The survival rate for childhood brain tumors has improved over the last several decades facilitated by continuing medical advances in diagnosis and treatment [1, 2]. Survival has not come without a cost, however, and the presence of late effects such as neurocognitive deficits are now well established [3, 4]. Evidence is similarly mounting for late effects in psychosocial adjustment, defined as social, emotional, and behavioral outcomes, demonstrated in a recent review of the literature published between 1969 and 1999 on childhood brain tumor survivors [5]. Marked social competency deficits were found among these survivors although inconsistencies in evidence were noted [5]. Since publication of this earlier review of psychosocial outcomes, there has been an escalation in studies focused on the social competence outcomes in childhood brain tumor survivors. There is no review that has yet focused specifically on the social competency outcomes among this population. It is critical that social competence outcomes among pediatric brain tumor survivors be reviewed because social competence is necessary to ensure optimal health and well-being. The absence of social competence is a significant risk factor for internalizing problems such as major depressive disorder as well as externalizing behavior problems including conduct disorder [6]. The purpose of the current review, therefore, was to evaluate this new cohort of studies focused on the social competency outcomes in childhood brain tumor survivors.

Definition of terms

Social competence has been defined as “the ability to achieve personal goals in social interaction while simultaneously maintaining positive relationships with others over time and across situations” [7]. Such a definition is too broad in scope, however, leading to difficulties in the operationalization and subsequent interpretation, application, and comparison of research findings. In response to criticisms about the breadth of the social competence construct, a more explicit definition of social competence has been offered by Cavell [8] who proposed that social competence be defined as the umbrella under which: (1) social adjustment, defined as the quality of interactions and the extent to which individuals are achieving developmentally appropriate, societally determined goals; (2) social performance on specific tasks, or the social exchange between individuals; and (3) social skills, the specific abilities needed to enable an individual to successfully perform social tasks, have been hypothesized to fall. These components are said to exist in a hierarchy as separate, but interrelated constructs, with social adjustment existing at the top, followed by social performance, and social skills forming the base of this hierarchy. Social competence outcomes in the current review will be evaluated based on Cavell’s [8] framework.

Consistent with the views of the National Cancer Institute, the term survivor has been used to describe anyone who has been diagnosed with cancer [9]. For the purposes of the current review, therefore, the term survivor will be used to describe any participant diagnosed with a brain tumor.

Determinants of social competence

The source of social competency deficits among child and adolescent brain tumor survivors is unclear. Although the previously conducted review on psychosocial outcomes considered correlates of psychosocial adjustment (e.g., demographic, clinical variables) [5] these were not explicit to social competency outcomes and therefore little was elucidated with respect to variables important to social competence outcomes. Potential determinants surrounding the source of social competence deficits, however, have been identified, including direct treatment effects such as surgery and/or cranial radiation therapy (CRT). CRT is known to affect cognitive abilities [3, 4], which may thereby affect social skills and subsequently social competence [10, 11].

Other variables in addition to disease/treatment factors may also influence social competence outcomes. The transactional stress and coping model posited by Thompson and Gustafson [12] emphasizes the association between chronic illness and psychological adjustment as a function of the transactions among child characteristics (e.g., age, intellectual functioning) in addition to biomedical variables (e.g., chronicity of disease). Moreover, Bronfenbrenner’s social ecological model [13] proposes that a multiplicity of personal and environmental factors contribute to child development and social outcomes. In order to better understand social competence outcomes in brain tumor survivors, child and developmental characteristics should be acknowledged, as well as familial/environmental factors [14–16]. There is little research that has been conducted among child and adolescent brain tumor survivors that has attempted to consider these external variables and the interrelationships among each in the context of social competence outcomes.

Objectives

This manuscript aimed to comprehensively evaluate the current literature on the social competence of child and adolescent brain tumor survivors. A new review of the literature on the social competence of brain tumor survivors was needed given the pervasiveness of social competence deficits among this population [5]. Social competence deficits present a risk for compromising long-term adjustment [6]. Greater understanding of these deficits and examination of factors that may contribute to these deficits are warranted so that appropriate intervention strategies may be developed. Focus was given to the literature published between 2000 and 2009. The way in which social competence was operationalized was given special consideration. In addition, consistent with the transactional stress and coping and the social ecological models [12, 13], disease/treatment factors, child, and environmental characteristics were examined as they related to these outcomes.

Methods

Articles published between 2000 and 2009 were accessed using PsycInfo and PubMed. The following medical subject headings terms were used as keywords in search criteria: “brain neoplasms” and “pediatrics” combined with terms thought to capture social competence: “social behavior”, “social adjustment”, and “social isolation”. In addition, reference lists from retrieved articles were reviewed to locate any other potentially relevant literature. Articles that were included for the current review met the following criteria: (1) study participants were diagnosed before the age of 18 years; and (2) study objectives included an assessment of children’s social competence following brain tumor diagnosis or treatment. A total of 25 were identified from the literature search, and 20 of those articles met the aforementioned criteria. These studies are summarized in Table 1.

Table 1 Studies investigating social outcomes in child and adolescent brain tumor survivors published between 2000 and 2009
Full size table

Results

Evidence for deficits in social competence

Social adjustment

Most studies evaluated social competence at the level of social adjustment [8]. Specifically, half of the studies characterized social competence as assessed by subscales of the parent report Child Behavior Checklist (CBCL) [17] which reflect social adjustment, although labeled “Social Competence” and “Social Problems”. The CBCL parent form is a standardized 118-item inventory summarized into total, internalizing (including social problems) and externalizing scores, as well as three different competence scales (including social competence). Internal consistency reliability ranges for the CBCL from 0.57 to 0.71 for internalizing, 0.70 to 0.86 for externalizing, and 0.69 to 0.82 for total problem behaviors, across ages [17]. Based on these scales, social adjustment was found to be compromised compared with normative values [18–22] or to children with juvenile rheumatoid arthritis (JRA) [23].

Similarly using the parent-reported CBCL, Barrera and colleagues [24] and Schultz and colleagues [25] conducted two large (e.g., greater than 100 participants) national studies to describe psychosocial outcomes among childhood cancer survivors including this population. The first, conducted in Canada [24], evaluated specific items of the CBCL and revealed that child and adolescent survivors of brain tumors were reported by parents to have fewer friends compared with other cancer survivors or a population-based control group. The other, conducted in the USA [25], utilized a subset of questions from the CBCL [26] and found that survivors of brain tumors had decreased social adjustment compared with their siblings [25].

Three studies administered the teacher report form (TRF) [17] of the CBCL to assess social adjustment with conflicting results [19, 20, 27]. Carey and colleagues [19] reported no significant difference to norms whereas Aarsen and colleagues [27] found teachers reported survivors of astrocytoma to have significantly higher social adjustment problems compared with population norms. The third study by Mabbott and colleagues [20], which was retrospective in nature, found teachers reported social adjustment to be within the normal range at first assessment, although social adjustment difficulties significantly increased in later assessments. Three studies reported social adjustment as assessed by the Youth Self-Report of the CBCL [17]; Aarsen and colleagues [27] and Poretti and colleagues [28] reported significantly greater social adjustment problems compared with population norms, whereas Carey and colleagues [19] found no significant difference between survivors and population norms.

Other measures of social adjustment include the social function subscale of the Pediatric Quality of Life Index (PedsQL) [29], a measure of health-related quality of life (HRQL). Using this subscale, Bhat and colleagues [30] found parent proxy reports and child self-reports showed significantly lower social adjustment compared with normative values and Palmer and colleagues [31] replicated these results compared with healthy controls. In Meeske et al.’s study [32], child and adolescent brain tumor survivors self-reported lower social adjustment compared with survivors of acute lymphoblastic leukemia (ALL). The types of treatments received by ALL survivors in this study (e.g., cranial radiation) were not included in this study. More than 10 years from diagnosis, Ribi and colleagues [33] and Poretti and colleagues [28] found parent and survivors of medulloblastoma and craniopharyngioma, respectively, also reported significantly lower scores on the social functioning subscale of the PedsQL compared with normative values. Using other HRQL questionnaires, Sands et al. [22] found parent proxies and survivors of craniopharyngioma reported “low average” social adjustment compared with normative values. Finally, other standardized questionnaires of social adjustment have yielded the same outcomes. Upton and Eiser [34] found parents and teachers reported significantly more difficulties with social adjustment as assessed through peer relationships compared with normative values and Bonner et al. [23] confirmed these findings compared with children with JRA based on parent report. No differences were found based on survivor self-reports.

Qualitative methodology has also been employed to assess social adjustment. Vance and colleagues [35] uncovered two consistent themes after conducting semi-structured interviews: (1) peer exclusion and being bullied; and (2) the discrepancy between their child’s social relationships prior to cancer treatment and the subsequent downfall with peers following the return to school. As well, in Upton and Eiser’s study, [34] almost half the mothers of children with brain tumors reported their child was socially isolated and half of these mothers felt their child’s behavior limited social opportunities. Interviews and focus groups conducted with brain tumor survivors, parents and siblings, Boydell and colleagues [36] similarly uncovered themes of social isolation.

In summary, with the exception of two studies based on self-reports [19, 23], social adjustment as assessed through a range of standardized questionnaires, semi-structured interviews and focus groups with multiple informants, has been found to be compromised in childhood brain tumor survivors compared with population norms, healthy controls, healthy siblings, children with other cancers, and children with other chronic illnesses.

Social performance

None of the studies reviewed assessed childhood brain cancer patients on aspects of social performance.

Social skills

Bonner and colleagues [23] conducted the only study that assessed social competence at the level of social skills. This study focused on the evaluation of social skills via cognitive assessment and measured facial expression recognition skills. They found that child survivors of brain tumors made significantly more errors interpreting adult facial expressions compared with children with JRA. Such skills are acknowledged to be essential for proficient social communication and interaction as well as to provide critical nonverbal social information to observers [10].

Determinants of social competence deficits

Disease/treatment factors

Where assessed, no relationship has been found between social competency outcomes with tumor type or tumor site [30, 37]. Among a population of craniopharyngioma survivors, tumor size did not appear to effect social adjustment outcomes [22].

With regard to treatment effects, the evidence is conflicting. Holmquist and Scott [38], Mabbott and colleagues [20] and Poggi and colleagues [18] found no effect of CRT on social competency outcomes when employing the CBCL social competence scale as an outcome. On the other hand, when employing the PedsQL, Bhat et al. [30] found that CRT with or without surgery was related to significantly lower social adjustment compared with no treatment, surgery only, or radiation and chemotherapy with or without surgery. The discrepancy in findings may be attributable to the fact that the PedsQL was designed for children with a chronic illness and may be more sensitive to differences in outcomes. Investigation at the level of social skills found a trend for the impact of CRT on the recognition of child facial expressions [23]. More research investigating the impact of CRT on social competence outcomes in brain tumor survivors is warranted.

Regarding chemotherapy effects, Holmquist and Scott [38] found children who received drugs such as vincristine, cytoxan, and VP16, reported greater social adjustment difficulties compared with those who did not receive chemotherapy regimes. Of note, however, there are a number of factors which may have confounded these results. Children receiving chemotherapy are more likely to have medulloblastoma or ST PNET tumors, which are larger volume tumors and similarly require craniospinal radiation followed by focal radiation, known to be a risk factor for hearing, vision, and neurocognitive deficits [39]. Children with brain tumors who do not receive chemotherapy, on the other hand, are more likely to have smaller volume tumors and are more likely to not to receive radiation [39]. Taking this into account, Kullgren et al. [37] examined the effects of having multiple treatments (i.e., surgery and radiation, chemotherapy and radiation, or surgery and chemotherapy) and found that this combined variable was also likely to predict difficulty with social adjustment 3 to 4 years following diagnosis. Disease recurrence was consistently associated with decreased social adjustment compared with survivors without disease recurrence according to Aarsen et al. [27] and Sands et al. [22]. Disease recurrence has also been associated with decreased social adjustment compared with those without disease recurrence in a study of survivors more than 10 years since treatment [28]. Finally, Bhat and colleagues [30] found the presence of a shunt to be related to social adjustment deficits, although this was not confirmed by Mabbott et al. [20].

Age at diagnosis was considered a variable of interest in five of the reviewed studies with contradicting evidence [20, 21, 23, 27, 37]. Kullgren et al. [37] and Mabbott et al. [20] found no relation between age at diagnosis and social adjustment. Aarsen and colleagues [27] found survivors with a diagnosis in adolescence reported lower social adjustment based on a HRQL measure compared with younger survivors. Foley and colleagues [21] found that younger age at diagnosis was related to poorer social adjustment based on father’s but not mother’s reports. Finally, Bonner et al. [23] found that the children who were younger at diagnosis and had CRT made more errors in facial recognition, compared with older children. Clearly, further research is needed to better understand the relationship between age at diagnosis and social adjustment in this population.

Time since diagnosis, on the other hand, has been found consistently to be related to social adjustment, with the longer the time since diagnosis, the worse the outcome. In Mabbott and colleagues’ [20] retrospective study, CBCL social adjustment was found to have significantly decreased over a median follow-up period of 4.17 years from diagnosis. In Kullgren and colleagues’ [37] longitudinal study, low CBCL social adjustment was found at 1 to 2 years post-diagnosis and these scores were predictive of lower CBCL social adjustment scores 3 to 4 years later. For self-reported social adjustment, no significant differences were found between survivors and population norms when time since diagnosis was 3.66 years [19]. Significantly worse outcomes, however, were found when time since diagnosis was 7.7 and 11 years [27, 28] providing further evidence that longer time since diagnosis may account for decrements in social adjustment over time.

Child factors

Few studies have evaluated the effect of gender or age at study on social competence, and among these no significant relationships have been found [18, 22, 30]. Lower body mass index (BMI), suggestive of underweight, was found to be related to poorer parent-reported social competence [40]. In the same study, examining the interrelationships among child variables (BMI, self-perception, IQ, and social competence), lower BMI was similarly related to self-perceptions of close friendships in the presence of lower IQ. These findings warrant further investigation to better understand the relationship of child characteristics and their interactions, and the social competence of these survivors.

With respect to the neurocognitive functioning of survivors, some associations with social competence have been found. Lower overall intellectual capabilities [18, 38] and lower nonverbal intelligence [19] were associated with decreased social adjustment. Conflicting reports exist with respect to verbal intelligence, however. Holmquist and Scott [38] found lower verbal memory and verbal fluency was associated with decreased social adjustment whereas, Carey and colleagues [19] found no relationship between verbal intelligence and social adjustment. Finally, Foley et al. [21] found survivor’s enrollment in special education was related to mother’s but not father’s reports of decreased social adjustment.

Environmental factors

Kullgren et al. [37] found socio-economic status (SES) to be related to social competence outcomes, although it did not contribute significantly to the variance of social competency when other variables were considered. Except for the examination of SES, there are no studies in the current review that considered the effect of family functioning on social competency outcomes.

Intervention efforts

Three studies have examined the effect of social competence interventions among children with brain tumors [41–43], though randomized control studies have yet to be conducted. DieTrill [41] and colleagues developed a social skills intervention completed by eight boys over 16 sessions. Improvements in social adjustment were reported based on satisfaction questionnaires administered to the families upon completion of the group. Without a baseline assessment, an absence of standardized questionnaires, and a sample of only eight boys, however, little can be derived from these results.

Barakat and colleagues [43] conducted a pilot study of a social skills intervention for 13 children with brain tumors. The intervention consisted of six sessions targeting social skills such as nonverbal communication and cooperation. Results were based on standardized measures (e.g., CBCL), 1 month prior to the intervention and 10 months following the intervention completed by survivors, their parents, and teachers. Improved social adjustment was reported. The small sample size, lack of control group, and follow-up assessments conducted 10 months post-intervention, however, make it difficult to attribute social adjustment changes to the intervention alone.

More recently, the preliminary outcomes of a social skills program with 32 participants was assessed [42]. The group intervention consisted of eight 2-h weekly sessions focused on social skills including friendship making and managing teasing. Using standardized questionnaires administered twice before the intervention, immediately following the intervention, and at 6-month follow-up, significant improvements in social adjustment were found. These improvements were maintained after 6-month follow-up [42]. No significant change was found for any standardized outcome measures completed by survivors. Although this study improved on previous methodology in terms of the sample size and repeated assessments before and after the intervention, it was limited by a lack of a control group.

Discussion

The literature on the social competence outcomes of child and adolescent survivors of brain tumors was reviewed based 20 identified studies. Consistent evidence was obtained for deficits in social competence. With one exception [23], these social competence deficits have been demonstrated at the level of social adjustment. Social adjustment deficits were found across a variety of data gathering techniques (standardized questionnaires, semi-structured interview), informants (parents, survivors, teachers), and study designs (cross-sectional, longitudinal, retrospective, comparisons to population norms, siblings, or healthy peers). Social adjustment deficits appear to endure into early adulthood and more than 10 years beyond diagnosis [28, 33]. Moreover, social adjustment difficulties seem specific to brain tumor populations as opposed to other cancer (e.g., ALL) [32, 38], or other illness (e.g., JRA) populations [25].

It was disappointing to find that by and large, most studies reporting social adjustment outcomes continue to rely primarily on comparisons to normative values and rarely use control or comparison groups. In fact, even healthy control groups appeared only twice as a comparison group among the studies reviewed [31, 44]. The use of normative values as a comparison in studies was also criticized in the earlier review paper [5] and almost one decade later, no improvement has been made in this area. Relying on normative data is problematic because normative data may overestimate the rates of social competence deficits in this population. Normative samples typically exclude children with adjustment problems and therefore do not represent the general population [45, 46]. At the very least, an increase in the use of comparison groups matched by age and gender (e.g., healthy peers, siblings, or children with other types of chronic illness), as opposed to population norms, is warranted to enhance validity of findings.

The current review has found an increase in the use of multi-informant assessment methodology (nine of 20 studies, 45%) compared with the previous review where only five studies (38%) included multi-informant assessments, of which the majority were comprised of parent proxy and teacher reports [5]. Teachers were only included as informants in the current review in three out of the 20 studies, each employing the TRF of the CBCL. For the most part, teachers reports tended to resemble that of parents. The increase in multi-informant reporting in the current review, therefore, is largely reflected by an increase in survivor self-reports.

The increase in survivor self-reports has highlighted greater discrepancies between child and parent and teacher reporting. There are several hypotheses for this discrepancy. Survivors of childhood cancer in general have been noted to underreport, or repress, their psychosocial difficulties, which may also account for the lack of significant differences [47, 48]. Within the literature on pediatric oncology, the discrepancy between survivors’ subjective reports and others (i.e., parents, teachers,) has been documented for outcome variables including competence (e.g., self-esteem) and distress (e.g., depression, anxiety, behavioral problems, general psychopathology, somatic distress) [48–52]. Alternatively, given cognitive late effects among this population, survivors may lack the insight to understand the basis for socially competent behavior [53]. Finally, parent reports may also be subject to inflation due to parental distress, particularly when dealing with a child with a chronic illness [32]. The discrepancies in reporting between parents (teachers, peers) and survivors in reports of social competence have not been investigated and warrant future research.

In comparing the recent studies of social competence reviewed in this study to the previous review [5], differences in data collection were evident. Specifically, greater utilization of quality of life measures (35% vs. 15%) and a decreased use of sociometric measures was observed. In fact, none of the studies in the current review employed sociometric measures, previously implemented by Noll and colleagues [54, 55], to examine peer relationships among children and adolescents with cancer, including brain tumor survivors. Although sociometric measures have been criticized for relying too heavily on popularity and likeability [56], peer reports are critical to our understanding of brain tumor survivors’ social competence, given the importance of returning to normal life after treatment.

There has been some consideration of the determinants of or contextual factors that may influence social competence outcomes. With respect to the effect of disease/treatment variables, only disease recurrence and longer time since diagnosis have been consistently shown to be related to greater social competence difficulties in these survivors [22, 27, 28]. According to Thompson and Gustafson [12], this may reflect the chronicity of illness and treatment demands on the child and family contributing to further isolation from family and friends [12]. Disease recurrence and time since diagnosis might also contribute to increased brain function impairment. Neurocognitive deficits in pediatric brain tumor survivors have similarly been found to worsen with time since diagnosis [57, 58].

Interestingly, the effects of CRT, speculated to be related to social competence difficulties as a result of its known effect on cognitive outcomes [59], have not been confirmed. None of the studies in the current review considered neurological or physical impairments. Again, more research is needed to identify specific disease/treatment variables related to social difficulties in these survivors.

With respect to child characteristics, higher intellectual capabilities, nonverbal intelligence, and verbal memory and verbal fluency were all associated with increased social function [18, 19, 38]. No consistent relationship emerged with gender, or age. There is little research that has explored other personal characteristics such as self-concept and social outcomes with this population. Further research on the associations of child characteristics with social outcomes is warranted.

Finally, there is little evidence with regard to the effects of family factors on social outcomes in childhood brain tumor survivors. This is an area that warrants greater attention in future research given recent findings highlighting the importance of the home environment (e.g., maternal age, maternal depression) of children with cancer who survive a stem cell transplantation for cognitive, educational and psychosocial outcomes [60].

As in the previous review, we found that studies herein remain largely cross-sectional with heterogeneous age at time of study and tumor population. In contrast, across the range of studies, mean age at diagnosis remained relatively consistent, ranging from 5 to 9 years of age, with the majority lying in the 6 to 7 years of age range. Only four studies, however, included age at diagnosis in analysis and the evidence for its effect is conflicting. Thus, more research is needed to better understand the relationship between the child’s development using age at diagnosis as an index of development and the effect of disease and treatment on social outcomes.

Critique and future directions

There remain some considerable limitations in the research conducted in this field. Although there is now sufficient evidence that social competence difficulties exist in this population, measurement of social competence has focused primarily at the level of social adjustment. Moreover, measures of social adjustment are limited largely to tools that have been designed for use in a healthy population. As highlighted in the previous review paper [5], measures designed for healthy populations have generally been criticized for their applicability in non-healthy children [46], which was similarly. There is a need for future research investigating the psychometric properties of these measures among a population of childhood brain tumor survivors to ensure their validity with this population.

With one exception [23], a major gap in the field is the lack of assessment at the more primary levels of social competence, mainly social performance and social skills. More research on these levels of social competence may help identify important targets for intervention (i.e., reading facial cues). It is time that the research in social outcomes of childhood brain tumors survivors move beyond the simple implementation of standardized questionnaires and instead employ more innovative methodologies to understand this construct.

Most studies lack a conceptual framework to understand the determinants of social competence in the context of a life changing illness such as a childhood brain tumor. There are a multiplicity of factors within the child’s environment that influence the developing child, before and after diagnosis and treatment [12–14]. Thus, when investigating social outcomes in children treated for a brain tumor, in addition to disease and treatment variables, the child and adolescents’ own characteristics, IQ, and self-esteem, as well as familial and environmental factors must be considered in order to generate a more comprehensive understanding of these social competence outcomes. It may be, for example, that although hypothesized in a number of studies, no consistent relationship has been identified with CRT and social outcomes because it has not yet been considered in conjunction with neurocognitive outcomes. Neurocognitive outcomes may moderate the effect of CRT on social competence when their interrelationships are considered. Currently, studies have examined some isolated child’s characteristics and disease factors, but rarely have their interrelations for mediation or moderation been explored.

Familial factors have seldom been studied in the context of social competence, which bears little improvement over the studies included in the earlier review [5]. There is still a lack of research evaluating SES, parental education, and other parental characteristics. Given the increased ethnic diversity in the Western world, ethnicity is also a factor that has yet to be included in studies of social competence, as cultural pattern, values and beliefs play an important role in social behavior and the development of social competence. In fact, not only has ethnicity rarely been investigated as a potential correlate, it was only documented as part of the sample demographics in four studies of the 20 studies reviewed [19, 23, 31, 37]. As well, SES and other parent characteristics need to be considered to better understand the role of these contextual factors and their interrelations to social outcomes in survivors of brain tumors in childhood and adolescence.

A major limitation for examining multiple factors as determinants of social competence outcomes are the small sample sizes available when studying participants with rare conditions such as childhood brain tumors. A small sample size remains a limitation in research conducted among this population, with the exception of two nationally based studies [24, 25]. Understandably, as the prevalence rate of child and adolescent brain tumors is minimal among the population, this is an issue that is difficult to circumvent. Nevertheless, a small sample restricts research not only in the evaluation of contextual factors, but also in the consideration of child and adolescent brain tumor survivors as a homogeneous group, despite in reality being a very heterogeneous group. Collaborations to conduct multi-site evaluations would address this problem. New research designs to investigate social competence outcomes in child and adolescent brain tumor survivors, therefore, should focus on collaborative consortiums from multiple research centers.

Finally, there remains a lack of adequate longitudinal studies to accurately evaluate survivors’ social competence over extended periods of time, and different developmental trajectories depending on the age of the child at diagnosis. Although challenging to conduct with high-risk populations, longitudinal studies are necessary to fully understand the long-term scope of the disease, its treatment and related factors. Conducted in the context of child and familial factors, longitudinal research would provide insight into potential predictive and protective mechanisms of long-term social outcomes in survivors of childhood brain tumors. Lastly, a longitudinal design would allow us to predict future social outcomes from early factors [61].

Conclusion

There is consistent evidence to conclude that child and adolescent survivors of brain tumors experience deficits in social competence at the level of social adjustment and that longer time since diagnosis and lower cognitive functioning are associated with social competence deficits. These deficits appear to persist into early adulthood and beyond and therefore, warrant early intervention. More research is needed, however, to better understand social deficits at the levels of social performance and social skills to uncover more specific sources of social difficulties. This research will guide the development of further evidence based social competence interventions tailored to the specific level of social deficit identified with this population.