Critical Reviews in Oncology/Hematology 95 (2015) 370–386

Psychosocial interventions for adolescents and young adult cancer patients: A systematic review and meta-analysis Diana Richter a,∗ , Michael Koehler b , Michael Friedrich a , Inken Hilgendorf c , Anja Mehnert a , Gregor Weißflog a a

University Medical Center Leipzig, Department of Medical Psychology and Medical Sociology, Philipp-Rosenthal-Straße 55, 04103 Leipzig, Germany b University Medical Center Magdeburg, Department of Hematology & Oncology, Leipziger Straße 44, 39120 Magdeburg, Germany c University Medical Center Jena, Clinic for Internal Medicine II, Department of Hematology and Internal Oncology, Erlanger Allee 101, 07747 Jena, Germany Accepted 7 April 2015

Contents 1. 2.

3.

4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Inclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Study selection and data extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Data analysis/synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Assessment of risk of bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Study selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Study characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Types of intervention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1. Technology-based interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2. Psychoeducational interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.3. Physical exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.4. Multi-component interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Quality of included studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Quantitative analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Implications for future research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Authors’ contribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reviewers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biographies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

371 373 373 373 373 373 374 374 374 375 379 379 380 380 380 381 381 381 383 383 384 384 384 384 384 384 386

Abstract Adolescent and young adult (AYA) cancer patients experience unique psychosocial needs and developmental challenges. A cancer diagnosis can stress this development and disrupt AYAs in their normal life. ∗

Corresponding author. Tel.: +49 341 9715426; fax: +49 341 9715419. E-mail address: [email protected] (D. Richter).

http://dx.doi.org/10.1016/j.critrevonc.2015.04.003 1040-8428/© 2015 Elsevier Ireland Ltd. All rights reserved.

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

371

The aim of this systematic review and meta-analysis was to assess the impact of psychosocial interventions on mental health in AYAs. A literature research was conducted, which resulted in twelve eligible studies. The standardized mean difference between intervention and control conditions was 0.13 (95% CI: −0.16 to 0.42) for quality of life, 0.27 (95% CI: −0.22 to 0.76) for cancer-related knowledge and −0.16 (95% CI: −0.73 to 0.42) on psychological distress indicating, small and non-significant effects for interventions improving mental health. This work strengthens the need for age-appropriated interventions in psycho-oncology. Future research should develop interventions more graduated by age. Randomized intervention studies with larger samples and focusing psychosocial outcomes are needed to establish evidence-based psycho-oncological interventions for AYAs. © 2015 Elsevier Ireland Ltd. All rights reserved.

Keywords: AYA; Cancer; Interventions; Psycho-oncology; Systematic review; Meta-analysis; Adolescents; Young adults

1. Introduction According to the definition by the National Comprehensive Cancer Network® (NCCN), adolescents and young adult cancer patients (AYAs) are generally defined as individuals aged between 15 and 39 years at their initial cancer diagnosis [1]. Nevertheless, there are inconsistencies in defining the age range for AYAs [2]. The specified age range was chosen because of typical issues and needs relevant to this group. In contrast to younger and older cancer patients, AYAs are different in their biological, psychosocial and socio-behavioral characteristics [3]. The incidence of cancer in AYAs has increased in Europe as well as in the USA in past decades [4]. In Germany, each year 15,000 patients aged between 15 and 39 years are diagnosed with cancer. Those are approximately 3% of all people diagnosed with cancer each year in Germany [5]. Survival depends on the specific cancer type, which means that the survival rate can be lower than in younger or older cancer patients e.g. for non-Hodgkin lymphoma or acute lymphoblastic leukemia [6]. AYAs with cancer represent a unique population in medical as well as in psychosocial terms [7]. While survival rates for younger children and older adults (age >40 years) have increased because of the progress of oncological treatments, there have only been small improvements in treatment in the AYA-group in the last three decades [8]. Biomedical differences between AYAs and other cancer patients arise as a result of a wide variety of cancer sites. Some cancer sites like Hodgkin’s lymphoma or testicular cancer occur most often in the AYA group [9]. Other cancer sites have higher incidence rates during older adulthood. In addition, the cancer incidence increases from one age cohort to the next and the distribution pattern of diagnoses (Fig. 1) changes continuously [10]. AYAs may tolerate more intensive therapies due to their decreased co-morbidity [1]. In turn, this could lead to several acute side-effects or long term consequences such as infertility, secondary malignancies or cardiac dysfunction [11]. AYAs are faced with specific psychosocial and sociobehavioral challenges. First of all, there are often several rapidly changing developmental issues ahead, such as education, career and planning a family. In addition, in adolescence

challenges may include identity/sexual development problems, struggles for autonomy from parents and issues related to dealing with a “new” body image [12]. A cancer diagnosis often means disruption to daily life including school or work and isolation from peers and family [13]. Additionally, in this age-group individuals have to tackle developmental tasks such as identity formation and career and family planning, all of which demand integration into normal life. Hence, coping with these developmental tasks alongside the experience of a cancer diagnosis and treatment, requires good support, sufficient coping strategies and increased sensitivity from others. A majority of AYAs become more dependent on their parents again because of increased care and financial needs [14], despite the growing desire to be autonomous [3]. Table 1 illustrates specific medical and psychosocial characteristics of AYA cancer patients. While most psycho-oncological studies of childhood cancer survivors or children and older adults with cancer focus on treatment, survival and interventions, few studies exist in the field of AYAs dealing with their specific needs and most prominent symptoms of this age-group [15–17]. Erickson [15,18] and Daniel [19] reported low physical activity including fatigue, nausea, pain and sleep disturbance as being the most common physical side-effects for AYAs. Furthermore, AYAs report a lower quality of life in comparison to their healthy peers [20] and have a higher risk of developing post-traumatic stress symptoms [21]. Yanez [16] showed in a study with 322 young adult (mean age 31.8) that cancer survivors had a higher level of cancerrelated distress. This distress was higher 13–24 months after treatment completion in comparison to the 25–60 months cohort (p < .01). Kwak [17] examined changes of distress over time in a study with 215 young adults between 14 and 39 years (mean age 23.6). He found higher levels of anxiety in AYAs at diagnosis, which increased further between 6 and 12 months post treatment (means anxiety scores of the Brief Symptom Inventory at baseline: 52.7, 6-months follow-up: 49.8 and 12-months follow-up: 51.5). Additionally, AYAs have reported specific unmet needs during treatment and in the transition into survivorship. Those most cited are psychosocial and information needs such as support from family and friends, psychological counseling,

372

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

Fig. 1. Incidence of the most occurring cancer types in AYAs compared to older age-groups (2010, Germany; source: RKI)

receiving age-appropriate information from services and Internet sites and meeting peers and cancer survivors [22]. The wishes of AYAs vary, including wanting sexuality interventions [23,24], cancer-related websites with peer interaction, games and information [25] and preparation for “life after disease” [26]. Younger survivors (18–40 years) have a greater desire for cancer-related information, information about the health system and sexuality issues [27]. The majority of the older AYAs (age >25 years up to 39 years) want to

have children after overcoming their disease, so information about fertility and fertility preservation methods is getting into the focus [28]. Previous explanations have emphasized the unique position of AYAs with their specific challenges, life goals and requirements distinguishing them from younger or older cancer patients. A comprehensive overview about interventions that are specifically designed for AYA to improve short and long-term psychosocial well-being (which includes

Table 1 Medical and psychosocial specifics of AYA cancer patients (adapted from NCCN AYA Guidelines® ). Topic

Specific

Spectrum of cancer types that affect the AYA population with distinct biology of diseases Treatment related/survivorship issues

a. Different from cancer types that affect the pediatric and older patients

Fertility and endocrine considerations

Psychosocial/behavioral considerations

Socioeconomic issues

a. Dose schedules (chemotherapy) b. Toxicities as long-term side effect (cardiovascular, pulmonary, neurological and endocrine complications, nephrotoxicity) c. Risk for a second primary malignancy a. Offering sperm banking b. Discussing embryo cryopreservation c. Oophoropexy d. Oocyte cryopreservation e. Menstrual suppression f. Risk for a child with impairment a. Treatment adherence (lack of participating in clinical trials) b. Chemobrain (both cognitive and emotional impairment) c. Involvement/interruption of school/work d. Relationships (peer, parents, partner, health care provider) e. Tobacco, alcohol or substance abuse (joy of experimenting this developmental stage) f. Sexual behavior/risks/concerns (e.g. frequently alternating partners in this age-group) g. Existential/spiritual matters (e.g. premature confrontation with mortality) a. Loss of employment b. Financial dependence vs. wish for autonomy c. Risk for financial loss/losing insurance d. Transportation/driver license e. Child care

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

several psychological and social self-report parameters) is still lacking. Therefore, the present study aims to systematically identify interventions from this field. The research questions are as follows: (a) Interventions Which AYA-specific intervention studies were conducted in the last decade? (b) Outcome/Effects Which outcomes representing psychosocial issues were measured? Which effects were reported? (c) Need for improvement Which suggestions for improvement or modification of the interventions were made?

2. Method 2.1. Inclusion criteria At the beginning of our systematic review process we formulated a protocol in accordance with the PICOS approach [29]. Consideration was given to studies that fulfilled the criteria in Table 2. The exclusion criteria were also reported in Table 2. The review was registered at “PROSPERO International prospective register of systematic reviews” (CRD42014007541). 2.2. Search strategy A systematic computer-based literature search was conducted in relevant databases on the basis of the PRISMAStatement [30]. Databases included were Pubmed, Web of Science, PsycINFO/Psyndex and the Cochrane Database of Systematic Reviews (CDSR) for the period from January 2003 to September 2013. The search was re-run immediately prior to analysis in February 2014 and further studies retrieved for inclusion, to ensure that the most current information is presented in this review. We used the search terms in combination presented in Fig. 2. In addition, we screened reference lists of all included studies and relevant review articles on related topics [31–34]. The Internet was also searched for websites of clinics or organizations with a special interest in AYA. For research in progress, we searched platforms with study protocols of registered trials http://clinicaltrials.gov, http://drks-neu.uniklinik-freiburg.de and http://www.who.int/ictrp/en. 2.3. Study selection and data extraction After eliminating duplicates using CITAVI (reference management program), all abstracts were independently screened by two reviewers based on title, keywords and abstract using the inclusion criteria. Any disagreements

373

regarding inclusion were resolved through discussion. In the case of any disagreement or doubt about potentially relevant articles, the full article was obtained. If all of the criteria were met after reading the full article, three reviewers decided about whether the study was included in the review. Following this, a data extraction form was used by two reviewers (DR and MK) to independently extract information on study characteristics, including the intervention type, study design, sample, instruments used and results. Any disagreements between the two primary authors were resolved by consulting a third reviewer. 2.4. Data analysis/synthesis Firstly, the included studies were categorized according to their study design and summarized in tables to give a descriptive overview of them. A meta-analysis of the outcomes was conducted when (a) sufficient data of outcome measures were provided and when (b) there was a comparison or control group. The Cochrane Collaboration’s analysis software “Review Manager 5.2” was used to carry out the meta-analyses. Only one effect per study for each outcome was included. When several instruments were used for the same construct (dependent measures) one instrument was excluded. The first exclusion criterion was the psychometric property of a measurement. When measurements had similar reliabilities we chose the instrument which was also used in another included study. To conduct the meta-analysis, Hedges’ g effect sizes (an adjusted version of Cohen’s d) and 95% confidence intervals were calculated for continuous data based on means and standard deviations for the intervention and control group (between-group) post treatment (first assessment/same period of time after intervention). Hedges’ g was used to measure the effect size, because other measures would not be appropriate for our purpose. The pooled standard deviation (used in Cohen’s d and Hedges’ g) accounts for different sample sizes. But if in cases with low sample sizes Cohen’s d would overestimate the standardized mean difference. Therefore it was reasonable to correct d with a factor – Hedges’ g [35]. Trials were pooled using a random effects model and displayed on forest plots [36]. The effect of Hedges’ g may be interpreted as small (0.2), medium (0.5), and large (0.8). It would be inappropriate to use random effects models only, when the test is significant, because of the low power of the test on account of the low sample size in each group. Nevertheless, we decided in favor of random effects models, because if T2 turns out to be zero, the random-effects analysis reduces to the fixed-effect analysis [35]. Whenever meta-analyses were conducted, heterogeneity was assessed by inspecting the forest plots, and using the I2 statistic. The varying influence of different instruments was minimized by distinguishing between different types of scales in the meta-analyses, which resulted in a small number of studies being included. To overcome the problem of differing

374

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

Table 2 Inclusion and exclusion criteria for selection of the articles. Inclusion criteria (1) Population Cancer patients (AYAs) aged between 15 and 39 years at the time of diagnosis and time of the intervention Intervention (2) All interventions which aimed to improve psychosocial wellbeing, also as a secondary outcome measure; irrespective of intensity, setting, duration and frequency Comparison (3) A comparison group in the intervention study was not required (4) Outcome Any self-reported psychosocial outcome; at least two measurement points (5) Study design All quantitative and mixed method studies allowing conclusions about the efficacy of the intervention - Written in English or German language - Published between January 2003 and February 2014 Exclusion criteria - Studies with focus on children or older cancer patients (age: 39 years), long term survivors (>3 years; temporal proximity to active treatment) and survivors of childhood cancer - No comments, editorials, case studies, purely qualitative research, poster, dissertation abstracts - No intervention studies solely focusing on improvement of physical outcomes and medication (interventions with a physical component and psychosocial component as secondary outcome were considered for this review) - Studies with mixed chronic disease samples (incl. cancer) - Studies providing assessment data, screening methods or descriptions of interventions - Unpublished study results

cancer OR tumor OR tumour OR tumors OR tumours OR carcinoma OR neoplasm OR neoplasms OR oncology OR oncological AND psychosocial OR psychological OR psycho oncological OR psychoeducaon OR psychotherapy OR behavior OR behaviour OR behaviours OR behaviors OR behavioral OR behavioural AND intervenon OR training OR therapy OR therapeucal OR treatment OR support OR counsel* AND adolescent OR adolescents OR adolescence OR "young adult" OR "young adulthood" OR AYA OR child OR children OR childhood Fig. 2. Search terms

scales directions within individual instruments the mean values of the corresponding studies were multiplied by minus one as proposed by Review Manager. A sensitivity analysis was conducted when heterogeneity exceeded I2 > 50%. When data for quantitative synthesis were not available, the analysis had to be restricted to only a narrative approach, because formal pooling of results would have been inappropriate. Narrative data synthesis was conducted by describing studies according to their intervention type (e.g. psychoeducational, exercise).

appropriate instrument for the evaluation of the internal validity of different study designs such as RCTS, before and after studies and cross-sectional studies. The tool has been shown to have good test-retest reliability and construct validity. The tool is available at http://www.ephpp.ca. The quality of a study will be classified as strong, moderate or weak based on the following six domains: (1) selection bias; (2) study design; (3) confounders; (4) blinding; (5) data collection method; and (6) withdrawals/dropouts. A weak assessment of two or more criteria of the quality domains resulted in a weak global rating.

2.5. Assessment of risk of bias Two independent reviewers assessed the internal validity of the studies using the Effective Public Health Practice Project Quality Assessment Tool for Quantitative Studies Risk of Bias (EPHPP) developed by the Effective Public Health Practice Project in Canada [37,38]. Any disagreements were resolved by a third reviewer. The EPHPP is an

3. Results 3.1. Study selection In total, 5084 records were identified through database searches and a further 31 articles through hand searching

Screening

Idenficaon

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

Records idenfied through database searching Cochrane: 271 Web of Science: 2787 Pubmed: 1179 PsyIndex/Info: 847 (n = 5084)

Addional records idenfied through other sources (n = 31)

Records aer duplicates removed (n = 3388)

Eligibility

Records screened on basis of tle and abstract (n = 3419)

Included

375

Full-text arcles assessed for eligibility (n = 214)

Publicaons included in synthesis (n = 12) studies included in meta-analysis (n=7)

Records excluded (n = 3205)

Full-text arcles excluded (n = 202), following reasons: survivor: 38 poster: 1 assessment study: 2 case study: 2 age 39y: 90 no cancer: 5 intervenon for other: 1 other language: 2 no intervenon: 58 qualitave study: 1 invalid outcome: 2

Fig. 3. Flow chart of study selection process

already existing reviews, related articles and clinical trials databases. After removal of duplicates, 3419 abstracts were screened on the basis of title and abstract using the above mentioned inclusion criteria. 214 full-text articles were assessed for eligibility. 202 publications did not meet the inclusion criteria and were excluded due to the reasons displayed in Fig. 3. Most of the excluded studies focused on cancer diseases in children (39 years; n = 90), survivors of childhood cancer (n = 38), or conducted no intervention (n = 58). Finally, 12 publications were included in our review. Two publications contain results of the same sample. 3.2. Study characteristics A summary of our findings of the 12 included publications is shown in Table 3. The year of publication of the studies

ranged from 2007 to 2014. All but two studies were conducted in the United States (Canada n = 1; Norway n = 1). The studies provided data for 737 patients (n = 421 males; n = 316 females; the sample of two publications was counted only once). The age of study participants ranged between 11 and 45 years. The majority of included tumor entities were hematological malignancies. The studies were heterogeneous in design: randomized controlled trials (RCT; n = 8), observational before and after studies (n = 4). Two of the observational studies used mixed methods. The time span for the first post-treatment measurement ranged up to three months; the longest follow-up was one year after baseline assessment. Two interventions were provided through a group setting, one study used a mixture of group and individual settings and the remaining nine studies provided individual interventions. Eight interventions were conducted in the hospital or another healthcare institution.

376

Table 3 Summary of included studies. Author

Type of intervention

Randomized controlled trials (RCT) Beale et al.a [39] Video game play Intervention game US “Re-Mission” vs. commercial game

Sample characteristics

Instruments

Results

Global rating

Multi-site Two-arm RCT t1: baseline t2: 1 month t3: 3 months

n = 360 at baseline (191 IG; 169 CG) t2: 172 IG; 149 CG t3: 164 IG; 141CG Age 13–29 years

Cancer knowledge test Acceptability and Credibility Rating Scale

• Knowledge test scores improved significantly over the follow-up periods in both groups • Scores of IG improved significantly more (P < .001); size of this effect was related to usage of Re-Mission • Credibility scores were negatively correlated with level of knowledge but not with change in knowledge level at t2 or t3 • Immediate follow-up measures suggest positive trends in IG for hope, spirituality, confidence/mastery and self-transcendence • Positive trends at 100 days include MOS, symptoms distress, defensive coping, spirituality and self-transcendence • IG participants also demonstrated gains in quality of life • Increased cancer-specific knowledge regarding sexual issues; improved body image; lessened anxiety about sexual and romantic relationships; and decreased overall level of psychological distress in IG (P < .001) • Gains were maintained through the 3-month follow-up

Strong

Technology-based 49% leukemia 34% solid tumors 17% lymphoma Burns et al.b [40] US

Therapeutic music video vs. audiobook group during HSCT 3 weeks, sessions twice weekly á 60 min

Two-arm RCT t1: baseline t2: after session 6 t3: 100 days post transplant

n = 12 at baseline (7 IG; 5 CG) t2: 10 t3: 9 Age 13–24 years

STAIC, Bodily Pain Scale and Mental Health Scale from CHQ, MOS, SDS, JCS, HHI, SPS, Confidence Scale of NHS, STS, RSES, IWB, LASA Uniscale

n = 21 at baseline (10 IG; 11 CG) t2: 19 t3: 17 Age 15–25 years

Sexual Knowledge Test, The Body Image Instrument, The Sexual Concerns Questionnaire, BSI, CARES

Technology-based

Canada et al. [47] US

Counseling sessions and workbook vs. waitlist group Two counseling sessions á 90 min

Two-arm RCT t1: baseline t2: after session t3: 3 months

Psychoeducational

Huyghe et al. [41] Canada

Interactive, computerized tool “Banking on Fatherhood” vs. waitlist group Technology-based

Two-arm RCT

IG: 81.8% hematologic 18.2% solid tumors CG: 60.0% hematologic 40.0% solid tumors n = 20 male AYA (10 IG; 10 CG) Age 14–45 years 40% hematologic 40% other 20% testicular

Knowledge test, DCS Evaluation form

• IG had significantly less decisional conflict about banking sperm than those who had not viewed the tool (P = .0065) • Knowledge scores were not significantly different between groups • All men rated the tool as easy to use, informative and addressing important psychosocial concerns

Moderate

Strong

Weak

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

Study design

Jones et al. [42]

Interactive CD-ROM vs. handbook group

Multi-site Two-arm RCT

Technology-based

t1: baseline t2: 3 months

US

US

Robb et al. [44] US

Video game play Intervention game “Re-Mission” vs. commercial game, technology-based

Therapeutic music video vs. audiobook group during HSCT 3 weeks, sessions twice weekly á 60 min

Multi-site Two-arm RCT t1: baseline t2: 1 month t3: 3 months

Two-arm RCT t1: baseline t2: after session 6 t3: 100 days post transplant

Technology-based Valle et al. [52] US

Unsupervised physical activity intervention via Facebook FITNET Facebook group vs. Facebook-based Self-help comparison group 12 weekly Multi-component

Two-arm RCT t1: baseline t2: 12 weeks

n = 371 at baseline (195 IG; 176 CG) t2: 181 IG; 153 CG t3: 164 IG; 140 CG Age 13–29 years 48.5% leukemia 35.1% solid tumors 16.4% lymphoma n = 113 at baseline (59 IG; 54 CG) t2: 41 IG; 40 CG t3: 30 IG; 37 CG Age 11–24 years 46.4% leukemia 28.6% solid tumors 25.0% lymphoma n = 86 at baseline (45 IG; 41 CG) t2: 32 IG; 34 CG Age 21–39 years IG: 51.1% other 31.1% hematologic 17.8% breast CG: 46.3% other 31.7% hematologic 22.0% breast

MHLC, POQOLS, KIDCOPE, Cancer Knowledge Test, Self-efficacy Self report: acceptability, use, satisfaction Acceptability survey for parents and health care professionals Primary outcomes: CDCI, MAS, MEMS, Clinic Visit Attendance, 6-MP blood assays, Secondary outcomes: Self-efficacy Scale, Cancer Knowledge Scale, PedsQL, FACT-G, MHLC, PSS SDS, MUIS, JCS, SPS, Perceived Social Support Scales, FACES II, PACS, Family Strengths Scale, HHI, STS, HARS

Primary outcomes: GLTEQ, Self report: adherence and feasibility Secondary outcomes: FACT-G, BMI, Weight

• Teens in IG were significantly more likely to increase their internal locus of control scores (P = .016) • No significant differences were observed on other measures • Acceptability in IG was higher, but not different between two parent groups • 92% of HCP would recommend CD-ROM • Antibiotic adherence was greater in the intervention group (P = .012), also oral chemotherapy adherence (P = .002) • Self-efficacy (P = .011) and knowledge also increased in IG compared to CG (P = .035) • Intervention did not affect self-report measures of adherence, stress, control, or quality of life. • at t2 significant improvement of courageous coping in the IG (P = .030) • at t3 significant better social integration (P = .028) and family environment in the IG (P = .008) • no change of defensive coping strategies in both groups

• Increase of light physical activity in IG (P = .032) and weight loss (P = .083) over time • no difference of adherence between groups • no QOL difference over time in both groups • Participants in both groups would recommend the program (46.9%; 61.8%) to other AYA survivors

Weak

Strong

Moderate

Moderate

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

Kato et al. [43]

n = 65 at t1 and (35 IG; 30 CG) Age 12–18 years

377

378

Table 3 (Continued) Author

Type of intervention

Rosipal et al. [51] US

Physical exercise Video gaming exercise with Nintendo Wii, stationary bicycle with video games or DDR dance pad Multi-component

Hauken et al. [50]

Goal-orientated rehabilitation program

Norway 3 weeks residential rehabilitation with follow-up Multi-component

Study design

Sample characteristics

Instruments

Results

Global rating

Non randomized Longitudinal

n = 10 at baseline Age 14–18 years

PedsQL Teen Version, PedsQL-MFS, GLTEQ, Fitnessgram

Weak

t1: baseline t2: after week 8 t3: post treatment t4: 3 months t5: 1 year

40.0% lymphoma 40.0% leukemia 10.0% CNS tumors 10.0% other

• Significant improvement in all measures across 16 weeks • Significant fitness changes in upper body strength and flexibility (P < .05) • QOL improvement at t5 (P < .01) • Feasibility of a theoretically-based PA intervention for adolescents with a history of cancer

Non randomized Longitudinal

n = 18 Age 19–25 years

6-MWT, TUG, BASES-C

Weak

t1: baseline t2: 2 days before HSCT t3: 2 days after HSCT t4: 7 days after HSCT t5: 14 days after HSCT Non randomized Mixed method Longitudinal

50.0% leukemia 33.0% lymphoma 17.0% other

• Nintendo Wii was preferred active video gaming equipment • Greatest amount of exercise time engaged in walking by 53% of exercise time • Neither functional capacity nor quality of life improved

COPM at t1 to t4, in-depth interviews at the end of the intervention

• Significant changes in COPM (P < .001) performance and satisfaction which interpreted as demonstrating increased participation in life • High levels of goal achievement revealed in interviews • Goals are helpful in the rehabilitation process • Three subthemes emerged being important to achieve goals: (1) Building capacity/finding the balance (2) Gaining new insight and (3) Follow-up process

Weak

t1: baseline t2: after week 3 t3: after 3 months t4: after 6 months

n = 16 Age 25–25 years 32% gynecological 30% other 19% lymphoma 19% breast

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

Non randomized, pre/post studies Physical activity (PA) Keats and Culos-Reed group intervention [49] 16 weeks á 90 min Canada Physical activity and 8 weeks á 30 min educational session

44% leukemia 31% other 25% solid tumor

Abbreviations: 6-MP, 6-mercaptopurine; BSI, Brief Symptom Inventory; CARES, Cancer Rehabilitation Evaluation System; CDCI, Chronic Disease Compliance Instrument; CHQ, Children Health Questionnaire; CG, Control Group; DCS, Decisional Conflict Scale; FACES, Family Adaptability/Cohesion Scale; FACT-G, Functional Assessment of Cancer Therapy-General; GLTEQ, Godin Leisure Time Exercise Questionnaire; HARS, Haase Adolescent Resilience Scale; HCP, health care professional; HHI, Herth Hope Index; HSCT, hematopoietic stem cell transplant; IG, intervention group; IWB, Index of Well-being; JCS, Jalowiec Coping Scale; KIDCOPE, testing coping skill in children; LASA, Linear Analogue Self-Assessment; MAS, Medication Adherence Scale; MEMS, Medication Event Monitoring System; MHLC, Multidimensional Health Locus of Control Scale; MOS, Medical Outcomes Study; MUIS, Mishel Uncertainty in Illness Scale; NHS, Nowotny Hope Scale; PACS, Parent-Adolescent Communication Scale; POQOLS, Pediatric Oncology Quality of Life Scale; PedsQL, Pediatric Quality of Life Inventory; PSS, Perceived Stress Scale; QOL, Quality of Life; RSES, Rosenberg Self-Esteem Scale; SDS, Symptom Distress Scale; Reed Self-Transcendence Scale; SPS, Reed Spiritual Perspective Scale; STAIC, State Trait Anxiety Inventory for Children; t, time of measurement; 6-MWR, 6-Minute Walk Test; BASES-C, Behavioral, Affective and Somatic Experiences Scale – Child Version; GLTEQ, Godin Leisure Time Exercise Questionnaire; HSCT, hematopoietic stem cell transplantation; PedsQL, Pediatric Quality of Life Inventory; PedsQL-MFS, Teen Pediatric Quality of Life Multidimensional Fatigue Scale; TUG, Timed-Up-and-Go test; COPM, Canadian Occupational Performance Measure; STAI, State Trait Anxiety Inventory; STAIC, State Trait Anxiety Inventory for Children; t, time of measurement. a This study is part of a larger study (Kato et al. [6]). b This is the pilot study for Robb et al. [7].

Pre/posttest Physical exercise

One session á 45 min US

Thygeson et al. [48]

Yoga class intervention

Non randomized Mixed method

n = 5 AYA 11 children (6–12 years) 5 adolescents (13–18 years) 33 parents

STAIC, STAI Interview: open-ended question, “What would you tell another parent or child about what it is like to take a yoga class?”

• Children had normal anxiety scores preclass that did not change • Adolescents (P = .04) and parents (P < .01) experienced significant decreases in anxiety scores • All cohorts gave positive feedback about the experience

Weak

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

379

3.3. Types of intervention A quantitative synthesis could not be performed with all studies due to the heterogeneity of the included studies and missing data. A narrative approach was used to describe the effects of the interventions by classification into intervention types. 3.3.1. Technology-based interventions Six studies [39–44] examined the effects of technologybased interventions. The majority of these studies used multiple outcome parameters. Beale et al. [39] and Kato et al. [43] contrasted a group playing a commercial video game with a group playing additionally a third person video game (“Re-Mission”) in which the participants act as an avatar destroying cancer cells and other enemies to enhance self-care during cancer treatment. Primary outcomes were cancer knowledge [39] and treatment adherence [43]. Kato et al. [43] assessed effects on mental health including quality of life and distress as secondary outcomes. Participants in the intervention group significantly improved in therapy adherence, self-efficacy and cancer-related knowledge. There was no correlation between acceptability and knowledge scores. Knowledge scores in the study of Beale et al. [39] increased significantly over time (group × time interaction F (1,302 = 4.07, p = .04, f = .013)). Secondary outcomes (control, cancer knowledge, quality of life, distress) in the sample of Kato et al. [43] showed no significant changes from baseline to follow-up. Burns et al. [40] assessed the feasibility of a therapeutic music video intervention (TMV) based on the Resilience in Illness Model (RIM; formerly called Adolescent Resilience Model ARM) developed by Haase et al. [45,46]. Participants were randomized into an intervention and a control group. The intervention started three days before stem cell transplantation. TMV included an active-to-passive structure tailored to the increasing side-effects which occur during hematopoietic stem cell transplant (HSCT). Participants created new lyrics and digitally recorded them with the help of a music therapist. In the passive part of the intervention participants designed their videos to finally present the project to families and friends. The control group listened to self-chosen audio books and afterwards discussed the impressions and thoughts about the book content with a therapist. Burns et al. [40] report a good consent rate of 63% which indicates a good feasibility of the TMV. A trend for an increased quality of life was observed in the intervention group. This feasibility study was the pilot for the study of Robb et al. [44]. Here, the same study design was used to further investigate primary (e.g. family functioning or courageous coping) and secondary outcomes (e.g. distress or resilience). Significant improvements were reported for social integration and family environment. There was no effect for distress. Jones et al. [42] created a CD-ROM to empower cancer patients to “beat” cancer and return to a productive life. The CD contains several topics (e.g. side effects, treatment

380

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

plan or peers talking about their cancer) especially designed for adolescents with cancer. The user can navigate easily throughout the CD. The control group received a handbook which includes the same topics as the CD-ROM. Pre-postquestionnaires measured the domains coping, health locus of control, quality of life, cancer knowledge, self-efficacy and acceptability. Participants ranked the CD as first choice for a learning tool which indicates a good acceptability. The intervention group scored higher in the Health Locus of Control Scale, but results showed no significant differences between the two groups in the above mentioned outcome measures. All participants who used the CD wished that they had got the CD at diagnosis. Huyghe et al. [41] pilot tested a computerized educational tool for patients (and professionals). A CD for patients covered the issues of male reproduction, impact of cancer treatment, cost-benefit analysis of fertility preservation methods and a decision aid “The Sperm Banking Thermometer”. Participants had to complete a knowledge test and the Decisional Conflict Scale, intervention group after viewing the CD and control group before viewing. Decisional conflict was significantly lower for participants of the intervention group, but knowledge scores did not differ between the two groups. 3.3.2. Psychoeducational interventions Canada et al. [47] implemented a psychoeducational intervention to enhance psychosexual development using an intervention and waitlist group. Participants of the intervention group took part in two individual counseling sessions delivered by a psychologist which cover themes like fertility, body image or sexual function. Additionally, participants received a workbook with information about the impact of cancer therapy on sexual development and fertility. The intervention group with immediate-start achieved significant improvement in all domains from baseline to 3-months follow-up, especially knowledge scores, body image, level of distress and dating situations. The level of distress increased in participants in the waitlist group over three months. 3.3.3. Physical exercise Thygeson et al. [48] explored the feasibility and effect of a single hatha yoga session together for children and adolescents with cancer and their parents on anxiety. Anxiety was assessed with the State Trait Anxiety Inventory before and after the yoga session. An open-ended question (“What would you tell another parent or child what it is like to take a yoga class?”) was also added. The consent rate for adolescents was only 42% and a significant decrease in anxiety levels was observed. This effect was also observed in the parent group. Anxiety levels remained the same in the children group. The open-ended question revealed that adolescents benefit from the yoga session by feeling more relaxed, calmer and managing stress more effectively. Keats and Culos-Reed [49] developed a physical activity intervention based on the “Theory of Planned Behavior”.

Participants took part in a weekly group session. The sessions consisted of 45 min of aerobic exercise, 15 min of core strength/flexibility training and 30 min of educational program (attitude, perceived behavior control, subjective norm, and intentions) in the first eight weeks. The following eight weeks included noncompetitive physical activities. The primary purpose of the intervention was to investigate its feasibility. Secondary outcomes examined the effects of the intervention on quality of life, fatigue and physical activity behavior. The adherence rate of 81.5% indicates a good feasibility and acceptance of the program. There was a significant improvement in upper body strength and flexibility and quality of life post treatment. This finding persisted at the 1-year follow-up. 3.3.4. Multi-component interventions Hauken et al. [50] conducted a mixed-methods study on a Goal-Oriented Rehabilitation Program for young adult cancer survivors. The program included five domains: (1) peer-to-peer support (training sessions together), (2) individual follow-up by a multidisciplinary team of professionals, (3) psychoeducation combining counseling, reading literature and homework (knowledge about cancer survivorship issues and using cognitive therapy as a coping tool), (4) physical activity (individual and group trainings tailored for each participant) and (5) next-of-kin weekend to visit participants. The Canadian Occupational Performance Measure (COPM) was used as an assessment tool for setting goals and to measure multidimensional performance in all areas of life. An individual face-to-face interview about experiences and achievements at the end of the program completed the data-assessment. Significant changes were made in the domains performance and satisfaction of prioritized goals. The qualitative findings also emphasize the high levels of goal achievement. An increasing number of participants returned to work during the intervention. Participants highlighted the multidimensional follow-up and greatly valued the peer-topeer support. Priority was given to achieve goals in physical activity. Rosipal et al. [51] provided an intervention to examine physical activity during HSCT and the use of active video gaming (AVG) in contrast to standard exercise activities like walking or basketball. Participants were asked to select at least one AVG equipment and record their daily activities. Physical performance testing was conducted before, during and after HSCT at five time points. Beside physical activity Rosipal et al. [10] examined patients’ quality of life which decreased during the transplantation phase and increased at discharge, though not statistically significantly. The average time spent daily engaging in physical exercise was 36.5 min and this decreased in the days after HSCT. Participants spent more time with standard exercise activity than AVG equipment. Valle et al. [52] compared two groups concerning physical activity and feasibility as primary outcomes and body weight and quality of life as secondary outcomes. The intervention

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

is designed to increase participants’ moderate-to-vigorous physical activity (MPVA) with a focus on walking. All participants received a pedometer in order to record their daily steps. Participants in the control group “Self-help comparison (SC)” joined a Facebook group in which they got weekly messages with basic information about physical activity. They had the opportunity to contact each other in the absence of a professional moderator. The intervention group “FITNET” received the same components as the control group but in addition Facebook messages included expanded behavioral lessons and group discussions were moderated by a study administrator. Furthermore “FITNET” participants could use a website to keep a diary about their physical activity goals and achievements. This design was based on social cognitive theory to enhance self-efficacy, behavioral capability, self-monitoring and social support. The comparisons of the measured outcomes showed only a few group differences post-treatment. MPVA was not different between the two groups. The values of light physical activity (e.g. easy walking, yoga) increased in the FITNET group. An unexpected finding was the weight loss over time in the FITNET group. 3.4. Quality of included studies Quality assessment was carried out for 12 articles. For a global rating of each included study see Table 3. Three studies were rated as methodologically strong, three were moderate, and six were weak. A weak rating resulted mostly from selection bias, a lack of outcome assessors/participants blinding or confounders. The quality of all pre/post studies assessed was weak due the missing information about confounders and blinding. 3.5. Quantitative analysis 7 out of 12 publications were included in the metaanalyses. The four observational studies were not considered for meta-analyses [48–51]. These observational studies were non-randomized pre/post studies. Therefore, control group data were not available for meta-analyses. Potential biases and heterogeneity would be greater for non-randomized studies compared with randomized controlled trials. The four studies were presented at the end of Table 3. One study reported insufficient data to compute an effect size [39]. The authors of the mentioned publication were contacted, but no further data could be obtained. The overall effect size across studies for intervention effects on quality of life showed no significant difference (4 trials, standardized mean difference, SMD: 0.13, 95% CI: −0.16 to 0.42) between intervention and control group after the intervention and overall heterogeneity was not given (Fig. 4a). Meta-analyses of the studies regarding knowledge about cancer and associated concerns showed no statistically significant differences between the two groups (4 trials, SMD: 0.27, 95% CI: −0.22 to 0.76) but a moderate to high heterogeneity (Fig. 4b). Sensitivity analysis revealed one study

381

[47] with a great impact on overall effect and heterogeneity. There was a smaller effect size and lower heterogeneity when eliminating this study (SMD: 0.07, 95% CI: −0.23 to 0.38; I2 = 31%). The interventions had no significant overall effect on psychosocial distress (4 trials, SMD: −0.16, 95% CI: −0.73 to 0.42) and the heterogeneity for the effects was moderate to high (Fig. 4c). One study [43] had an impact on heterogeneity and overall effect size. After excluding this study from the calculation, heterogeneity was lower and effect size higher (SMD: −0.35, 95% CI: −0.77 to 0.06; I2 = 0%). Table 4 displayed effect sizes when analyzing intervention types separately. Small but significant effects were found in the psycho-educational intervention by Canada et al. [47], according an increasing knowledge and decreasing psychosocial distress.

4. Discussion AYAs are individuals who suffer from particularly high levels of distress during the different stages of their cancer (diagnosis, treatment and survivorship). Therefore, efforts are needed to develop interventions covering the specific needs of this age-group. A systematic review was undertaken in order to identify interventions for AYAs to improve their psychosocial wellbeing. To our knowledge, this is the first systematic review and meta-analysis to date examining the effects of psychosocial interventions on adolescent and young cancer patients. We included 12 publications in our review. However, the number of interventions is likely to be higher, remaining unidentified due to a lack of scientific evaluation or publication. One reason for the small study sizes could be difficulties recruiting patients to the AYA-group. Most participants of the included studies were contacted directly while in hospital, or through leaflets and bulletins. People between 15 and 39 years are very active and mobile. This makes it even more difficult to contact this age-group. Caldwell et al. found that parents influence the decision to decline participation in a clinical trial as a matter of safety [53]. AYAs want to be autonomous and choose themselves whether they take part in a clinical trial. Most clinical trials are RCTs, where AYAs are unable to choose a participant group [54]. This could decrease the motivation to take part in a study, because the patient could be assigned to an undesired treatment group [55]. Furthermore, cancer patients of the AYA-group have lower interest in study participation due to engagement in other activities [56,57]. Some studies however met these recruitment challenges, using alternative strategies. Hendricks-Ferguson et al. [58] reported good communication between study personnel as being beneficial to recruitment, while Rabin et al. [59] approached most of the AYAs through mailings. Harlan et al. [60] suggest combining a mixture of recruitment strategies using population-based cancer registries alongside social media networking, as well as websites describing the study, and using incentives.

382

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

Fig. 4. (a) Effect of interventions on quality of life. (b) Effect of interventions on knowledge. (c) Effect of interventions on psychosocial distress.

Canada [47] stated the stigma associated with mental health treatment can act as recruitment barrier (counseling was provided by a clinical psychologist). Training nurses and physical assistants to be able to offer psychological interventions may be a solution to this problem. Our quantitative analysis indicates that technologybased as well as psycho-educational, multi-component and

physical exercise interventions focused on AYAs have very small, non-significant effects. When analyzing the overall effect size using different outcomes, patients achieved the biggest improvement in cancer-related knowledge. The vast majority of the included studies had difficulties in recruiting an acceptable sample size with adequate power to detect significant differences between groups. Larger samples

Table 4 Effect sizes according to intervention type. Outcome

Type of intervention

Hedges’ g

95% CI

I2 (%)

P

k

Quality of life

Technology-based Multi-component Technology-based Psychoeducational Technology-based Psychoeducational

0.08 −0.01 0.07 1.43 0.07 −0.95

−0.27 to 0.44 −0.49 to 0.47 −0.23 to 0.38 0.38 to 2.47 −0.43 to 0.56 −1.92 to 0.02

0.0 Not applicable 31 Not applicable 62 Not applicable

.60 .96 .23 .007 .07 .06

3 1 3 1 3 1

Knowledge Distress

Abbreviations: I2 , measure of heterogeneity; k, number of studies.

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

are necessary to reach significant effect sizes. It is also questionable whether interventions with a non-randomized study design can reveal the intervention as the main cause for the measured effect, even if moderating variables are controlled for. Regardless, it cannot provide hard evidence for the effectiveness of one intervention over another when comparing non-randomized studies. Additionally, effects should be evaluated based on the research context and their practical or clinical value [61]. These findings highlight that it has not been possible to initialize randomized controlled study designs with a high external validity. Neither has it been possible to implement adequate healthcare models for AYAs which allow the accomplishment of research-based interventions. We cannot make any general statements about which intervention could be prioritized in the AYA-group but trends become apparent. A small superiority of the technologybased interventions could be demonstrated when analyzing all outcome measures and follow-ups. But multimedia-based interventions can only target patients with a computer, laptop, etc. As in the study of Valle et al. [52], patients without a Facebook account or no Internet access could not be reached, although Internet use is widely disseminated in the AYAgroup [25]. Psychosocial parameters were barely examined as primary outcomes. Several of the included studies focused on an increase of cancer-related knowledge or improvement of physical activity and strength. These improvements may claim to have an indirect positive impact on mental health. Hauken [50] hypothesized that setting physical goals is easier to measure, perhaps providing a good starting point for rehabilitation. Later on, patients could then focus more on psychosocial needs. The authors of the included studies suggested several modifications or improvements for future studies in order to improve psychosocial outcomes in AYAs. In Robb et al.’s study [44] patients underwent an intervention during HSCT – a time of high distress. Perhaps such interventions should focus more on relaxation and uncertainty management. Thygeson [48] yielded a significant decrease of anxiety in the AYA-group and their parents in a single yoga session. These results are in line with findings from other research groups [62,63]. However, Thygeson [48] only examined the effect of a single yoga session, so definitive conclusions could not to be drawn. In general, the sustainability and evidence of interventions is questionable and could be covered by follow-up measures or booster sessions after a certain amount of time. Keats [49], for example, encouraged examining the impact of long-term adherence to a physically active lifestyle after a physical exercise intervention. The preference for physical group activities is an important finding of Rosipal’s [51]. Walking groups (or if requested multiplayer games online) should be arranged to support the exchange between peers especially when receiving inpatient treatment. This is associated with the commonly expressed

383

desire for more socialization and peer support during hospitalization [64]. Finally, defining an age range fitting to AYAs remains difficult. The main concern of the present review was searching for interventions especially designed for cancer patients between 15 and 39 years. Only five of the included studies explicitly used the term of “AYA”. The remaining studies included “adolescents” or “young adults”. This indicates the difficulty in setting up an intervention incorporating a wide range of developmental age-groups. In general, the majority of studies offer interventions for children and/or adolescents up to 25 years. But patients up to 30 years old were not specifically considered and were often integrated in interventions for “older” adults. This is reflected in all but two of the included studies which offered interventions on participants only up to the age of 29 years. 4.1. Implications for future research In the future, AYA research should focus on psychosocial interventions with psychosocial functioning as the primary outcome measure. Half of our included studies examined the effects on psychosocial functioning as a secondary outcome. These studies concentrated on the improvement of knowledge, physical function, and therapy adherence. In this regard, it would be interesting to analyze the interdependence of different outcomes to see which complex interactions can be identified. This might include whether an increase of cancerrelated knowledge through psycho-educational interventions leads to a better quality of life or to what extent physical activity influences levels of distress in AYAs, something which has previously been demonstrated in other cancer patients [65,66]. Researchers could most benefit from RCTs that stratify according to age, when examining wide age-range. However, a detailed inquiry is needed into what is important to cancer patients between 15 and 39 years and what specific unmet needs they have. Additionally, recruitment strategies need reconsidering to attain more representative age samples in future studies. Future research should also evaluate age-appropriated interventions regarding the long-term effects and their impact of these on the survival time of AYAs. Due to the lack of studies dealing with AYAs in Germany, the investigation of interventions for this age-group appears to be an important area for future research. It also implies a need for stronger and closer networking within research groups in Germany and the USA to achieve synergies and create real progress in this field. 4.2. Limitations Our review has several methodological limitations. Due to the small number of included studies the results of the effect sizes of the meta-analysis must be interpreted with caution. Some researchers might criticize meta-analyses as inappropriate when including fewer than 20 studies [67]. However,

384

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

no fixed minimum number of studies is required. The results are also subject to some bias. By including only studies published in English or German an upward bias could have been raised because of a loss of information. To detect this reporting bias, funnel plots would have had to be analyzed. But we discarded the option of interpreting funnel plots due to the small study size. Similarly, moderator or subgroup analysis by existing heterogeneity of studies seemed unsuitable. Another problem occurred when a study had several outcome measures for the same construct. These measures are dependent and therefore the resulting effect sizes are likewise dependent. We solved the problem by eliminating the outcome measure which is less reliable or differed most from measures in other studies. There are other approaches to dealing with dependence in a methodologically proper way like calculating the covariance between different measures, which requires knowing the size of the correlation of the raw scores. But the necessary data for calculating these values were not available. We have to consider that some part of the differences between effect sizes is not only due to sampling error (it is the assumption under fixed effect models), but also a consequence of different interventions and furthermore of instruments of measure. There is a need for evidence-based quality assessment tools that are adaptable to different study designs such as case studies, feasibility studies and other non-randomized controlled studies. The majority of tools were developed to rate exclusively randomized controlled trials [68,69]. They usually contain a rating for blinding or confounding which is improper for studies without control groups, making rating difficult and potentially resulting in a weak assessment of the outcome measures. Practical and ethical considerations can often also make blinding impossible due to the nature of the intervention being used, resulting in increased bias [70]. The effects of uncontrolled single group studies should be interpreted with caution due to the absence of a control group. Finally follow-up data were not considered in the meta-analyses, so it is not possible to make statements about any long-term effects.

5. Conclusion The present systematic review and meta-analysis strengthens the need for age-appropriate interventions in the field of psycho-oncology. However, due to the limited number of available studies, caution is needed when interpreting the effects. But trends can be distinguished: psycho-educational interventions and the use of multimedia are suited for adolescent and young cancer patients. The included studies may be considered suitable as “start-up” studies for AYA-related topics. Future interdisciplinary research is needed for interventions aimed at more specific age-groups. Randomized intervention studies with larger samples and a focus on psychosocial outcomes are also necessary to assure more reliable results.

Authors’ contribution DR conducted the literature search and wrote the paper. MK devised the protocol and made a decision on study inclusion and data extraction together with DR. MF and DR conducted the meta-analysis. GW and DR investigated risk of bias of the included studies. GW, IH and AM were advisors in the review. All authors provided comments and approved the final manuscript.

Conflict of interest The authors declare no conflict of interest or competing financial interests.

Reviewers Elisabeth Andritsch, Clinical Division of Oncology, Medical University of Graz, Auenbruggerplatz 15, AT-8036 Graz, Austria.

Acknowledgements This study received no specific grant from any funding agency. We would like to thank Christiane Hofmann for instructions for the literature search, Catherine Shneerson for the English language edit and also Dr. Jochen Ernst for his comments on the manuscript.

References [1] Coccia PF, Altman J, Bhatia S, et al. Adolescent and young adult oncology. Clinical practice guidelines in oncology. J Natl Compr Cancer Netw 2012;10(9):1112–50. [2] Treadgold CL, Kuperberg A. Been there, done that, wrote the blog: the choices and challenges of supporting adolescents and young adults with cancer. J Clin Oncol 2010;28(32):4842–9. [3] McGoldrick D, Gordon P, Whiteson M, Adams H, Rogers P, Sutcliffe S. Awareness and advocacy for adolescents and young adults with cancer. Cancer 2011;117(S10):2311–5. [4] van Gaal JC, Bastiaannet E, Schaapveld M, et al. Cancer in adolescents and young adults in north Netherlands (1989–2003): increased incidence, stable survival and high incidence of second primary tumours. Ann Oncol 2008;20(2):365–73. [5] GEKID. Gesellschaft der epidemiologischen Krebsregister in Deutschland e.V. (GEKID): Atlas der Krebsinzidenz und–mortalität in Deutschland. http://www.ekr.med.uni-erlangen.de/GEKID/ [accessed 03.04.14]. [6] Veal GJ, Hartford CM, Stewart CF. Clinical pharmacology in the adolescent oncology patient. J Clin Oncol 2010;28(32):4790–9. [7] Ramphal R, Meyer R, Schacter B, Rogers P, Pinkerton R. Active therapy and models of care for adolescents and young adults with cancer. Cancer 2011;117(S10):2316–22.

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386 [8] Fern LA, Whelan JS. Recruitment of adolescents and young adults to cancer clinical trials—international comparisons, barriers, and implications. Semin Oncol 2010;37(2):e1–8. [9] Bleyer A, Barr R, Hayes-Lattin B, Thomas D, Ellis C, Anderson B. The distinctive biology of cancer in adolescents and young adults. Nat Rev Cancer 2008;8(4):288–98. [10] Hilgendorf I, Freund M, Kropp P, et al. Onkologische Erkrankungen bei Jugendlichen und jungen Erwachsenen – Zahlen aus dem Klinischen Krebsregister Rostock als Arbeitsgrundlage für eine interdisziplinäre Herausforderung. Tumor Diagn Ther 2011;32(2):85–92. [11] Woodward E, Jessop M, Glaser A, Stark D. Late effects in survivors of teenage and young adult cancer: does age matter? Ann Oncol 2011;22(12):2561–8. [12] Zebrack B, Isaacson S. Psychosocial care of adolescent and young adult patients with cancer and survivors. J Clin Oncol 2012;30(11): 1221–6. [13] D’Agostino NM, Penney A, Zebrack B. Providing developmentally appropriate psychosocial care to adolescent and young adult cancer survivors. Cancer 2011;117(S10):2329–34. [14] Pritchard S, Cuvelier G, Harlos M, Barr R. Palliative care in adolescents and young adults with cancer. Cancer 2011;117(S10):2323–8. [15] Erickson JM, MacPherson CF, Ameringer S, Baggott C, Linder L, Stegenga K. Symptoms and symptom clusters in adolescents receiving cancer treatment: a review of the literature. Int J Nurs Stud 2013;50(6):847–69. [16] Yanez B, Garcia SF, Victorson D, Salsman JM. Distress among young adult cancer survivors: a cohort study [accessed 28.01.14]. [17] Kwak M, Zebrack BJ, Meeske KA, et al. Trajectories of psychological distress in adolescent and young adult patients with cancer: a 1-year longitudinal study. J Clin Oncol 2013;31(17):2160–6. [18] Erickson JM, Beck SL, Christian BR, et al. Fatigue, sleep-wake disturbances, and quality of life in adolescents receiving chemotherapy. J Pediatr Hematol Oncol 2011;33(1):e17. [19] Daniel LC, Brumley LD, Schwartz LA. Fatigue in adolescents with cancer compared to healthy adolescents. Pediatr Blood Cancer 2013;60(11):1902–7. [20] Smith AW, Bellizzi KM, Keegan THM, et al. Health-related quality of life of adolescent and young adult patients with cancer in the United States: The Adolescent and Young Adult Health Outcomes and Patient Experience Study. J Clin Oncol 2013;31(17):2136–45. [21] Kwak M, Zebrack BJ, Meeske KA, et al. Prevalence and predictors of post-traumatic stress symptoms in adolescent and young adult cancer survivors: a 1-year follow-up study. Psychooncology 2013;22(8):1798–806. [22] Dyson GJ, Thompson K, Palmer S, Thomas DM, Schofield P. The relationship between unmet needs and distress amongst young people with cancer. Support Care Cancer 2012;20(1):75–85. [23] Kelly D. Developing age appropriate psychosexual support for adolescent cancer survivors: a discussion paper. J Sex Med 2013;10:133–8. [24] Murphy D, Klosky JL, Termuhlen A, Sawczyn KK, Quinn GP. The need for reproductive and sexual health discussions with adolescent and young adult cancer patients. Contraception 2013;88(2):215–20. [25] Schiffman JD, Csongradi E, Suzuki LK. Internet use among adolescent and young adults (AYA) with cancer. Pediatr Blood Cancer 2008;51(3):410–5. [26] Hauken MA, Larsen TMB, Holsen I. Meeting reality. Cancer Nurs 2013;36(5):E17–26. [27] Hall AE, Boyes AW, Bowman J, Walsh RA, James EL, Girgis A. Young adult cancer survivors’ psychosocial well-being: a cross-sectional study assessing quality of life, unmet needs, and health behaviors [accessed 28.01.14]. [28] Geue K, Richter D, Schmidt R, et al. The desire for children and fertility issues among young German cancer survivors. J Adolesc Health 54 (5), 2014, 527-535. [29] Richardson WS, Wilson MC, Nishikawa J, Hayward RS. The wellbuilt clinical question: a key to evidence-based decisions. ACP J Club 1995;123(3):A12–3.

385

[30] Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ 2009;339(July (1)):b2700. [31] Ranmal R, Prictor M, Scott JT. Interventions for improving communication with children and adolescents about their cancer. Cochrane Database Syst Rev 2008;(4). [32] Sansom-Daly UM, Peate M, Wakefield CE, Bryant RA, Cohn RJ. A systematic review of psychological interventions for adolescents and young adults living with chronic illness. Health Psychol 2012;31(3):380–93. [33] Seitz DC, Besier T, Goldbeck L. Psychosocial interventions for adolescent cancer patients: a systematic review of the literature. Psychooncology 2009;18(7):683–90. [34] Scott JT, Harmsen M, Prictor MJ, Sowden AJ, Watt I. Interventions for improving communication with children and adolescents about their cancer. Cochrane Database Syst Rev 2003;(3):CD002969. [35] Borenstein Introduction to Meta Analysis. New York, NY: John Wiley and Sons Ltd; 2009. [36] Ried K. Interpreting and understanding meta-analysis graphs – a practical guide. Aust Fam Phys 2006;35(8):635–8. [37] Thomas B, Ciliska D, Dobbins M, Micucci S. A process for systematically reviewing the literature: providing the research evidence for public health nursing interventions. Worldviews Evid-Based Nurs 2004;1(3):176–84. [38] Armijo-Olivo S, Stiles CR, Hagen NA, Biondo PD, Cummings GG. Assessment of study quality for systematic reviews: a comparison of the Cochrane Collaboration Risk of Bias Tool and the Effective Public Health Practice Project Quality Assessment Tool: methodological research. J Eval Clin Pract 2012;18(1):12–8. [39] Beale IL, Kato PM, Marin-Bowling VM, Guthrie N, Cole SW. Improvement in cancer-related knowledge following use of a psychoeducational video game for adolescents and young adults with cancer. J Adolesc Health 2007;41(3):263–70. [40] Burns DS, Robb SL, Haase JE. Exploring the Feasibility of a Therapeutic Music Video Intervention in Adolescents and Young Adults During Stem-Cell Transplantation. Cancer Nurs 2009;32(5):E8. [41] Huyghe E, Martinetti P, Sui D, Schover LR. Banking on Fatherhood: pilot studies of a computerized educational tool on sperm banking before cancer treatment. Psychooncology 2009;18(9):1011–4. [42] Jones JK, Kamani SA, Bush PJ, Hennessy KA, Marfatia A, Shad AT. Development and evaluation of an educational interactive CD-ROM for teens with cancer. Pediatr Blood Cancer 2010;55(3):512–9. [43] Kato PM, Cole SW, Bradlyn AS, Pollock BH. A video game improves behavioral outcomes in adolescents and young adults with cancer: a randomized trial. Pediatrics 2008;122(2):e305–17. [44] Robb SL, Burns DS, Stegenga KA, et al. Randomized clinical trial of therapeutic music video intervention for resilience outcomes in adolescents/young adults undergoing hematopoietic stem cell transplant: A report from the Children’s Oncology Group. Cancer 2014;120(6):909–17. [45] Haase JE, Kintner EK, Monahan PO, Robb SL. The resilience in illness model, Part 1. Cancer Nurs 2014:1. [46] Haase JE. The adolescent resilience model as a guide to interventions [accessed 11.03.14]. [47] Canada AL, Schover LR, Li Y. A pilot intervention to enhance psychosexual development in adolescents and young adults with cancer. Pediatr Blood Cancer 2007;49(6):824–8. [48] Thygeson MV, Hooke MC, Clapsaddle J, Robbins A, Moquist K. Peaceful play yoga: serenity and balance for children with cancer and their parents. J Pediatr Oncol Nurs 2010;27(5):276–84. [49] Keats Culos-Reed SN. A community-based physical activity program for adolescents with cancer (Project TREK) – program feasibility and preliminary findings. J Pediatr Hematol Oncol 2008;30(4): 272–80. [50] Hauken MA, Holsen I, Fismen E, Larsen TMB. Participating in life again. Cancer Nurs 2014:1.

386

D. Richter et al. / Critical Reviews in Oncology/Hematology 95 (2015) 370–386

[51] Rosipal NC, Mingle L, Smith J, Morris GS. Assessment of voluntary exercise behavior and active video gaming among adolescent and young adult patients during hematopoietic stem cell transplantation. J Pediatr Oncol Nurs 2013;30(1):24–33. [52] Valle CG, Tate DF, Mayer DK, Allicock M, Cai J. A randomized trial of a Facebook-based physical activity intervention for young adult cancer survivors. J Cancer Surviv 2013;7(3):355–68. [53] Caldwell PH, Butow PN, Craig JC. Parents’ attitudes to children’s participation in randomized controlled trials. J Pediatr 2003;142(5):554–9. [54] Broome ME, Richards DJ. The influence of relationships on children’s and adolescents’ participation in research. Nurs Res 2003;52(3):191–7. [55] Bleyer A, Budd T, Montello M. Adolescents and young adults with cancer: the scope of the problem and criticality of clinical trials. Cancer 2006;107(7 Suppl.):1645–55. [56] Gattuso J, Hinds P, Tong X, Srivastava K. Monitoring child and parent refusals to enrol in clinical research protocols. J Adv Nurs 2006;53(3):319–26. [57] Ward-Smith P, Hamlin J, Bartholomew J, Stegenga K. Quality of life among adolescents with cancer. J Pediatr Oncol Nurs Off J Assoc Pediatr Oncol Nurses 2007;24(3):166–71. [58] Hendricks-Ferguson VL, Cherven BO, Burns DS, et al. Recruitment strategies and rates of a multi-site behavioral intervention for adolescents and young adults with cancer [accessed 19.03.14]. [59] Rabin C, Horowitz S, Marcus B. Recruiting young adult cancer survivors for behavioral research. J Clin Psychol Med Settings 2013;20(1):33–6. [60] Harlan LC, Lynch CF, Keegan THM, et al. Recruitment and follow-up of adolescent and young adult cancer survivors: the AYA HOPE Study. J Cancer Surviv 2011;5(3):305–14. [61] Durlak JA. How to select, calculate, and interpret effect sizes. J Pediatr Psychol 2009;34(9):917–28. [62] Park CL, Cho D, Wortmann JH. The impact of Yoga upon young adult cancer survivors [accessed 19.03.14]. [63] Buffart LM, van Uffelen JGZ, Riphagen II, et al. Physical and psychosocial benefits of yoga in cancer patients and survivors, a systematic review and meta-analysis of randomized controlled trials. BMC Cancer 2012;12(1):559. [64] Nass SJ, Patlak M. Identifying and addressing the needs of adolescents and young adults with cancer, workshop summary; 2013. [65] Buffart LM, Thong MSY, Schep G, et al. Self-reported physical activity: its correlates and relationship with health-related quality of life in a large cohort of colorectal cancer survivors. PLoS ONE 2012;7(5):e36164.

[66] Zhao G, Li C, Li J, Balluz L. Physical activity, psychological distress, and receipt of mental healthcare services among cancer survivors. J Cancer Surviv 2013;7(1):131–9. [67] Koschack J. Weder heiliger Gral noch Teufelswerk: eine methodenkritische Bewertung der Meta-Analyse. Dtsch med Wochenschr 2009;134(48):2465–8. [68] Clark HD, Wells GA, Huët C, et al. Assessing the quality of randomized trials: reliability of the Jadad scale. Control Clin Trials 1999;20(5):448–52. [69] Higgins JPT, Altman DG, Gotzsche PC, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials [accessed 07.04.14]. [70] Higgins J, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. [S.l.]: The Cochrane Collaboration; 2011.

Biographies Diana Richter, M.A., is a social scientist and research fellow at the Department of Medical Psychology & Medical Sociology, University of Leipzig (Germany). Her main field of scientific interest is the life situation of chronically ill adolescents and young adults with the focus on quality of life, family planning and psycho-social interventions. She is a member of the workgroup “Psycho-Oncology” of the German Cancer Society (DKG). Since 2014 she has been in training as a child and youth psychotherapist. Gregor Weißflog, PhD, is a psychologist and research fellow at the Department of Medical Psychology & Medical Sociology, University of Leipzig (Germany). He is head of the working group “Interventions in Psycho-Social Oncology”. His main fields of scientific interest are coping with cancer, psycho-social interventions in cancer patients including psychotherapeutic processes. Beyond this, he is engaged in the investigation of long-term mental and physical health consequences after political persecution.