The aim of this study was to examine neurodevelopment, psychological adjustment, and health-related quality of life (HRQoL) in adolescents after bypass surgery for congenital heart disease (CHD) during early childhood.
The aim of this study was to examine neurodevelopment, psychological adjustment, and health-related quality of life (HRQoL) in adolescents after bypass surgery for congenital heart disease (CHD) during early childhood.
Fifty-nine adolescents (34 females, 25 males) with CHD were examined at a median age of 13 years 8 months (range 11y 5mo–16y 11mo). Outcome was assessed with the Wechsler Intelligence Scale for Children, (fourth edition); the Beery Test of Visual-Motor Integration; the Rey–Osterrieth Complex Figure Test; the Zurich Neuromotor Assessment; the Strengths and Difficulties Questionnaire; and the KIDSCREEN questionnaires. Results were compared with those of 40 age- and sex-matched healthy comparison individuals.
Outcome with regard to full-scale IQ, perceptual reasoning, and the working memory scale was poorer in patients with CHD than in the comparison group (all p≤0.001). Visual perception, visuomotor integration (p≤0.001), and executive functions (Rey figure copy: p=0.05) were also affected. Patients with CHD also had lower scores on all motor domains (p<0.02) except static balance. Psychological adjustment was affected only in the ‘peer relationship’ domain (p=0.05). Quality of life was similar to that of typically developing peers.
Adolescents with CHD may manifest persistent cognitive and motor impairments, while psychological adjustment and self-reported HRQoL are mostly typical. Thus, long-term neurodevelopmental evaluations are necessary to provide early educational and therapeutic support.
Congenital heart disease
Health-related quality of life
Rey–Osterrieth Complex Figure test
Strengths and Difficulties Questionnaire
Zurich Neuromotor Assessment
A variety of studies have shown that children with congenital heart disease (CHD) who undergo open heart surgery are at risk of a broad spectrum of neurodevelopmental impairments.[1-4] The majority of studies report outcomes in early childhood and at early school age while the literature on long-term neurodevelopmental consequences of CHD is limited. However, it is important to know whether the deficits detected in early childhood persist into adolescence, indicating a permanent cerebral deficit. This would have implications for academic achievement in adulthood. Only one recently published study (Bellinger et al.) presents outcome data for adolescents with transposition of the great arteries, who were part of the Boston Circulatory Arrest Trial. They were examined at the age of 16 years and compared with healthy comparison individuals. Adolescents with transposition of the great arteries were more likely to present with various problems in neurodevelopmental function, visual–spatial skills, and attention. Despite this, little is known regarding cognitive function in adolescents with a broader spectrum of CHD diagnoses. As there is no literature on neurodevelopmental impairments in adolescents with CHD, we sought to detect abnormalities in a wide range of developmental domains. The expected difficulties were based on reported long-term outcomes in other ‘at-risk’ populations, such as adolescents born preterm. Furthermore, we wanted to know whether the described functional deficits affect quality of life and other aspects of behaviour. Health-related quality of life (HRQoL) has been shown to be impaired in 10-year-old children after heart surgery.
Thus, the aim of this study was to determine neurodevelopment, psychological adjustment, and HRQoL in adolescents after bypass surgery for CHD and to examine predictors of adverse outcomes. We hypothesized that we would find a broad spectrum of neurodevelopmental deficits, as well as poorer psychological adjustment and impaired HRQoL.
We examined 59 adolescents (34 females, 25 males; median age 13y 8mo; range 11y 5mo–16y 11mo) with different types of CHD. Participants had taken part in their first neurodevelopmental assessment at a median age of 10 years 5 months (range 6y 6mo–16y 10mo). These children were recruited from a sample of 200 children who underwent their first bypass surgery at the University Hospital Zurich between 1995 and 1998 (median age at first surgery 11mo; range <1mo–5y 7mo), whose parents had a good command of the German language and who were between 6 and 16 years of age at the time of examination. After excluding children with a diagnosis of a chromosomal or genetic syndrome and those with a diagnosis of congenital or acquired neurological disease, 117 children were examined (for details see von Rhein). At the assessment at 14 years of age, 87 of these 117 children were younger than 17 years and thus were eligible for the study. Twenty-three adolescents refused to participate and five were lost to follow-up. The final sample comprised 59 adolescents (68%). No selection criteria were applied to the heart defect or to the age at first surgery. Cardiac diagnoses are presented in Table 1. Demographic and surgical characteristics did not differ significantly between participants and those lost to follow-up. Based on the results of the first assessment, those we examined represented a subgroup with fewer neurological abnormalities (33% vs 60%, p<0.001) and a higher IQ than the 28 non-participants (IQ at 10y: 93.2 [SD 18.4] vs 86.1 [SD 14.1], p=0.02). Neuromotor performance, HRQoL, and psychological adjustment were similar. Forty healthy children (median age 13y 11mo; range 9y 0mo–16y 11mo; 22 females, 18 males) served as a comparison group. Comparison individuals were either enrolled for this study or had participated as healthy volunteers in a previous study. Participants with CHD and comparison children were similar in terms of age, sex, and socio-economic status. They all went to regular school and did not suffer from any chronic medical or neurological disease or any learning disability. HRQoL and psychological adjustment were not assessed in all the comparison children.
|Characteristic||n (%)||Median (range)|
|Preterm birth||9 (15.3)||–|
|Sex: male; female||25 (42.4); 34 (57.4)||–|
|Acyanotic CHD||29 (49)||–|
|Atrial septal defect, ventricular septal defect||20||–|
|Cyanotic CHD||30 (51)||–|
|Tetralogy of Fallot||7||–|
|Transposition of the great arteries||10||–|
|Double inlet left ventricle||6||–|
|Total anomalous pulmonary venous connection||2||–|
|Birthweight (g)||3180 (940–4380)|
|Five-minute Apgar score||–||9 (7–10)|
|Age at first surgery (y)||–||0.9 (0–5.6)|
|Weight at surgery (kg)||–||7.8 (2.8–20.3)|
|Duration of extracorporeal circulation (min)||–||91 (5–189)|
|Duration of aortic cross-clamping (min)||–||40 (13–83)|
|Circulatory arrest time (min)||–||0 (0–18)|
|EEG anomalies/focal neurological symptoms post-operatively||–||11 (18.6)|
|Length of hospital stay (d)||–||16 (7–72)|
|Total number of surgical operations||–||1 (1–4)|
A large number of potential risk factors were retrieved from the patients' records and parental interviews to determine their association with outcomes. These included birth, cardiac, and peri- and intraoperative variables.
A wide range of cognitive and motor functions was evaluated using a comprehensive neurodevelopmental assessment.
A standardized neurological examination was performed and the outcome was classified according to severity, ranging from mild to severe (mild, abnormalities in muscle tone or tendon reflexes; moderate, abnormalities in tone and stretch reflexes but without cerebral palsy [CP]; and severe, CP). Healthy comparison individuals were not examined by a physician.
IQ was evaluated using the Wechsler Intelligence Scale for Children, 4th edition, (WISC-IV), providing total IQ scores and subscores including verbal comprehension, perceptual reasoning, working memory, and processing speed. IQ was compared with comparison individuals and low IQ was defined as being more than 1SD below the comparison mean.
Visuomotor and visuoperceptive skills were assessed with the Beery Test of Visual-Motor Integration, a non-verbal standardized test (mean 100, SD 15) with higher scores indicating better performance.
Executive function was tested with the Rey–Osterrieth Complex Figure Test (ROCFT). The test consists of two parts, first copying a complex geometric figure and then reproducing the figure from memory 15 minutes later. Higher scores indicate better performance (maximum score 72).
Motor function was assessed with the Zurich Neuromotor Assessment (ZNA).[11, 12] This is a standardized procedure for assessing specific motor tasks in terms of speed (timed performance) and movement quality (intensity of associated movements) in the following domains: repetitive, alternating and sequential movements of feet, hands and fingers; adaptive skills (pegboard, dynamic balance); and static balance. Based on timed measurements, block components were formed: pure motor, adaptive fine motor, adaptive gross motor, and balance component. Movement quality was scored by classifying the duration (score 0–10) and degree (score 0–3) of associated movements. Associated movements were scored from video recordings by a trained paediatrician (CS) who was blinded to all medical and developmental characteristics. Performance on the ZNA was expressed as a z score.
Psychological adjustment was evaluated with the Strengths and Difficulties Questionnaire (SDQ) self-report and proxy report, a screening test measuring emotional symptoms, conduct problems, hyperactivity–inattention, peer interaction, and prosocial behaviour. The SDQ is a screening tool with good reliability and validity that has been previously used in this participant group and in other German studies. Higher scores reflect more problems. Results were compared with UK norms (www.sdqinfo.org) because Swiss or German norms were not available.
HRQoL was measured with the KIDSCREEN, a multidimensional standardized assessment addressing patients as well as their parents. The KIDSCREEN questionnaires for parents consist of 27 items regarding physical well-being, psychological well-being, parent relation and autonomy, social support, peers, and school environment (http://www.kidscreen.org). Similar questions concerning those five dimensions are presented to the children. The intensity or frequency of a certain feeling or behaviour within a period of a week is evaluated on a five-point Likert scale. Scale scores are transformed into T values based on reference data from a community sample (mean 50, SD 10), with higher scores demonstrating better quality of life. Results were compared to Swiss reference norms.
Socio-economic status was calculated by means of a six-point scale of both paternal occupation and maternal education with a possible range from 2 to 12. This measure was developed and has been used in Switzerland.
Participants with CHD were examined by a developmental paediatrician (MvR), who was aware of the medical condition of the patients. Comparison individuals were assessed with the same tests as the participants with CHD (WISC-IV, Beery VMI, ROCFT, and ZNA), except for psychological adjustment and HRQoL. The study was approved by the ethics committee of the University Children's Hospital Zurich. Parents and adolescents gave written informed consent to participation in the study.
All data were analysed using spss statistics software, version 18.0 (IBM Corporation, New York, USA). Descriptive statistics are presented as the mean (SD), median and ranges for non-normally distributed variables, or frequencies. All outcome measures were used as continuous total or continuous subscale scores. χ2 tests or t-tests were used to compare outcome variables and normative data if a normal distribution was present. The Wilcoxon–Mann–Whitney test was used to compare groups of non-normally distributed variables. A stepwise linear regression analysis was conducted to test the predictive value of selected risk factors for adverse outcomes based on the literature (sex, cyanotic CHD, extracorporeal circulation time, post-operative neurological abnormalities, length of hospital stay, and socio-economic status).[16, 17] Potential predictors were included if they correlated with a test score at a p value of <0.10 in univariate analyses. Predictors that met this criterion were included in a stepwise backward analysis in which the criterion for inclusion was a p value of <0.05 and the criterion for exclusion was a p value of ≥0.1. Variables entered into the multiple regression models were cyanotic CHD, number of surgeries, age at first surgery, extracorporeal circulation time, post-operative neurological or electroencephalographic abnormalities, duration of hospital stay, sex, and socio-economic status. All statistical tests were two sided and a p value of ≤0.05 was considered statistically significant.
Fifty-nine individuals with CHD (34 females, 25 males; median age 13y 8mo; range 11y 5mo–16y 11mo) and 40 age- and sex-matched comparison individuals participated in this study. The median age at examination was 13 years 7 months (range 11y 5mo–16y 11mo) and median socio-economic status was 7.9 (range 2–12). Detailed medical characteristics for the individuals with CHD are provided in Table 1 (for further information, see also von Rhein et al.). Participants with CHD were not significantly different from the healthy comparison group with respect to age, sex, and socio-economic status (all p>0.2).
Neurodevelopmental outcome was poorer in participants with CHD than in the comparison children on all tested domains (Table 2). Particular weaknesses were detected in the areas of perceptual reasoning, working memory, visuomotor integration, visual perception, and executive function. Neurodevelopmental outcomes were comparable in children with cyanotic and acyanotic CHD. Although the mean IQ of participants with CHD was in the normal range, the proportion of participants in whom IQ was more than 1SD below the comparison mean was significantly higher than in the comparison group (full-scale IQ 42.4% vs 15.8%, p=0.01; perceptual reasoning 52.5% vs 15.8%, p<0.001; working memory 44.1% vs 18.4%, p=0.02). The proportions of participants in whom verbal comprehension and processing speed was more than 1SD below the comparison mean was also higher in the CHD group than in the comparison group (32.8% vs 18.4% and 32.2% vs 21.1% respectively), but not significantly so.
|Patients (n=59), mean (SD)||Comparison children (n=40), mean (SD)||p|
|Wechsler Intelligence scale for children|
|Full scale IQ||103.10 (16.49)||112.68 (10.43)||0.001|
|Verbal comprehension||109.14 (18.83)||115.21 (15.96)||0.11|
|Perceptual reasoning||103.66 (14.77)||113.82 (8.83)||<0.001|
|Working memory||93.42 (13.64)||103.76 (12.19)||<0.001|
|Processing speed||101.46 (14.19)||105.39 (11.20)||0.15|
|Rey-Osterrieth Complex Figure Test score|
|Copy||69.03 (8.65)||71.49 (3.04)||0.05|
|Memory||48.86 (12.64)||53.57 (13.29)||0.09|
|Beery Test of Visual-Motor Integration|
|Visual perception||93.78 (11.36)||103.05 (7.72)||<0.001|
|Motor coordination||98.76 (14.25)||102.89 (8.53)||0.79|
|Visuomotor integration||93.54 (12.37)||101.49 (8.03)||0.001|
|Zurich Neuromotor Assessment|
|Pure motor function||−0.06 (1.08)||0.51 (1.24)||0.016|
|Adaptive fine motor||−0.74 (1.30)||0.45 (1.07)||<0.001|
|Adaptive gross motor||−1.03 (2.93)||0.89 (1.59)||<0.001|
|Static balance||−0.19 (0.73)||0.37 (0.57)||0.123|
|Associated movements||−0.60 (0.92)||−0.09 (0.90)||0.008|
Parents reported school problems in 16% of participants with CHD (handwriting, reading, and mathematics), with 14% of participants needing extra tutoring in mathematics and 20% needing extra tutoring in reading and writing. Of the participants with CHD, 88% (n=51) went to regular school, whereas all comparison individuals did. Seven participants with CHD (12%) were in special education and 50% had a history of therapeutic support (such as physiotherapy, early education, speech therapy, and occupational therapy), some still on-going. No information was available for the comparison group regarding therapeutic support.
On neurological evaluation, 20 participants with CHD (34%) had mild to severe abnormalities and none had CP. Motor performance was poorer in all domains, particularly with respect to adaptive fine and gross motor skills and associated movements (Table 2).
Self-reported ratings of emotional symptoms and conduct problems were lower than UK norms; however, proxy report did not reveal such differences. Both patients and their parents reported more peer relationship problems. Overall, total difficulties scores were not different when self-rated, but parent-reported total difficulties scores were higher than UK norms (Table 3). The HRQoL of the study group was similar to community norms (Table 4).
|Emotional symptoms||1.9 (1.7)||2.8 (2.1)||0.001||2.3 (2.0)||1.9 (2.0)||0.13|
|Conduct problems||1.5 (1.2)||2.2 (1.7)||0.001||1.7 (1.5)||1.5 (1.7)||0.33|
|Hyperactivity/inattention||3.9 (2.0)||3.8 (2.2)||0.43||3.6 (2.3)||3.2 (2.6)||0.26|
|Peer relationship problems||1.9 (1.7)||1.5 (1.4)||0.06||2.0 (1.9)||1.5 (1.7)||0.05|
|Total difficulties score||9.3 (4.2)||10.3 (5.2)||0.15||9.7 (4.9)||8.2 (5.8)||0.03|
|Prosocial behaviour||8.2 (1.6)||8.0 (1.7)||0.47||8.2 (1.6)||8.6 (1.6)||0.10|
|Self-rated t-values, mean (SD)||p||Proxy-rated t-values, mean (SD)||p||Self vs proxya p|
|Physical well-being||48.7 (10.9)||0.25||47.3 (10.6)||0.07||0.49|
|Psychological well-being||51.1 (10.6)||0.40||49.2 (12.0)||0.25||0.31|
|Autonomy||51.6 (10.3)||0.49||50.0 (11.3)||0.20||0.24|
|Social support||50.6 (11.2)||0.30||47.9 (12.5)||0.23||0.16|
|School||50.1 (9.7)||0.67||49.7 (11.2)||0.70||0.71|
Using stepwise linear regression analyses we found that, of a variety of potential factors (see statistics), only socio-economic status was predictive of IQ (β=0.33, p=0.01) and score on the ROCFT memory task (β=0.31, p=0.02). Post-operative neurological abnormalities were predictive of performance on the ZNA dynamic balance task (β=0.27, p=0.04), and the post-operative length of hospital stay was predictive of the neuroscore (β=0.32, p=0.01). Extracorporeal circulation time was predictive of conduct problems (self-report: β=0.33, p=0.01) and hyperactivity (proxy report: β=0.37, p=0.006), assessed by SDQ. Neurodevelopmental outcome was not different between cyanotic and acyanotic CHD. Adolescents with a cyanotic CHD had more proxy-reported emotional symptoms than those with an acyanotic CHD (p=0.01). No other difference was found for CHD severity. For the sum score of the self-rated HRQoL, no significant predictor was found in the multiple regression analysis. However, when examining at the level of domains of HRQoL, cyanotic CHD was predictive of physical well-being (β=−0.27, p=0.05) and longer extracorporeal circulation time was predictive of poorer autonomy (β=−0.31, p=0.03). For the proxy-reported HRQoL, longer extracorporeal circulation time predicted worse psychological well-being (β=−0.28, p=0.03) and poorer autonomy (β=−0.31, p=0.02), and longer length of hospital stay predicted less social support (β=−0.30, p=0.03).
In this study, we report the outcome of adolescents with a wide range of CHDs who underwent full-flow bypass surgery during early childhood. In line with our hypothesis, we found that adolescents manifested a wide range of neurodevelopmental impairments in the domains of cognitive, visuo-perceptual, and motor functioning. A relatively large proportion of adolescents received educational and therapeutic support. In contrast to our hypothesis, most dimensions of psychological adjustment and self-rated HRQoL were comparable to those of healthy adolescents.
The relatively high IQ in our participants with CHD might be explained by their high socio-economic status as socio-economic status is known to be strongly correlated with IQ. In addition, the attrition rate of lower-functioning participants may also contribute to this effect. By comparing cognitive outcome of participants with CHD with age-, sex-, and socio-economic status-matched peers, we were able to show that participants with CHD performed more poorly than comparison individuals. Accordingly, the rate of poor performance (>1SD below the mean IQ of comparison individuals) in most domains of intellectual performance was higher in participants with CHD than in comparison individuals. This highlights the importance of comparing neurodevelopmental outcome with peers or comparison individuals with similar socio-economic status, as test norms may underestimate true deficits. The cognitive areas most affected were visuospatial performance, visuomotor performance, perceptual reasoning, and working memory, all important functions for higher academic achievement. Our findings are in line with those of Bellinger et al., who reported deficits in visuospatial functions, memory, and executive functions, as well as reading and mathematics problems in 16-year-old adolescents with operated d-transposition of the great arteries.
We found that adolescents with CHD had significantly more motor problems in pure fine and gross motor functioning than comparison individuals. Furthermore, movement quality, measured by the degree of associated movements, was also impaired. Our findings are in line with studies reporting motor deficits in younger children with CHD.[19, 20]
Problems in psychological adjustment occurred less frequently in the adolescents in our study than in studies of younger children,[1, 21] in which hyperactivity and problems in attention and emotion, as well as more internalizing and externalizing behaviour, were described in school-age children with CHD. Interestingly, in our study, adolescents with CHD reported significantly fewer emotional symptoms and conduct problems than the UK norms. However, importantly, both adolescents with CHD and their parents reported more peer relationship problems. It may be postulated that psychological adjustment improves with age, while peer relationship problems persist. This finding is important as it could represent reduced social competence (i.e. less experience in reading social signs) or limited social experience as a consequence of either behavioural problems at school age, leading to less successful or fewer social interactions. Furthermore, the heart disease itself could be associated with a higher rate of anxiety making it less likely that the affected child and peers will interact.
On the other hand, self-reported psychological adjustment may underestimate the true problems of these adolescents, as indicated by the significantly elevated parent-rated total difficulties score. Our findings are in contrast to those of Bellinger et al., who reported that in almost 20% of adolescents Connors Index scores were in the ‘clinical concern’ range. Importantly, in their study, behaviour was reported by parents and not by the adolescents themselves, and this may explain the higher rate of behavioural problems in their study than in ours.
Previous studies have shown that HRQoL can be impaired in younger children with CHD,[6, 22] whereas other studies have reported a normal HRQoL in school-aged children with CHD. In our current study, HRQoL of participants with CHD was reported to be similar to the HRQoL of community norms. Self-reported HRQoL tended to be better than parent-rated HRQoL. This phenomenon has been observed in children with CHD and other populations and reflects the different perspectives of participants and their parents, which need to be valued independently. Goldbeck and Melches found significant interaction effects between parental QoL and parental proxy reports of their children's QoL, especially in families affected by social stress. The good HRQoL of adolescents and young adults with chronic diseases may, in general, be the result of an adjustment process in which the individual lowers his or her expectations regarding academic and/or social achievements and success. This may be reflected by the fact that parents in our study reported that their children experienced more problems in the ‘physical well-being’ domain than the norm, whereas adolescents rated their physical well-being as not different from norms. This finding suggests that children with CHD might lack insight into the seriousness of their condition.
Our second aim was to examine predictors of long-term adverse outcomes. In our sample, only socio-economic status, extracorporeal circulation time, and post-operative neurological abnormalities were predictive of adverse intellectual and neuromotor performance, as well as poorer psychosocial adjustment and HRQoL (self-report and proxy report). These findings are in line with those of previous studies.[16, 17] The relatively small role of medical risk factors in predicting neurodevelopmental outcome may be related to the fact that, over time, medical risk factors lose their predictive strength and are replaced by environmental and socio-demographic factors. Interestingly, severity of CHD was not related to neurodevelopmental outcome. This may be due to the small sample size or the selected participant sample. However, a study by Majnemer also demonstrated that more severe CHD was not associated with poorer neurodevelopmental outcome. In contrast, cyanotic CHD was predictive of the physical well-being domain of the self-rated HRQoL. This is analogous to the findings by Goldbeck and Melches in a cohort of cardiac participants undergoing rehabilitation. It is worth mentioning that self-rated and proxy-rated HRQoL were related more to medical and perioperative variables than to neurodevelopmental outcome.
The presented study has several limitations worth mentioning. Our study population consisted of a subgroup of participants with mild functional deficits, as those more affected were more likely to refuse participation (mean IQ of participants at 10y: 93 vs 86 in non-participants). This introduces a bias such that neurodevelopmental impairments are most likely to be more frequent in the general population of adolescents undergoing bypass surgery for CHD and that psychological adjustment and quality of life may be poorer than here described. Assessment bias may have occurred as the examiner was not blinded to participants' medical history. Our sample was heterogeneous as it consisted of patients with varied CHD diagnoses. Owing to the small sample size, the study may have been underpowered to detect risk factors for specific CHD diagnoses. In addition, patients with more complex CHD diagnoses, in particular univentricular CHD, were under-represented in our cohort, as not all of these defects were surgically treated at that time. Behaviour and quality of life were compared with reference values, which may introduce bias in the sense that an underestimation of true problems may have occurred. Furthermore, we did not obtain information on schooling and therapeutic support in the comparison group, limiting the comparison between participants with CHD and typically developing individuals regarding this aspect. The Rey–Osterrieth Complex Figure Test may not be a good measure to assess executive function alone, as patients with CHD often show difficulties in visuomotor integration, limiting the information on executive functions, and we did not use other tests or questionnaires to assess executive functions in our study. Moreover, we assessed only HRQoL. Future studies should address other outcome domains, such as self-esteem or self-efficacy.
Adolescents with CHD after bypass surgery may demonstrate a wide range of cognitive and motor impairments. Importantly, self-rated psychological adjustment and quality of life are good. Our findings support the notion that neurodevelopmental problems described in younger children with CHD may persist and are not outgrown as a result of maturational processes. Our findings underline the importance of long-term outcome assessment in all individuals with CHD after bypass surgery. Resources need to be made available for these comprehensive examinations, as well as for counselling and therapies.
This study was funded by the Swiss Heart Foundation and Else Kröner-Fresenius-Stiftung.