How does parental cancer affect adolescent and young adult offspring? A systematic review

Accepted Manuscript Title: How does parental cancer affect adolescent and young adult offspring? A systematic review of recent literature Authors: Adam Walczak, Fiona McDonald, Pandora Patterson, Katherine Dobinson, Kimberley Allison PII: DOI: Reference:

S0020-7489(17)30197-9 http://dx.doi.org/10.1016/j.ijnurstu.2017.08.017 NS 3008

To appear in: Received date: Revised date: Accepted date:

9-1-2017 25-8-2017 28-8-2017

Please cite this article as: Walczak, Adam, McDonald, Fiona, Patterson, Pandora, Dobinson, Katherine, Allison, Kimberley, How does parental cancer affect adolescent and young adult offspring? A systematic review of recent literature.International Journal of Nursing Studies http://dx.doi.org/10.1016/j.ijnurstu.2017.08.017 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Manuscript Type:

Review article

Title:

How does parental cancer affect adolescent and young adult offspring?: A systematic review of recent literature

Running Title:

Review: Impact of cancer on offspring

Authors:

Adam Walczak1, Fiona McDonald1,2, Pandora Patterson1,2, Katherine Dobinson1, Kimberley Allison1

Affiliations:

1. Canteen Australia, NSW Australia 2. Cancer Nursing Research Unit, University of Sydney, NSW Australia

Corresponding Author: Adam Walczak Research, Evaluation and Social Policy CanTeen Australia GPO Box 3821, Sydney, NSW 2001 [email protected] +612 9007 0232

Abstract: Objectives: To i) identify and synthesise evidence published since 2007 regarding the impact of parental cancer on adolescent and young adult offspring, ii) identify methodological and evidence gaps addressed during this period and iii) highlight those requiring further attention. Design: A systematic review and thematic synthesis of peer reviewed literature regarding the impact of parental cancer upon AYA offspring. Data Sources: Online searches of CINAHL, Embase, Medline, PsychInfo and Scopus databases were conducted. Reference lists of included articles were screened and additional searches by prominent authors were performed. Review Methods: Study selection, data extraction and quality analysis was undertaken by three independent researchers. Extracted study data was iteratively reviewed and discussed to achieve consensus regarding thematic synthesis of included studies. Results: Database and hand-searching yielded 1730 articles, 54 of which were included in the final synthesis. Included studies are discussed with respect to the following themes: i) study design and Page 1 of 64

quality; ii) measurement and sampling; iii) positive and negative aspects of parental cancer; iv) needs; v) communication and information; vi) coping strategies; vii) interventions; and viii) family functioning and other predictors. Twenty-nine studies reported negative impacts related to parental cancer, while eight identified positive outcomes related to post-traumatic growth. Five returned null or mixed findings. Unmet needs were frequently explored and a new validated measure developed. Communication and information were particularly important for offspring, though these needs were often unmet and parents wanted guidance regarding discussions with their children. Offspring may adopt a variety of coping strategies, some of which appear maladaptive, and may cycle between different approaches. Few evaluations of interventions were identified, and further work in this area is needed. Further evidence has emerged that poorer family functioning and other family and illnessrelated factors predict worse psychosocial outcomes for offspring, however evidence for other predictors such as age and gender remain mixed. Conclusions: Additional evidence for the negative psychosocial impact of parental cancer on adolescent and young adult offspring, their needs, and factors predicting psychosocial outcomes has emerged in the last decade. However, substantial gaps and methodological issues remain and evidence for the development, efficacy or implementation of interventions for this population is very limited. There is also a clear need for greater focus on bereaved and young adult offspring and those from non-western cultural groups, who remain under-represented in research conducted to date.

Key words:

adolescent and young adult; offspring; outcomes; parental cancer; psychosocial

What is already known about the topic? •

The psychosocial impact of a cancer diagnosis extends to the family of patients; adolescent and young adult offspring appear to be particularly affected.

•

Previous reviews largely discuss negative impacts and potential protective/risk factors, although there are inconsistencies between study findings.

•

Earlier reviews identified substantial methodological limitations in the included studies. Page 2 of 64

What this paper adds •

Review of 54 papers from previous decade indicates that parental cancer may have positive and negative impacts for offspring.

•

Evidence for moderators varies in consistency, with mixed support for previously identified influences.



Overall study quality appears to have improved, although limitations remain - there is a particular lack of research on bereavement, non-Western cultural groups, and interventions to support offspring

Introduction Approximately 28,500 parenting age Australian adults were diagnosed with cancer in 2008, 72% of whom were estimated to have adolescent or young adult (AYA – 12-24 years old) offspring(1-3). Approximately 21,000 young Australians are therefore expected to be newly impacted by their parent’s cancer each year(1-3). Parental cancer during the AYA years can be particularly challenging due to the dynamic nature of this life stage, characterised by significant cognitive, social and emotional development(4). A young person’s efforts to establish individual identity, pursue education and employment, form intimate relationships, and gain financial independence may be hampered(5) and additional responsibilities at home may draw them further into their family at a time when they are seeking greater autonomy (6). The psychosocial impact of parental cancer on AYAs has consequently been the subject of considerable research, which has been summarised in broad reviews by Grabiak and colleagues (6) and Osborn (7) in 2007, as well as in several subsequent literature reviews with narrower foci (8-12).

Page 3 of 64

Grabiak and colleagues’ review(6) discussed the impact of parental cancer on adolescent offspring (10-20 years old) in relation to emotions and behaviours, perceptions and knowledge of parental cancer, role changes and coping. It concluded that parental cancer had a largely negative emotional and behavioural impact on AYA offspring, with internalising problems being common and externalising problems associated with poor family functioning and communication(6). Offspring viewed diagnosis and treatment as the most difficult stages of parental cancer, with timely and trustworthy information being desirable soon after diagnosis and throughout the cancer trajectory(6). Coping strategies adopted by AYA offspring included emotion- and problem-focused coping(6). By contrast, Osborn’s review focused on psychosocial difficulties arising from early-stage parental cancer in offspring aged 5-18(7). Inconsistencies in the included literature led this review to draw the cautious conclusion that offspring were at “slightly increased risk” of internalising problems compared to reference groups(7). Whilst disease and treatment factors appeared unrelated to offspring psychosocial difficulties, family communication, parental mood and adjustment, and gender played an important role. Adolescent daughters impacted by parental cancer were found to be more negatively impacted in psychosocial domains, and difficulties self-reported by offspring exceeded those reported by parents(7).

More recently, narrower systematic reviews have been conducted looking at the well-being of children (0-18 years) affected by parental cancer (10), the impact of advanced parental cancer on adolescents (12-18 years) (9), and predictors of psychosocial adjustment to parental cancer in offspring of all ages (8). Krattenmacher and colleagues’ review reaffirmed previously identified familial risk and protective factors for offspring psychosocial adjustment, which was positively associated with family functioning and negatively associated with parental depressive mood (8). Phillips’ review highlighted the vulnerability of offspring and identified both negative impacts and positive gains resulting from parental cancer, including greater appreciation for family and increased independence(9). Morris and colleagues’ review highlighted that children were significantly impacted Page 4 of 64

by their parent’s cancer, with the gender of the child appearing to influence how this impact presented(10). Parental characteristics and family functioning were also noted as mediators of the impact of parental cancer on the child(10).

Later, Huang and colleagues’ meta-synthesis identified common themes from qualitative studies on non-terminal parental cancer, including disruption and adjustment, information and support seeking, and emotional impacts and concerns (12). Additionally, Inhestern and colleagues’ review identified interventions available to families affected by parental cancer, and highlighted barriers and facilitators for involvement (11). Finally, Ellis and colleagues’ review of children’s (0-18 years) psychosocial needs and available interventions recommended that those developing future interventions should consider the need for age-appropriate information, support in communicating with family and healthcare professionals, and supportive environments that normalise their experiences (13).

These previous reviews have identified substantial methodological limitations within the existing literature. These have variously included largely inadequate sample sizes, cross-sectional design limitations, conflicting or absent definitions of the studied age ranges, inconsistent measurement approaches, and a dearth of knowledge regarding the impact on AYAs of parental cancers other than breast cancer(6-10). More recent reviews have had narrower foci, covering only psychosocial adjustment(8), advanced cancer (9), children(10), qualitative studies (12) and interventions (11, 13). This review therefore sought to update the findings of the two more comprehensive reviews published in 2007 with a focus on a broad AYA age range and parental disease continuum, as well as to integrate all relevant findings for this population.

AYAs affected by parental cancer are particularly vulnerable to negative psychosocial outcomes as they may be more aware of the seriousness and implications of a cancer diagnosis, but may not yet Page 5 of 64

have developed the ability to cope with the situation (14). However, despite previous indications of this heightened impact, the distress and needs of offspring may be overlooked by both family members and medical professionals (15). The aim of this systematic review was therefore to identify and synthesise evidence published since 2007 regarding the impact of parental cancer on AYA offspring. Secondary aims were to provide a broad overview of outcomes and risk/protective factors for clinical attention, assess potential avenues for intervention, and highlight the emergence and resolution of methodological and evidence gaps including those requiring further attention.

Methods: Search Strategy Database and manual searches were undertaken to identify relevant English language articles. Searches of CINAHL, Embase, Medline, PsychInfo and Scopus were performed on 28 July and 12 August 2015 using the following terms: (Cancer OR Malignan* OR Neoplasm) AND (offspring OR son OR daughter OR child*) AND (unmet needs OR distress OR impact OR bereave* or deceased). Search results were updated on 21 October 2016 using the same terms. Search results were limited to peerreviewed research in humans published since 2007. Reference lists of included articles were screened and additional searches by author were performed if two or more relevant articles by the same author were identified. Inclusion Criteria Included studies had to be peer-reviewed original research focussing in part or in full on the psychosocial impact of parental cancer on AYA offspring (12-24 years old), and could include AYAs with living or deceased parents with cancer. Studies could use quantitative and/or qualitative methods. Reviews, commentaries, study protocols without results presented, editorials, poster or dissertation abstracts, case reports, books and conference proceedings were excluded. Study Selection and Data Extraction

Page 6 of 64

Search results were exported to Endnote X5 and duplicates removed. Titles and abstracts were screened by 3 authors (AW, KA and KD) using inclusion criteria to identify potentially appropriate papers. This was followed by full text screening to arrive at the final sample of included papers. Discrepancies in inclusion decisions were resolved by iterative discussions between AW and KD. Key study details were extracted using a descriptive table to facilitate study comparison and synthesis. Quality Analysis Quality analysis was conducted using the QualSyst Standard Quality Assessment Criteria(16). Qualitative and quantitative studies received a rating between 0 and 1 based on manualised scoring criteria, facilitating comparison between studies with different methodologies. Ten articles were assessed by 2 authors (AW and KD) with differences resolved by iterative discussion to ensure consistency in ratings. Remaining articles identified in 2015 searches were assessed by a single author (AW). Articles identified in the 2016 search were assessed by a single author (KA) with uncertainties resolved by discussion with a second author (AW). Synthesis Following data extraction and quality analysis, key features and findings of included studies were independently reviewed and iteratively discussed by four authors (AW, FM, KA and KD) to establish the thematic structure for synthesis of study findings. Included studies are discussed below with respect to the following themes: i) study design and quality; ii) measurement and sampling; iii) positive and negative aspects of parental cancer; iv) needs; v) communication and information; vi) coping strategies; vii) interventions; and viii) family functioning and other predictors.

Results: Database searches produced 1626 unique results. One hundred and four additional articles were identified from other sources. Title and abstract screening yielded 106 full-text articles suitable for screening and 54 were included in the final synthesis (see Figure 1). INSERT FIGURE 1 HERE Page 7 of 64

Twenty studies used qualitative methods(14, 15, 17-34), thirty-two used quantitative methods(3566) and two were mixed methods studies(67, 68). See Table 1 for key information about each included study. INSERT TABLE 1 HERE Study Design and Quality The most common quantitative study design was cross sectional (n=14) (35, 37-39, 41, 44-46, 48, 49, 52, 53, 56, 58, 62), followed by case-comparison (n=12)(36, 40, 42, 43, 47, 51, 54, 55, 60, 61, 63, 64), and instrument validation (50) . One study(39) included both a retrospective cross-sectional component and a prospective longitudinal component to facilitate comparison of findings between the two methodologies. Most quantitative studies examined associations between variables using survey methodology. However, one used Rorschach tests (60), while three evaluated interventions for offspring affected by parental cancer (59, 65, 66).The majority of qualitative studies (14, 17-33) utilised single time-point semi-structured interviews, although three employed focus group methodology (15, 30, 34). Of the two mixed methods studies, one (67) used data from qualitative interviews and questionnaires to inform the development of an offspring needs measure, while the other (68) used qualitative interviews to explore initial quantitative survey findings on financial hardships associated with parental cancer in more detail.

Study quality ranged from 0.57 to 1.00 with a mean of 0.88 (see Table 1 for individual ratings). Amongst qualitative studies, ratings were typically diminished by insufficient detail regarding sampling strategies, data collection, analysis methods and demographics, as well as insufficient discussion of factors influencing data interpretation. Similar limitations were seen in the intervention studies, with intervention and control conditions also being inadequately described. Amongst quantitative studies, ratings were variously diminished by limitations in recruitment methods, which may have biased study samples, insufficient description of sample characteristics, sample size issues and a lack of evidence that potential confounders were considered in analyses. Some of these Page 8 of 64

limitations were also reflected in the included mixed methods studies. The evidence synthesis presented below should be considered with these limitations in mind.

Measurement and Sampling The constructs explored in the included studies, the measures used to assess them and the studies using each of the various measures are presented in Table 2 with a total of sixty-two measures targeting twenty-nine constructs. Over 85% of these measures were validated and most quantitative studies exclusively utilised validated scales. Twenty-four different validated measures of psychological functioning were used, representing various constructs including internalising and externalising problems, distress, depression, anxiety, stress, post-traumatic stress, worry, coherence, emotional reactions and general mental health. Fifteen measures related to family and social environments were used, variously measuring parental role satisfaction, family functioning, marital functioning, relationships, social support, and social functioning and parent behaviour. Additionally, six measures of health-related quality-of-life were used, three of coping, two each of needs, selfharm, communication and socioeconomic issues, and one each of eating disorders, performance status, temperament and experiences. Two intervention fidelity measures were used and one new measure for adolescent and young adult offspring was identified; the Offspring Cancer Needs Instrument(67), which was used in its un-validated form in two studies(67, 69) and subsequently validated(50). Un-validated measures were used exclusively in two studies(35, 68) and in tandem with validated measures in five studies(43, 54, 56, 59, 67). These un-validated measures included two related to family and social environment, a modified version of the Deliberate Self-Harm Inventory, one regarding imminent parental death, modified parent consumer satisfaction and fidelity measures, one focussed on parent-child communication and one related to recent experiences. Qualitative studies variously used semi-structured interview, focus group, unstructured autoethnographic and open-ended survey approaches targeting a variety of issues. INSERT TABLE 2 HERE Page 9 of 64

The included studies were conducted in the USA (n=14), Germany (n=7), the Netherlands (n=7), Australia (n=5), UK (n=4), Norway (n=3), Sweden (n=3), Finland (n=2), Canada (n=1), Denmark (n=1), France (n=1), Israel (n=1), South Korea (n=1) and Turkey (n=1). Two multinational studies were identified which utilised data collected in six European countries. Studies predominantly focussed on adolescents, with twenty-six studies restricting the sample to this age group (15, 17, 18, 21, 24, 2830, 33, 37-45, 47, 51, 56-59, 62, 70), seventeen studies spanning childhood and adolescence(19, 20, 26, 31, 32, 34, 52-55, 60, 61, 63-66, 68), seven using a combined adolescent and young adult sample(21, 22, 25, 49, 50, 67, 71) and four including only young adults (14, 23, 27, 35).

The age ranges attached to the descriptors child, adolescent and young adult varied between studies, though most considered adolescents to include those between 11 and 18 years of age and those 18 and over to be young adults. An upper limit for young adulthood was not specified in most studies; however some included an upper age limit for the study of 24 years(22, 48, 50, 67). Six studies also divided their sample into younger (e.g. 12-17 years) and older (e.g. 18 to 24 years) adolescents for the purposes of analysis, however the age ranges utilised were not consistent (22, 37, 44, 51, 67, 71).

The majority of studies included mothers and fathers with heterogeneous cancer diagnoses and both male and female AYA participants, though mothers were over-represented: on average, 76% of AYA samples were affected by maternal cancer. Six studies did not indicate whether parents with cancer were mothers or fathers. Ten studies only included mothers with cancer; nine restricted participation to mothers with breast cancer and their male or female offspring(25, 33, 36, 37, 42, 54, 55, 61, 65), two restricted participation to adolescent females and their mothers with breast cancer(25, 42) and one included male and female offspring and mothers with any cancer diagnosis(41). None restricted the sample to fathers or male offspring only. Adolescent and young adult samples were predominantly female (mean = 61%), with all but eight studies(17, 18, 24, 28, 35, 57, 63, 65)having a majority female sample and three studies(23, 25, 42) having a sample comprised entirely of female Page 10 of 64

adolescent or young adult participants. One qualitative study relied solely on parental accounts, and did not report the gender of participants’ offspring (31).

Twenty-one studies collected data only from AYA offspring, eight from AYA offspring and their parent with cancer, eleven from AYA offspring and both their parent with cancer and their well parent, and the remaining three collected data from support organisation staff (67), teachers and student peers(54), or parents only (31). Self-reports of the young person’s functioning were compared to reports of their functioning obtained from their parents in sixteen studies(25, 26, 38, 40, 41, 44, 49, 52, 53, 55-57, 59, 62-64), with the AYA self-reports suggesting worse outcomes for the young person than parent reports.

Positive and negative aspects of parental cancer Twenty-nine studies discussed negative psychosocial impacts of parental cancer. Offspring were typically found to have increased levels of internalising and externalising problems (37, 41, 44, 49, 53, 63, 70), distress(14, 20, 25, 27, 32, 40, 42, 51, 67), anxiety(45, 58, 60), depression(45, 51, 58, 60), stress(37, 40, 41), and cancer-related worry(18, 20, 23, 25, 27-29), as well as experiencing poorer health-related quality-of-life(70) and self-esteem(60). Qualitative research further suggested that many offspring face feelings of isolation (14, 17, 18, 29, 39), loss of control (27, 29), and the incomprehensibility of their situation (14, 17), but that the impact on offspring was often overlooked by family members and medical staff (15). However, not all studies indicated negative impacts; four studies suggested that offspring impacted by parental cancer did not differ from controls on measures of psychological, social, and family functioning(43, 47, 54, 70). Another study returned mixed findings, suggesting that offspring of cancer patients tend to have similar or lower levels of internalising and externalising problems compared to population norms, but were significantly more likely to have clinically-elevated scores on these measures (64).

Page 11 of 64

Eight studies identified positive impacts and experiences associated with the offspring cancer experience(15, 17, 20, 27-30, 38), with two of these papers explicitly using Tedeschi and Calhoun’s model of post-traumatic growth as a theoretical framework (72). Four of the five domains of posttraumatic growth (72) were evidenced in these studies: strengthened interpersonal relationships, especially amongst the family(20, 27, 28, 30, 38); enhanced appreciation of life and other people(20, 27, 28, 30); personal growth and development(15, 17, 27-30); and positive changes in goals and priorities(17, 28), including an increased interest in health and improved health behaviours(27, 30).

Needs Five studies primarily explored the needs of offspring, although needs were frequently reflected in other papers. The majority of these studies were conducted in Australia and reported on the development(67), validation(50), and predictive value (69) of the 47-item Offspring Cancer Needs Instrument (OCNI). The seven domains of this measure cover the major needs reported by offspring impacted by parental cancer, which were echoed in other studies. The most commonly reported unmet needs concerned support from friends (14, 21, 68) and other young people in similar situations (15, 21, 50), practical assistance (e.g. in educational or vocational settings) (18, 68), and in dealing with feelings(17, 21). Other key domains of need included information (19, 21, 26, 68), family support (14, 21), and time out from the cancer situation(21). These findings were largely reflected in studies of offspring bereaved by parental cancer, although that population appeared to have additional needs related to being informed of the ill parent’s prognosis (35) and having space to grieve (22). Predictive modelling demonstrated that higher levels of unmet needs were associated with increased distress (50, 69) and impaired psychosocial functioning (67). Additionally, offspring who have a father with cancer and those with poorer family functioning tended to report more unmet needs, and needs appear to decrease with increased time since diagnosis(69).

Communication & Information Page 12 of 64

Despite being identified as important factors for offspring, there was considerable variation in the amount of communication and information that young people desired and received (19, 29). While one qualitative study found that most families communicated with openness, trust, and honesty (18), two quantitative studies found that communication was worse in families affected by parental cancer than controls (27, 36). Familial attitudes, beliefs, and comfort in discussing the cancer situation may be influential factors (18) with both parents (19, 26, 31) and children (32) reporting difficulty in talking about the illness and wanting guidance on how to approach the subject (26). Family communication challenges may contribute to discrepancies between young people’s desired and received levels of information: offspring considered being well-informed about the cancer to be important (26, 33), but a substantial proportion still reported unmet needs in this domain (22, 67). Kennedy (19) noted that offspring require information about all aspects of the cancer experience, including details of treatment, the parent’s prognosis, and how they can help. Offspring primarily rely on their parents for this information, although health-care professionals are also valued information sources (19, 68). These needs were greatest immediately post-diagnosis, and decrease with time since diagnosis (19, 69). Finally, two studies explored the impact of communication and information on psychosocial outcomes, finding that poor family communication is associated with increased internalising, externalising, and stress responses (41, 47).

Coping Strategies Young people reported the need to find different ways of coping with the threat of parental cancer and its impact on their lives (17). Seeking support was one of the most frequently reported strategies (14, 15, 21, 28, 29, 44, 58), as was spending more time with the family (15, 17, 21, 24, 25, 29, 33). These support-seeking strategies were generally perceived to be effective coping mechanisms, and appear to be associated with better mental health outcomes (44). By contrast, avoidance-based strategies such as familial (25, 29, 58) and social withdrawal (44) were associated with poorer psychosocial outcomes (44, 58), although one qualitative study suggested that this “turning inwards” Page 13 of 64

may be done to avoid burdening others (33). Offspring who are able to accept the situation and their reactions to it (58) and discuss this with others (21, 28) may have better outcomes (44, 58) than those who cope through denial (17), suppress their emotions (18, 26), or avoid discussing the situation (17, 26). Distraction-based coping approaches such as thinking about other things (15, 18, 21, 26, 28, 33) and spending time away (20, 21) have less clear effects. Studies variously suggested that they may reflect maladaptive avoidance strategies (26) or adaptive attempts to preserve normality (20, 21, 68). Other coping strategies involve turning to religion or spirituality (73), thinking positively (21, 26, 28), or active problem-solving (44). The choice of strategy may be based on perceived efficacy, with the frequency of use of these approached being correlated with ratings of their effectiveness (44). Several studies reported that young people cycle between active and avoidant approaches, suggesting that they are unlikely to stick to one strategy or even one class of strategies (25, 29).

Interventions Six publications were identified which presented data on five interventions for offspring affected by parental cancer (31, 32, 34, 59, 65, 66). Two interventions primarily aimed to improve familial communication and relationships. Davey and colleagues (59) reported on a five-session culturallyadapted family intervention for African-American families affected by parental cancer, although little detail of the content and structure of the intervention and psychoeducational control were provided. Parents who completed the intervention reported significantly better family communication and children reported higher satisfaction than the control group; however, no changes were seen in participants’ depression or anxiety levels. Bugge and colleagues’ Family Support Program (31, 32) similarly aimed to improve communication, understanding, security and planning in families affected by a parent’s terminal cancer diagnosis. Qualitative feedback from participating parents and children indicated that communication had improved, both within the family and with external support

Page 14 of 64

networks (31, 32); children additionally reported improved knowledge about the illness, and identified the benefits of psycho-education about coping with emotions (32).

Other interventions had different foci; for example, John and colleagues’ (65) residential rehabilitation program aimed to prevent the development of emotional and behavioural problems in offspring affected by maternal breast cancer. Compared to the pre-intervention control period, larger improvements in both mothers’ and children’s psychological (and particularly, emotional) functioning were observed. However in all of these studies, post-intervention evaluation occurred soon after the completion of the final session and it is not clear whether the observed benefits would endure over time. By contrast, Tucker and colleagues’ (34) evaluation of the On Belay program involved focus groups with children (and their parents) who had participated in the program recently and more distally. While this intervention didn’t explicitly aim to meet the psychosocial needs of offspring, participants enjoyed the opportunity to take time out from the cancer situation and connect with others with similar experiences. Additionally, through mastering challenges in the adventure-based program, offspring were able to reflect on their strengths and how these could be used to cope with the impact of their parent’s cancer. However, as with the Family Support Program (31, 32), the lack of quantitative data makes it difficult to evaluate whether there was a significant improvement in psychosocial outcomes for families participating in these interventions – and if so, which program components were responsible. Lastly, Schmitt and colleagues (66) solicited feedback from parents involved in their familial counselling service. The majority reported that these sessions should be incorporated into the standard treatment of parents newly diagnosed with cancer and appeared to also agree that these services should begin in the months soon after diagnosis. However, the lack of clarity in reporting of their methods and results complicates the interpretation and application of these findings.

Family Functioning and Other Predictors Page 15 of 64

Family functioning has been explored both as a domain affected by parental cancer, and a predictor of the psychosocial impact of parental cancer on offspring. Specifically, offspring who rate their family as more functional, with better communication and cohesion and less conflict, tend to report significantly fewer internalising and externalising problems (37, 47, 49, 53, 70), lower distress (69), decreased stress responses (37), fewer unmet needs (69), and better health-related quality-of-life (70). In several studies, family dysfunction was the best predictor of negative impacts (37, 49, 53, 70). Offspring outcomes also appear to be associated with the wellbeing of other family members, with the impact of parental cancer variously predicted by parental depression, (37, 53, 57) anxiety (41, 57), adjustment (55), and general physical and psychosocial functioning (47, 62, 63). Similarly, parental marital distress appears to be significantly associated with internalising problems in offspring (41). Attachment styles within the family may also contribute to the psychosocial impact of cancer, with maternal anxious attachment and daughters’ avoidant attachment both significantly associated with increased levels of distress(42). Moreover, these family factors are likely to be interdependent; Schmitt and colleagues (52) noted that depressive symptoms in the parent with cancer were the best predictor of impaired family functioning, suggesting a potential dual impact on offspring.

Findings regarding the impact of AYA offspring age and gender were somewhat inconsistent. Most studies reported that female offspring fared significantly worse, particularly with respect to internalising symptoms (37, 39-41, 44, 56, 58, 60, 69). However, two studies suggested that males were more affected across a number of psychosocial domains (45, 49). Another two studies found no gender differences (38, 70), while one returned mixed findings depending on whether the analyses used adolescents’ self-reports or parent reports (44). Research on age differences in outcomes is similarly inconclusive. Studies variously report that the impact of parental cancer increases (41, 51) or decreases with age (49). Others find no significant age-effects (38, 40, 44, 56) or that effects vary depending on the outcome examined (70). Few studies investigated the impact of the ill parent’s Page 16 of 64

gender, and those which did returned conflicting results about whether the offspring of ill fathers (53, 69) or ill mothers fared worse (60). One found that the daughters of ill fathers fared worst of all (53), while two found no significant effects of parent gender (47, 56).

Disease factors may also predict psychosocial outcomes. While sampling and reporting issues limit the ability to identify the effects of cancer type, the effects of disease stage have been investigated. Offspring of parents whose disease has recurred may be significantly more negatively impacted (41, 56), although this has not been consistently replicated (38). Offspring of parents receiving palliative care also appear to have significantly better psychosocial functioning that children of those at other stages of their disease (70). Findings are similarly inconclusive regarding whether outcomes improve (40, 64, 69) or remain stable over time (38, 41, 56, 57), with one study returning mixed results depending on the outcome examined (39).

Discussion: Despite inconsistencies in the specific aspects of psychosocial functioning measured in the existing literature, it is clear that a parental cancer diagnosis can have a substantial and highly disruptive impact on their AYA offspring (14, 18, 20, 23, 25, 27-29, 37, 40-42, 44, 45, 49, 51, 53, 58, 60, 63, 64, 67, 70). It must be noted however that these findings are by no means universal, with several studies identifying positive psychosocial outcomes related to post-traumatic growth(15, 17, 20, 27-30, 38) and five returning null or mixed findings(43, 47, 54, 64, 70). Consistent with previous reviews, additional evidence has emerged between 2007 and 2016 reinforcing the relationship between poorer family functioning and worse psychosocial outcomes for AYAs(37, 46, 47, 49, 53, 71). Familyrelated factors, such as the wellbeing of other family members and parental attachment styles, also had predictive value(37, 41, 47, 49, 53, 55, 57, 62, 63, 69, 70), supporting earlier suggestions that offspring psychosocial outcomes may be more reflective of how families are coping (7, 74).

Page 17 of 64

Beyond this, findings regarding the specific character and level of impact of parental cancer on AYAs and the predictive value of factors such as age, gender and parental illness characteristics are less consistent. In contrast to previous reviews (6, 7) where the included studies predominantly sampled North American mothers with breast cancer, the majority of studies presented in this review included mothers and fathers with heterogeneous cancer diagnoses, both male and female AYA participants, and patients with both curable and later stage diagnoses. These studies were conducted in fourteen different countries, including two multinational studies surveying six European nations each. Perhaps as a result of this broader and more representative sample, some findings from the previous reviews were not consistent with the recent literature.

While additional evidence has emerged of a greater negative emotional impact experienced by adolescent daughters due to parental cancer(37, 39-41, 44, 48, 53, 56), several studies challenged this finding. These variously indicated that males may fare worse than females(34,42), that gender has no effect upon offspring’s psychosocial outcomes(26,33,35) or that there may be an interaction between parent and offspring gender(53). The impact of age similarly requires further investigation with mixed findings regarding whether older or younger offspring fare better (41, 48, 49, 51). These inconsistencies may in part be driven by considerable variations in measures used across studies, with a total of sixty-one outcome and predictor variables used. The most commonly assessed area was psychological functioning, which was variously characterised as internalising or externalising problems, distress, depression, anxiety, stress, post-traumatic stress, worry, coherence, emotional reactions or general mental health. These constructs were measured by twenty-four different validated scales, complicating efforts to draw specific conclusions regarding how parental cancer impacts AYA offspring. Additionally, the referent age ranges for adolescents and young adults varied considerably and the field of adolescent and young adult health as a whole would greatly benefit from international consistency and a clear rationale regarding the age range of this population.

Page 18 of 64

Similarly, indications that medical and treatment variables had little predictive value appear to be challenged by the finding that offspring of parents with palliative stage cancer had better functioning than offspring of parents with earlier stages of cancer(46), perhaps as a result of the holistic and family oriented approach to care offered by palliative care services, which may include the provision of psychosocial care to offspring. This finding requires further exploration and replication given that it contrasts with other findings that communication about impending parental death is difficult (19) and often limited (35), and that poor communication and uncertainty are associated with worse coping and psychosocial outcomes (26, 39, 41). Indications that cancer recurrences had a greater effect than primary diagnoses are consistent with previous reviews (7, 8), perhaps because these offspring have a better understanding of the severity and implications of the disease (75). However, clarifying the impact, predictors and sequelae of parental cancer on AYA offspring is somewhat challenging given inconsistencies in the reporting of parental cancer progression. While most studies gave time since diagnosis or staging information (n=39) this was frequently expressed as a broad range or descriptive stage (e.g. ‘diagnosed within the last 1-5 years’ or ‘with advanced cancer’).

Survivorship was only exclusively investigated in one study, however it did not clarify whether the impact on offspring is different during survivorship as compared to acute treatment, advanced or bereavement settings (42). The results of studies on the impact of advanced parental cancer suggested that family members’ main concern was not having enough time together(24), and that older adolescents living with a parent with advanced cancer may experience higher distress and depression than AYAs in the general population(51). However, several qualitative studies of these offspring suggested that their experiences were largely similar to those of offspring with parents with cancer at others stages (19, 20, 28, 29), and that AYAs whose parent had palliative stage cancer may actually have better outcomes (46). Only five studies focussed exclusively or in part on bereaved offspring (14, 17, 22, 27, 35); the majority of these were small-scale qualitative investigations, making it difficult to draw strong conclusions. The sole population-based survey of bereaved Page 19 of 64

offspring did, however, indicate that communication regarding the parent’s imminent death is often poor and young people are often not given adequate time to prepare for their parent’s death (35). This is consistent with findings from two qualitative studies evaluating an intervention for offspring affected by terminal parental cancer, which suggested that these children may benefit from interventions aiming to improve communication and coping (31, 32).

Few evaluations of interventions for AYAs impacted by parental cancer were identified in this systematic review (31, 32, 34, 59, 65, 66), and relatively little detail was provided regarding program structure and content. Evaluation methods tended to be rather limited, largely relying on qualitative or un-validated measures. Study designs made it difficult to assess whether observed outcome improvements were significant, if these benefits were maintained over time, and which features of the interventions contributed to these effects. Future evaluations should adopt more rigorous approaches to study design, utilise validated measures and explore long-term efficacy. Such evidence will be vital in seeking resources to support implementation of effective interventions. A protocol describing the evaluation of a manualised, group-based acceptance and commitment therapy program has also been published (76), however results are yet to be reported. Given the substantial impact of parental cancer on offspring, there is a clear need for interventions and clinical practice guidelines to help healthcare services and healthcare professionals to identify these vulnerable young people and provide age-appropriate information and support. Relative to Inhestern and colleagues’ recent review of interventions for families affected by parental cancer (11), the small number of publications meeting this paper’s inclusion criteria highlights the lack of contemporary, age-appropriate, and demonstrably effective interventions for AYA offspring.

As recommended by Ellis and colleagues (13), the development of interventions for this population should take into account their psychosocial needs. Development, validation and assessment of the predictive value of the Offspring Cancer Needs Instrument has provided much needed systematic Page 20 of 64

insight into the specific needs of young people living with a parent’s cancer (50, 67, 69), which may useful in formulating targeted interventions. Rigorous and repeatable identification of unmet needs has promising clinical implications, with validated instruments providing opportunities for healthcare professionals to better assist these young people, and to identify gaps and successes in service delivery over time. The relationships identified between needs, distress, demographic factors and family functioning will be useful for identifying at-risk offspring and providing support where it is most likely to improve psychosocial outcomes (50, 67, 69). A measure of needs for bereaved offspring currently in development will be a valuable counterpart to this measure (77).

Some previously identified methodological criticisms have been addressed in the recent literature. In comparison to the previous reviews where small sample sizes were noted as a limitation, 27 of the 54 studies included in this review had a sample size greater than 100. Several methodological quality issues were nonetheless apparent in this body of literature. The majority of studies were conducted exclusively in hospital or clinical settings (n=33), with only twelve conducted in community settings. Seven utilised both clinical and community-based recruitment methods and two did not clearly indicate where the study was conducted. This bias towards hospital-based recruitment may limit opportunities to generalise findings beyond this context and greater focus on community-based studies may yield useful insights into the impact of parental cancer on young people who do not attend the treatment centre with their ill parent.

Response rates were not reported in more than half of the included studies (25 of 54) and particularly poorly reported for qualitative studies, with only 6 of 22 including this information. Assessing the potential for response biases is thus challenging. In 7 of the 25 studies for which it was reported, the response rate was 50% or less, with some reporting rates under 30%. Studies conducted in hospital settings appeared to yield good response rates, while studies in communitybased settings reported comparatively lower rates. This may, however, be indicative of the Page 21 of 64

challenges inherent in engaging a population spread across a broad geographical area through nonface-to-face recruitment modalities rather than shortcomings in study methodology. Lastly, while not uncommon in psychosocial research(78), study populations were heavily gender biased, with a mean of 61% female participants, making generalisation of findings to male AYA offspring challenging.

There is substantial culturally-based heterogeneity in how patients and families respond to severe or life-limiting illness, and their health outcomes are often worse than those of the dominant cultural group(79). However, while the included studies spanned fourteen countries, the majority of these were Western and no studies presented data substantially drawn from minority cultural or linguistic groups. Moreover, the variability in methods and measures used complicates comparisons between countries, making it unclear whether observed differences in results – for example, that Turkish males had worse outcomes than females (45), counter to most studies in Western countries – are attributable to cultural differences or methodological variation. Models of care developed from the existing literature to address the needs of AYAs with a parent with cancer may have limited relevance in non-western countries or in western countries with increasing cultural diversity; however, it is difficult to establish this conclusively without comparative data. While this review indicates some progress has been made, greater efforts are still crucial to ensure culturally and linguistically diverse perspectives are represented in future research and models of care.

The variety of outcomes and measures used in the included studies created difficulty in integrating the findings from this area and precluded conducting a meta-analysis. Efforts to establish a consensus on a core set of outcomes and measures to be used in future studies could facilitate future meta-analyses and the establishment of large-scale cooperative data-sets. Additionally, only five longitudinal studies were identified in this context (39, 40, 43, 57). The absence of further longitudinal studies is a significant gap in the literature and this field would benefit from exploration of the impact of parental cancer on offspring over time. While predictors of AYA offspring outcomes Page 22 of 64

have been established to varying degrees in a research context, the clinical utility and significance of these findings is yet to be explored.

Finally, substantial challenges were faced in identifying relevant literature due to inconsistencies in terminology, keywords and indexing of studies. Various studies used the terms offspring and children interchangeably to refer to progeny of parents diagnosed with cancer. Others used the term children to refer to minors less than 12 or 18 years of age. The inclusion of the term “child*” in the current search strategy substantially increased the identification of irrelevant literature related to the impact of cancer on patients who were minors. Greater consistency in terminology, keyword assignment and indexing will assist with future reviews.

Conclusion The literature published since 2007 provides additional evidence for the negative psychosocial impact of parental cancer on adolescent and young adult offspring and the predictive value of family functioning and other family and illness-related factors on these outcomes. Nonetheless, substantial gaps and methodological issues remain and evidence for the development, efficacy or implementation of interventions for this population is very limited. There is also a clear need for greater focus on bereaved and young adult offspring and those from non-western cultural groups, who have been under-represented in research conducted to date. Clinical practice guidelines to assist healthcare professionals in identifying and supporting young people impacted by parental cancer are also clearly needed, along with evidence regarding how best to provide this support in healthcare and community contexts. It will be essential to work towards a consensus and greater consistency regarding the measurement of key outcomes such as psychosocial well-being to facilitate future meta-analyses and establishment of large, co-operative datasets. Finally, with the growth of AYA offspring-focussed research beyond the North America, the establishment of international

Page 23 of 64

working groups is recommended to establish a clear research agenda and drive improvements in policy and provision of support to this vulnerable and underserved population.

Page 24 of 64

References: 1. Australian Institute of Health and Welfare A. Australian Cancer Incidence and Mortality Books. Canberra: AIHW; 2011. 2. de Vaus D. Diversity and change in Australian families: Statistical profiles. Melbourne, Australia: Australian Institute of Family Studies; 2004. 3. Australian Bureau of Statistics (ABS). Census QuickStats.: ABS; 2011 [Available from: http://www.censusdata.abs.gov.au/census_services/getproduct/census/2011/quickstat/0. 4. Palmer S, Patterson P, Thompson K. A national approach to improving adolescent and young adult (AYA) oncology psychosocial care: the development of AYA-specific psychosocial assessment and care tools. Palliat Support Care. 2014;12(3):183-8. 5. Zebrack B. Psychological, Social, and Behavioral Issues for Young Adults With Cancer. Cancer. 2011;117(10):2289-94. 6. Grabiak BR, Bender CM, Puskar KR. The impact of parental cancer on the adolescent: An analysis of the literature. Psychooncology. 2007;16(2):127-37. 7. Osborn T. The psychosocial impact of parental cancer on children and adolescents: A systematic review. Psychooncology. 2007;16(2):101-26. 8. Krattenmacher T, Kuhne F, Ernst J, Bergelt C, Romer G, Moller B. Parental cancer: Factors associated with children's psychosocial adjustment-A systematic review. Journal of Psychosomatic Research. 2012;72(5):344-56. 9. Phillips F. Adolescents living with a parent with advanced cancer: A review of the literature. Psychooncology. 2014;23(12):1323-39. 10. Morris JN, Martini A, Preen D. The well-being of children impacted by a parent with cancer: An integrative review. Support Care Cancer. 2016;24:3235-51. 11. Inhestern L, Haller A-C, Wlodarczyk O, Bergelt C. Psychosocial interventions for families with parental cancer and barriers and facilitators to implementation and use - A systematic review. PLoS ONE. 2016;11(6):e0156967. 12. Huang X, O'Connor M, Lee S. School-aged and adolescent children's experience when a parent has non-terminal cancer: A systematic review and meta-synthesis of qualitative studies. Psychooncology. 2014;23:493-506. 13. Ellis SJ, Wakefield CE, Antill G, Burns M, Patterson P. Supporting children facing a parent's cancer diagnosis: A systematic review of children's psychosocial needs and existing interventions. Eur J Cancer Care. 2017;26:e12432. 14. Karlsson E, Andersson K, Ahlström BH. Loneliness despite the presence of others Adolescents' experiences of having a parent who becomes ill with cancer. European Journal of Oncology Nursing. 2013;17:697-703. 15. Davey MP, Tubbs CY, Kissil K, Niño A. 'We are survivors too': African-American youths' experiences of coping with parental breast cancer. Psychooncology. 2011;20:77-87. 16. Kmet LM, Lee RC, Cook LS. Standard quality assessment criteria for evaluating research papers from a variety of fields. Alberta: Canada: Alberta Heritage Foundation for Medical Research. 2004. 17. Dehlin L, Martensson L. Adolescents' experiences of a parent's serious illness and dealth. Palliative and Supportive Care. 2009;7:13-25. 18. Finch A, Gibson F. How do young people find out about their parent's cancer diagnosis: A phenomonological study. European Journal of Oncology Nursing. 2009;13(3):213. 19. Kennedy VL, Lloyd-Williams M. Information and communication when a parent has advanced cancer. Journal of Affective Disorders. 2009;114(1):149-55. 20. Kennedy VL, Lloyd-Williams M. How children cope when a parent has advanced cancer. Psychooncology. 2009;18:886-92. 21. Maynard A, Patterson P, McDonald FEJ, Stevens G. What is helpful to adolescents who have a parent diagnosed with cancer? J Psychosoc Oncol. 2013;31(6):675-97. Page 25 of 64

22. Patterson P, Rangganadhan A. Losing a parent to cancer: a preliminary investigation into the needs of adolescents and young adults. Palliat Support Care. 2010;8(3):255-65. 23. Puterman J, Cadell S. Timing is everything: The experience of parental cancer for young adult daughters - A pilot study. J Psychosoc Oncol. 2008;26(2):103-21. 24. Sheehan DK, Draucker CB. Interaction patterns between parents with advanced cancer and their adolescent children. Psychooncology. 2011;20(10):1108-15. 25. Stiffler D, Barada B, Hosei B, Haase J. When Mom has breast cancer: adolescent daughters' experiences of being parented. Oncology Nursing Forum. 2008;35(6):933-40. 26. Thastum M, Johansen MB, Gubba L, Olesen LB, Romer G. Coping, social relations, and communication: a qualitative exploratory study of children of parents with cancer. Clinical Child Psychology and Psychiatry. 2008;13(1):123-38. 27. Wong ML, Cavanaugh CE, MacLeamy JB, Sojourner-Nelson A, Koopman C. Posttraumatic growth and adverse long-term effects of parental cancer in children. Families, Systems, & Health. 2009;27(1):53-63. 28. Phillips F. The experience of adolescents who have a parent with advanced cancer: A phenomenological inquiry. Palliat Support Care. 2015;13(4):1057-69. 29. Phillips F, Lewis FM. The adolescent's experience when a parent has advanced cancer: A qualitative inquiry. Palliat Med. 2015;29(9):851-8. 30. Kissil K, Niño A, Jacobs S, M. D, Tubbs CY. "It has been a good growing up experience for me": Growth experiences among African American youth coping with parental cancer. Families, Systems and Health. 2010;28(3):274-89. 31. Bugge KE, Helseth S, Darbyshire P. Parents' experience of a family support program when a parent has incurable cancer. J Clin Nurs. 2009;18(24):3480-8. 32. Bugge KE, Helseth S, Darbyshire P. Children's experiences of participation in a family support program when their parent has incurable cancer. Cancer Nurs. 2008;31(6):426-34. 33. Clemmens DA. The significance of motherhood for adolescents whose mothers have breast cancer. Oncology Nursing Forum. 2009;36(5):571-7. 34. Tucker AR, Sugerman D, Zelov R. On Belay: Providing connection, support, and empowerment to children who have a parent with cancer. Journal of Experiential Education. 2013;36(2):93-105. 35. Bylund-Grenklo T, Kreicbergs U, Uggla C, Valdimarsdottir UA, Nyberg T, Steineck G, et al. Teenagers want to be told when a parent's death is near: A nationwide study of cancer-bereaved youths' opinions and experiences. Acta Oncologica. 2015;54(6):944-50. 36. Cho OH, Yoo YS, Hwang KH. Comparison of parent-child communication patterns and parental role satisfaction among mothers with and without breast cancer. Applied Nursing Research. 2015;28(2):163-8. 37. Edwards L, Watson M, St. James-Roberts, Ashley S, Tilney C, Brougham B, et al. Adolescent’s stress responses and psychological functioning when a parent has early breast cancer. Psychooncology. 2008;17:1039–47. 38. Gazendam-Donofrio SM, Hoekstra HJ, Van der Graaf WT, Van de Wiel HB, Visser A, Huizinga GA, et al. Family functioning and adolescents' emotional and behavioural problems: when a parent has cancer. Annals of Oncology. 2007;18(12). 39. Gazendam-Donofrio SM, Hoekstra HJ, van der Graaf WTA, van de Wiel HB, Visser A, Huizinga GA, et al. Adolescents' emotional reactions to parental cancer: Effect on emtional and behavioral problems. Journal of Pediatric Psychology. 2011;36(6):346-59. 40. Huizinga GA, Visser A, van der Graaf WTA, Hoekstra HJ, Gazendam-Donofrio SM, HoekstraWeebers JEHM. Stress response symptoms in adolescents during the first year after a parent's cancer diagnosis. Support Care Cancer. 2010;18(11):1421-8. 41. Huizinga GA, Visser A, Van der Graaf WT, Hoekstra HJ, Stewart RE, Hoekstra-Weebers JE. Family-oriented multilevel study on the psychological functioning of adolescent children having a mother with cancer. Psychooncology. 2011;20(7):730-7. Page 26 of 64

42. Inbar C, Ety B, Ayala H, Tamer P. The mental health of breast cancer survivors and their adolescent daughters. Psychooncology. 2013;22(6):1236-41. 43. Jantzer V, Gross J, Stute F, Parzer P, Brunner R, Willig K, et al. Risk behaviors and externalizing behaviors in adolescents dealing with parental cancer - a controlled longitudinal study. PsychOncology. 2013;22:2611-6. 44. Krattenmacher T, Kühne F, Führer D, Beierlein V, Brähler E, Resch F, et al. Coping skills and mental health status in adolescents when a parent has cancer: A multicenter and multi-perspective study. Journal of Psychosomatic Research. 2013;74:252-9. 45. Küçükoglu S, Çelebioglu A. Identification of psychological symptoms and associated factors in adolescents who have a parent with cancer in Turkey. European Journal of Oncology Nursing. 2013;17:75-80. 46. Kuhne F, Krattenmacher T, Bergelt C, Ernst JC, Flechtner HH, Klitzing K, et al. Parental palliative cancer: Psychosocial adjustment and health-related quality of life in adolescents participating in a German family counselling service BMC Palliative Care. 2012;11(1):21. 47. Lindqvist B, Schmitt F, Santalahti P, Romer G, Piha J. Factors associated with the mental health of adolescents when a parent has cancer. Scandinavian Journal of Psychology. 2007;48:34551. 48. McDonald FEJ, Patterson P, White KJ, Butow PN, Costa DSJ, Kerridge I. Correlates of unmet needs and psychological distress in adolescent and young adults who have a parent diagnosed with cancer. Psychooncology. 2015;2015. 49. Möller B, Barkmann C, Krattenmacher T, Kühne F, Bergelt C, Beierlein V, et al. Children of cancer patients: Prevalence and predictors of emotional and behavioral problems. Cancer. 2014;120:2361-70. 50. Patterson P, McDonald FEJ, Butow P, White KJ, Costa DSJ, Pearce A, et al. Psychometric evaluation of the Offspring Cancer Needs Instrument (OCNI): an instrument to assess the psychosocial unmet needs of young people who have a parent with cancer. Support Care Cancer. 2013;21(7):1927-38. 51. Rainville F, Dumont S, Simard S, Savard M. Psychological distress among adolescents living with a parent with advanced cancer. J Psychosoc Oncol. 2012;30:519-34. 52. Schmitt F, Piha J, Helenius H, Baldus C, Kienbacher C, Steck B, et al. Multinational study of cancer patients and their children: Factors associated with family functioning. J Clin Oncol. 2008;26(36):5877-83. 53. Thastum M, Watson M, Kienbacher C, Piha J, Steck B, Zachariae R, et al. Prevalence and predictors of emotional and behavioural functining of children where a parent has cancer. Cancer. 2009;115:4030-9. 54. Vannatta K, Grollman JA, Noll RB, Gerhardt CA. Impact of maternal breast cancer on the peer interactions of children at school. Psychooncology. 2008;17(3):252-9. 55. Vannatta K, Ramsey RR, Noll RB, Gerhardt CA. Associations of child adjustment with parent and family functioning: Comparison of families of women with and without breast cancer. Journal of Developmental and Behavioral Pediatrics. 2010;31(1):9-16. 56. Visser A, Huizinga GA, Hoekstra HJ, van der Graaf WTA, Hoekstra-Weebers JEHM. Temperament as a predictor of internalising and externalising problems in adolescent children of parents diagnosed with cancer Support Care Cancer. 2007;15(4):395-403. 57. Götze H, Ernst J, Brahler E, Romer G, von Klitzing K. Predictors of quality of life of cancer patients, their children, and partners. Psychooncology. 2015;24(7):787-95. 58. Costas-Muñiz R. Hispanic adolescents coping with parental cancer. Support Care Cancer. 2012;20:413-7. 59. Davey MP, Kissil K, Lynch L, Harmon L, Hodgson N. A culturally adapted family intervention for African American families coping with parental cancer: Outcomes of a pilot study. Psychooncology. 2013;22:1572-80. 60. Flahault C, Sultan S. On being a child of an ill parent: A Rorschach investigation of adaptation to parental cancer compared to other illnesses. Rorschachiana. 2010;31:43-69. Page 27 of 64

61. Foran-Tuller K, O'Hea EL, Moon S, Miller SJ. Posttraumatic stress symptoms in children of mothers diagnosed with breast cancer. J Psychosoc Oncol. 2012;30(1):41-56. 62. Gazendam-Donofrio SM, Hoekstra HJ, van Der Graaf WTA, Pras E, Visser A, Huizinga GA, et al. Quality of life of parents with children living at home: When one parent has cancer. Support Care Cancer. 2008;16(2):133-41. 63. Krattenmacher T, Kühne F, Halverscheid S, Wiegand-Grefe S, Bergelt C, Romer G, et al. A comparison of the emotional and behavioural problems of children of patients with cancer or a mental disorder adn their association with parental quality of life. Journal of Psychosomatic Research. 2014;76:213-20. 64. Visser A, Huizinga GA, Hoekstra HJ, van der Graaf WT, Gazendam-Donofrio SM, HoekstraWeebers JE. Emotional and behavioural problems in children of parents recently diagnosed with cancer: A longitudinal study. Acta Oncologica. 2007;46:67-76. 65. John K, Becker K, Mattejat F. Impact of family-oriented rehabilitation and prevention: An inpatient program for mothers with breast cancer and their children. Psychooncology. 2013;22:268492. 66. Schmitt F, Manninen H, Santalahti P, Savonlahti E, Pyrhönen S, Romer G, et al. Children of parents with cancer: A collaborative project between a child psychiatry clinic and an adult oncology clinic. Clinical Child Psychology and Psychiatry. 2007;12(3):421-36. 67. Patterson P, Pearce A, Slawitschka E. The initial development of an instrument to assess the psychosocial needs and unmet needs of young people who have a parent with cancer: piloting the Offspring Cancer Needs Instrument (OCNI). Support Care Cancer. 2011;19(8):1165-74. 68. Torp S, Thoresen L, Grønninsæter AB, Grov EK, Gustavsen K. Financial and social effects on children and adolescents when a parent is diagnosed with cancer. Child and Adolescent Social Work. 2013;30(4):293-310. 69. McDonald FEJ, Patterson P, White KJ, Butow PN, Costa DSJ, Kerridge I. Correlates of unmet needs and psychological distress in adolescent and young adults who have a parent diagnosed with cancer. Psychooncology. 2016;25(4):447-54. 70. Kühne F, Krattenmacher T, Bergelt C, Ernst JC, Flechtner H-H, Führer D, et al. Parent palliative cancer: Psychosocial adjustment and health-related quality of life in adolescents participating in a German family counselling service. BMC Palliative Care. 2012;11:21. 71. McDonald FEJ, Patterson P, White K, Butow PN, Costa DSJ, Kerridge I. Correlates of unmet needs and psychological distress in adolescent and young adults who have a parent diagnosed with cancer. Psychooncology. 2015. 72. Tedeschi R, Calhoun L. The post-traumatic growth inventory: Measuring the positive legacy of trauma. Journal of Traumatic Stress. 1996;9:45-471. 73. . !!! INVALID CITATION !!! {Costas-Muñiz, 2012 #1626;Jones, 2013 #520}. 74. Watson M, St. James-Roberts I, Ashley S, Tilney C, Brougham B, Edwards L, et al. Factors associated with emotional and behavioural problems among school age children of breast cancer patients. Br J Cancer. 2006;94:43-50. 75. Visser A, Huizinga GA, Hoekstra H, Van Der Graaf WTA, Klip EC, Pras E, et al. Emotional and behavioural functioning of children of a parent diagnosed with cancer: A cross-informant perspective. Psychooncology. 2005;14:746-58. 76. Patterson P, McDonald FEJ, Ciarrochi J, Hayes L, Tracey D, Wakefield CE, et al. A study protocol for Truce: A pragmatic controlled trial of a seven-week acceptance and commitment therapy program for young people who have a parent with cancer. BMC Psychology. 2015;3:31. 77. McDonald FEJ, Patterson P, Costa DSJ. Assessing unmet needs and distress in adolescents and young adults who have a parent or sibling die from cancer. 18th International Psycho-Oncology Society Congress; Dublin, Ireland2016, October. 78. Hoyt MA, Rubin LR. Gender representation of cancer patients in medical treatment and psychosocial survivorship research: Changes over three decades. Cancer. 2012;118(19):3824-4832.

Page 28 of 64

79. Anderson LM, Scrimshaw SC, Fullilove MT, Fielding JE, Normand J, Services TFoCP. Culturally competent healthcare systems: A systematic review. American Journal of Preventive Medicine. 2003;24(3S):68-79.

Page 29 of 64

Figure 1: PRISMA Study Flow Diagram

Records identified through database searches (n = 1720)

Additional records identified through other sources (n = 104)

Records after duplicates removed (n = 1626 + 104)

Records screened (n = 1730)

Records excluded (n = 1624)

Full-text articles assessed for eligibility (n = 106)

Full-text articles excluded (n = 52)

Studies included in qualitative synthesis (n = 54)

Page 30 of 64

Table 1: Key Study Information Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Bugge 2008 Norway

0.85

Design: Qualitative evaluation of a family-centred preventative support intervention Data Source: Interviews

(-) Demographics (-) In-depth, open-ended interviews about the experiences of children living with parental terminal cancer and their experiences with the Family Support Program – looking particularly at the program’s impact on psychosocial problems and coping

Bugge 2009 Norway

0.80

Design: Qualitative evaluation of a family-centred preventative support intervention Data Source: Interviews

 Offspring’s primary concerns were the fear that the ill parent would die, feeling unable to discuss the situation with parents, distress at the deterioration of the ill parent’s health, and misconceptions and fears about cancer as an illness.  Offspring who participated in the FSP appreciated the opportunity to talk to someone outside the family, and reported increased knowledge about the illness and improved communication with family and peers. Psycho-education about coping with emotions was identified as helpful.  The FSP helped and supported parents in preparing and support their children – which they reported previously wanting to do, but lacked knowledge and experience of how to do so.  Parents gained greater insight into their children’s thoughts and emotions through improved familial communication.  Parents appreciated the opportunity to talk through and plan future situations, and clarify supports available to the family.

BylundGrenklo 2014 Sweden

1.00

Design: Population-based survey Data Source: Questionnaires (nationwide population based survey)

Sample:  12 offspring (6-16yrs, Mage = 9 yrs; 33% male) with a parent with terminal cancer (58% fathers) Response Rate: 55% of families Recruitment Setting: Families who had completed the Family Support Program (FSP) Sample:  4 parents with terminal cancer (Mage = 42yrs; 3 months – 8yrs post-diagnosis, M = 2.3yrs) and 9 healthy parents/caregivers (total sample 46% male) with children aged 5-18yrs. Response Rate: 55% of families Recruitment Setting: Families who had completed the Family Support Program (FSP) Sample:  622 bereaved young adult offspring (18-26yrs, 50% male) whose parent died 7-10yrs prior Response Rate: 73% Recruitment Setting: Community recruitment via Swedish National Causes of Death Register

(-) Demographics, illness and treatment details (for parents with cancer) (-) Interviews with participating parents and caregivers about why they chose to attend the FSP, whether the programme met their needs, how it could be improved, and other types of support the family had received.

(-) Demographics (-) Purpose-designed questionnaire targeting disclosure of imminent parental death

 98% of participants indicated that teenage children should be informed of parent’s imminent death  59% had been informed of parent’s imminent death (37% by parent, 7% by parent & health care professional together, 8% by health care professional alone)  Reasons for non-disclosure of imminent death; either one or both of the parents “pretended illness was not that serious” or parents not aware of imminent death  43% of participants unaware of their parent’s imminent death in the hours preceding the death

Page 31 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Cho 2015 South Korea

1.00

Design: Case-comparison Data Source: Questionnaires

Sample:  51 mother (100% breast cancer, 18% <1yr, 36% 1-5yrs & 36% >1yr post surgical resection)/ offspring dyads (Mage=21.9yrs, 48% male)  50 healthy mother/offspring dyads (Mage=21.5yrs, 43% male) Response Rate: 73% breast cancer sample/71% healthy sample Recruitment Setting: University Hospital

(-) Demographics (+) Parent-Adolescent Communication Scale (Korean version) (+) Parental Role Satisfaction Tool

 A closed communication style was more common than an open communication style in both cancer affected and healthy parent/adolescent dyads, but was significantly higher in children of women with breast cancer than in children of healthy women  Overall Parental Role Satisfaction, satisfaction with parental role conflicts and satisfaction support from children was significantly lower in women with breast cancer than in healthy women

Page 32 of 64

Clemmens 2009 USA

0.85

Design: Qualitative Data Source: Semi-structured interviews

Sample: 11 adolescents (13-19yrs) whose mother had been diagnosed with breast cancer in the last year. Response Rate: 100% Recruitment Setting: Cancer clinics

(-) Semi-structured interviews exploring the experience of having a mother diagnosed with breast cancer

 Participants reported life-changing impacts of their mothers’ diagnoses, which interfered with their other activities.  Participants sought to focus on things other than the cancer situation, frequently “turning inwards” in order to avoid burdening or distressing their mothers.  Participants expressed strong desires to spend time with their mothers, although it was distressing to see their health decline.  Participants described moments of “normality” when their mothers had good days, and wished for a return to normal after treatment was completed.  Participants valued open communication about the situation; when uninformed, they worried and imagined the worst.

Page 33 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Costas-Muniz 2012 USA

0.73

Design: Cross-sectional Data Source: Questionnaires

Sample: 51 Puerto Rican adolescents (1218yrs, Mage = 15.5; 45% male) with a parent living with cancer (20% early phase, 28% advanced, 52% in remission) Response Rate: Approximately 83% Recruitment Setting: “Relay for Life” events organised by the American Cancer Society of Puerto Rico

(-) Demographics and medical information (+) Centre of Epidemiological Studies Depression scale (Spanish version) (+) Spielberger State Trait Anxiety Inventory (Spanish version) (+) Brief COPE (Spanish version)

 64% of participants reported clinical levels of depression  44% of participants reported high levels of anxiety  Most commonly used coping strategies: seeking emotional support, acceptance, and religion  Girls reported higher levels of depression than boys; no gender differences in anxiety or coping strategies.  Depression positively related to maladaptive coping strategies (denial, behavioural disengagement) and negatively related to positive coping strategies (acceptance, humour).  Anxiety negatively related to emotion-focused strategies (active coping, planning. acceptance) and positively related to dysfunctional coping strategies (denial, substance use, behavioural disengagement.  Dysfunctional coping strategies accounted for the most variance in depression (31%) and anxiety (29%) scores.

Page 34 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Davey 2011 USA

0.90

Design: Qualitative Data Source: Focus groups

Sample: 9 African-American families:  9 parents diagnosed with Stage I-III breast cancer in the last 2yrs (34-56yrs; 11% male)  12 adolescent offspring (1118yrs; 25% male) Response Rate: Not reported Recruitment Setting: Oncology clinics

(-) Demographics, including questions assessing spirituality and religiosity (-) Focus groups exploring the essential and culturally sensitive elements that should be incorporated into a familybased intervention: current concerns, support networks, family roles, coping with the cancer diagnosis, relationship with ill parent, changes due to parental cancer, and how others can help.

 Participants reported many ways of coping with their emotional reactions to the news of their parent’s diagnosis. These included avoidance/distraction, engaging in positive behaviours, self-care, seeking social support, and spirituality.  Participants felt overlooked by other family members and medical staff, who in focusing on the parent with cancer often missed the concerns of adolescent offspring.  Participants reported changes in family functioning, including the ill parent and family roles and interactions.  Participants reported some post-traumatic growth – finding strengths through coping with parental cancer situation.  When asked how clinicians could help, participants suggested a group where they could connect with others in the same situations, which would provide comfort and companionship. They identified food, games, parental encouragement, and opportunities for fun as incentives for participation.

Davey 2013 USA

0.75

Design: Non-randomized quasiexperimental pilot of a culturally-adapted family intervention Data Source: Questionnaires

Sample: 12 African-American families:  12 parents/guardians with cancer (8% male; 43% Stage I, 14% Stage II, 43% Stage III)  19 school-aged children (eligible age 10-18yrs; 42% male,) Response Rate: Not reported Recruitment Setting: Four urban oncology clinics; also advertised through cancer news outlets and church support groups

(-) Demographics (-) Attendance and treatment fidelity (-) Modified version of a Parent Consumer Satisfaction measure (+) Interaction Behaviour Questionnaire (-) New measure of general parent-child communication (+) Children’s Depression Inventory (+) Revised Children’s Manifest Anxiety Scale (+) Centre for Epidemiological Studies Depression Scale

 Parents/guardians who completed the culturally-adapted family intervention rated parent-child communication better, relative to those who completed the psycho-education intervention. There were no differences in child ratings.  Children who completed the culturally-adapted family intervention were more satisfied than those who completed the psycho-education intervention. There were no differences in parents’ ratings.  There were no significant changes in psychosocial distress (parent depression, offspring depression and anxiety, views of relationships) following the intervention, for either group.  NB. a cut-off of p ≤ .1 was used to determine significance.

Page 35 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Dehlin 2009 Sweden

1.00

Design: Qualitative Data Source: Semi-structured interviews

Sample:  5 bereaved adolescents (1618yrs at interview, 14-17 at bereavement; 60% male) whose parent died 0.67-3yrs prior Response Rate: 83% Recruitment Setting: Gynaecological/oncology clinics

(-) Semi-structured interview covering adolescents experiences of their parent’s death including the period of the parent’s illness, the time around the death and the time afterwards

 Parent’s illness seen as a strong threat, with adolescents understanding that their own/their family’s lives would be greatly changed by the illness/death.  Incomprehensibility of parent’s serious illness/death was threatening on its own, & adolescents sought to make sense of the illness/death to understand what was happening.  Adolescents described the necessity of finding different ways of relating to/managing the threat (e.g. restoring order, seeking closeness, adapting, gaining control, avoiding talking about the illness, not accepting & considering parent dead before they actually died)  The adolescents described feelings of being alone or alienated, even though they were close to family/friends & did not actively seek support.  Adolescents’ lives were changed by their experiences, beyond their bereavement over the parent - they felt they had become more mature than their friends & that there had been a change in their thinking about life, changes in values & changes in their views of relationships with other people.

Page 36 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Edwards 2008 UK

0.95

Design: Cross-sectional Data Source: Questionnaires

Sample:  56 adolescent offspring (1117yrs, Mdnage=15yrs, 32% male)  53 mothers (100% stage I/II breast cancer, 28% on chemo, time since diagnosis Mdn=8.5mths) Response Rate: 84% adolescents/79% mothers Recruitment Setting: Outpatient clinic at cancer treatment centre

(-) Demographics (+) Child Impact of Events Scale – C-IES (+) Youth Self-Report-Child Behaviour Checklist – YSR (+) Mental Health subscale of the Child Health Questionnaire – CHQ-MH; CF87Child Form (+) McMaster Family Assessment Device (FAD) (+) Family Environment Scale (FES) (+) Beck Depression Inventory (BDI-II) (+) SF8 version of the Medical Outcomes Health Survey

 41% of adolescents had high stress levels (males=33%, females=45%)  Females had significantly higher intrusion & avoidance scores  Psychological problems reported in 28% males & 32% females vs 16% in general American population  Poor family functioning & cohesiveness associated with adolescent emotional & behavioural problems & poor family cohesiveness is also associated with increased stress responses, specifically: Stress Response (C-IES)  Sig Univariate Predictors: poor family cohesion & general functioning predict higher intrusion/total scores, mothers being employed predict higher avoidance/total scores & caseness  Sig Multivariate Predictors: Total scores predicted by female gender, poor family cohesions & lower socioeconomic status, intrusion predicted by low family cohesion, avoidance predicted by female gender, being younger, low family cohesion & mothers being employed Internalising & Externalising Problems (YSR & CHQ-MH)  Sig Univariate Predictors: poor general family functioning & higher maternal depression predicted internalising problems, low cohesion & poor general family functioning predicted externalising problems & mothers being employed predicted externalising problems  Sig Multivariate Predictors: family functioning problems associated with adolescent psychopathology, poorer cohesion had strongest association with total problems, cohesion associated with externalising problem cases. General family functioning most important predictor of internalising problem cases & maternal depression independent predictor

Page 37 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Finch 2009 UK

0.90

Design: Qualitative Data Source: Semi-structured interview

Sample:  7 adolescent offspring (14-18yrs; 57% male) with a parent with cancer (43% fathers; diagnosed 4-11mths prior) Response Rate: Not reported Recruitment Setting: Cancer Centre

(-) Semi-structured interview covering recall of 1st learning of parent’s diagnosis, how it was assimilated & understood, what it meant to be told a parent has cancer, preparation for the news, communication about cancer in the family, impact of cancer on home life & relationships, supportive needs

Flahault 2010 France

0.79

Design: Case comparison Data Source: Rorschach tests

Sample:  52 children of parents with cancer (8-18yrs, Mage = 12.5yrs; 46% male; 49% ill fathers)  50 children with a parent with respiratory insufficiency or diabetes (8-18yrs, Mage = 12.2yrs; 56% female; 58% ill fathers) Response Rate: Not reported Recruitment Setting: Medical institutes in France

(+) Rorschach Comprehensive System

 Six dimensions of the experience of learning about a parental cancer diagnosis emerged: o Hearing about a parent’s diagnosis – perceived as serious & threatening, leading to worry about parental death o Vulnerability of self and others – self, parents and family seen as vulnerable but they avoided thinking about the cancer to protect themselves and remained emotionally strong to protect others. Cancer was a threat to the security of the family through the loss of the parent-child relationship and worry about parental death raised existential issues o Communication within the family – seen as dependent on family attitudes, beliefs and comfort in talking about cancer with most experiencing openness, trust and honesty but one describing his family as avoidant. o Feeling supported in experience – degree to which participants felt supported varied, with one feeling isolated, while others felt supported by friends & another described benefit in identifying with those with similar experience o Experience and support of school – most felt that school support could have been better with some wanting more acknowledgement and better education cancer o Experience and support of hospital – fear & anxiousness were common in relation to visiting hospital with most seeing no role for themselves in the hospital and no emotional or psychosocial element of hospital care  Participants advocated honesty and openness in family communication about cancer.  Participants demonstrated strength and resourcefulness in adapting to their parent’s cancer  Higher depression and anxiety, and lower self-esteem, in offspring of cancer patients (relative to offspring of parents with respiratory insufficiency or diabetes)  Difficulties more pronounced in girls and offspring of ill mothers

Page 38 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Foran-Tuller 2012 USA

0.86

Design: Case comparison Data Source: Questionnaires

(+) Child Behavior Checklist (+) UCLA PTSD Index (+) Patient Health Questionnaire

 Mothers with cancer perceived their children to have significantly more PTSD and internalising symptoms than mothers without cancer  No differences in maternal perceptions of externalising symptoms

GazendamDonofrio 2007 Netherlands

0.86

Design: Case-comparison Data Source: Questionnaires

Sample:  40 mothers with breast cancer (Mage = 45.74yrs; within 8 wks of diagnosis) and a child aged 6-18 (Mage = 12.45 yrs)  39 mothers without breast cancer (Mage = 42.20yrs), but with a child aged 6-18 (Mage = 10.73yrs)  47% children were male; no reporting of child gender by maternal health status. Response Rate: Not reported Recruitment Setting: Large suburban pediatric office Sample:  221 adolescent offspring (Mage=15.5, 48% male)  138 parents with cancer (19% males; Mage=45.4yrs; time since diagnosis diagnosed 1-5yrs prior, M = 2.8yrs; 22% with recurrence, 78% no evidence of current disease)  114 spouses (81% male; Mage=46.5yrs) Response Rate: 44% Recruitment Setting: University Medical Centre

(-) Demographics Dutch language versions of: (+) Child Behaviour Checklist-CBCL (+) Youth Self-Report-Child Behaviour Checklist-YSR (+) Impact of Events Scale-IES (+) Family Environment Scale

 Overall, families with parental cancer (1-5years earlier) displayed positive functioning  Parents reported their families as more expressive, less conflictual, better organised, less controlling & more social (than norm group)  Parental Family Relationship Index (FRI) scores significantly higher than norm (t=5.9, p=0.30), cohesion and Family Structure Index (FSI) scores not significantly different to the norm  Adolescents reported their families as more cohesive, expressive, organised, social, & fewer conflicts (than norm group)  Adolescent offspring FRI and FSI scores significantly higher than norm scores for adolescents (no significant diffs for ‘family values’ or ‘control’)  No differences based on gender or age of children/adolescents  Comparing parents’ scores to adolescents scores: parents scores were significantly more positive for all scales except ‘family values’  Illness-related factors (time since diagnosis, treatment intensity, disease status) not significantly related to family environment or children’s functioning

Page 39 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

GazendamDonofrio 2008 Netherlands

1.00

Design: Cross-sectional Data Source: Questionnaires

Sample:  304 adolescent offspring (Mage=14.5yrs, 45% male)  166 parents with cancer (Mage= 44.7yrs; 22% male; diagnosed 15yrs prior, M = 2.76yrs; 86% intensive treatment, 22% recurrence)  166 healthy spouses (Mage= 45.8yrs; 78% male) Response Rate: 44% Recruitment Setting: University Medical Centre in/outpatient cancer clinics

(-) Demographics Dutch language version of: (+) RAND-36 Quality of Life measure (+) Child Behaviour Checklist – CBCL (+) Youth Self-Report-Child Behaviour Checklist – YSR

 Patient health-related quality of life (HRQoL) scores not significantly different to norms on five domains but were clinically and/or statistically lower than norms on social functioning, role limitations & vitality  Male spouses’ HRQoL was not significantly different to norms but female spouses reported significantly better physical functioning & more social problems than norms  HRQoL varied by cancer type, treatment intensity & recurrence  Significant relationship found between patients’ psychosocial/physical functioning & spouses’ psychosocial functioning  Significant weak-moderate correlations between patient psychosocial functioning & their reports of their younger & adolescent children’s total, internalising & externalising problems  Significant weak correlation between patient physical functioning & adolescents’ total and internalising problems  Significant weak-moderate correlation between patient physical functioning & adolescent self-report total, internalising & externalising problems & patient psychosocial functioning & adolescent total & internalising problems  Spouses’ psychosocial functioning significantly weakly correlated with their reports of adolescent total & internalising problems & adolescent reports of internalising problems

Page 40 of 64

GazendamDonofrio 2011 Netherlands

0.95

Design: Retrospective cross sectional & longitudinal prospective Data Source: Questionnaires

Sample: Retrospective:  221 adolescent offspring (1118yrs, Mage=15.5yrs; 48% males)  138 parents with cancer (Mage=45.4yrs; 19% males; ongoing treatment=82%, surgery=17.5%, time since diagnosis M=35.6mths) Prospective:  70 adolescent offspring (1118yrs, Mage=14yrs; 44% males)  70 parents with cancer (37% males): o At baseline: Mage= 42.5yrs; ongoing treatment=77%, surgery=23%, time since diagnosis M=2.3mths o At T2: ongoing treatment=82%, surgery=11%, completed treatment=5%, time since diagnosis M=7.7mths; o At T3: ongoing treatment=1%, surgery=0%, completed treatment=99%, time since diagnosis M=13.8mths) Response Rate: Retrospective=44%/prospective= 50% Recruitment Setting: University medical centre

(-) Demographics (+) Situation-specific Emotional Reaction Questionnaire-child – SSERQ-c (+) Youth Self Report-Child Behaviour Checklist – YSR

 Prospective Study: adolescent uncertainty & helplessness significantly decreased over time, loneliness & positive emotions stable  Adolescents’ emotional reactions to parent’s cancer similar in both studies but prospective participants reported less uncertainty, helplessness & loneliness  Emotional reactions unrelated to parent gender in both studies  Daughters reported more helplessness than sons in retrospective study & at 1st follow-up in prospective study & more positive emotions at follow-up in prospective study  Children of parents who received surgery reported more positive emotions than those receiving longer-term Treatment in retrospective study, but not in prospective study  For adolescents in retrospective study: o Uncertainty significantly related to total & internalising problems o Loneliness significantly related to total, internalising & externalising problems o Daughters’ helplessness significantly related to total & internalising problems o Positive emotions unrelated to functioning  For adolescents in prospective study: o Uncertainty significantly related to; daughters’ total & internalising problems at all measurement points & sons’ internalising problems at baseline & externalising at 1st follow-up o Loneliness significantly related to daughters’ internalising problems at baseline & total, internalising & externalising problems at 1st & 2nd follow-up but unrelated to sons’ functioning o Helplessness & positive emotions unrelated to functioning o Baseline functioning predicted functioning at all followups o Emotional reaction at baseline did not explain variance in functioning at either follow-up but changes from baseline to 1st follow-up significantly accounted for variance in daughters’ total & internalising problems at 1st follow-up o Changes in emotional reactions did not significantly account for variance in daughters’ externalising problems o Increased uncertainty between baseline & 1st follow-up increased total & internalising problems o Changes in emotions not significantly related to functioning in sons

Page 41 of 64

Reference

Quality

Study Design & Data

Götze 2015 Germany

0.77

Design: Longitudinal prospective Data Source: Questionnaires

Huizinga 2010 Netherlands

0.86

Design: Longitudinal prospective Data Source: Questionnaires

Sample & Context

Sample:  110 offspring (11-18yrs, Mage=14.3yrs; 55% males)  161 parents with cancer (2355yrs, Mage=40.4yrs; 23% males)  115 healthy partners (24-61yrs, Mage=40.7yrs; 70% males) Response Rate: 74% initial; 77% T2 Recruitment Setting: Hospitals Sample:  49 test adolescent offspring (1118yrs, Mage=14.6yrs; 43% males)  37 parents with cancer (38-55yrs, Mage= 44.8yrs; 38% males; diagnosed 0.2-4mths prior, M=2.2mths)  37 healthy spouses (37-56yrs, Mage=45.1yrs; 62% males)  174 reference adolescent offspring (Mage=15.2yrs; 44% males) with a parent diagnosed with cancer 1-5yrs prior Response Rate: 50% Recruitment Setting: Universal Medical Centre

Measures

(+) Short Form-8 Health Survey - SF-8 (+) European Organization for Research and Treatment of Cancer 30-Item Core Quality of Life Questionnaire (EORTC QLQC30) (pt) (+) KIDSCREEN-27 (offspring) (+) Hospital Anxiety and Depression Scale (HADS) (+) Oslo Social Support Scale (OSS) (+) Centre for Epidemiological Studies Depression Scale for Children (CES-DC) (-) Demographics (+) Impact of Events Scale – IES (+) Youth Self Report – YSR (+) Child Behaviour Checklist – CBCL

Findings o Changes in emotional reactions from baseline to 2nd follow-up significantly accounted from variance in daughters’ total & internalising problems & sons’ internalising problems o Increase in daughters’ uncertainty associated with increase in total problems and increase in sons’ helplessness associated with increase in internalising.  In offspring, all areas of QOL showed small (non-significant) improvements over time  In the regression model, offspring quality of life was associated with depressive symptoms (-), and ill parents’ anxiety and depression (-)

 Clinically elevated SRS were found in 29% of test adolescents at T1, 16% at T2, and 14% at T3 and 29% in reference adolescents  Test adolescents reported most symptoms at baseline with significant decrease in intrusion, avoidance & total distress over time  Reference adolescents reported significantly more intrusion at baseline, less avoidance at T2 and less intrusion & total distress at T3 & daughters reported significantly less intrusion at T3  Daughters seemed more at risk than sons. Adolescents’ age, patient’s gender, and intensity and duration of treatment did not significantly affect SRS.  Adolescents with more SRS reported having more emotional/behavioural problems. Parents observed fewer problems in those adolescents.  Initial SRS affected later SRS and emotional problems

Page 42 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Huizinga 2011 Netherlands

0.86

Design: Cross-sectional Data Source: Questionnaires

Sample:  271 adolescent offspring (Mage=14.6yrs; 45% males)  128 mothers diagnosed with cancer 1-5yrs prior (Mage=44.3yrs, time since diagnosis M=2.6yrs, chemo/radiotherapy/multimodal treatment=82%)  96 male spouses (Mage=47.1yrs) Response Rate: 43% (hospital recruited)/81% (externally recruited) Recruitment Setting: University Medical Centre & community outreach

(-) Demographics & Medical Information (+) Impact of Events Scale (+) Youth Self-Report – Child Behaviour Checklist (+) Spielberger State-Trait Anxiety Inventory (+) Maudsley Marital QuestionnaireSatisfaction (+) Parent-Adolescent Communication Scale

Inbar 2013 Israel

0.77

Design: Case-comparison Data Source: Questionnaires

(-) Demographics (+) Experiences in Close Relationships inventory-Hebrew version (+) Mental Health Inventory-Hebrew version

Jantzer 2013 Germany

0.91

Design: Longitudinal casecomparison Data Source: Questionnaires

Sample:  58 female adolescent offspring / mother dyads (100% breast cancer survivors, 8mths post treatment)  48 female adolescent offspring / healthy mother dyads  Daughters aged 13-18yrs (Mage= 17.09yrs); mothers aged 38-61yrs (Mage= 48.73yrs); no reporting of age by maternal health status. Response Rate: 83% (survivors)/87.5% (healthy) Recruitment Setting: Oncology Clinic Sample:  74 adolescents with a parent diagnosed with cancer <5yrs earlier (Mage=15.6yrs; 45% male)  75 adolescents with healthy parents (Mage=15.9yrs; 36% male) Response Rate: Not reported Recruitment Setting: Clinical settings & public outreach

 Being female was best predictor of adolescents’ stress response symptoms (SRS) & internalising problems  Age was best predictor of externalising problems with older adolescents reporting more problems  Higher parental trait anxiety associated with more SRS & externalising problems in adolescents  Parental SRS significantly related to adolescents’ SRS but not internalising/externalising problems  Parental marital distress related to adolescent internalising problems  Parent-adolescent relationship significantly related to adolescent internalising/externalising problems  Poorer communication between parents and adolescents related to SRS, internalising & externalising problems  Intensive treatment significantly related to externalising problems  Recurrent disease significantly related to SRS and internalising problems  Time since diagnosis not related to adolescent outcomes  Daughters of breast cancer survivors report higher levels of distress than daughters of controls if they had avoidant attachment style  Daughters of breast cancer survivors reported higher levels of distress than daughters of controls if their mothers had high anxious attachment  Breast cancer survivors reported more distress and lower well-being than healthy controls, however this difference was only significant among BC survivors with high scores for attachment anxiety

(-) Demographics (+) Strengths & Difficulties Questionnaire – Short Form Self-Report 11-16yrs (+) SCOFF eating disorders questionnaire (+) Child & Youth Survey KiGGS (-) Modified Deliberate Self-Harm Inventory (+) Paykel Suicide Scale

 Having a parent with cancer did not significantly alter eating disorder, drug use, smoking, alcohol consumption, truancy, deliberate self-harm, suicidal behaviour, conduct problem, hyperactivity or peer problem outcomes over time for adolescents

Page 43 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

John 2013 Germany

0.82

Design: Non-randomized pilot of a manualised 3-week residential rehabilitation program using pre/post measures and a withinsubject control Data Source: Questionnaires

(+) European Organization for Research and Treatment of Cancer 30-Item Core Quality of Life Questionnaire (EORTC QLQC30) (pt) (+) Inventory for Quality of Life in Children and Adolescents (ILC) (+) Strengths and Difficulties Questionnaire

 Larger improvements in mothers’ emotional functioning, children’s psychological health, and children’s emotional symptoms were observed from pre- to post-intervention, relative to the change between the two pre-intervention tests.  This was true for all subscales of the maternal QoL measure; for all subscales of the children’s QoL measure except School; and for children’s conduct problems and hyperactivity symptoms, but not for peer problems or prosocial behaviour.

Karlsson 2013 Sweden

0.85

Design: Qualitative Data Source: Narrative interviews

Sample:  116 mothers diagnosed with primary, non-metastasized breast cancer (29-57yrs, M = 41.1yrs) with children aged 3-14 yrs (M = 7.4yrs; 47% female) Response Rate: 95.8% consented; 51% met criteria regarding questionnaire timing Recruitment Setting: Mothers who had completed the rehabilitation program Sample:  6 young adult offspring (20-22yrs; 16% male; 5 bereaved) with a parent with cancer (age at parent’s diagnosis 14-16yrs, age at parent’s death 15-20yrs) Response Rate: Not reported Recruitment Setting: Outreach through university

(-) Semi-structured narrative interview covering the diagnosis of participants’ parents’ cancer

 The main message that the young adults communicated in the interviews was interpreted as the overarching theme ‘Loneliness despite the presence of others’.  Two domains with three categories each emerged: o Distance, comprising a feeling of loneliness, lacking the tools to understand, and grief and anger o Closeness, comprising belief in the future, comfort and relief, and a need for support  The young adults felt a loneliness that they had never experienced before, and they lacked the tools to understand the situation.  They felt grief and anger over what the cancer had caused.  They had still managed to regain faith in the future.  They found comfort and relief in the thought that this would not necessarily happen to them again, and they gained support from talking to family and friends

Page 44 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Kennedy (2) 2009 UK

0.75

Design: Qualitative Data Source: Semi-structured interviews

Sample:  11 children (8-18yrs; 18% males)  10 parents with advanced cancer (37-60yrs; 30% males; time since diagnosis M=12.4mths)  7 healthy parents (31-73yrs; 57% males) Response Rate: Not reported Recruitment Setting: Cancer Centre, Hospices, GP practices & Gynaecological Team

(-) Demographics (-) Semi-structured interview covering the impact of parental advanced cancer on children and how they cope with the diagnosis

Kennedy 2009 UK

0.80

Design: Qualitative Data Source: Semi-structured interviews

Sample:  11 children (8-18yrs; 18% males)  10 parents with advanced cancer (30% male; time since diagnosis M=12.4mths)  7 healthy parents (57% males) Response Rate: Not reported Recruitment Setting: Cancer Centre, Hospices, GP practices & Gynaecological Team

(-) Demographics (-) Semi-structured interview covering children’s’ information needs and preferences for gaining information when a parent is diagnosed with cancer

 Four themes emerged regarding response to diagnosis, coping mechanisms, life changes & positive outcomes: 1. Response to diagnosis: Children described being distressed by their parent’s diagnosis and having concerns related to their parents and their own health & postbereavement life. Parents did not recognise the impact on children to the same degree as described by children 2. Coping mechanisms: Distraction & maintaining normality were described as the dominant strategies of coping for children. Parents focused on limiting the impact of their cancer on their children by maintaining normality 3. Life changes: increased responsibilities and decreased social activity were considered to be the most noticeable of life changes 4. Positive outcomes: Positive aspects described by children and parents included strengthening of relationships and learning to value family members and the important things in life.  Four themes identified by children regarding information & communication: 1. Need and type of information: info on parent’s illness, how to help parents, treatment and tests info, length of hospital stays, child’s own health. Girls more concerned about own health. Need for information changes depending on illness trajectory (more required in early stages). 2. Sources of information: parents, health care professionals, books, leaflets, internet, media. Parents viewed as main source of info 3. Communication with parents: desire to communicate with parents with varying degree of openness between childparent dyad 4. Communication with others: friends, siblings, general acquaintances, teachers, health care professionals  Two main themes identified by parents regarding information & communication: 1. Telling the children: the most difficult issue was telling children about their advanced cancer 2. Children’s questions: children presented with many questions surrounding the parent’s treatment, tests, and side effects. Most difficult questions were regarding death

Page 45 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Kissil 2010 USA

0.85

Design: Qualitative Data Source: Focus groups

Sample:  12 African American adolescents (11-18yrs, Mage=14.5yrs; 25% male)  9 parents (34-56yrs, Mage =44yrs; 11% male) with Stage IIII breast cancer Response Rate: Not reported Recruitment Setting: 2 urban oncology clinics

(-) Semi-structured focus groups with six broad question prompts concerning current concerns, social support, coping practices, family roles, and personal and family changes, lead by two facilitators. Post-traumatic growth was only discussed if spontaneously reported by participants

Krattenmacher 2013 Germany

1.00

Design: Cross-sectional Data Source: Questionnaires

Sample:  214 adolescent offspring (1118yrs; young adolescents Mage=12.4yrs, 56.8% male; older adolescents Mage=16.2yrs, 46.1% male)  144 parents with cancer (3062yrs, Mage = 44.0yrs; 20% males; 67% mothers/69% fathers diagnosed <1yr prior)  116 healthy parents (32-67yrs, Mage =45.0yrs; 67.2% males) Response Rate: Not reported Recruitment Setting: University Oncology Centres (multi-centre study)

(-) Demographics (+) Socio-economic status (+) Kidscope (+) Strengths & Difficulties Questionnaire (German version) (+) KIDSCREEN

Cancer deeply affected adolescent offspring’s lives - however all adolescents suggested the experience had generated positive changes in them - domains from Tedeschi & Calhoun (2004)  Greater appreciation for life  Increased intimacy in relationships (with diagnosed parent, other family members, those with similar experiences)  Sense of increased personal strength - more mature, enhanced sense of agency (including responsibility)  Considering new possibilities and paths, and spirituality, were not really mentioned by this sample (despite being modelled)  Changes in health-related attitudes and behaviours  97% of adolescents reported good health-related quality of life (HRQoL)  71% within normal range of symptomology  29% exhibited psychopathological symptoms (15.5 subclinical/ 13.5 clinical range)  Adolescents reported worse functioning than parental reports of their functioning indicated (p<0.001)  Girls reported more internalisation than boys (p=0.004) o parent reports did not identify this pattern o mothers reported more internalising in boys (p=0.042)  No gender related differences found for overall problems or HRQoL from adolescent or parent points of view  No significant effect for age or gender on coping  Frequency of coping strategies reported by adolescents was mediated by perceived efficacy (higher efficacy associated with more frequent use, p<0.001)  Problem-focussed and approach-oriented coping associated with better mental health  Avoidance-oriented coping associated with worse mental health  Emotion focussed coping associated with both lower and higher mental health

Page 46 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Krattenmacher 2014 Germany

1.00

Design: Case-comparison Data Source: Questionnaires

Sample:  67 families affected by parental cancer - 62 children (0-10yrs, Mage = 6.4; 48% males), 46 adolescents (11-18yrs, Mage = 14.6; 56% males), 67 ill parents (Mage = 42.9yrs; 33% males) and 51 healthy parents (Mage = 44.8yrs; 67% female)  69 families affected by parental mental illness - 61 children (010yrs, Mage = 5.3; 34% males), 54 adolescents (11-18yrs, Mage = 15.6; 43% males), 69 ill parents (Mage = 39.4yrs; 20% males), and 36 healthy parents (Mage = 42.7yrs; 72% males) Response Rate: Not reported Recruitment Setting: University Medical Centre

(+) SF-8 (Parental HRQoL) - Physical Component Summary and Mental Component Summary (+) Child Behaviour Checklist and Youth Self-Report (adolescents)

Küçükoğlu 2013 Turkey

0.89

Design: Cross-sectional Data Source: Questionnaires

Sample:  90 adolescent offspring (1318yrs, Mage=15.64yrs; 48% males) with a parent diagnosed with cancer 0.5-5yrs prior (Mage = 48.62yrs; 33.3% fathers) Response Rate: Not reported Recruitment Setting: Radiation/Medical Oncology Clinics at a University Hospital

(-) Demographics (+) Brief Symptom Inventory – BSI (+) Scale of Perceived Social Support from the Family & Friends

 Prevalence of sub-clinically relevant symptoms in offspring affected by parental cancer was 14.9% (ill parent report), 11.1% (healthy parent report) and 30.6% (adolescent selfreport); prevalence of clinically relevant symptoms was 24.3% (ill parent report), 27.8% (healthy parent report) and 19.4% (self-report).  Prevalence of sub-clinically relevant symptoms in offspring affected by parental mental illness was 11.1% (ill parent report), 8.0% (healthy parent) and 32.4% (adolescent selfreport); prevalence of clinically relevant symptoms was 40.9% (ill parent report), 38.0% (healthy parent report) and 17.6% (adolescent self-report).  Prevalence of emotional and behavioural problems was higher in both groups compared to population norms (by both parent and self-reports).  No significant differences between families with cancer and families with mental illnesses once sample differences were controlled.  Ill parents reported more internalisation in girls than in boys, especially in adolescents (ns in children).  Healthy parents’ MCS and PCS predicted children’s internalising and total problems by parent ratings, but not by adolescent self-ratings.  Psychological symptom scores were generally high (M=60.11), with depression scores being the highest (20.22), followed by anxiety (M=13.83), Negative identity (M=11.53), hostility (M=7.30) and somatisation (M=7.22).  Mean anxiety, depression, negative identity, somatization, hostility and total psychological symptom scores significantly higher for males than females  No significant correlation between perceived social support from family or friends & psychological symptoms

Page 47 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Kühne 2012 Germany

0.91

Design: Cross-sectional Data Source: Questionnaires

Sample:  86 adolescent offspring (1121yrs, Mage=13.78 yrs, 44% males) from 66 families with a parent with cancer (29-65yrs, Mage =44.31yrs; 43% ill fathers; 54% diagnosed >12mths ago) Response Rate: Not reported Recruitment Setting: Multicentre research project

(-) Demographics (+) Strengths and Difficulties Questionnaire-SDQ (+) KIDSCREEN-10 (+) Kiscope (+) Hospital, Anxiety & Distress ScaleHADS (+) McMaster Family Assessment DeviceFAD

Lindqvist 2007 Finland

0.91

Design: Case-comparison Data Source: Questionnaires

Sample:  54 adolescent offspring (44% male, Mage=14.0yrs; 56% female, Mage=13.5yrs) from canceraffected families (24% ill fathers; 88% 1st diagnosis; 26.8% curable)  49 adolescents (61% male, Mage=14.1yrs; 39% female, Mage=13.7yrs) from healthy families Response Rate: Not reported Setting: University Hospital Oncology Clinic (cancer affected)/elementary & secondary schools (control)

(-) Demographics (+) SF-8 Health Survey (+) McMaster Family Assessment DeviceFAD (+) Youth Self Report-Child Behaviour Checklist-YSR

 Adolescents with palliative cancer stage parents displayed significantly less total problems and better overall healthrelated quality of life (HRQoL) than children whose parents were at other disease stages  Parents disease status and general family functioning were the best predictors of psychosocial problems and HRQoL  Adolescents with parents in other disease stages reported significantly worse HRQoL than the norm group, however there were no significant differences between HRQoL of adolescents with a palliative parent and norm group  No significant differences in psychosocial problems as a result of adolescent’s age or gender  Older age indicated significantly lower HRQoL  Poorer family functioning predicted lower HRQoL and increased psychosocial problems  No significant differences between adolescents with a parent diagnosed with cancer and controls regarding levels of perceived psychological distress and healthy family functioning  No effects for gender of either the parent or adolescent upon the adolescent’s mental health  Significant negative correlation between the ill parent’s physical health and the adolescent’s mental health score on the YSR, such that as the parent’s physical health deteriorates, scores for internalizing and externalizing problems among adolescent’s increases  Healthy family functioning associated with fewer psychological symptoms in adolescent participants  Strongest predictors for adolescent psychological symptoms: communication (internalizing problems), affective involvement (externalizing problems), & problem solving (total psychological problems)

Page 48 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Maynard 2013 Australia

0.95

Design: Qualitative Data Source: Semi-structured interviews

Sample:  15 adolescents (14-18yrs, Mage= 15.9yrs; 40% male) with a parent with non-terminal cancer (40% male, time since diagnosis M=2.3yrs) Response Rate: Not reported Recruitment Setting: Adolescent & Young Adult Cancer Support Organisation

(-) Demographics (-) Semi-structured interview covering finding out about parental cancer, things that helped with coping, what would help other young people, helpful/not helpful things parents did, what they wished parents had understood

McDonald 2015 Australia

1.00

Design: Cross sectional Data Source: Questionnaires

Sample:  255 adolescents/young adults (12-24yrs, Mage =16.1yrs, at diagnosis Mage =14.0yrs; 1217yrs, 26% males; 18-24yrs, 19% males) with a living parent diagnosed with cancer in the previous 5yrs (31% male, time since diagnosis M=16.7mths) Response Rate: Not reported Recruitment Setting: Adolescent & Young Adult Cancer Support Organisation

(-) Demographics (+) Family Relationships Index-FRI (+) Offspring Cancer Needs InstrumentOCNI (+) Kessler 10-K10

Young people with a parent diagnosed with cancer helped by:  Parental interpersonal and instrumental behaviours (coping, availability to children, inviting child to talk, maintaining normalcy, awareness of child’s behaviour, quality time with children, methods of disclosing information, letting child meet doctor/see treatment setting, not withholding information)  Personal strategies including coping strategies and strengthening the family (time away, distraction, outlets for emotion, informing people at school, sourcing information, thinking positively, time alone, spending time with ill parent, helping with housework)  Community support including an extended social network (friends, extended family and peers with similar experiences)  Higher information needs domain scores associated with higher family conflict & cohesion, less time since diagnosis & lower family expressiveness  Higher family needs domain scores associated with having father with cancer & lower levels of expressiveness  Higher practical assistance needs domain scores associated with lower family expressiveness, less time since diagnosis, cancer having relapsed, & higher levels of family conflict  Higher Time Out/Recreation needs domain scores associated with lower family expressiveness & cancer having relapsed  Higher Feelings needs domain scores associated with having father diagnosed with cancer, having cancer relapse and lower family expressiveness  Higher support from friends needs domain scores associated with having father diagnosed with cancer, less time since diagnosis, greater family conflict and less family expressiveness  Higher support from other young people needs domain scores associated with being younger, cancer having relapsed, less time since diagnosis, greater family conflict and less family expressiveness  Having a father with cancer, a shorter time since diagnosis and poor family functioning were associated with increased unmet needs  Being female and older, having more unmet cancer needs and poorer family function were associated with increased distress  Having a parent relapse with cancer associated with unmet needs in domains of practical assistance, time out and support from other young people who have been through something similar

Page 49 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Möller 2014 Germany

0.82

Design: Cross-sectional Data Source: Questionnaires

(-) Demographics (+) Strengths & Difficulties Questionnaire (+) SF-8 (+)Freiberg Questionnaire of Coping with Illness – FKV (+) Hospital Anxiety and Depression ScaleHADS (+) Winkler Index Socioeconomic Status (+) Family Assessment Device (FAD) (+) WHO-ECOG performance status

 Children of cancer patients had slightly higher levels of mean emotional & behavioural symptoms than norms  General family dysfunction was best predictor of emotional & behavioural problems  No illness-related factors associated with child adjustment  Healthy parents reported significantly more emotional & behavioural problems in latency-aged sons (similar but non-sig reports from ill parents) & healthy parents reported more emotional & behavioural problems in 6-10 yr olds than 11-13 yr olds  Younger adolescents (11-13 yrs) reported more symptoms than older adolescents (14-21yrs)

Patterson 2010 Australia

0.95

Design: Qualitative Data Source: Qualitative questionnaire

Sample:  324 adolescents/young adults (11-21 yrs, Mage=11.9yrs; 45% males) from 235 families  235 parents with cancer (Mage = 43.4yrs; 26% fathers; time since diagnosis M=24.8mths; 25% mothers/40% fathers newly diagnosed)  167 healthy parents Response Rate: 80% Recruitment Setting: University medical centres & hospitals Sample:  62 bereaved offspring (12-24yrs, 55% 12-17yrs, 45% 18-24yrs; 19% male) whose parent died of cancer (31% ill fathers, length of bereavement M=10.6mths) Response Rate: 30.5% Recruitment Setting: Adolescent & Young Adult Cancer Support Organisation

(-) Demographics (-) Open-ended questionnaire eliciting needs, why needs are important, whether needs were being met & the impact of the need on how the participant feels

 Seven themes emerged regarding needs of bereaved offspring (the number of participants reporting needs in each category & the number for who needs were unmet is included in brackets): o Support & understanding (reported=56, unmet=25) o Help coping with feelings (reported=32, unmet=14) o To talk to people who have had a similar experience (reported=28, unmet=11) o Information (reported=11, unmet=5) o To have a break/have fun (reported=8, unmet=1) o Space & time to grieve (reported=7, unmet=2) o Help with household responsibilities (reported=6, unmet=4)

Page 50 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Patterson 2011 Australia

0.85

Design: 2 phase measure development study – qualitative & cross sectional Data Source: Semi-structured interviews & questionnaires

Sample: Phase 1:  14 adolescents/young adults with a parent with cancer (12-24yrs, 43% male)  26 support organisation staff

Phase 1: (-) Demographics (-) Open/close ended interview questions about young person’s needs arising from parental cancer (-) Young person’s endorsement of needs identified in literature (-) Open/close ended survey questions on support organisation staff perceptions of the needs of your people with a parent diagnosed with cancer

 Eight qualitative needs domains identified – information , peer support, expressing & coping with feelings, respite & recreation, support for carers, supportive education and work environment, family factors & access to support services  97% of young people endorsed ≥1 need, 90% endorsed ≥10 needs, 25% endorsed ≥50 needs, 87% had ≥1 unmet need, 43% had ≥10 unmet needs & 25% had ≥20 unmet needs  Mean unmet needs=12.82 (SD=13.33, range 0-54), of those who had ≥10 unmet needs, 47% indicated those needs had been met  Highest endorsed needs in peer support (96%), respite & recreation(94%) & expressing & coping with feelings (92%) domains, 73% had unmet peer support needs, 57% had unmet education & work environment needs & 56% had unmet expressing & coping with feelings needs  Internal reliability of domains ranged from α=0.92-0.64 with feedback indicating good face & content validity  Sig positive correlation for 12-17 year olds between SDQ-TD and OCNI scores (r=0.33, p<0.001) and for 18-24 year olds between OCNI and DASS-21 depression (r=0.77, p=0.000), anxiety (r-0.66, p=0.002) & stress scores (r=0.56, p=0.010)  50% of 12-17 year olds & 45% of 18-24 year olds had atrisk/clinically elevate scores on psychological distress measures

qualitative

0.90 quantitative

Phase 2:  116 adolescents/young adults with a parent with cancer (1224yrs, 84% 12-17yrs, 16% 1824yrs; 34% male, Mage=15yrs; 66% female, Mage=16yrs; 43% parents finished treatment, time since diagnosis 42.5%<2yrs, 45.1% 2-5yrs, 12.4%>5yrs) Response Rate: Phase 1: 22% young people/46% staff Phase 2: 30% Recruitment Setting: Adolescent & Young Adult Cancer Support Organisation

Phase 2: (-) Demographics (-) Offspring Cancer Needs Instrument – OCNI (+) Strengths and Difficulties Questionnaire – SDQ (+) Depression, Anxiety & Stress Scale-21 – DASS21

Page 51 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Patterson 2013 Australia

1.00

Design: Instrument validation Data Source: Questionnaires

Sample:  256 adolescents/young adults (12-24 yrs, Mage=16.1yrs; 27% male) with a parent diagnosed with cancer in the last 5yrs (38% male; 62.3% on treatment; time since diagnosis M=16.7mths) Response Rate: Not reported Recruitment Setting: Adolescent & Young Adult Cancer Support Organisation

(-) Demographics (-) Offspring Cancer Needs Instrument – OCNI (+) Kessler 10 – K10

Phillips 2015 USA

1.00

Design: Qualitative Data Source: Semi-structured elicitation interviews

Sample: 7 adolescents from 6 families (1115yrs, Mage = 13.6yrs; 28% males) with a parent with Stage IV cancer (33% males) Response Rate: Not reported Recruitment Setting: Outpatient clinics

(-) Single-occasion, semi-structured elicitation interviews covering issues, concerns, management behaviour, and types and sources of tension or disconnection between the ill parent and the adolescent about the cancer.

 Revealed 7 domains identified from the final OCNI measure: information, family issues, practical assistance, time out, feelings, support (friends) and support (other young people)  Respondents appeared to use the scale consistently (confirmed with Rasch analysis)  Several items removed from the measure following EFA and Rasch analysis (due to poor fit), consequently re-organising domains (18 items removed, items 19 and 48 shifted to new domains) Excellent psychometric properties:  Evidence for strong construct validity (significant correlations with K10 in related domains e.g. ‘feelings’ domain)  Excellent internal consistency, with the lowest domain displaying α=0.89  Test-retest coefficient = 0.73, hence good reliability  Evidence for good content validity: minimal number of unmet needs listed by participants not within the scope of the OCNI (those listed were highly specific e.g. genetic concerns about the person’s own risk of cancer)  Majority of participants (80.9%) reported the OCNI was easy to understand (agree or strongly agree)  Core construct - Weaving a Normal Life with Cancer - involving five major domains: o Feeling the weight of the world on my shoulders worrying about the effects of the cancer; dealing with lack of control over the disease/outcomes; feeling alone/unsupported o Cancer changes everything - impact of cancer on the lives of everyone close - ongoing/chronic impact o Confronting or getting away from the cancer - coping strategies - often cycling between avoidance and involvement o Talking about it - mixed needs for family communication about the cancer o Cancer was a positive for me.. it taught me - lessons learned and positive effects of parental character - shaping growth, development and self image; fostering maturity; increased independence; more health-aware

Page 52 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Phillips 2015 USA

0.90

Design: Qualitative Data Source: Semi-structured interviews

Sample: 10 adolescents (14-17yrs, 3-16yrs at diagnosis; 70% male) living with a parent diagnosed with advanced/ metastatic cancer (stage 3+) Response Rate: Not reported Recruitment Setting: Non-profit organisation providing psychosocial support for families facing a parent’s life-threatening illness

(-) Semi-structured interviews covering their experience of advanced parental cancer - information, changes in (relationship with) the ill parent, positives, coping strategies, impact, communication and support

Puterman 2008 Canada

0.85

Design: Qualitative Data Source: Auto-ethnographical interviews

Sample: 5 young adult females offspring (21-32yrs) with a parent diagnosed/re-diagnosed with cancer (80% ill mothers, 40% ill fathers; currently on treatment or on treatment within previous 5yrs) Response Rate: Not reported Recruitment Setting: Public cancer treatment centre

(-) Unstructured, interactive autoethnographic interviews covering how young adult women experience a parent’s cancer diagnosis

 Four essential themes (with 16 subthemes) emerged: o Life interrupted - the experience of growing up with cancer and having it shape their lives. subthemes: (1) mis/ conceptions about cancer; (2) struggling with changes changes in parent/ availability; (3) limiting family activities; (4) fear, worry and anxiety; (5) juggling life and cancer o Being there - contributing to family wellbeing/helping out: (1) being a “good kid”; (2) caring for their family. o Managing emotions - coping strategies: (1) distraction; (2) talking about it (cancer and reactions); (3) thinking positively; (4) social support o Positive prevails - growth and transformation: (1) personal growth; (2) increased appreciation of life; (3) enhanced relationships; (4) changed views of life  The issue of time emerged as an overriding theme in the interviews in addition to a number of subthemes: 1. Unexpected timing – Parental illness was unexpected at the young adult’s life stage 2. Uncertainty/waiting – the challenges of living with uncertainty was prominent and made some participants feel as though life was on hold or in limbo 3. Being present – parental illness required young adults to refocus their life on the present, was both coping mechanism & challenge 4. Concerns for future – concerns including susceptibility to cancer & the absence of parental input in their lives were raised along with sadness for parents in missing major life events 5. Acceleration of time – participants considered accelerating their lives so that their parents could be part of important events & time felt accelerated by taking on increased responsibility for parents 6. illness progression – time became structured by illness progression & treatment & firsts & lasts in life events, acute awareness of passing time when waiting for test results 7. holidays – celebrations (e.g. birthdays, Christmas) take on increased significance, are altered by parental illness & mark passing of time  Daughters’ perception of time was altered by parental illness and it also highlighted the significance of parental illness during young adulthood.

Page 53 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Rainville 2012 Canada

0.91

Design: Case-comparison Data Source: Questionnaires

(-) Demographics (+) Indice de détresse psychologique de Santé Québec – IDPSQ-14

 Adolescents living with a parent with advanced cancer had significantly higher psychological distress and depressive state than adolescents in the general population o further analyses suggested this was only the case for late adolescents (15-18yrs) who had significantly higher psychological distress than same age group in the general population o Difference attributable to significantly higher anxious state & irritability in late adolescents  Late adolescents experienced significantly more psychological distress and anxious state than early adolescents (12-14yrs)  Late adolescents experience marginally significantly higher irritability, but are not different from early adolescents on depressive state or cognitive problem measures

Schmitt 2007 Finland

0.57

Design: Quantitative evaluation of a family counselling service Data Source: Unclear

Sample:  28 adolescent offspring (1218yrs, Mage=14.86yrs; 36% male) with a parent with advanced cancer (39% fathers; 79% palliative treatment; time since advanced cancer diagnosis M=250days)  2346 comparison adolescents (12-18yrs, Mage=14.48yrs; 49% males) Response Rate: Parent with cancer sample: 72%, Comparison sample: Not reported Recruitment Setting: Tertiary care hospitals Sample: 85 families affected by parental cancer, including:  37 cancer patients (Mage = 45.2yrs, SD = 6.9yrs; 14% male) with children aged 4-17yrs (M = 10.7yrs, SD = 3.9yrs; 43% male) who had received counselling  48 cancer patients (Mage = 41.5yrs, SD = 5.7yrs; 35% male) with children aged 4-17yrs (M = 10.6yrs, SD = 3.8yrs; 54% male) Response Rate: Not reported Recruitment Setting: Counselling service associated with oncology clinic

(-) Demographics (-) Questions about counselling preferences

 Counselled families had significantly older parents, and were more likely to be single-parent families and less likely to be reconstituted families than those who refused counselling.  Families who refused counselling most often had multiple reasons, such as spousal/offspring refusal, patient exhaustion, travel difficulties, small children, or economic problems.  81% of families started counselling 1-7 months after diagnosis, with 30% starting around 3 months post-diagnosis  57% of patients reported that counselling started at the right time; 8% wanted counselling earlier (eg. immediately after diagnosis), and 35% had no preference.  76% reported that family counselling sessions should be incorporated into the treatment regime of new cancer patients with children.  5% considered the counselling service a waste of time/money.

Page 54 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Schmitt 2008 International:

1.00

Design: Cross-sectional Data Source: Questionnaires

Sample:  489 child/adolescent offspring (11-17yrs; 47% males, Mage=10.8yrs; 52% females, Mage=11.7yrs)  372 parents with cancer (23% males, Mage=45.3yrs; 77% females, Mage=43.4yrs; time since diagnosis M=25.2mths)  258 healthy parents (71% male, Mage= 44.4yrs; 29% female, Mage=41.3yrs) Response Rate: 55% Germany, 70% Finland, 63% UK, Switzerland, Austria, Denmark unknown Recruitment Setting: Not reported (conducted in multiple countries)

(-) Demographics (+) Family Assessment Device (+) Beck Depression Inventory (+) Short form Medical Outcomes Health Survey (SF-8)

0.90

Design: Qualitative Data Source: Semi-structured interviews

Sample:  10 adolescent offspring (1218yrs; 60% male)  9 parents with advanced cancer (38-73yrs; 33% male)  7 healthy spouses/partners Response Rate: Not reported Recruitment Setting: Hospice

(-) Individual semi-structured interview covering the cancer diagnosis, family relationships, family communication & shared understanding of cancer and prognosis & descriptions of interactions between family members

 Children reported non-sig more impairment in family functioning than parents & tended to perceive more family functioning impairment when mother was ill (except on behaviour control which children perceived to be worse when father was ill)  Depression prevalence was 35% for ill/28% for healthy mothers and 28% for ill/13% for healthy fathers  The ill parent’s depression was significantly associated with impaired family functioning on 5 of 7 FAD subscales (general function, problem solving, roles, communication & affective involvement) in analyses including parent & children  Older age associated with increased perception of impairment in affective responsiveness  Family role significantly associated with perceived impairment in family functioning for 5 of 7 FAD subscales (problem solving, behaviour control, communication, affective responsiveness, affective involvement)  Perception of impairment of family functioning not associated with parental depression in analyses with children  Poorer physical status of the ill parent was significantly associated with impairment on roles & communication  The ill parent’s depression was the most significant factor associated with impairment in family functioning  Family members’ main concern was not having enough time together.  In response to the threat of limited time, a four-stage dynamic process for optimizing the time they had left together was described: 1. Coming to know time together is limited 2. Spending more time together 3. Extending time together 4. Giving up time together to end the suffering  The adolescents and their ill parents did not change their interaction patterns until they realized their time together was limited by the advanced cancer. Then they spent more time together to make things easier for each other.

Germany, Switzerland, UK, Finland, Austria & Denmark

Sheehan 2011 USA

Page 55 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Stiffler 2008 USA

0.80

Design: Qualitative Data Source: Open-ended interview

Sample:  8 adolescent daughters of women with stage 0-III breast cancer (13-24yrs; 10-15yrs at diagnosis, M=13.5yrs) Response Rate: Not reported Recruitment Setting: Various clinical/community sites

(-) Open-ended interview with minimum prompts covering the experience of being parented when the adolescent’s mother has cancer.

 8 major themes were found: o World turned upside down – daughters could vividly describe period of finding out about mother’s cancer, experienced sense of chaos, conflicting emotions, denial & fear mother would die, realisation that cancer impacts whole family & sense that life was changing & nothing could be taken for granted o Need to get away/reluctant to leave – felt pulled toward & need to push away from the situation o Mom can’t die – belief that treatment will save mother, fear of the unknown & worry about what will happen if mother dies o Hole where mom used to be – mothers struggled to retain the maternal role & daughters saddened by loss of their mother as they had known them o Filling the hole where mom used to be – daughters felt a responsibility to replace the mother’s missing input into the family by doing household tasks o Being there for/managing mom – some felt responsibility to manage their mother & her care, providing nurse-like support, being protective & putting others ahead of self o Managing own reactions/being selfish or difficult – struggle to find reaction acceptable to self & others, vacillating between positive & negative reactions or withdrew when they couldn’t manage o Guarded relief – sense of guarded relief when mother had finished cancer treatment, looked forward to returning to normal life but still aware that mother was not completely safe, worry that cancer might return  Daughters were struggling with changes going on in their lives, found themselves in reversed roles with their mothers & felt conflicted between meeting their mothers’ needs & their own.  Daughters required help in communicating their needs to their mothers and other

Page 56 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Thastum 2008 Denmark

0.90

Design: Qualitative Data Source: Semi structured interviews

Sample: 21 child/adolescent offspring (815yrs, Mage=11yrs; 38% male) with a parent with cancer (13% male; time since diagnosis 2-153mths) from 15 families Response Rate: Not reported Recruitment Setting: University Hospital

(-) Demographics (-) Semi-structured interview about children’s experiences of having a parent with cancer & parents experiences of the needs of their children in relation to the cancer

 Children were aware of the name & general facts of the parent’s illness & expressed how important it was to have precise knowledge of the illness  Parents expressed a wish for guidance on how to inform their children about the illness  There was limited emotional communication between the generations though children were very observant of both ill & healthy parent’s emotional condition.  Five coping strategies used by children were identified: o Helping others – including practical assistance at home & emotional help to parents o Parentification – some children helped parents by concealing or suppressing their own emotions o Distraction – actions to divert negative thoughts & emotions arising from the parent’s illness (form of denial) o Keeping it in the head – active attempt to not tell anyone about illness-related thoughts (mostly younger children) o Wishful thinking – self-comfort action – by thinking positively, nothing bad will happen  Children advised other children to talk to others about the parent’s illness, help parents & consider the parent’s needs  Both adaptive & destructive examples of parentification found  Communication patterns & parental coping seemed to be highly related to the child’s coping repertoire  Even though most children seemed to manage rather well, all children were strongly affected by the illness  The ‘healthiest’ adaptation related to factors within the family system, which has implications for the provision of help

Page 57 of 64

Thastum 2009 International: Germany, Switzerland, UK, Finland, Austria & Denmark

0.91

Design: Cross-sectional Data Source: Questionnaires

Sample:  352 children (4-17 yrs, Mage=11.7yrs;44% male)  350 parents with cancer (21% fathers, Mage=44.5yrs; 79% mothers, Mage=44.0yrs ; time since diagnosis M=1.9yrs)  250 healthy parents (73% male, Mage=45.6yrs; 27% female, Mage= 44.3yrs) Response Rate: Not reported Recruitment Setting: Not reported (conducted in multiple countries)

(-) Demographics (+) Child Behaviour Checklist – CBCL (+) Youth Self Report-Child Behaviour Checklist – YSR (+) Beck Depression Inventory (+) SF8 version of the Medical Outcomes Health Survey (+) General functioning subscale of the Family Assessment Device – FAD-GF

 Ill parents reported significantly more internalizing, externalizing, and total problems in male children & rated significantly higher proportion of sons as having abnormal scores in internalizing (31.9%) & total problems (30.6%) compared to normative population (17%).  Healthy parents reported significantly more internalizing & total problems in male children, significantly more internalizing, externalizing, & total problems in female children, & rated male children as having significantly higher proportion of abnormal externalizing (32.7%) & total problems (27.3%). Healthy mothers reported significantly more total problems in children than did healthy fathers.  Ill parents reported significantly more internalizing problems in adolescent sons & daughters & rated adolescent sons as having a significantly higher proportion of abnormal internalizing problems (34.8%) than norms. Ill fathers reported significantly more total & internalizing problems in adolescent daughters than ill mothers.  Healthy parents reported significantly more internalizing & total problems in adolescent daughters & rated adolescent daughters as having a significantly higher proportion of abnormal internalizing (44.8%), externalizing (26.9%), and total problems (37.3%). Healthy mothers reported significantly more internalizing problems in adolescent daughters than healthy fathers.  Adolescent self-reported emotional & behavioural problems not significantly different to norms  More internalising externalising and total psychological problems reported when fathers rather than mothers were ill  Adolescent daughters whose father was ill reported significantly more internalizing & total problems than the norms & significantly higher internalizing & total problems were found among adolescent daughters whose father rather than mother was ill o unclear whether this difference was due to parents’ different diagnoses, gender or other factors  Breast cancer patients reported significantly lower total problem scores for children than other patients & adolescents of breast cancer patients had lower self-report total problem scores other adolescents.  Breast cancer patients reported significantly lower internalizing & externalizing scores in their latency-aged & adolescent children & adolescent self-reported YSR scores were significantly lower in breast cancer families than other families.

Page 58 of 64

Reference

Torp 2013 Norway

Quality

1.00 Quantitative

0.60 Qualitative

Tucker 2013 USA

0.70

Study Design & Data

Sample & Context

Measures

Design: Mixed methods Data Source: Questionnaires (quant) and in-depth interviews (qual)

Sample: Quantitative study:  386 cancer patients (27-63yrs, Mage=45.8yrs; 31% male) living at home with at least one child under 18yrs Qualitative study:  10 children (10-16yrs; 70% 1013yrs; 30% 16-18yrs; 30% male) from 9 families. Response Rate: Quantitative: 54% to total survey 29% were eligible for this study Qualitative: Not reported Recruitment Setting: Norwegian cancer registry

(-) Demographics and cancer questions (-) Questions about financial situation (-) In-depth interviews regarding experiences associated with family’s socio-economic situation; participants were allowed to discuss other cancerrelated topics if they desired.

Design: Qualitative evaluation of an adventure-based support group for offspring Data Source: Focus groups

Sample:  12 youth (10-15yrs; 42% male), who had a parent with cancer (67% in remission, 8% on treatment, 17% in hospice care, 8% deceased)  9 parents (11% male; 44% cancer survivors, 56% spouses) Response Rate: Not reported Recruitment Setting: Previous participants of the On Belay program (attended at least twice in preceding three years)

(-) Semi-structured focus groups about the On Belay program, and how participation affected feelings of support, personal power, and the ability to communicate about and process the parental cancer experience.

Findings  The best predictor of internalizing problems in children and adolescents was parental depression  The best predictor of externalizing problems in children and adolescents was family dysfunction  Most cancer patients reported good (60%) or neutral (32%) financial situation, with only 8% reporting (very) poor situations  29% reported household financial hardship attributed to cancer - most attributed this to reduced income (51%), or a combination of reduced income and increased expenses (40%)  Half of affected families had to reduce expenses on goods/activities for their children  Risk factors for financial hardships included unemployment, metastasis, chemotherapy; single parenthood and comorbidity.  Parents reported trying to shield their children from the effects of financial hardship, with apparently limited success.  Children reported (desire for) shift to financial independence e.g. getting/seeking part time work, not asking for money  Children reported importance of maintaining friendships and a “normal” life - those who had been isolated during serious parental illness experienced negative psychosocial impacts  Needs for support and accommodations at school were raised, esp. by younger (10-13) children  Logistic and material difficulties of hospital stays - lack of privacy, and distress at seeing other sicker patients (dying)  Emphasised importance of parents and healthcare professionals providing accurate and honest information about the cancer, treatment and prognosis  Participants found a sense of community with those who had had similar experiences of parental cancer, and who were understood what it was like. Friendships were made that extended beyond the program.  Participants enjoyed the opportunity to have a fun experience that allowed them to escape from the immediate parental cancer situation.  Mastery of challenges in the adventure-based programs allowed participants to reflect on their strengths and how these might be used to cope with their situation. The use of the challenge course was considered integral to the impact of the program.

Page 59 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Vannatta 2008 USA

0.91

Design: Case-comparison Data Source: Questionnaires

Sample:  60 children of women with breast cancer (8-16yrs; 48% male)  58 teachers  1138 classmates (60 selected as comparison group matched for race, gender & age) Response Rate: Children=81%, teachers= Not reported, classmates=90% Recruitment Setting: Medical oncology practices & schools

(+) Revised Class Play (RCP) (-) Best friend nominations (-) Peer acceptance ratings

Vannatta 2010 USA

0.86

Design: Case-comparison Data Source: Questionnaires

Sample:  40 women with breast cancer (stage I-III; Mage=43.5yrs) & their child (Mage=12.2)  40 comparison mother (Mage= 40.9yrs)/child (Mage=12.4) dyads unaffected by cancer  45% of children male; no reporting of child gender by maternal health status Response Rate: 67% (breast cancer sample), 73% (healthy sample) Recruitment Setting: Medical oncology practices

Child report: (+) Children’s Depression Inventory (+) Children’s Report of Parent Behaviour Inventory – Revised Parent report: (-) Demographics (+) Child Behaviour Checklist (+) Symptom Checklist 90 – Revised (+) Parent’s Report of Parent Behaviour Inventory – Revised (+) Dyadic Adjustment Scale (+) Family Environment Scale

 No differences in indicators of peer acceptance or friendships at school found between children with or without mothers with breast cancer  Indications that boys with a mother with cancer were perceived to be more sensitive-isolated than others by peers & teachers (however their self-perceptions did not reflect this social isolation pattern observed by others)  Children of mothers with breast cancer did not view themselves differently to those with healthy parents on dimensions of social behaviour (positive and negative dimensions)  Children of mothers with cancer perceived by teachers to be significantly less aggressive/disruptive than comparison peers  Conclude: it appears that maternal cancer did not have a negative impact upon social behaviour, peer acceptance and friendships with the exception of the gender difference mentioned above  Ill parental dyads (mothers with BC + spouses) reported higher levels of distress than comparison parents  An inverse association was found between child internalizing problems and parental adjustment in both the BC and control groups (note: magnitude of correlations between ‘parental distress’ and ‘child adjustment’ was not significantly different between the two groups)  No significant differences found between children whose mothers had BC and those with a healthy mother for: 1. Self-reported depressive symptoms 2. Parental ratings of internalizing/externalizing behaviour, and total competence  No significant differences between groups regarding family functioning  ‘Warm & supportive’ parenting style by both mother and father was associated with lower levels of child internalizing behaviour, ONLY for children whose mother had BC  Conclude: the difference in distress was present between parental groups, but not between the BC and control children

Page 60 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Visser 2007 Netherlands

0.94

Design: Cross-sectional Data Source: Questionnaires

Sample:  340 adolescent offspring (1118yrs, Mage=14.9yrs; 44% male)  212 parents with cancer (Mage = 45.4yrs; 20% fathers; time since diagnosis M=2.6yrs) Response Rate: 43% approached by study team, 81% self referral Recruitment Setting: University medical centre & media outreach

(-) Demographics (+) Revised Early Adolescent Temperament Questionnaire – EATQ-R (+) Youth Self Report-Child Behaviour Checklist – YSR (+) Child Behaviour Checklist – CBCL (-) Questionnaire of Recently Experienced Events

Visser 2007 Netherlands

0.95

Design: Longitudinal casecomparison Data Sources: Questionnaires

Sample:  69 ill parents (33-55yrs, Mage = 43yrs; 33% male), diagnosed in previous 12 months  57 spouses (27-56yrs, Mage = 43yrs; 61% male)  57 primary school children (411yrs, Mage = 8yrs; 44% male)  66 adolescent children (1218yrs, Mage = 15yrs; 47% male)  Comparison group of families with a parent who had been diagnosed with cancer 1-5yrs ago - reported in another paper Response Rate: 50% for longitudinal participants; 43% for retrospective comparison group. Retention was 66% at T2 (6m) and 62% at T3 (12m) Recruitment Setting: University Medical Centre

(+) Child Behavior Checklist and Youth Self Report

 Daughters of parents with cancer had significantly more internalising problems than sons (based on ill parent- and selfreports)  Prevalence of problems not significantly related to child age or education or parent age, gender or education  Sig more internalising problems reported when parent had recurrent rather than primary disease  Time since diagnosis not associated with internalising or externalising problems even after controlling for recurrence  Number of negative life events in last year significantly positively correlated with internalising and externalising problems experienced by children  All temperament dimensions significantly related to children’s self-reported internalising problems  Shyness, pleasure intensity & fear/worry significantly related to parent reported internalising problems  Effortful control, frustration & fear/worry significantly associated with parent/child-reported prevalence of externalising problems in children  Temperament seemed to be a more important predictor of problems reported by children than parents  Participants who dropped out tended to report more problems in children - internalising (spouses, adolescents), externalising (ill parents, spouses) and total (spouses, adolescents)  Generally, children were reported to have comparable or fewer internalising, externalising, and total problems compared to both the norm group the retrospective cohort.  Adolescents also self-reported fewer externalising problems (sons, compared to norm group) and internalising and total problems (daughters, compared to retrospective cohort)  Primary school sons were more likely to have internalising problems than the norm group (24% vs. 10%); adolescent sons had more internalising (32% vs.9?%) and total (26% vs. 9%) problems compared to the norm population; adolescent daughters had more internalising (26%), externalising (20%) and total (23%) problems, compared to the norm group (89%).  Prevalence of clinically significant problems decreased to approximately normative levels by T3, although adolescent sons still had clinically elevated risk of internalising problems.  Internalising, externalising, and total problems decreased significantly over time, although problems were quite stable.

Page 61 of 64

Reference

Quality

Study Design & Data

Sample & Context

Measures

Findings

Wong 2009 USA

0.90

Design: Qualitative Data Source: Semi-structured interviews

Sample:  27 young adults (18-38yrs, Mage =23.7yrs; 30% male,) who had a parent diagnosed with cancer during childhood (44% fathers; 40% deceased at time of interview) Response Rate: Not reported Recruitment Setting: Community outreach

(-) Demographics (-) Open-ended interview targeting how parental cancer had affected the participant’s life

 Open-ended interviews revealed both positive and negative long-term outcomes evolving from having lived through one’s parents’ cancer journey (inclusion criteria allowed for adults whose parent was either still alive or had passed away due to cancer)  44% of participants expressed posttraumatic growth experiences  Four main themes of posttraumatic growth were revealed: 1. Improved character (maturing early, more compassionate) 2. Increased appreciation for life (increased gratitude) 3. Strengthened personal relationships (closer to family) 4. Increased interest in cancer issues (seeking career in oncology)  59% of participants expressed adverse consequences  Four main domains of negative impacts/adverse consequences of having a parent with cancer were revealed: 1. Feelings of loss and void (following parent’s passing) 2. Concern for their own health (concerns for genetic vulnerability) 3. Negative changes in outlook on life (difficulty trusting others, fear of loss, lack of control) 4. Negative effects on personal relationships (difficulty with family communication  3 participants reported the impact of their parent’s cancer had been neither positive nor negative

Page 62 of 64

Table 2: Constructs, Measures and Utilisation in Quantitative Studies Construct

Measure

Studies

Psychological Functioning > Internalising/Externalising Problems

+ Youth Self Report (YSR)

+ Symptom Checklist 90-Revised (SCL-90-R) + Rorschach Comprehensive System + Patient Health Questionnaire (PHQ)

Edwards, 2008 (1); Gazendam-Donofrio, 2007 (2); GazendamDonofrio, 2008 (3); Gazendam-Donofrio, 2011 (4); Huizinga, 2010 (5); Huizinga, 2011 (6); Krattenmacher, 2014 (7); Lindqvist, 2007 (8); Thastum, 2009 (9); Visser, 2007a (10); Visser, 2007b (11) Edwards, 2008(1); ForanTuller, 2012 (12); Gazendam-Donofrio, 2007 (2); GazendamDonofrio, 2008 (3); Huizinga, 2010 (5); Krattenmacher, 2014 (7); Lindqvist, 2007 (8); Thastum, 2009 (9); Vannatta, 2010 (13); Visser, 2007a (10); Visser, 2007b (11) Janzter, 2013 (14) Edwards, 2008 (1) ; Gazendam-Donofrio, 2007 (2); Huizinga, 2010 (5); Huizinga, 2011 (6) McDonald, 2016 (15); Patterson, 2013 (16) Rainville, 2012 (17) Küçükoğlu, 2013 (18) Edwards, 2008 (1) ; Schmitt, 2008 (19); Thastum, 2009 (9) Davey, 2013 (20); Vannatta, 2010 (13) Costas-Muñiz, 2012 (21); Götze, 2015 (22) Costas-Muñiz, 2012 (21); Huizinga, 2011 (6) Davey, 2013 (20) Götze, 2015 (22); Kühne, 2012 (23); Möller, 2014 (24) Patterson, 2011 (25) Foran-Tuller, 2012 (12) Gazendam-Donofrio, 2011 (4) Inbar, 2013 (26) Janzter, 2013 (14); John, 2013 (27); Krattenmacher, 2013 (28); Kühne, 2012 (23); Möller, 2014 (24); Patterson, 2011 (25) Vannatta, 2010 (13) Flahault, 2010 (29) Foran-Tuller, 2012 (12)

+ Parental Role Satisfaction Tool

Cho, 2015 (30)

+ Child Behaviour Checklist (CBCL)

> Distress

+ KiGGS Child and Youth Health Survey + Impact of Events Scale (IES)/Child Impact of Events Scale (CIES)

+ Kessler-10 (K10)

> Depression

+ Indice de détresse psychologique de Santé Québec (IDPSQ-14) + Brief Symptom Inventory (BSI) + Beck Depression Inventory (BDI-II)

+ Children’s Depression Inventory (CDI) + Centre for Epidemiological Studies Depression Scale (CES-DC) > Anxiety

+ Spielberger State-Trait Anxiety Inventory

> Depression, Anxiety & Stress

+ Revised Children’s Manifest Anxiety Scale + Hospital Anxiety & Distress Scale (HADS)

> Post-Traumatic Stress > Emotional Reactions > General Mental Health

Family & Social Environment > Parental Role Satisfaction

+ Depression, Anxiety & Stress Scale (DASS21) + UCLA PTSD Index + Situation-specific Emotional Reaction Questionnaire – Child (SSERQ-c) + Mental Health Inventory + Strengths & Difficulties Questionnaire (SDQ)

Page 63 of 64

Construct > Family Functioning

Measure + McMaster Family Assessment Device (FAD)

+ Family Relationships Index (FRI) + Family Environment Scale (FES)

> Marital/Partnership Functioning > Relationships > Social Support > Social Functioning

> Parent Behaviour Health-Related Quality-of-life

+ Maudsley Marital Questionnaire – Satisfaction + Dyadic Adjustment Scale + Experiences in Close Relationships Inventory + Interaction Behaviour Questionnaire (IBQ) + Scale of Perceived Social Support from Family & Friends + Oslo Social Support Scale (OSS) + Revised Class Play (RCP) - Best Friend Nominations - Peer Acceptance Ratings + Children’s Report of Parent Behaviour Inventory – Revised + Parent’s Report of Parent Behaviour - Revised + Child Health Questionnaire (CHQ) + SF8 medical outcomes health survey (SF8)

+ KIDSCREEN Parent & Adolescent Versions

+ RAND-36 Quality of life measure + EORTC 30-item core quality of life questionnaire (QLQ-C30)

Coping

+ Inventory for Quality of Life in Children and Adolescents (ILC) + KIDCOPE

Needs

+ Freiberg Questionnaire of Coping with Illness (FKV) + Brief COPE + Offspring Cancer Needs Instrument (OCNI)

- Disclosure of imminent parental death – views & experiences Eating Disorders Self-harm Performance Status Communication Temperament Experiences Socioeconomic Issues

Intervention Fidelity & Satisfaction

+ SCOFF eating disorders questionnaire + Paykel Suicide Scale - Modified Deliberate Self Harm Inventory + WHO-ECOG Performance Status + Parent-Adolescent Communication Scale - Measure of general parent-child communication + Revised Early Adolescent Temperament Questionnaire (EATQR) - Questionnaire of Recently Experienced Events + Winkler Index Socioeconomic Status - Financial situation questionnaire - Attendance & Fidelity - Modified Parent Consumer Satisfaction measure

Studies Edwards, 2008 (1); Kühne, 2012 (23); Lindqvist, 2007 (8); Möller, 2014 (24); Schmitt, 2008 (19); Thastum, 2009 (9) McDonald, 2016 (15) Edwards, 2008 (1); Gazendam-Donofrio, 2007 (2); Vannatta, 2010 (13) Huizinga, 2011 (6) Vannatta, 2010 (13) Inbar, 2013 (26) Davey, 2013 (20) Küçükoğlu, 2013 (18) Götze, 2015 (22) Vannatta, 2008 (31) Vannatta, 2008 (31) Vannatta, 2008 (31) Vannatta, 2010 (13) Vannatta, 2010 (13) Edwards, 2008 (1) Edwards, 2008(1); Götze, 2015 (22); Krattenmacher, 2014 (7); Lindqvist, 2007 (8); Möller, 2014 (24); Schmitt, 2008 (19); Thastum, 2009 (9); Götze, 2015 (22); Kühne, 2012 (23); Krattenmacher, 2013 (28) Gazendam-Donofrio, 2008 (3) Götze, 2015 (22) John, 2013 (27) John, 2013 (27) Kühne, 2012 (23); Krattenmacher, 2013 (28) Möller, 2014 (24) Costas-Muñiz, 2012 (21) McDonald, 2016 (15); Patterson, 2011 (25); Patterson, 2013 (16) Bylund-Grenklo, 2015 (32) Janzter, 2013 (14) Janzter, 2013 (14) Janzter, 2013 (14) Möller, 2014 (24) Cho, 2015 (30) Davey, 2013 (20) Visser, 2007 (11) Visser, 2007 (11) Möller, 2014 (24); Krattenmacher, 2013 (28) Torp, 2013 (33) Davey, 2013 (20)

Page 64 of 64