coloured Progressive Matrices; DSM-IV Diagnostic and Statistical Manual of Mental Disorders, 4th Edition; EDI, Early Developmental Instrument; FDS, Forward digit span test; FSIQ, Full Scale Intelligence-Quotient; GDMS, General Decision Making Scale; GDS, Gesell Developmental Schedule; GMDS, Griffith’s Mental Development Scale; GMFCS, General Motor Function Classification Score; HAWIVA-III, Hannover-Wechsler Intelligence Scale, 3rd edition; IQ, Intelligence Quotient; K-ABC, Kaufmann Assessment Battery for Children (Korean Version); KET-KID, Kognitiver Entwicklungstest fu ¨ r das Kindergartenalter; MAND, McCarron Assessment of Neuromuscular Development; MRI, Magnetic Resonance Imaging; OWLS, Oral and Written Language; PDMS, Peabody Developmental Motor Scales; PIQ, Performance Intelligence-Quotient; PPVT, Peabody Picture Vocabulary Test; PSLE, Primary School Leaving Examination; PVF, Phonemic Verbal Fluency test; SDMT, Symbol Digit Modality Test; SON-R, 2.5e7yr test Hogrefe/Snijders-Oomen Non-Verbal Intelligence Test; SVF, Semantic Verbal Fluency test; VIQ, Verbal Intelligence-Quotient; WAIS Wechsler Adult Intelligence Scale; VMI Beery-Buktenica Developmental Test of Visual Motor Integration, 5th Edition; WPPSI-III, Wechsler Preschool and Primary Scale of Intelligence, 3rd Edition; WPPSI-R, Wechsler Preschool and Primary Scales for Intelligence: revised.
Clausen et al.
China 2010
-
Study design
10
Country of origin and publication yr
Study ID
Neurocognitive outcomes in children after general anaesthesia
‘Development’ was assessed in 35 studies. One-third of these used the Bayley Scales of Infant Development (BSID) (2nd or 3rd edition) or parts hereof. Tests evaluating ‘intelligence/ cognition’ were used in 21 studies, with seven of these being the Wechsler Preschool and Primary Scale of Intelligence. In 15 studies assessing ‘academics/achievement’, this was either the sole measure or the primary outcome. The domain of ‘diagnosis/screening’, covering neuropsychological disorders and learning disability (LD), was used as an outcome measure in 14 studies. Magnetic resonance imaging (MRI) was performed in seven studies. Eight studies assessed miscellaneous outcomes including: biomarkers, mortality, neurological examination, measurement of head circumference or impairment of vision: visual acuity, refractive error, and thickness of retinal nerve fibre layer. The main features of included studies are summarised in Table 2. Complete study characteristics (study identification, design, anaesthesia protocol, type of surgical/diagnostic procedure, cohort selection, number of exposed/non-exposed individuals, age at exposure, gender, age at follow-up, and outcome measures) are presented as Supplementary Material (Supplementary 2). Overall, studies originated from 19 different countries published from 1990 to 2017. The annual publication rate has increased over time, ranging from one publication in 1990 to 15 in 2016. The prevalent study designs were retrospective and prospective observational studies. Number of exposed individuals ranged from n¼15 to n¼37 231 (mean n¼1454). In more than half of the studies (38 out of 67) the outcome was compared between exposed and unexposed individuals. Anaesthetic drugs and procedures were specified in 28 (42%) studies. Outcome was assessed at one point in time in 53 (79%) studies. Age at exposure and outcome assessment ranged from less than 30 days to age 20 yr, respectively (mean¼37 months). For the youngest infants, age at exposure seemed to cluster around 1e3 months (Fig. 2). In many studies investigating outcome later than 2 yr of age, infants and toddlers were around 36e48 months old at time of exposure (Fig. 3).
Quality assessment Risk of bias evaluation according to NOS assessment tool for cohort studies is illustrated in Table 3. Overall, studies were rewarded three to eight stars (mean, six) out of a maximum of nine stars. As a result of a different design, one study was rewarded a ‘low risk of bias’ in three out of four domains according to the Cochrane risk of bias tool for randomised controlled trials.4 Exposed individuals are representative of the particular community in 57 of the studies. According to the NOS manual, ‘community’ is defined as the study population (of exposed individuals).15 Twenty-nine studies (43%) did not evaluate an outcome in non-exposed controls, which introduced selection bias. If ‘ascertainment of exposure’ was based on a civil register, this was judged of equal quality as hospital files or journals, even though the data were collected for administrative purposes.77 In total, 55 studies (82%) were labelled to have a low risk of bias in this category. In less than half of the studies (n¼22) the relevant outcome was measured before exposure. This is not considered an increased risk of bias if surgery and GA was conducted in early infancy as testing before exposure would have been nonfeasible.
-
11
Fig 2. Age at exposure <2 yr of age (mean¼5.2 months).
Fig 3. Age at exposure >2 yr of age (mean¼70 months).
As a result of the observational, prospective, or retrospective design of most studies internal validity could not easily be optimised during the design phase. Naturally, most studies adjust for potential confounding factors using stratification of relevant groups, propensity matching, or regression analysis. A majority of studies detected outcome by independent assessment or linkage of (hospital) records. Only few blinded their assessors. As 29 (43%) studies did not include nonexposed controls, blinding would not have added any additional value and ‘non-blinding’ is not considered a significant risk of bias in these studies. Whether length to follow-up was adequate, was evaluated individually in each study depending on the outcome measure. For instance, in studies assessing academic achievement at age 15e16 yr following exposure before age 4 yr, this was considered ‘adequate’, based on the assumption that adolescents possess all cognitive skills necessary to sit a standardised test. Additionally, as the BSID is validated for young ages, assessment of infants using this scale was considered adequate. This domain was rewarded a ‘low risk of bias’ in 53 (79%) studies. In comparison, only 37 (55%) studies were rewarded a star for ‘adequacy of follow-up’; either studies followed up on 70% or less of exposed individuals or did not provide a description of those lost.
Discussion Numerous studies have investigated long-term cognitive consequences of GA, surgery, or both, in neonates and infants. Judged by the increasing numbers of papers published each year, the interest in this topic is growing, presumably because
12
-
Clausen et al.
Table 3 Risk of bias evaluation; A-I represents rating categories according to Newcastla-Ottawa Scale for assessing quality of non-randomised studies. A, Representativeness of the exposed cohort; B, Selection of the non-exposed cohort; C, Ascertainment of exposure; D, Outcome of interest was not present at start of study; E, Comparability of cohorts - major factor controlled for; F, Comparability of cohorts - any additional factor controlled for; G, Assessment of outcome; H, Follow-up long enough for outcome to occur? I, Adequacy of follow-up of cohorts. Study ID 3
Glatz and colleagues Nestor and colleagues16 Hu and colleagues17 Conrad and colleagues18 Birajdar and colleagues19 Terushkin and colleagues20 Lap and colleagues21 Yazar and colleagues22 Sun and colleagues2 Seltzer and colleagues23 Davidson and colleagues4 O’Leary and colleagues24 Djurhuus and colleagues25 Clausen and colleagues75 Poor Zamany Nejat Kermany and colleagues26 Hoffman and colleagues27 Hansen and colleagues28 Graham and colleagues29 Doberschuetz and colleagues30 Diaz and colleagues31 de Heer and colleagues32 Aun and colleagues33 Bakri and colleagues34 Taghon and colleagues35 Naguib and colleagues36 Ko and colleagues37 Hansen and colleagues38 Gano and colleagues39 Backeljauw and colleagues40 ckova and colleagues41 Petra Yan and colleagues42 Yin and colleagues43 Morriss and colleagues44 Ing and colleagues45 Gaynor and colleagues46 Garcia Guerra and colleagues47 Cheng and colleagues48 Andropoulos and colleagues49 Ko and colleagues50 Minutillo and colleagues51 Hansen and colleagues52 Fan and colleagues53 Bong and colleagues54 Yang and colleagues55 Andropoulos and colleagues56 Walker and colleagues57 Sprung and colleagues58 Sananes and colleagues59 Rocha and colleagues60 Long and colleagues61 Long and colleagues62 Ing and colleagues5 Filan and colleagues63
A B
C D E
F
G H I
+ + + + + + + + +
+ + + + + +
+ + + + + + + + +
+ +
+ + + + + + + + +
+ + + + + + +
+ + + + + + +
+ + + + + + + +
+ + + + + +
+ + + +
+ + + +
+ + + +
+ + +
+ + + +
+ + +
+ + +
+ + + +
+ + +
+ +
+ + +
+ + +
+ + + +
+ + + +
+ + +
+ +
+ + + + + + + + + + + + +
+ + + + + + + + + + + +
+ + + + + + + + + + + + + + +
+ + + + + +
+ + + + + + + + +
+ + + + + + + + + + + + + + +
+ + + + + + + + + +
+ + + + + + + + + + + +
+ + + + + + + +
+ + + + + + + + + + + + + + + + + + +
+ + +
+ + + + + + +
+ + + +
+ + +
+ + + + + +
+ + + + +
+ + + + +
+ + + +
+ + + + + + +
+ + + +
+ + + + + + +
+ +
+ + + + + +
+ + + + + +
+ + + + + + + +
+ + +
Continued
Table 3 Continued Study ID
A B
C D E
F
G H I
Hansen and colleagues64 Flick and colleagues65 DiMaggio and colleagues7 Walker and colleagues57 Fan and colleagues66 DiMaggio and colleagues67 Wilder and colleagues6 Majnemer and colleagues68 Bartels and colleagues69 Kalkman and colleagues70 Kayaalp and colleagues71 The Victorian Infant Collaboration72 Ludman and colleagues73 Ludman and colleagues74 Representativeness of the exposed cohort Selection of the non-exposed cohort Ascertainment of exposure Outcome of interest was not present at start of study Comparability of cohortsdmajor factor controlled for Comparability of cohortsdany additional factor controlled for Assessment of outcome Follow-up long enough for outcomes to occur? Adequacy of follow-up of cohorts
+ + + + + + + + + + + +
+ + + + + + + +
+ + + + + + + + + +
+ + + + + + + +
+ + + + + + + + +
+ + + + + + + +
+ + + + + + +
+ + + + + + + + + +
+ + + + + + +
+ + + + + + + + + + + A B C D
E
F
G H I
of the lack of a robust and reproducible phenotype. Investigators have used a variety of specific assessment tools, or parts hereof, to evaluate different criteria within development, cognition, intelligence, academic achievement, and neuropsychological diagnosis. Comparatively, few studies have assessed outcomes such as neuroimaging,18,35,48,49,56,63 mortality, or serum-biomarkers.18,19,22,38,39,42,44,48 Previously, it has been questioned which criteria are suitable outcome measures in studies on long-term neurocognitive consequences following GA and surgery.78 So far, epidemiologic studies have shown contradictory results. Register-based studies are unable to demonstrate any causative associations. Although this is largely made up for by the opportunity to analyse large sample sizes, epidemiologic studies have been criticised and the call for randomised clinical trials has been repetitively stated. Based on the findings in this review, this criticism may not be entirely reasonable, as age at exposure has not been assessed systematically. Rather, focus has been on infants or young children between 1 and 3 months or 2 and 4 yr, respectively. This clustering can probably be explained by the assumption that development of the infant brain is most vulnerable to potential toxins during peak synaptogenesis of neurons (i.e. period from 3rd trimester to 12e24 months of age).79 In comparison, the effect of GA exposure during age 3 months to 2 yr or >4 yr has been investigated scarcely. As brain development in young children is dynamic and on-going, associations in other domains might have been missed. Furthermore, it may be reasonable to assume, that infants
Neurocognitive outcomes in children after general anaesthesia
undergoing surgery in the first months of life suffer from significant underlying conditions, which might confound the effect of exposure to GA on cognitive outcome. The studies seem to resemble each other regarding design (i.e. retrospective cohorts), but vary immensely in all other characteristics highlighted in this review (type of, and indication for, surgery; sample size; age at exposure and follow-up; as well as outcome measures).
Cognition and neurodevelopment Intelligence, covering several cognitive domains (memory, reasoning, spatial ability, executive function, and processing speed), is considered a relatively stable and constant trait throughout life.80 Cognition in male adolescents was tested in the Vietnam Experience Study, a cohort of soldiers involved in military service abroad,81 and these individuals were reassessed at middle-age. A high correlation between scores in young- and middle-aged individuals was found. In a Scottish birth cohort, intelligence was assessed in 11-yr-old children who were followed-up 66 yr later. It was concluded that mental abilities are stable across a human lifespan unless a person suffers severe diseases.80 Similar findings have been confirmed in a multitude of similar longitudinal studies regarding intelligence. Assessment of development has focused on the domains of language, perception, and motor development. It has long been recognised that infant neuro-behaviour follows an age specific pattern. In 1931, Gesell82 described the ‘Developmental morphology of infant behaviour pattern’ as ‘significant consistencies in the tempo, the forms and sequences of behaviour growth which point to the presence of a lawful developmental mechanic’. Unlike intelligence, abilities within these domains change over time. Development within the norm is evaluated based on whether the child reaches predefined ‘milestones’. Deviations from or delay in progression might be as a result of underlying disease or external factors.83 Interestingly, development and overall cognitive ability has been shown to correlate. In a 34-yr follow-up of a sample of 307 children identified in the Copenhagen Perinatal Cohort comprising 9125 deliveries from late 1959 to 1961, an association between several milestones of motor development and intelligence in adulthood was demonstrated.84 In a French mother-child cohort, it was shown that milestones before age 2 yr predicted intelligence in 5e6-yr-old children.85 In summary, sensorimotor development relates well to cognition/intelligence and psychiatric disorders. Further, as milestones are well defined, development can be assessed even in the youngest ages. At the same time, development can be influenced both positively and negatively by external factors (e.g. by interaction or with parents or severe diseases necessitating frequent treatment or hospitalisation, xenobiotics, etc.), which compromises the stability of development as an outcome measure. Intelligence is a stable factor across lifetime, but assessment is not feasible until the child has achieved some basic cognitive skills. Hence, assessing children younger than school age is not recommended.
Neuropsychological diagnosis LD, autism spectrum disorders (ASD), and attention deficit hyperactivity disorder (ADHD) have also been used as outcome measures in the current context.
-
13
LD comprises dyscalculia and dyslexia. Both disorders are common (with prevalence ~5% for each item). Their origin is multifactorial, and their clinical presentation heterogeneous and changing over time.86,87 LD has been associated with psychiatric diseases and maladaptive traits.88 However, if LD is diagnosed and supportive interventions appointed, deficits can be attenuated. Additionally, differences in scores of verbal intelligence and performance intelligence have been shown to predict LD.89 Importantly, a diagnosis of LD is not necessarily equivalent to lower intelligence scores. ASD is characterised by impaired neurodevelopment and considered to be a lifelong condition.90,91 Children with ASD have difficulties interacting and communicating socially and present a behavioural pattern shaped by strict routine.81 Severity of symptoms varies among individuals and with development,92 which aggravates the clinical diagnosis. However, the course of sensorineural development relates to the clinical presentation and hence predicts the seriousness of the condition.93 ADHD is a mental development disorder clinically presented by hyperactivity and impaired ability to focus and concentrate. ADHD has been linked with neurocognitive development.94 While the natural course of ADHD cannot be predicted by neurocognitive assessments, the causal pathway of the disorder is known to consist of multiple factors, including an inherited genetic disposition.95 The prevalence varies with gender, ethnicity, countries, states96 and there prevails some disagreement whether ADHD should be diagnosed by detection of core symptoms only or whether emotional and executive functions should be assessed as well.97 Several studies have used academic achievement (i.e. scores in standardised exams). Not surprising, intelligence and academic achievement are linked together.98,99 However, academic performance is influenced by many other factors unrelated to GA/surgical exposure,25 for example, altered relationships with peers at school and low self-esteem because of childhood obesity,100 as well as specific life-style factors (longer sleep duration101 and physical activity102). In a recent study by our group, achievement at the Danish national 9th grade exam was compared between 558 children undergoing surgery for isolated oral clefts and randomly selected controls. The type of oral cleft, parental age, and length of education were more important than number and timing of exposures.75
Miscellaneous outcomes Neuroimaging by MRI offers measurement of gross brain structures. MRI has shown promising results in the early (<2 yr of age) identification of ASD. Children who were diagnosed with ASD later in life presented an atypical pattern of connectivity of brain regions compared with non-exposed children; microstructural properties of white matter fibre tracts were characterised by fractional anisotropy and axial diffusivity. Further, children with high risk of ASD presented blunted trajectories consistent with reduced fractional anisotropy.103 Among studies identified in this review, few have used MRI to look for changes after exposure to GA in childhood.35,40,63 As was shown, GA itself could not be associated with any reduction in brain tissue, exposed children had lower scores of performance and language intelligence, and lower scores were associated with decreased volume of grey matter in posterior
14
-
Clausen et al.
brain regions.40 However, 53 of 81 types of surgery were ear, nose, throat (ENT) related procedures. This (particular) population of children has been suspected to suffer language and speech problems and achieve poor academic results because of prolonged hearing loss caused by, for example, otitis media.104,105 One study compared MRI scans of premature infants (born <30 weeks of gestation) and neonates with low birth weight (<1250 g), both groups either exposed or non-exposed to GA and surgery at 39e42 weeks corrected age. Additionally, development was assessed at age 2 yr. Brain volume was decreased and mental developmental index score (one of many items included in the BSID III) lowered in exposed children.35 As these results were based on 15 exposed and nonexposed individuals, they suffer from a comparatively small sample size. The serum-concentration of S100B following GA was assessed as a secondary outcome in one study.42 In traumatology, S100B was introduced as a marker of severity of brain damage after traumatic brain injury.106,107 It has also been used for follow-up of patients with malignant melanoma.108 However, S100B is not brain-tissue specific109,110 and its utility as a marker for potential neuronal damage following GA is unknown.
Methodological considerations Psychometric tests are used for commercial, research, and clinical purposes.111 Crucial for evaluating the quality of a test are reliability (consistency of measurements of a test) and validity (degree to which a test measures, what it claims to measure). Internal reliability reflects the interrelation between items contained in a test.112 Validity is a crucial feature of a test.113 Validity does not apply to the test itself but resembles a property of a test score. When a test has been ‘validated’ (i.e. from one language or country to another), its ability to measure the construct of interest (intelligence, behaviour, etc.) has been correlated with tests of the same construct (‘convergent validity’) or has shown poor correlation with a test of a different construct (‘discriminative validity’). Validity is not a static factor but must be considered in each specific study context (study population, test users, overall setting). For instance, concurrent validity of the Bayley Developmental Index and the Cognitive Adaptive Test/Clinical Linguistic and Auditory Milestone Scale (CAT/CLAMS) was determined in a primary care health setting assessing children 12- and 30-months-old without suspected developmental delay.113 The authors concluded that the CAT/CLAMS had reasonable concurrent validity with the Bayley Scales and was easy and time efficient to use by professionals (e.g. paediatricians) without special training.114,115 As shown in our data, sample sizes vary tremendously, with the lowest numbers being 1535 and 21.23,36,41,55 Small sample sizes are problematic, as they reduce the power of a study. This eventually reduces reliability of findings because of low probability of finding the true effects, reduced positive predictive value (i.e. the probability of a true association in case of a significant result) and ‘effect inflation’.116,117 Considering the great number of children exposed to surgery and GA each year worldwide, the effect size of potential neurotoxicity is most likely to be minimal, as the clinical presentation or potential phenotype has not yet been defined. Evaluation of study effect size is not within the scope of this review. Nevertheless, reduced power because of small sample
and effect size remains a major concern within the field of anaesthesia-related neurotoxicity research. Few studies (n¼22) have reported on ‘baseline’ values before exposure, that is, demonstrated that the outcome of interest (reduced intelligence, delay of development, etc.) was not present before exposure (Table 3). However, inclusion of individuals prone to neurodevelopmental delay might lead to overestimation of an adverse effect of GA/surgery on neurocognitive outcome. Several studies have evaluated a selected study group knowingly associated with impaired neurodevelopment.118,119 These comprised children with hypoplastic left heart syndrome undergoing the Norwood procedure,28 congenital diaphragmatic hernia, or gastroschisis.51,60 Children undergoing major surgery might suffer from other underlying diseases/conditions, which confound the effect estimate and blur a ‘true’ association. Confounding by indication and selection bias is a concern in 29 (43%) studies, which did not compare outcome between exposed and non-exposed controls. Similarly, the potential confounding by co-morbidity cannot be entangled in studies assessing outcome after exposure for miscellaneous types of surgery [n¼45 (67%)]: minor surgery, for instance laser surgery for vascular malformations20 or inguinal hernia repair,7,64 compared with major surgery, is less likely to associate with severe underlying disease. Some outcomes either vary with age (behaviour, development) or show a heterogenic clinical presentation (ASD, LD, ADHD). Detecting an impaired score within these ‘unstable’ domains might overestimate an effect size that the child might ‘catch up’ later. ‘Loss-to-follow-up’ or missing details of those lost are a concern in 29 (43%) studiesdwhether this loss underestimates or overestimates the true effect depends on the reason(s) why individuals cannot be assessed. As these studies either followed-up on less than 70% of respective study populations or do not account for reasons and characteristics of those lost, specifics cannot be eluded. In concordance with our findings, previously published data reported a loss-to-follow-up because of migration of approximately 30%. Within the framework of observational studies based on birth cohorts, increasing mobility was considered to represent a major challenge. However, the authors did not consider differences between migrants and non-migrant (i.e. prevalence of congenital defects, maternal age and level of education, compliance to attend prenatal visits, ethnicity or single parent status) to influence their results on the prevalence on LD in their specific cohort.120 This may be a significantly flawed approach.
Quality assessment using the NOS We chose (to use) the NOS for quality assessment of the included studies, as this scale has been recommended for cohort studies and endorsed by the Cochrane collaboration.121 Evaluator burden has been described as less compared with other tools because of ease of use and limited time-consumption.122 These findings do not coincide completely with our experience. Some decision rules for quality rating of domains are less comprehensive and prone to misinterpretations. Evaluating (e.g. ‘comparability’) turned out to be difficult. Neurodevelopment in children is recognised as a complex and multifactorial phenomenon. Focusing on a single most
Neurocognitive outcomes in children after general anaesthesia
important factor is frequently not possible. In studies evaluating academic achievements, ‘male gender’ was a major confounder, as boys perform poorer than girls. When studies investigated outcome after neonatal cardio-thoracic surgery, ‘underlying conditions’ also constituted a major confounder; the prognosis differed depending on the specific cardiac malformation necessitating surgery (for instance uni-ventricular vs bi-ventricular cardiac malformation). We followed the same rule of decision-making for studies investigating outcomes after miscellaneous surgeries, as the impact of underlying condition is assumed to be pivotal for development. Recently, 13 papers were reviewed and rated according to NOS as part of a meta-analysis, discussing outcome measurements after a single GA administered before 3 and 4 yr.123 All 13 papers were included in the current study. The overall number of stars rewarded varied slightly from ours. Because subgroup labelling is not stated, any variation in decision-making cannot be specifically compared. Based on these considerations, we agree that inter-rater reliability of the NOS is a concern.124 To the best of our knowledge, we identified all studies relevant to this review according to inclusion criteria. As a result of exclusion of papers not published in English, we might have missed relevant work.
Conclusion The effect(s) of GA (and surgery) in young children have been assessed by various outcome measures. This variability in test items used reflects the problem, that the clinical presentation of a potential damaging impact on young children’s brains following GA/surgery is not properly defined. Studies within this topic continue to be inconsistent regarding outcome measures as well as other important factors. Consequently, future observational studies are only considered meaningful if they are conducted in cohorts of considerable sample size and comparable regarding type of surgery, comorbidity, age at exposure, follow-up, indication for intervention, and outcome. As development of neurocognition is multifactorial, the investigation of a single causative factor might be too simplistic. Neurotoxicity of anaesthetics might be clinically detectable only if several factors occur in combination. Regarding the choice of outcome measurement, the prevailing use of development tests is considered feasible in infants. In older age groups (i.e. >6 yr), intelligence tests seem reasonable, because intelligence is a stable characteristic and covers cognitive functions, as do academic achievements. The majority of studies included in this review have a low to moderate risk of bias in the domains of selection, comparability, and outcome assessment. However, some concerns exist about the feasibility of NOS. To avoid confusion, it is recommended to specify quality-rating rules.
Authors’ contributions Designed and composed the review protocol: all authors Independently screened title and abstracts and full text of articles: N.G.C., S.K. Extracted data: N.G.C. Checked extractions for completeness and correctness: S.K. Conducted quality assessment of studies by the NewcastleOttawa Scale: N.G.C. Quality assessment of studies re-assessed: S.K.
-
15
Drafted the paper: N.G.C. Critically revised the article repetitively: S.K., T.G.H. Approved the final version of the paper: all authors
Acknowledgements We are grateful to Research Librarian, Mrs Eriksen MB at The Medical Research Library, University Library of Southern Denmark for her kind assistance with design of the elaborate database searches.
Declaration of interest The authors have no conflicts of interests to declare. No ethical approvals were required for conductance of this study.
Funding The study is part of N.G.C.’s PhD project, which was awarded with the European Society of Anaesthesia (ESA) MAQUET GRANT at the annual meeting of the European Society of Anaesthesia 2015. N.G.C.’s project is further supported by funds provided by the Department of Anaesthesia and Intensive Care Medicine, University Hospital Odense.
Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.bja.2017.11.107.
References 1. Hansen TG. Use of anaesthetics in young children: consensus statement of the European Society of Anaesthesiology, the European Society of Paediatric Anaesthesiology, the European Association of Cardiothoracic Anaesthesiology and the European Safe Tots Anaesthesia Research Initiative. Eur J Anaesthesiol 2017; 34: 327e8 2. Sun LS, Li G, Miller TL, et al. Association between a single general anesthesia exposure before age 36 months and neurocognitive outcomes in later childhood. JAMA 2016; 315: 2312e20 3. Glatz P, Sandin RH, Pedersen NL, Bonamy AK, Eriksson LI, Granath F. Association of anesthesia and surgery during childhood with long-term academic performance. JAMA Pediatr 2017; 171: e163470 4. Davidson AJ, Disma N, de Graaff JC, et al. Neurodevelopmental outcome at 2 years of age after general anaesthesia and awake-regional anaesthesia in infancy (GAS): an international multicentre, randomised controlled trial. Lancet 2016; 387: 239e50 5. Ing C, DiMaggio C, Whitehouse A, et al. Long-term differences in language and cognitive function after childhood exposure to anesthesia. Pediatrics 2012; 130: e476e85 6. Wilder RT, Flick RP, Sprung J, et al. Early exposure to anesthesia and learning disabilities in a populationbased birth cohort. Anesthesiology 2009; 110: 796e804 7. DiMaggio C, Sun LS, Li G. Early childhood exposure to anesthesia and risk of developmental and behavioral disorders in a sibling birth cohort. Anesth Analg 2011; 113: 1143e51
16
-
Clausen et al.
8. FDA U. FDA Drug Safety Communication: FDA review results in new warnings about using general anesthetics and sedation drugs in young and pregnant women. 2016. Available from: https://wwwfdagov/Drugs/DrugSafety/ucm532356htm [Accessed 10 June 2017] 9. Bilotta F, Evered LA, Gruenbaum SE. Neurotoxicity of anesthetic drugs: an update. Curr Opin Anaesthesiol 2017; 30: 452e7 10. Disma N, Mondardini MC, Terrando N, Absalom AR, Bilotta F. A systematic review of methodology applied during preclinical anesthetic neurotoxicity studies: important issues and lessons relevant to the design of future clinical research. Paediatr Anaesth 2016; 26: 6e36 11. Ringer SK, Ohlerth S, Carrera I, et al. Effects of hypotension and/or hypocapnia during sevoflurane anesthesia on perfusion and metabolites in the developing brain of pigletsda blinded randomized study. Paediatr Anaesth 2016; 26: 909e18 12. McCann ME, de Graaff J. Current thinking regarding potential neurotoxicity of general anesthesia in infants. Curr Opin Urol 2017; 27: 27e33 13. Weiss M, Hansen TG, Engelhardt T. Ensuring safe anaesthesia for neonates, infants and young children: what really matters. Arch Dis Child 2016; 101: 650e2 14. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and metaanalyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009; 151: W65e94 15. Wells GA, Shea B, O’Connell D, Peterson J, Welch V, Losos M, Tugwell P. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available from: http://wwwohrica/ programs/clinical_epidemiology/oxfordasp [Accessed 22 August 2016 & 3 March 2017]. 16. Nestor KA, Zeidan M, Boncore E, et al. Neurodevelopmental outcomes in infants undergoing general anesthesia. J Pediatr Surg 2017; 52: 895e900 17. Hu D, Flick RP, Zaccariello MJ, et al. Association between exposure of young children to procedures requiring general anesthesia and learning and behavioral outcomes in a population-based birth cohort. Anesthesiology 2017; 127: 227e40 18. Conrad AL, Goodwin JW, Choi J, Block RI, Nopoulos P. The relationship of exposure to anesthesia on outcomes in children with isolated oral clefts. J Child Neurol 2017; 32: 308e15 19. Birajdar S, Rao S, McMichael J. Neurodevelopmental outcomes of neonates undergoing surgery under general anesthesia for malrotation of intestines. Early Hum Dev 2017; 109: 32e6 20. Terushkin V, Brauer J, Bernstein L, Geronemus R. Effect of general anesthesia on neurodevelopmental abnormalities in children undergoing treatment of vascular anomalies with laser surgery: a retrospective review. Dermatol Surg 2017; 43: 534e40 21. Lap CC, Bolhuis SW, Van Braeckel KN, et al. Functional outcome at school age of children born with gastroschisis. Early Hum Dev 2017; 106e107: 47e52 22. Yazar S, Hewitt AW, Forward H, et al. Early anesthesia exposure and the effect on visual acuity, refractive error, and retinal nerve fiber layer thickness of young adults. J Pediatr 2016; 169: 256e9
23. Seltzer L, Swartz MF, Kwon J, et al. Neurodevelopmental outcomes after neonatal cardiac surgery: role of cortical isoelectric activity. J Thorac Cardiovasc Surg 2016; 151: 1137e42 24. O’Leary JD, Janus M, Duku E, et al. A population-based study evaluating the association between surgery in early life and child development at primary school entry. Anesthesiology 2016; 125: 272e9 25. Djurhuus BD, Hansen TG, Pedersen JK, Faber CE, Christensen K. School performance in cholesteatomaoperated children in Denmark: a nationwide population-based register-study. Acta Otolaryngol 2016; 136: 663e8 26. Poor Zamany Nejat Kermany M, Roodneshin F, Ahmadi Dizgah N, Gerami E, Riahi E. Early childhood exposure to short periods of sevoflurane is not associated with later, lasting cognitive deficits. Paediatr Anaesth 2016; 26: 1018e25 27. Hoffman GM, Brosig CL, Bear LM, Tweddell JS, Mussatto KA. Effect of intercurrent operation and cerebral oxygenation on developmental trajectory in congenital heart disease. Ann Thorac Surg 2016; 101: 708e16 28. Hansen JH, Rotermann I, Logoteta J, et al. Neurodevelopmental outcome in hypoplastic left heart syndrome: impact of perioperative cerebral tissue oxygenation of the Norwood procedure. J Thorac Cardiovasc Surg 2016; 151: 1358e66 29. Graham MR, Brownell M, Chateau DG, Dragan RD, Burchill C, Fransoo RR. Neurodevelopmental assessment in kindergarten in children exposed to general anesthesia before the age of 4 years: a retrospective matched cohort study. Anesthesiology 2016; 125: 667e77 30. Doberschuetz N, Dewitz R, Rolle U, Schlosser R, Allendorf A. Follow-up of children with gastrointestinal malformations and postnatal surgery and anesthesia: evaluation at two years of age. Neonatology 2016; 110: 8e13 31. Diaz LK, Gaynor JW, Koh SJ, et al. Increasing cumulative exposure to volatile anesthetic agents is associated with poorer neurodevelopmental outcomes in children with hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 2016; 152: 482e9 32. de Heer IJ, Tiemeier H, Hoeks SE, Weber F. Intelligence quotient scores at the age of 6 years in children anaesthetised before the age of 5 years. Anaesthesia 2017; 72: 57e62 33. Aun CS, McBride C, Lee A, et al. Short-term changes in postoperative cognitive function in children aged 5 to 12 years undergoing general anesthesia: a cohort study. Medicine 2016; 95: e3250 34. Bakri MH, Ismail EA, Ali MS, Elsedfy GO, Sayed TA, Ibrahim A. Behavioral and emotional effects of repeated general anesthesia in young children. Saudi J Anaesth 2015; 9: 161e6 35. Taghon TA, Masunga AN, Small RH, Kashou NH. A comparison of functional magnetic resonance imaging findings in children with and without a history of early exposure to general anesthesia. Paediatr Anaesth 2015; 25: 239e46 36. Naguib AN, Winch PD, Tobias JD, et al. Neurodevelopmental outcome after cardiac surgery utilizing cardiopulmonary bypass in children. Saudi J Anaesth 2015; 9: 12e8
Neurocognitive outcomes in children after general anaesthesia
37. Ko WR, Huang JY, Chiang YC, et al. Risk of autistic disorder after exposure to general anaesthesia and surgery: a nationwide, retrospective matched cohort study. Eur J Anaesthesiol 2015; 32: 303e10 38. Hansen TG, Pedersen JK, Henneberg SW, Morton NS, Christensen K. Neurosurgical conditions and procedures in infancy are associated with mortality and academic performances in adolescence: a nationwide cohort study. Paediatr Anaesth 2015; 25: 186e92 39. Gano D, Andersen SK, Glass HC, et al. Impaired cognitive performance in premature newborns with two or more surgeries prior to term-equivalent age. Pediatr Res 2015; 78: 323e9 40. Backeljauw B, Holland SK, Altaye M, Loepke AW. Cognition and brain structure following early childhood surgery with anesthesia. Pediatrics 2015; 136: e1e12 ckova I, Zach J, Borsky J, et al. Early and late opera41. Petra tion of cleft lip and intelligence quotient and psychosocial development in 3e7 years. Early Hum Dev 2015; 91: 149e52 42. Yan J, Li YR, Zhang Y, Lu Y, Jiang H. Repeated exposure to anesthetic ketamine can negatively impact neurodevelopment in infants: a prospective preliminary clinical study. J Child Neurol 2014; 29: 1333e8 43. Yin J, Wang SL, Liu XB. The effects of general anaesthesia on memory in children: a comparison between propofol and sevoflurane. Anaesthesia 2014; 69: 118e23 44. Morriss Jr FH, Saha S, Bell EF, et al. Surgery and neurodevelopmental outcome of very low-birth-weight infants. JAMA Pediatr 2014; 168: 746e54 45. Ing CH, DiMaggio CJ, Malacova E, et al. Comparative analysis of outcome measures used in examining neurodevelopmental effects of early childhood anesthesia exposure. Anesthesiology 2014; 120: 1319e32 46. Gaynor JW, Ittenbach RF, Gerdes M, et al. Neurodevelopmental outcomes in preschool survivors of the Fontan procedure. J Thorac Cardiovasc Surg 2014; 147: 1276e82. discussion 82e83 e5 47. Garcia Guerra G, Robertson CM, Alton GY, et al. Neurotoxicity of sedative and analgesia drugs in young infants with congenital heart disease: 4-year follow-up. Paediatr Anaesth 2014; 24: 257e65 48. Cheng HH, Wypij D, Laussen PC, et al. Cerebral blood flow velocity and neurodevelopmental outcome in infants undergoing surgery for congenital heart disease. Ann Thorac Surg 2014; 98: 125e32 49. Andropoulos DB, Ahmad HB, Haq T, et al. The association between brain injury, perioperative anesthetic exposure, and 12-month neurodevelopmental outcomes after neonatal cardiac surgery: a retrospective cohort study. Paediatr Anaesth 2014; 24: 266e74 50. Ko WR, Liaw YP, Huang JY, et al. Exposure to general anesthesia in early life and the risk of attention deficit/ hyperactivity disorder development: a nationwide, retrospective matched-cohort study. Paediatr Anaesth 2014; 24: 741e8 51. Minutillo C, Rao SC, Pirie S, McMichael J, Dickinson JE. Growth and developmental outcomes of infants with gastroschisis at one year of age: a retrospective study. J Pediatr Surg 2013; 48: 1688e96 52. Hansen TG, Pedersen JK, Henneberg SW, Morton NS, Christensen K. Educational outcome in adolescence following pyloric stenosis repair before 3 months of age: a nationwide cohort study. Paediatr Anaesth 2013; 23: 883e90
-
17
53. Fan Q, Cai Y, Chen K, Li W. Prognostic study of sevoflurane-based general anesthesia on cognitive function in children. J Anesth 2013; 27: 493e9 54. Bong CL, Allen JC, Kim JT. The effects of exposure to general anesthesia in infancy on academic performance at age 12. Anesth Analg 2013; 117: 1419e28 55. Yang HK, Chungh DS, Hwang JM. The effect of general anesthesia and strabismus surgery on the intellectual abilities of children: a pilot study. Am J Ophthalmol 2012; 153: 609e13 56. Andropoulos DB, Easley RB, Brady K, et al. Changing expectations for neurological outcomes after the neonatal arterial switch operation. Ann Thorac Surg 2012; 94: 1250e5. discussion 5e6 57. Walker K, Halliday R, Holland AJ, Karskens C, Badawi N. Early developmental outcome of infants with infantile hypertrophic pyloric stenosis. J Pediatr Surg 2010; 45: 2369e72 58. Sprung J, Flick RP, Katusic SK, et al. Attention-deficit/ hyperactivity disorder after early exposure to procedures requiring general anesthesia. Mayo Clin Proc 2012; 87: 120e9 59. Sananes R, Manlhiot C, Kelly E, et al. Neurodevelopmental outcomes after open heart operations before 3 months of age. Ann Thorac Surg 2012; 93: 1577e83 60. Rocha G, Azevedo I, Pinto JC, Guimaraes H. Follow-up of the survivors of congenital diaphragmatic hernia. Early Hum Dev 2012; 88: 255e8 61. Long SH, Galea MP, Eldridge BJ, Harris SR. Performance of 2-year-old children after early surgery for congenital heart disease on the bayley scales of infant and toddler development, third edition. Early Hum Dev 2012; 88: 603e7 62. Long SH, Harris SR, Eldridge BJ, Galea MP. Gross motor development is delayed following early cardiac surgery. Cardiol Young 2012; 22: 574e82 63. Filan PM, Hunt RW, Anderson PJ, Doyle LW, Inder TE. Neurologic outcomes in very preterm infants undergoing surgery. J Pediatr 2012; 160: 409e14 64. Hansen TG, Pedersen JK, Henneberg SW, et al. Academic performance in adolescence after inguinal hernia repair in infancy: a nationwide cohort study. Anesthesiology 2011; 114: 1076e85 65. Flick RP, Katusic SK, Colligan RC, et al. Cognitive and behavioral outcomes after early exposure to anesthesia and surgery. Pediatrics 2011; 128: e1053e61 66. Fan XC, Ye M, Li DZ, Shi Y, Xu Y. Cognitive function in congenital heart disease after cardiac surgery with extracorporeal circulation. World J Pediatr 2010; 6: 268e70 67. DiMaggio C, Sun LS, Kakavouli A, Byrne MW, Li G. A retrospective cohort study of the association of anesthesia and hernia repair surgery with behavioral and developmental disorders in young children. J Neurosurg Anesthesiol 2009; 21: 286e91 68. Majnemer A, Limperopoulos C, Shevell MI, Rohlicek C, Rosenblatt B, Tchervenkov C. A new look at outcomes of infants with congenital heart disease. Pediatr Neurol 2009; 40: 197e204 69. Bartels M, Althoff RR, Boomsma DI. Anesthesia and cognitive performance in children: no evidence for a causal relationship. Twin Res Hum Genet 2009; 12: 246e53 70. Kalkman CJ, Peelen L, Moons KG, et al. Behavior and development in children and age at the time of first anesthetic exposure. Anesthesiology 2009; 110: 805e12
18
-
Clausen et al.
71. Kayaalp L, Bozkurt P, Odabasi G, et al. Psychological effects of repeated general anesthesia in children. Paediatr Anaesth 2006; 16: 822e7 72. Group TVICS. Surgery and the tiny baby: sensorineural outcome at 5 years of age. The victorian infant collaborative Study group. J Paediatr Child Health 1996; 32: 167e72 73. Ludman L, Spitz L, Lansdown R. Intellectual development at 3 years of age of children who underwent major neonatal surgery. J Pediatr Surg 1993; 28: 130e4 74. Ludman L, Spitz L, Lansdown R. Developmental progress of newborns undergoing neonatal surgery. J Pediatr Surg 1990; 25: 469e71 75. Clausen NG, Pedersen DA, Pedersen JK, et al. Oral clefts and academic performance in adolescence: the impact of anesthesia-related neurotoxicity, timing of surgery, and type of oral clefts. Cleft Palate Craniofac J 2017; 54: 371e80 76. Ing C, Wall MM, DiMaggio CJ, et al. Latent class analysis of neurodevelopmental deficit after exposure to anesthesia in early childhood. J Neurosurg Anesthesiol 2017; 29: 264e73 77. Lynge E, Sandegaard JL, Rebolj M. The Danish national patient Register. Scand J Public Health 2011; 39: 30e3 78. Davidson AJ, Becke K, de Graaff J, et al. Anesthesia and the developing brain: a way forward for clinical research. Paediatr Anaesth 2015; 25: 447e52 79. Lagercrantz H. Connecting the brain of the child from synapses to screen-based activity. Acta Paediatr 2016; 105: 352e7 80. Deary IJ, Penke L, Johnson W. The neuroscience of human intelligence differences. Nat Rev Neurosci 2010; 11: 201e11 81. American Psychiatric Association. Diagnostic and statistical manual of mental Disorderse5. 5th Edn. Washington: DAPA; 2013 82. Gesell A. The developmental morphology of infant behavior pattern. Proc Natl Acad Sci U S A 1932; 18: 139e43 83. Diepeveen FB, Dusseldorp E, Bol GW, OudesluysMurphy AM, Verkerk PH. Failure to meet language milestones at two years of age is predictive of specific language impairment. Acta Paediatr 2016; 105: 304e10 84. Flensborg-Madsen T, Mortensen EL. Infant developmental milestones and adult intelligence: a 34-year follow-up. Early Hum Dev 2015; 91: 393e400 85. Peyre H, Charkaluk ML, Forhan A, Heude B, Ramus F. Do developmental milestones at 4, 8, 12 and 24 months predict IQ at 5e6 years old? Results of the EDEN motherchild cohort. Eur J Paediatr Neurol 2017; 21: 272e9 86. Lagae L. Learning disabilities: definitions, epidemiology, diagnosis, and intervention strategies. Pediatr Clin North Am 2008; 55: 1259e68 87. Dykens EM. Psychopathology in children with intellectual disability. J Child Psychol Psychiatr 2000; 41: 407e17 88. Emerson E. Prevalence of psychiatric disorder in children and adolescents with and without intellectual disability. J Intellect Disabil Res 2003; 47: 51e8 89. Drummond CR, Ahmad SA, Rourke BP. Rules for the classification of younger children with nonverbal learning disabilities and basic phonological processing disabilities. Arch Clin Neuropsychol 2005; 20: 171e82 90. American Psychiatric Association Task Force on DSM-IV, Diagnostic and statistical manual of mental disorders, DSMIV-TR, 4. edition, text revision. Washington, D.C.: American Psychiatric Association; 2000. xxxvii, 943 s.ill
91. Taylor MJ, Gillberg C, Lichtenstein P, Lundstrom S. Etiological influences on the stability of autistic traits from childhood to early adulthood: evidence from a twin study. Mol Autism 2017; 8: 5 92. Jones RM, Lord C. Diagnosing autism in neurobiological research studies. Behav Brain Res 2013; 251: 113e24 93. Lord C, Risi S, DiLavore PS, Shulman C, Thurm A, Pickles A. Autism from 2 to 9 years of age. Arch Gen Psychiatry 2006; 63: 694e701 94. van Lieshout M, Luman M, Twisk JW, et al. Neurocognitive predictors of ADHD outcome: a 6-year followup study. J Abnorm Child Psychol 2017; 45: 261e72 95. Thapar A, Cooper M, Eyre O, Langley K. What have we learnt about the causes of ADHD? J Child Psychol Psychiatr 2013; 54: 3e16 96. Cuffe SP, Moore CG, McKeown RE. Prevalence and correlates of ADHD symptoms in the national health interview survey. J Atten Disord 2005; 9: 392e401 97. Zhou X, Reynolds CR, Zhu J, Kamphaus RW, Zhang O. Evidence-based assessment of ADHD diagnosis in children and adolescents. Appl Neuropsychol Child 2017: 1e7 98. Cattell RB, Barton K, Dielman TE. Prediction of school achievement from motivation, personality, and ability measures. Psychol Rep 1972; 30: 35e43 99. Barton K, Dielman TE, Cattell RB. Personality and IQ measures as predictors of school achievement. J Educ Psychol 1972; 63: 398e404 100. Wang F, Veugelers PJ. Self-esteem and cognitive development in the era of the childhood obesity epidemic. Obes Rev 2008; 9: 615e23 101. Chaput JP, Gray CE, Poitras VJ, et al. Systematic review of the relationships between sleep duration and health indicators in school-aged children and youth. Appl Physiol Nutr Metab 2016; 41: S266e82 102. Poitras VJ, Gray CE, Borghese MM, et al. Systematic review of the relationships between objectively measured physical activity and health indicators in school-aged children and youth. Appl Physiol Nutr Metab 2016; 41: S197e239 103. Wolff JJ, Gu H, Gerig G, et al. Differences in white matter fiber tract development present from 6 to 24 months in infants with autism. Am J Psychiatr 2012; 169: 589e600 104. Roberts JE, Rosenfeld RM, Zeisel SA. Otitis media and speech and language: a meta-analysis of prospective studies. Pediatrics 2004; 113: e238e48 105. Fougner V, Korvel-Hanquist A, Koch A, et al. Early childhood otitis media and later school performanceda prospective cohort study of associations. Int J Pediatr Otorhinolaryngol 2017; 94: 87e94 106. Thelin EP, Jeppsson E, Frostell A, et al. Utility of neuronspecific enolase in traumatic brain injury; relations to S100B levels, outcome, and extracranial injury severity. Crit Care 2016; 20: 285 107. Dadas A, Washington J, Marchi N, Janigro D. Improving the clinical management of traumatic brain injury through the pharmacokinetic modeling of peripheral blood biomarkers. Fluids Barriers CNS 2016; 13: 21 108. Pham N, Fazio V, Cucullo L, et al. Extracranial sources of S100B do not affect serum levels. PLoS One 2010; 5(9): e12691. https://doi.org/10.1371/journal.pone.0012691 109. Steiner J, Schiltz K, Walter M, et al. S100B serum levels are closely correlated with body mass index: an important caveat in neuropsychiatric research. Psychoneuroendocrinology 2010; 35: 321e4
Neurocognitive outcomes in children after general anaesthesia
110. Sorci G, Riuzzi F, Arcuri C, et al. S100B protein in tissue development, repair and regeneration. World J Biol Chem 2013; 4: 1e12 111. Strauss ES, Elisabeth MS, Spreen O. A compendium of neuropsychological tests: administration, norms and commentary. Oxford: Oxford University Press; 2006 112. Cronbach LJ, Meehl PE. Construct validity in psychological tests. Psychol Bull 1955; 52: 281e302 113. Tieman BL, Palisano RJ, Sutlive AC. Assessment of motor development and function in preschool children. Ment Retard Dev Disabil Res Rev 2005; 11: 189e96 114. Wachtel RC, Shapiro BK, Palmer FB, Allen MC, Capute AJ. CAT/CLAMS. A tool for the pediatric evaluation of infants and young children with developmental delay. Clinical Adaptive Test/Clinical Linguistic and Auditory Milestone Scale. Clin Pediatr (Phila) 1994; 33: 410e5 115. Voigt RG, Brown 3rd FR, Fraley JK, et al. Concurrent and predictive validity of the cognitive adaptive test/clinical linguistic and auditory milestone scale (CAT/CLAMS) and the Mental Developmental Index of the Bayley Scales of Infant Development. Clin Pediatr (Phila) 2003; 42: 427e32 116. Button KS, Ioannidis JP, Mokrysz C, et al. Power failure: why small sample size undermines the reliability of neuroscience. Nat Rev Neurosci 2013; 14: 365e76 117. Ioannidis JP. Why most discovered true associations are inflated. Epidemiology 2008; 19: 640e8
-
19
118. Sistino JJ. Attention deficit/hyperactivity disorder after neonatal surgery: review of the pathophysiology and risk factors. Perfusion 2013; 28: 484e94 119. Bevilacqua F, Rava L, Valfre L, et al. Factors affecting short-term neurodevelopmental outcome in children operated on for major congenital anomalies. J Pediatr Surg 2015; 50: 1125e9 120. Katusic SK, Colligan RC, Barbaresi WJ, Schaid DJ, Jacobsen SJ. Potential influence of migration bias in birth cohort studies. Mayo Clin Proc 1998; 73: 1053e61 121. Tools for assessing methodological quality or risk of bias in non-randomized studies. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions; 2011. Version 5.1.0. Available from: www. cochrane-handbook.org. 122. Hartling L, Milne A, Hamm MP, et al. Testing the Newcastle Ottawa Scale showed low reliability between individual reviewers. J Clin Epidemiol 2013; 66: 982e93 123. Zhang H, Du L, Du Z, Jiang H, Han D, Li Q. Association between childhood exposure to single general anesthesia and neurodevelopment: a systematic review and metaanalysis of cohort study. J Anesth 2015; 29: 749e57 124. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010; 25: 603e5
Handling editor: J.G. Hardman