Research gaps in the neurodevelopmental assessment of children with complex congenital heart defects: a scoping review

Background Children with congenital heart defects (CHD) are at risk for a range of developmental disabilities that challenge cognition, executive functioning, self-regulation, communication, social-emotional functioning, and motor skills. Ongoing developmental surveillance is therefore key to maximizing neurodevelopmental outcome opportunities. It is crucial that the measures used cover the spectrum of neurodevelopmental domains relevant to capturing possible predictors and malleable factors of child development. Objectives This work aimed to synthesize the literature on neurodevelopmental measures and the corresponding developmental domains assessed in children aged 1−8 years with complex CHD. Methods PubMed was searched for terms relating to psycho-social, cognitive and linguistic-communicative outcomes in children with CHD. 1,380 papers with a focus on complex CHD that reported neurodevelopmental assessments were identified; ultimately, data from 78 articles that used standardized neurodevelopmental assessment tools were extracted. Results Thirty-nine (50%) of these excluded children with syndromes, and 9 (12%) excluded children with disorders of intellectual development. 10% of the studies were longitudinal. The neurodevelopmental domains addressed by the methods used were: 53% cognition, 16% psychosocial functioning, 18% language/communication/speech production, and 13% motor development-associated constructs. Conclusions Data on social communication, expressive and receptive language, speech motor, and motor function are underrepresented. There is a lack of research into everyday use of language and into measures assessing language and communication early in life. Overall, longitudinal studies are required that include communication measures and their interrelations with other developmental domains.


Introduction
One in 100 newborns is affected by a congenital heart defect (CHD) (1).Twenty-five percent of them have a severe CHD that requires early corrective heart surgery within the first year of life (1).In about one third of these severely affected children, a genetic-syndromic disease causes the heart defect(s) (2).Since surgical and cardiological pediatric therapy protocols have improved over the last few years, around 80%-90% of children with CHD now survive to adulthood (3,4).
Despite the extensive knowledge gained about the developmental profiles of children with CHD, little is known about predictive and moderating developmental factors related to quality of life and psychosocial well-being.While only 5% of the variance in cognitive outcomes can be explained by surgical factors (15-17) and 1% by the choice of a cardiopulmonary bypass (18), up to 33% is determined by innate patient-and family-related variables (19,20).Numerous studies have shown that the total length of stay in hospital is another predictor of cognitive outcome (21).However, with regard to developmental trajectories, a large proportion of the variance currently remains unexplained.There is great diversity in children with CHD, for instance, in the type of heart defect and its pathophysiological consequences, the palliation needed and its cardiocirculatory consequences, the variety of etiologies and the family resources available.
In 2012, the American Heart Association/American Academy of Pediatrics (5) highlighted the increased developmental risk for children with CHD and the need for ongoing developmental surveillance to maximize neurodevelopmental outcome opportunities.In 2020, the Cardiac Neurodevelopmental Outcome Collaborative established a consensus-based, standardized battery for the content and timing of neurodevelopmental assessments for children with complex CHD with the goal of promoting consistent neurodevelopmental care and quality improvement (22,23).The recommendations include core and extended versions of agespecific assessment batteries.
This scoping review aims to provide a concise snapshot of the measures used to investigate neurodevelopmental domains of children with complex CHD between the ages of one and eight years.The research question was: To what extent do the standardized neurodevelopmental assessment tools cover the spectrum of neurodevelopmental domains relevant for capturing possible predictors and malleable factors of child development?

Materials and methods
The methodological framework by Arksey and O'Malley (24) was used for this scoping review: (1) identify the research question; (2) identify relevant studies; (3) select studies; (4) chart the data; and (5) collect, summarize and report results.
The data synthesis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.
Studies were identified by searching the electronic database PubMed.Initial searches were conducted to identify relevant literature.Each search used the keywords "congenital heart disease" and one or more of the keywords from the following list: "mental health," "quality of life," "psychosocial outcome," "neurodevelopmental outcome," "social communication," "parentchild interaction," "language," "social cognition," "family," "selfesteem," "anxiety," "depression", "functional outcome." The quality of the studies was assessed using an adapted version of the Newcastle-Ottawa Scale (25) with a maximum score of 7 points.One point was allocated to each of the following categories: a) representativeness of the sample (e.g., all syndromes were included, no exclusion due to ethnicity, language or cognitive function); b) sample size (adequate); c) adequate control group available; d) clear definition of the neurodevelopmental measures used; e) clear description of the domains assessed f) the study reported on outcomes; and g) appropriate statistical analysis was conducted and included.
Original articles were included in the review if all of the following criteria were met: (i) study participants were diagnosed with complex CHD and (ii) children aged between 1 and 8 years were included; (iii) the article was in English, peer-reviewed and from the period 1980-2021; (iv) standardized quantitative methods were used to evaluate neurodevelopment; and (v) the study had a fair to good quality [adapted Newcastle-Ottawa Scale (NOS) score >2].Studies based exclusively on questionnaires or interviews without direct assessment were excluded.
Four authors (Raphael Oberhuber (RO), Nikoletta Deluggi (ND), Regina Wiltsche (RW), Marina Blum (MB)) conducted the initial searches, and after initial exclusion at the title and abstract levels, 270 articles remained.These 270 articles were again independently screened at the title, abstract and text levels.The remaining 170 articles were then reviewed by two researchers [Johannes Hofer (JH) and MB] at the text level, which resulted in a final selection of 78 articles with 100% agreement.The reference lists of the selected articles were examined for the possibility of additional relevant studies; however, no additional studies were found.Figure 1 shows the flow diagram of the review process.
A standardized data extraction form was used to extract data from the included articles.For each article, the following data were extracted: (i) study title, (ii) names of authors, (iii) Digital Object Identifier, (iv) year of publication, (v) journal, (vi) details of the study population, (vii) number of participants, (viii) percentage of male participants, (ix) ethnicity of participants, (x) inclusion criteria, (xi) exclusion criteria, (xii) study type, (xiii) outcome measures, (xiv) methods/tests used, (xv) main findings of the study.The standardized neurodevelopmental measures extracted were categorized into four main domains: cognition, psychosocial functioning, language/communication/speech production and motor functioning.Table 1 lists the domains assigned to the set constructs.

Results
Seventy-eight articles met the inclusion criteria of the study.The majority of these (57/78; 73%) were published in the United States, and the remaining 21 in other high-income countries.No study was conducted in a middle-or low-income country.
The publication years of the articles ranged from 1995 to 2021, with a median of 2013 (SD = 5.7).All studies were published within the past 28 years, and 60% in the last ten years.
Most studies (58%) used a retrospective, cross-sectional, or mixed study design, while the remaining 42% used a prospective and/or longitudinal design.The adapted NOS scores for the studies included ranged from 3 to 7 for a maximum of seven   3).Forty-seven (59%) articles used data taken from these 13 cohorts.Overall, eight longitudinal cohorts were included in the articles reviewed, with follow-ups conducted after 1-5 years.
The neurodevelopmental domains measured were distributed as follows: Cognition-associated constructs (53%) formed the main domain tested, followed by speech/language/communication (18%), psychosocial functioning (16%) and motor development (13%).Overall, a variety of standardized measures were used to describe neurodevelopmental functioning; however, the Bayley scales most prominent.
Tables 4A-C lists all extracted measures that were used in three or more publications according to domain, construct and target age group.
Thirty-nine of the measures (54%) were used in only one study.Thirty-nine (50%) of the 78 articles excluded children with syndromes, and nine (12%) excluded children with cognitive disabilities.

Discussion
The goal of this review was to generate a concise overview of the methodology used to document neurodevelopment in young    Seventy-eight studies met the inclusion criteria, all of which were conducted in high-income countries within the past 28 years.That no study from a low-or middle-income country was found demonstrates the compelling need to extend rigorous science and innovative clinical practice by focusing on stepped care processes at the global level.
The quality of the studies included, assessed by means of an adapted Newcastle-Ottawa Scale, ranged from fair to excellent.Most studies (58%) used a retrospective, cross-sectional, or mixed study design, while the remaining 42% used a prospective and/or longitudinal design.Less than half of the studies (30/78, 38%) included representative samples that did not exclude severely affected children (28-31, 33-35, 37-45, 48-52, 56, 60-63, 74, 79, 82, 84, 85).Notably, 50% of the 78 articles excluded children with various syndromes, and 12% excluded children with cognitive disabilities.This is especially true for the few larger longitudinal studies (26, 27): Children with chromosomal changes and/or additional impairments were excluded.Insufficient recruitment of children with poor cardiac outcomes and disproportionate inclusion of privileged children and families are common.Including children with special needs requires additional time and knowledge.It is challenging to combine ageappropriate testing with testing that is appropriate to the individual child's level of development.This finding is not surprising and is well known for other well-studied patient cohorts, for instance, children with hearing loss (86) and children with autism spectrum disorders (87).Epidemiological study designs are needed to ensure that all children with CHD are included in our understanding of neurodevelopment and neurodevelopmental trajectories and their potential malleability.
Longitudinal data on the neurodevelopment of children with CHD are restricted to 8 different study cohorts, where follow-ups were conducted over a maximum of five years.Despite their high quality and their scene-setting impact, these studies did not use cohorts that were broadly representative.Considering the recommendations from the Cardiac Neurodevelopmental Outcome Collaborative (22), the measures chosen for neurodevelopmental trajectories were not sufficiently fine-tuned to allow malleable predictors of good neurodevelopmental outcomes to be identified.
Overall, trajectories of structural language development, especially social communication and speech production, have been less researched than other domains of cognitive functioning.
With increasing age of the children studied, published data on (i) language and communication and (ii) speech motor and motor function decrease, and neurodevelopmental measures within the cognitive domain become even more predominant.
Considering the predominant use of the Bayley Scales for early infant development and the current literature pointing towards an overestimation of neurodevelopment using the Bayley III scales and possible underestimation by using the Bayley II scales (88)(89)(90)(91)(92), there is a need for supplementary measures of early childhood neurodevelopment.However, the Bayley III language scales provide a good estimate of language development (88,91,92).
There remains a lack of research into measures of language and communication early in life and into everyday use of language, although social communication is expected to impact language development, social cognition, peer interaction, and mental health later in life.In addition, standardized measures of child self-regulation and parent-child interaction are missing and need to be addressed in future studies, as they represent important, potentially malleable predictors.
Overall, a variety of standardized measures have been used to describe neurodevelopmental functioning, with the Bayley scales dominating.Comparing the results of this scoping review to the recent recommendations for neurodevelopmental assessment (22,23) highlights the importance of a clear and standardized guidance for neurodevelopmental assessments.
Given the known risks associated with CHD and the demonstrated benefit of early intervention in other populations (93)(94)(95)(96), regular monitoring and periodic neurodevelopmental assessment are critical throughout childhood in order to optimize the neurodevelopmental outcomes of patients with CHD.
Standard application of well-balanced neurodevelopmental assessment batteries across cardiac neurodevelopmental sites holds enormous promise for both clinical care and research within the CHD population.
The findings discussed above must be considered in the light of specific limitations of this scoping review and the limitations of scoping reviews in general: Including studies only by searching the electronic database PubMed may exclude further relevant published literature and grey literature, which leads to potential bias in the findings.This can result in an incomplete representation of the evidence available.The used search terms lack infant specific formulations like regulation or attention.
In conclusion, our systematic review of the neurodevelopmental assessment tools used in children with complex CHD identified the following research gaps: ➢ No data on low-or middle-income countries, ➢ There is a lack of representative studies of the whole cohort of children with severe CHD: children with syndromes or intellectual disability are in most cases excluded, ➢ Need for longitudinal studies that focus on a balanced use of measures for all important neurodevelopmental domains, ➢ Data on social communication, expressive and receptive language, speech motor, and motor function are underrepresented, ➢ Presently, there remains a lack of (i) research into the everyday use of language and language and communication measures early in life and (ii) tools to measure early social communication skills.

FIGURE 1 PRISMA
FIGURE 1 PRISMA 2020 flow chart of the process of study selection.

TABLE 1 Domains and assigned constructs.
MDI, mental development index; MFQ, mood and feelings questionnaire; long, longitudinal; Peds-QL, pediatric quality of life inventory; ps, prospective; PDMS, peabody developmental motor scale; PPVT, peabody picture vocabulary test; rs, retrospective; RVPAS, right ventricle-to-pulmonary artery shunt; SCPC, superior CavoPulmonary connection; SV Physiology, single ventricle; SGA, small for gestational AGE; SON, Snijders-Omen nonverbal test of intelligence; SRS, social responsiveness scale; SVR, pediatric heart network single ventricle reconstruction trial; TAPQOL, Netherlands Organisation for Applied Scientific Research Academic Medical Centre Preschool Children Quality of Life; Tx, transplant; TFS, total functional score; TOF, tetralogy Of fallot; UVH, univentricular heart; VLBW, very-low birth weight; VSD, ventricular septal defect; VABS, vineland adaptive behavior scales; WISC, Wechsler intelligence scale; WISC, Wechsler intelligence scale for children; WRAVMA pegboard, wide range assessment of visual motor abilities; WPPSI, Wechsler preschool and primary scale of intelligence; WeeFIM, functional independence measure for children.
ABAS, adaptive behavior assessment system; ABFW, child language test; ACC, agenesis of the corpus callosum; ADSI, ankara developmental screening inventory; ASQ, ages and stages questionnaire; BAS II, British ability scales II; BASC, behavior assessment system for children; BRIEF, behavior rating inventory of executive function-preschool; Beery-VMI, Beery-Buktenica developmental test of visual-motor integration; BMI, body mass index; BNP, B-type natriuretic peptide measurements; BSI, brief symptom inventory; BSID, bayley scale of infant development; Conners, Conners' parent rating scale-revised-short form; CBCL, child behavior checklist; CCC-II, children's communication checklist; CDH, congenital diaphragmatic hernia; CHD, congenital heart defect; CHQ-PF50, child health questionnaire 50 items; CNS, central nervous system; CPB, cardiopulmonary bypass; CSF, CerebroSpinal fluid; c/s, cross-sectional; DD, delayed development; ECMO, ExtraCorporeal membrane oxygenation; EF, executive functions; FS-II (R), functional status II ® ; GMDS, functional status II ® ; GSI, general severity index; GTF, gastrostomy tube feeding; HC, healthy control; HLHS, hypoplastic left heart syndrome; HRQoL, health-related quality of life; HT, heart transplantation; ITSEA, infant-toddler social and emotional assessment; K-ABC, Kaufman assessment battery for children; MBTS, modified blalock-taussig shunt; MRI, magnetic resonance imaging; ND, neurodevelopmental; NICU, Neonatal Intensive Care Unit; NEPSY, a developmental NEuroPSYchological assessment; OMES-E, orofacial myofunctional assessment protocol; OR, odds ratio; PedsQL, pediatric quality of life inventory; SES, socio-economic status; PDI, psychomotor development index; points, with a mean of 5.4.Table 2 lists all longitudinal studies included and all other studies with a NOS score >4.A total of 12.125 participants, ranging in age from 1 month to 15 years, were included across the articles.All extracted measures referred to the age group 1-8 years.The number of participants ranged from 10 to 1.770, with a mean of 155 (SD = 236.5).Thirteen different patient cohorts were found (Table

TABLE 3
Summary of CHD cohorts.

TABLE 4 (
A)-(C) extracted standardized measures used in a minimum of 3 of the extracted publications.