Abstract
Abstract
Although the whole family is affected by a parent's palliative disease, palliative care research does not yet routinely consider patients' minor children. Children's and adolescents' psychosocial functioning may be impaired during prolonged parental disease with poor prognosis. Therefore, more and more health care providers are establishing clinical initiatives for families of palliative patients with minor children. However, the number of these family interventions, as well as their theoretical and empirical backgrounds and evidence base, has yet to be determined. The purpose of this study was to systematically review structured and published interventions for this target group, as well as empirical studies on these interventions. The evidence base and impact of interventions on families were considered. Literature published between 1980 and present focusing on psychosocial family-, child- or parent-centered interventions during palliative care was retrieved from PsycINFO®, Embase, MEDLINE®, CINAHL®, and PSYNDEX databases. Five interventions met the inclusion criteria. Programs focused on different populations, had diverse empirical and theoretical backgrounds and features, and were evaluated by studies of varying methodological quality. This systematic review illustrates the lack of well designed and elaborated intervention concepts and evaluation studies in this field, highlighting the necessity of conceptual and methodological rigor to inform clinical practice on a sustainable basis in the future.
Introduction
Anticipated parental death is a crisis for every family member. 9 Both non-ill 10 and seriously ill parents 11 experience confusion concerning their parental competence. Child-centered parenting by healthy parents can decrease with lengthy disease of the partner. 12 In turn, both influence the adaptation of children to parental loss.12,13 Poorer patient prognosis, more intensive treatment, worse functional impairment, reduced mental health, and perception of parental illness as more serious are associated with negative self-reported psychosocial functioning of children of physically ill parents.14–17
Children have less cognitive capacity and coping strategies than adults to deal with such a crisis. 18 The predictability and stability of everyday life can disappear. Due to psychosocial stress, healthy spouses may have difficulty adequately meeting their children's needs, which may in turn exacerbate their children's situation. 12 Children fear the ill parent's death and worry about the healthy parent‘s well-being. 19 They are afraid of the unknown as well as of not being able to effectively cope with the situation. 20 The association between disease severity and child functioning warrants further exploration, 14 and as parental palliative disease seems to play a major role in child adjustment, reviews should account for this stage of the illness. 7 Although of importance, other reviews explicitly exclude parental palliative disease.3,16
Against this background, our review aims (1) to identify structured psychosocial interventions published and evaluated for families of palliative patients with minor children, (2) to analyze the evidence base of those interventions, and (3) to evaluate the impact of the interventions on families (if applicable).
Methods
Search strategy
When developing our concept of child-oriented medical family counseling towards a manualized interventional concept including special modules for families in palliative care in the context of a German multisite research project, 21 this review was part of a literature search of related international interventions.
A systematic review approach22–24 was adapted to observational studies and descriptions of intervention concepts. English and German publications since 1980 were identified searching the databases PsycINFO, PSYNDEX, Embase, MEDLINE, and CINAHL in March 2011. The first search included the keywords (child* OR parent* OR family OR spous*) AND (terminal* OR palliative) AND (intervention* OR counsel* OR therap*). Because couple interventions were thought to possibly contain child-related issues as well, this keyword was further integrated. To cover interventions beginning in palliative care and lasting throughout bereavement, the subsequent search added the keywords (parent* death OR death of a parent OR dying parent OR bereaved child* OR child bereav*).
Hand searches were conducted in April and May 2011 through palliative care and psychooncology journals, reference lists of relevant books, included articles, and relevant review articles. An online search of relevant websites was carried out to identify evaluated intervention concepts. The full search strategy is available upon request from the first author.
Inclusion criteria and assessment of study quality
Articles were included based on these criteria:
In accordance with our study questions, publications exclusively on child bereavement were excluded. Dissertation abstracts, case studies, anecdotal reports, and comments were excluded as well.
After the first elimination of studies based on title and abstract (see Fig. 1), the raters included/excluded and assessed publications independently, based on the full texts and according to a checklist requiring a decision based on each inclusion criterion (yes/no). The two trained raters were the first author (FK) and a master's psychology student with advanced methodological skills (JCG). If there were divergent assessments, consensus was achieved by discussing inclusion criteria.

Flow chart of the literature search procedure.
According to a systematic review of tools for quality assessment of observational studies, 25 critical appraisal checklists provided by Cardiff University 26 were utilized. They were chosen because different study types are evaluated according to one common and comparable procedure. Depending on study design, raters assess specific aspects of study aims/focus, method, population, and results; they provide concluding ratings according to predefined questions. In our review, both raters independently evaluated whether or not there was evidence for the respective questions or whether there was unclear evidence. Afterwards, discordant assessments were discussed and a concluding decision was made.
Study quality will be reported, but due to small sample sizes of heterogeneous studies, no minimum-quality criterion was set for inclusion. Two independent raters assessed study quality, then discussed their findings. Interrater reliability reached a Kappa coefficient of K=.75, which can be categorized as good. 27 Data extraction and report are based on prominent publications in the field.16,28,29
Results
While in total, 5 interventions described in 24 publications fulfilled the inclusion criteria and were included in this review, many interventions identified during the literature search did not fulfill the inclusion criteria. Exclusions involved not being structured by time and/or content,30–47 not being focused on palliative care or exclusively focused on bereavement,48–71 focused on early-stage cancer72–73 or recurrent breast cancer, 74 not including minor children/parents,75–78 or not being evaluated in an empirical study79–83 (listed regarding first inclusion criterion that did not fit). For reasons and corresponding numbers of exclusions dependent on the stage of the process, see Figure 1.
Psychosocial family interventions
Table 1 gives an overview of intervention concepts including brief characterizations. Interventions were developed in the United States,84,85 mainly in Australia, 86 or predominantly in Europe. 87 During a conjoint European project (COSIP), three independent modifications of an intervention were developed in collaboration between partner centers.88–90
Interventions addressed adolescents 89 or children and adolescents.84,87–90 They focused on parental physical illness, 88 cancer,89,90 terminal cancer,84,97 terminal illness, 86 or HIV/AIDS. 91 Although one program was predominantly developed for adult patients and their families, we included it because one of its topics involves the “needs of younger children.” 91 The COSIP interventions were not exclusively designed for families in palliative care.88–90 Yet as in most other interventions,84,86,91 they include anticipatory/adaptive grief as one intervention goal. This and further intervention aims (parenting competence, active coping, communication) capture the concerns of families mentioned above. All authors indicated that developmental issues are a crucial determinant of counseling.
Theoretical and empirical backgrounds were diverse, and intervention development more or less explicated (see Table 1 and Fig. 2). Important indications were illness severity,84,87 family functioning, 86 or disease disclosure/parental death in the course of the intervention. 91 The most frequent recurrent intervention aims were parenting competence,86–91 anticipatory/adaptive grief,84,86,88–91 active coping,86–91 and communication.84,88–90 Intervention goals are also related to important risk and protective factors of psychosocial adaptation of minor children in cancer families, such as familial communication and functioning, parental psychological functioning, and coping strategies.14–16, 93, 94

Features of intervention concepts (multiple answers).
Programs were described as needs oriented,86–91 preventive,84,86–88 or brief and focused84,86,88–90 by their authors. Regarding number of sessions, the length of interventions varied between short86–90 and long.84,91 In all cases, individual and group sessions were offered, except in one case 86 of exclusively family sessions. All interventions were face-to-face, one provided regular telephone contacts, 84 and four were also conducted in the family home.84,86,89,90 One intervention did not include the ill parent on a regular basis. 84
In three interventions, an unpublished manual, 87 verbal descriptions, 84 or a published practical guide 88 describing the intervention were mentioned. If the authors had published a treatment manual,86,91 intervention techniques and tailoring to the treatment unit were elaborated, and both training and supervision were mandatory. Otherwise, authors listed training 87 or regular supervision.84,88,89 Therapists' backgrounds were mainly in social work, family therapy, and psychology or psychiatry (see Fig. 2).
Related evaluation studies
Table 2 summarizes characteristics of evaluation studies and their main results. Studies were mainly quantitative and cross-sectional (see Fig. 3). Three studies87,89,95 did not incorporate a control group, three compared the intervention with standard care/absence of counseling,90,96,97 and one with telephone monitoring. 98 Sampling was either convenience87,89,90,95 or randomized.96–98 Recruitment took from six months 95 to five years,96,98 and participation varied from 44%–57%87,95,96,98 to 72%–89%.89,90,97 There were studies with smaller,87,90,95 medium sized,89,98 and larger samples.96,97

Features of evaluation studies (outcomes, design, approach; multiple answers; RCT, randomized controlled trial).
A-adolescents, C-children, CG-control group, IG-intervention, Pt-patients, T-therapists, WP-well parents.
Two research projects investigated therapist fidelity.97,99 Participants versus nonparticipants or completers versus noncompleters were compared by three research groups.89,90,100 Reasons for nonparticipation were family- or disease-related burden or refusal of individual family members. 89 Parents in the counseling group were older, more often single, 92 and reported more emotional distress. 90
Study quality was diverse. Independent of design, and referring to the critical appraisal checklists, 26 there were four studies of higher89,90,96,97 and three of lower87,95,98 methodological quality.
Studies considered contentment, experiences with, or implementation of the intervention more often than psychosocial outcomes (see Fig. 3). Children mentioned fear of parental death and an inability to talk to their parents about their concerns. 87 Adolescents especially experienced communication with healthy fathers as problematic. 98 When families participated in a program supporting parenting, children perceived improved parental competence and communication. 98
Altogether, participating families indicated that they were satisfied with counseling.90,95,97,98 In one study, 95 therapists were more critical than family members, whereas in another study 91 there were no differences between therapists' and family members' ratings. In a further study, medical staff gave a more cautious evaluation. 89
In one study, mothers reported higher levels of depressiveness 90 at baseline. In the pre-post counseling comparison, parent- and child-reported depressiveness decreased. 90 Another study 96 highlighted that relationships with children were worse in dysfunctional families with parents with higher depressiveness scores. Female and older adolescents of HIV/AIDS patients reported more emotional distress. 97 Parental emotional distress predicted emotional distress in adolescents, and, if the mother was ill, adolescents reported more emotional distress and risky sexual behavior. 97 Parental physical health predicted adolescent problem behavior and parental death was associated with more emotional distress and adolescent problem behavior. 97
According to a one-group pre-post measurement, parent-reported family functioning improved after counseling. 90 In family-focused grief therapy (FFGT), families had different benefits from therapy.96,101 Precategorized “sullen” and “intermediate” families tended to benefit more from FFGT, whereas members of “hostile” families benefited more from individual psychotherapy. Furthermore, the most distressed family members improved significantly. Correspondingly, Christ and colleagues 102 proposed a categorization of high-risk, vulnerable, and resilient families based on their clinical experience.
In counseling for parents with AIDS/HIV and their adolescent children,103,97 although both groups profited in various domains, positive effects seemed to last for the first two years, primarily for the program's duration. At six years 104 there were still positive effects in the intervention group, although only evident in secondary outcomes.
Discussion
Few interventions with various backgrounds
There is a small number of structured, published, and evaluated interventions for families of palliative patients with minor children. Their theoretical and empirical background varies, as well as the comprehensibility of their design. Furthermore, in the field of interventions for children of cancer patients, programs are seldom developed according to evidence-based or theory-driven frameworks. 16 Our review reveals that successful interventions begin in palliative care and last throughout bereavement.84,86,91 Publishing manual-based interventions is a means of quality management, as authors describe important aspects of their interventions more thoroughly. Number of treatment sessions varied in accordance with different populations and conceptualizations. One research group offered booster sessions, but did not control for participation. 97 Therefore, our findings support the conclusion of other authors who stress the necessity of an investigation of effective “doses” of support in family caregiver palliative care research, 8 also with due regard to minor children.
In accordance with empirical results regarding risk factors,7,15,16 participating parents were more often single, 89 emotionally more distressed, 90 and children mentioned fear of parental death. 87 Therefore, it is promising to support the most distressed subgroups and to respect resilience in well-functioning families.89,105 It needs to be defined whether psychosocial burden means symptomatology, limited functioning, decreased health-related quality of life, or family functioning. 89 Kissane and colleagues (2007) 100 propose routine screening of family functioning to facilitate recognition of high-risk families. The authors prefer individual therapy for members of hostile families in the light of potential negative effects within the scope of a preventive model.
Defining the appropriate time point for screening is very important, because parental palliative disease is not a single event but a phase of familial transition. 10 Self-assessed need hints at individual shock and psychosocial burden, and should coincide with offering support. 89 Periodic assessment and clinical experience are important further ways of determining distress. 89 Consistent with the fact that “there is no one-size service that fits all”106(p. v), research should unravel effective aspects of treatment. 29
Diverse evidence base
As there is only a handful of methodologically sound studies,89,90,96,97 the evidence base of child-centred interventions in families with parental palliative disease is rather low. This also applies to studies on children of cancer patients in general, where intervention studies are mostly poorly constructed.28,94 Methodological shortcomings include widely used cross-sectional designs, lack of control groups, heterogeneous questionnaires that cannot be easily compared, inconsistent treatment exposure, and overall, rather low participation rates.
In one study, patients refused participation because of organizational and disease-related burdens, 89 which is perhaps a core feature of the population studied and may require flexibility on the part of therapists and researchers. Sampling bias arises from convenience sampling87,89,90,95 and clinical personnel that potentially screened out more dysfunctional families. 86 Furthermore, certain subgroups participated less, such as avoidant families, 86 or fathers and Spanish-speaking mothers in another study. 97 Depending on parental disease, samples were of higher86,89 or lower 97 socioeconomic status.
Researchers relied on self-report data and often used additional self-established measures. A great deal of effort was invested in large-scale studies, but most comparisons were nonsignificant, perhaps due to poor sensitivity of the applied measures. Given the deteriorating physical condition of the palliative patient, the focus on improvement of most instruments should perhaps be reconsidered, since psychosocial stabilization is an important treatment goal in itself in these families. Likewise, resilience is still rarely taken into account in studies on families of palliative patients with minor children.
Most participants were satisfied with the interventions.90,95–98 Although social desirability is quite likely, contentment with a program is an important prerequisite for building a therapeutic working alliance. The evaluation studies included in this review indicate that the identified interventions for families of palliative patients and their minor children tend to improve communication, depressiveness, and family functioning.90,96–98 Nevertheless, due to theoretical and methodological weaknesses, these findings should be considered in the context of the respective evidence base.
Conclusions
Parental palliative disease can be a particularly critical situation for dependent children, and long-lasting illness is associated especially with internalizing child symptoms.7,16 Although there is a variety of family, child and parent-centered concepts, only five structured and published interventions met our inclusion criteria. As presented, successful interventions already begin in palliative care and last throughout bereavement, transferring interventions into manual-based forms adds to quality management, and supporting the most distressed subgroups presents an adequate approach.
As with research on interventions for cancer families with minor children,3,4,15,28 there is a lack of well-designed and elaborated intervention concepts and evaluation studies in palliative care. Future empirical studies are facing the tasks of screening for high-risk families and individuals according to self-assessed need and clinical assessment; of study recruitment against the background of physical deterioration of the patient; of defining appropriate outcome criteria (e.g., stability, resilience); and of selection and development of sensitive outcome measures. Stringent development of specific interventions for families of palliative patients with minor children is necessary and needs to be theoretically clear, empirically substantiated, clinically sensible, and supported by methodologically sound intervention studies. This can help to explain and establish what and why psychosocial professionals do what they do.
Footnotes
Acknowledgments
We would like to thank Doreen Kopper for her assistance with literature search and acquisition, and Monica Blotevogel for English copyediting. The study is part of the German multisite research project “Psychosocial Services for Children of Parents with Cancer” supported by German Cancer Aid (Deutsche Krebshilfe, grant #108303). In this multisite project the following institutions and principal investigators are collaborating:
Author Disclosure Statement
No competing financial interests exist.
